Namespaces
Variants
Actions

Difference between revisions of "Functional equation, methods of solution of a"

From Encyclopedia of Mathematics
Jump to: navigation, search
(Importing text file)
 
(latex details)
 
(One intermediate revision by one other user not shown)
Line 1: Line 1:
 +
<!--
 +
f0420701.png
 +
$#A+1 = 201 n = 0
 +
$#C+1 = 201 : ~/encyclopedia/old_files/data/F042/F.0402070 Functional equation, methods of solution of a
 +
Automatically converted into TeX, above some diagnostics.
 +
Please remove this comment and the {{TEX|auto}} line below,
 +
if TeX found to be correct.
 +
-->
 +
 +
{{TEX|auto}}
 +
{{TEX|done}}
 +
 
Methods for finding exact or approximate solutions of specific or abstract functional equations, that is, equations of the form
 
Methods for finding exact or approximate solutions of specific or abstract functional equations, that is, equations of the form
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420701.png" /></td> <td valign="top" style="width:5%;text-align:right;">(1)</td></tr></table>
+
$$ \tag{1 }
 +
P ( x) =  y
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420702.png" /> is a certain, generally speaking non-linear, operator transforming the elements of a <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420703.png" />-space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420704.png" /> (or a space of some other type) into elements of a space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420705.png" /> of the same type (see [[Functional equation|Functional equation]]). Exact solutions in the form of analytic expressions have been obtained only for a few types of functional equations, therefore approximate methods of solution are especially valuable.
+
where $  P ( x) $
 +
is a certain, generally speaking non-linear, operator transforming the elements of a $  B $-
 +
space $  X $(
 +
or a space of some other type) into elements of a space $  Y $
 +
of the same type (see [[Functional equation|Functional equation]]). Exact solutions in the form of analytic expressions have been obtained only for a few types of functional equations, therefore approximate methods of solution are especially valuable.
  
 
A number of methods have evolved to find solutions of general functional equations of the form (1). For example, the method of infinite power series, the method of successive approximations, the [[Galerkin method|Galerkin method]] (the method of moments), the method of tangent hyperbolas, Chebyshev's method (of tangent parabolas), the Newton–Kantorovich method and its modifications, the method of steepest descent (cf. [[Steepest descent, method of|Steepest descent, method of]]), etc., as well as the method of variation of the parameter (direct, iterative and combined, cf. [[Parameter, method of variation of the|Parameter, method of variation of the]]) of particular types and its various modifications, including successive approximations to the inverse operator. General methods have been applied to solve various specific functional equations in mathematical analysis. In addition there are special methods for solving specific functional equations, including numerical methods, for example, the [[Grid method|grid method]], etc. The method of variation of the parameter, the Newton–Kantorovich method and certain other methods indicated also have theoretical significance, since they can be used to draw conclusions about the existence, uniqueness and the domain of location of a solution of a functional equation without finding the solution itself, which is sometimes no less important then the actual value of the solution. The direct method of variation of the parameter is considered below as an example.
 
A number of methods have evolved to find solutions of general functional equations of the form (1). For example, the method of infinite power series, the method of successive approximations, the [[Galerkin method|Galerkin method]] (the method of moments), the method of tangent hyperbolas, Chebyshev's method (of tangent parabolas), the Newton–Kantorovich method and its modifications, the method of steepest descent (cf. [[Steepest descent, method of|Steepest descent, method of]]), etc., as well as the method of variation of the parameter (direct, iterative and combined, cf. [[Parameter, method of variation of the|Parameter, method of variation of the]]) of particular types and its various modifications, including successive approximations to the inverse operator. General methods have been applied to solve various specific functional equations in mathematical analysis. In addition there are special methods for solving specific functional equations, including numerical methods, for example, the [[Grid method|grid method]], etc. The method of variation of the parameter, the Newton–Kantorovich method and certain other methods indicated also have theoretical significance, since they can be used to draw conclusions about the existence, uniqueness and the domain of location of a solution of a functional equation without finding the solution itself, which is sometimes no less important then the actual value of the solution. The direct method of variation of the parameter is considered below as an example.
  
In the Banach space <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420706.png" /> of linear bounded operators, let a non-linear operator ordinary differential equation be given on the infinite interval:
+
In the Banach space $  [ X] = [ X \rightarrow X] $
 +
of linear bounded operators, let a non-linear operator ordinary differential equation be given on the infinite interval:
 +
 
 +
$$ \tag{2 }
 +
 
 +
\frac{dx }{d \lambda }
 +
  = \
 +
x ( I - Ax) \ \
 +
(=  ( I - xA) x),\ \
 +
x ( 0)  =  x _ {0} ,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420707.png" /></td> <td valign="top" style="width:5%;text-align:right;">(2)</td></tr></table>
+
where  $  A, x _ {0} , I \in [ X] $(
 +
$  A $,
 +
$  x _ {0} $
 +
are given,  $  I $
 +
is the identity operator), and  $  x ( \lambda ) $
 +
is an abstract function with values in  $  [ X] $(
 +
$  X $
 +
is a  $  B $-
 +
space). The problem of constructing the inverse operator  $  A  ^ {-1} $,
 +
for an invertible operator  $  A $,
 +
the solution of a linear operator equation of the form  $  I - Ax = 0 $,
 +
etc., are all reduced to solving equation (2) by the direct method of variation of the parameter. If the spectrum of the operator  $  Ax _ {0} $(
 +
respectively,  $  x _ {0} A $)
 +
is situated in the right half-plane, then problem (2) has a unique solution  $  x ( \lambda ) $,
 +
$  x ( 0) = x _ {0} $,
 +
$  \lambda \in [ 0, \infty ) $.
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420708.png" /> (<img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f0420709.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207010.png" /> are given, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207011.png" /> is the identity operator), and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207012.png" /> is an abstract function with values in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207013.png" /> (<img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207014.png" /> is a <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207015.png" />-space). The problem of constructing the inverse operator <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207016.png" />, for an invertible operator <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207017.png" />, the solution of a linear operator equation of the form <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207018.png" />, etc., are all reduced to solving equation (2) by the direct method of variation of the parameter. If the spectrum of the operator <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207019.png" /> (respectively, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207020.png" />) is situated in the right half-plane, then problem (2) has a unique solution <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207021.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207022.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207023.png" />.
+
An analytic application to the Cauchy problem (2) of methods for the numerical integration of ordinary differential equations leads to a whole class of direct methods of variation of the parameter for solving (2), and consequently also to the solutions of problems that reduce to (2). For example, the [[Euler method|Euler method]] with irregular step  $  h _ {k} $
 +
leads to the following method for constructing  $  A  ^ {-1} $:
  
An analytic application to the Cauchy problem (2) of methods for the numerical integration of ordinary differential equations leads to a whole class of direct methods of variation of the parameter for solving (2), and consequently also to the solutions of problems that reduce to (2). For example, the [[Euler method|Euler method]] with irregular step <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207024.png" /> leads to the following method for constructing <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207025.png" />:
+
$$ \tag{3 }
 +
x _ {k + 1 }  = \
 +
x _ {k} + h _ {k} x _ {k} ( I - Ax _ {k} ) \ \
 +
( =  x _ {k} + h _ {k} ( I - x _ {k} A) x _ {k} ),
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207026.png" /></td> <td valign="top" style="width:5%;text-align:right;">(3)</td></tr></table>
+
$$
 +
= 0, 1 , .  . . .
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207027.png" /></td> </tr></table>
+
The steps  $  h _ {k} $
 +
in (3) are chosen in various ways. When the spectrum of  $  I - Ax _ {0} $(
 +
respectively,  $  I - x _ {0} A $)
 +
is on the real axis in an interval  $  [- \rho _ {0} , 0] $,
 +
$  0 < \rho _ {0} < \infty $,
 +
then the following method for choosing  $  h _ {k} $
 +
is very effective:
  
The steps <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207028.png" /> in (3) are chosen in various ways. When the spectrum of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207029.png" /> (respectively, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207030.png" />) is on the real axis in an interval <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207031.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207032.png" />, then the following method for choosing <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207033.png" /> is very effective:
+
$$ \tag{4 }
 +
h _ {k}  = \
 +
{
 +
\frac{1}{1 + \rho _ {k} }
 +
} ,\ \
 +
\rho _ {k + 1 }  = \
 +
\rho _ {k} L _ {k} ,\ \
 +
4L _ {k}  = \
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207034.png" /></td> <td valign="top" style="width:5%;text-align:right;">(4)</td></tr></table>
+
\frac{\rho _ {k} }{1 + \rho _ {k} }
 +
,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207035.png" /></td> </tr></table>
+
$$
 +
= 0, 1 \dots \  h _ {k + 1 }  =
 +
\frac{4h _ {k} }{( 1 + h _ {k} )  ^ {2} }
 +
.
 +
$$
  
Here, the rate of convergence in (3) is of order higher than <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207036.png" />, and the norm of the discrepancy decreases according to the law <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207037.png" />. The case of an interval <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207038.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207039.png" />, can be reduced, with advantage, to the one-step method (3) considered, where
+
Here, the rate of convergence in (3) is of order higher than $  2  ^ {k} $,  
 +
and the norm of the discrepancy decreases according to the law $  \rho _ {k + 1 }  = \rho _ {k}  ^ {2} /4 ( 1 + \rho _ {k} ) $.  
 +
The case of an interval $  [ 0, \overline \rho \; _ {0} ] $,  
 +
$  0 < \overline \rho \; _ {0} < 1 $,  
 +
can be reduced, with advantage, to the one-step method (3) considered, where
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207040.png" /></td> </tr></table>
+
$$
 +
= \
 +
{
 +
\frac{1}{1 - \overline \rho \; _ {0} }
 +
} ,\ \
 +
\rho _ {0}  = \
 +
 
 +
\frac{\overline \rho \; {} _ {0}  ^ {2} }{4 ( 1 - \overline \rho \; _ {0} ) }
 +
.
 +
$$
  
 
Regarding the convergence of (3), (4), there are some results, stated for specific operator equations, based on general facts and depending on which space is taken as a base.
 
Regarding the convergence of (3), (4), there are some results, stated for specific operator equations, based on general facts and depending on which space is taken as a base.
Line 33: Line 115:
 
The problem of constructing projections of the form
 
The problem of constructing projections of the form
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207041.png" /></td> </tr></table>
+
$$
 +
\left . P ( Ax _ {0} ) \ \
 +
( \textrm{ or }  P ( x _ {0} A)),\ \
 +
P ( K)  = \
 +
e ^ {- K \lambda }
 +
\right | _ {\lambda = \infty }  ,\ \
 +
K \in [ X],
 +
$$
 +
 
 +
also reduces to the solution of problem (2) (the case when the spectrum of  $  I - Ax _ {0} $(
 +
respectively,  $  I - x _ {0} A $)
 +
is in  $  [- \rho _ {0} , 1] $)
 +
and so does the problem of constructing pseudo-inverse operators  $  x  ^ {+} $(
 +
respectively,  $  {}  ^ {+} x $)
 +
for the operator  $  A $
 +
such that
 +
 
 +
$$
 +
I - Ax  ^ {+}  = \
 +
P ( Ax _ {0} ),\ \
 +
I - {}  ^ {+} xA  = \
 +
P ( x _ {0} A),\ \
 +
x  ^ {+}  = {}  ^ {+} x.
 +
$$
  
also reduces to the solution of problem (2) (the case when the spectrum of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207042.png" /> (respectively, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207043.png" />) is in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207044.png" />) and so does the problem of constructing pseudo-inverse operators <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207045.png" /> (respectively, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207046.png" />) for the operator <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207047.png" /> such that
+
For a direct construction of the projections  $  P ( Ax _ {0} ) $(
 +
respectively, $  P ( x _ {0} A) $),
 +
the method (3), (4) can be rewritten in the form
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207048.png" /></td> </tr></table>
+
$$
 +
P _ {k + 1 }  = \
 +
P _ {k} + h _ {k} P _ {k} ( P _ {k} - I),\ \
 +
k = 0, 1 \dots
 +
$$
  
For a direct construction of the projections <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207049.png" /> (respectively, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207050.png" />), the method (3), (4) can be rewritten in the form
+
$$
 +
P _ {0}  = I - Ax _ {0} \  (= I - x _ {0} A).
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207051.png" /></td> </tr></table>
+
Depending on the location of the spectrum and the properties of  $  A $
 +
one can take, for example, the operators  $  I $,
 +
$  A $,
 +
$  A  ^ {*} $,
 +
etc., in place of  $  x _ {0} $.
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207052.png" /></td> </tr></table>
+
Using the direct method of variation of the parameter one can reduce to the abstract linear functional ordinary differential equation  $  ( 0 \leq  \lambda \leq  \infty ) $
  
Depending on the location of the spectrum and the properties of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207053.png" /> one can take, for example, the operators <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207054.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207055.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207056.png" />, etc., in place of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207057.png" />.
+
$$ \tag{5 }
  
Using the direct method of variation of the parameter one can reduce to the abstract linear functional ordinary differential equation <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207058.png" />
+
\frac{dy }{d \lambda }
 +
  = \
 +
x _ {0} ( b - Ay),\ \
 +
y ( 0)  = y _ {0} ,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207059.png" /></td> <td valign="top" style="width:5%;text-align:right;">(5)</td></tr></table>
+
where  $  b, y _ {0} \in X $,
 +
$  x _ {0} , A \in [ X] $
 +
and  $  y ( \lambda ) $
 +
is an abstract function with values in  $  X $,
 +
the problem of constructing directly the pseudo-solutions  $  y  ^ {+} $
 +
of a linear functional equation of the form
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207060.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207061.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207062.png" /> is an abstract function with values in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207063.png" />, the problem of constructing directly the pseudo-solutions <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207064.png" /> of a linear functional equation of the form
+
$$ \tag{6 }
 +
b - Ay  ^ {+}  = \
 +
P ( Ax _ {0} )
 +
( b - Ay _ {0} )
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207065.png" /></td> <td valign="top" style="width:5%;text-align:right;">(6)</td></tr></table>
+
or of the functional equations  $  b - Ay  ^ {+} = 0 $
 +
if  $  P ( Ax _ {0} ) ( b - Ay _ {0} ) = 0 $.  
 +
Other problems also reduce to (5), including problems for ordinary and partial differential equations, integral equations, etc. The formula
  
or of the functional equations <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207066.png" /> if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207067.png" />. Other problems also reduce to (5), including problems for ordinary and partial differential equations, integral equations, etc. The formula
+
$$
 +
y ( \lambda )  = \
 +
e ^ {- x _ {0} A \lambda }
 +
y _ {0} + r ( \lambda )
 +
x _ {0} b,\ \
 +
r ( \lambda )  = \
 +
\int\limits _ { 0 } ^  \lambda 
 +
e ^ {- x _ {0} A ( \lambda - s) }  ds,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207068.png" /></td> </tr></table>
+
gives the unique solution to equation (5) satisfying  $  y ( 0) = y _ {0} $.
  
gives the unique solution to equation (5) satisfying <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207069.png" />.
+
For example, applying Euler's method with forward step  $  h _ {k + 1 }  $
 +
to problem (5) leads to the following method for constructing pseudo-solutions  $  y  ^ {+} $
 +
to (6):
  
For example, applying Euler's method with forward step <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207070.png" /> to problem (5) leads to the following method for constructing pseudo-solutions <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207071.png" /> to (6):
+
$$ \tag{7 }
 +
y _ {k + 1 }  = \
 +
y _ {k} + h _ {k + 1 }
 +
x _ {0} ( b - Ay _ {k} ),\ \
 +
k = 0, 1 , .  . . .
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207072.png" /></td> <td valign="top" style="width:5%;text-align:right;">(7)</td></tr></table>
+
The steps  $  h _ {k + 1 }  $
 +
are chosen, for example, in terms of the roots of the [[Chebyshev polynomials|Chebyshev polynomials]]  $  T _ {N} ( t) $:
  
The steps <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207073.png" /> are chosen, for example, in terms of the roots of the [[Chebyshev polynomials|Chebyshev polynomials]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207074.png" />:
+
$$ \tag{8 }
 +
h _ {j}  = \
 +
{
 +
\frac{2}{[ M + m - ( M - m) t _ {j} ] }
 +
} ,\ \
 +
t _ {j} = \cos  \left [
 +
{
 +
\frac{( 2 \kappa _ {j} - 1) \pi }{2N}
 +
} \right ] ,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207075.png" /></td> <td valign="top" style="width:5%;text-align:right;">(8)</td></tr></table>
+
$$
 +
= 1 \dots N,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207076.png" /></td> </tr></table>
+
when  $  x _ {0} A $
 +
is a [[Self-adjoint operator|self-adjoint operator]] with non-zero part of the spectrum in  $  [ m, M] $,
 +
$  0 < m \leq  M $.
 +
Here,  $  P _ {N} = ( \kappa _ {1} \dots \kappa _ {N} ) $
 +
is a certain permutation of  $  1 \dots N $
 +
in order that the computation be stable and  $  N $
 +
is the degree of the required polynomial. The method (7), (8) gives an optimum estimate of convergence only at the  $  N $-
 +
th step. When  $  N = 2  ^ {i} $(
 +
or  $  \approx 2  ^ {i} $)
 +
the choice of steps in (7) is very effective if they depend successively on the roots of the Chebyshev polynomials
  
when <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207077.png" /> is a [[Self-adjoint operator|self-adjoint operator]] with non-zero part of the spectrum in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207078.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207079.png" />. Here, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207080.png" /> is a certain permutation of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207081.png" /> in order that the computation be stable and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207082.png" /> is the degree of the required polynomial. The method (7), (8) gives an optimum estimate of convergence only at the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207083.png" />-th step. When <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207084.png" /> (or <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207085.png" />) the choice of steps in (7) is very effective if they depend successively on the roots of the Chebyshev polynomials
+
$$
 +
T _ {2} ( t) - T _ {0} ( t),\
 +
T _ {1} ( t), T _ {1} ( t) \dots
 +
T _ {2 ^ {i - 2 }  } ( t), T _ {2 ^ {i - 2 }  } ( t),
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207086.png" /></td> </tr></table>
+
instead of on  $  T _ {N} ( t) $;  
 +
this substantially simplifies the problem of ordering the steps  $  h _ {j} $
 +
and increases the effectiveness of the computation, especially for large  $  N $.
 +
In this case the error decreases optimally after using all the ordered roots of each polynomial, which is important for a simplification of the control of the computation. If  $  P ( Ax _ {0} ) ( b - Ay _ {0} ) = 0 $,
 +
then  $  P ( x _ {0} A) ( y  ^ {+} - y _ {0} ) = 0 $;  
 +
moreover, if  $  P ( x _ {0} A) $
 +
is an orthogonal projection operator, then  $  y  ^ {+} - y _ {0} $
 +
is the normal solution of the functional equation  $  A ( y  ^ {+} - y _ {0} ) = b - Ay _ {0} $.
 +
On the other hand (7) implies (for  $  k = 0, 1 , . . . $)
  
instead of on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207087.png" />; this substantially simplifies the problem of ordering the steps <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207088.png" /> and increases the effectiveness of the computation, especially for large <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207089.png" />. In this case the error decreases optimally after using all the ordered roots of each polynomial, which is important for a simplification of the control of the computation. If <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207090.png" />, then <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207091.png" />; moreover, if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207092.png" /> is an orthogonal projection operator, then <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207093.png" /> is the normal solution of the functional equation <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207094.png" />. On the other hand (7) implies (for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207095.png" />)
+
$$
 +
b - Ay _ {k + 1 }  = \
 +
U _ {k + 1 }  ( b - Ay _ {0} ),\ \
 +
U _ {k + 1 }  = \
 +
\prod _ {i = 1 } ^ { {k }  + 1 }
 +
( I - h _ {i} Ax _ {0} ).
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207096.png" /></td> </tr></table>
+
If  $  P ( Ax _ {0} ) $
 +
is a projection operator, then
  
If <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207097.png" /> is a projection operator, then
+
$$
 +
\| U _ {k} - P ( Ax _ {0} ) \|  \leq  \
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207098.png" /></td> </tr></table>
+
\frac{2q  ^ {k} }{( 1 + q  ^ {2k} ) }
 +
  \rightarrow  0,
 +
$$
  
as <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f04207099.png" /> <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070100.png" />.
+
as $  k \rightarrow \infty $
 +
$  ( q = ( \sqrt M - \sqrt m )/( \sqrt M + \sqrt m )) $.
  
There are also direct methods of variation of the parameter of Euler type using the recurrence relations for Chebyshev, and closely related, polynomials (without using the roots of these polynomials explicitly); for these the error decreases at each step. These are multi-step methods and have an increased rate of convergence. Applying Euler's method with backward step <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070101.png" /> to problem (5) gives the following effective class of methods:
+
There are also direct methods of variation of the parameter of Euler type using the recurrence relations for Chebyshev, and closely related, polynomials (without using the roots of these polynomials explicitly); for these the error decreases at each step. These are multi-step methods and have an increased rate of convergence. Applying Euler's method with backward step $  h _ {k + 1 }  $
 +
to problem (5) gives the following effective class of methods:
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070102.png" /></td> </tr></table>
+
$$
 +
y _ {k + 1 }  = \
 +
y _ {k} + h _ {k + 1 }
 +
( I + h _ {k + 1 }  x _ {0} A)  ^ {-1}
 +
x _ {0} ( b - Ay _ {0} ),\ \
 +
k = 0, 1 \dots
 +
$$
  
 
whence
 
whence
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070103.png" /></td> </tr></table>
+
$$
 +
b - Ay _ {k + 1 }  = \
 +
W _ {k + 1 }
 +
( b - Ay _ {0} ),
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070104.png" /></td> </tr></table>
+
$$
 +
W _ {k + 1 }  = \prod _ {i = 1 } ^ {k  + 1 } ( I + h _ {i} Ax _ {0} )  ^ {-1} ,\  k = 0, 1 , .  . . .
 +
$$
  
Moreover, if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070105.png" /> is a projection operator and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070106.png" /> is such that for any <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070107.png" />,
+
Moreover, if $  P ( Ax _ {0} ) $
 +
is a projection operator and $  h _ {i} > 0 $
 +
is such that for any $  i $,
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070108.png" /></td> </tr></table>
+
$$
 +
\| ( I - P ( Ax _ {0} )) ( I + h _ {i} Ax _ {0} )  ^ {-1} \|
 +
\leq  \rho _ {i}  < 1,
 +
$$
  
 
then
 
then
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070109.png" /></td> </tr></table>
+
$$
 +
\| W _ {k} - P ( Ax _ {0} ) \|  \leq  \
 +
\prod_{i=1}^ { k }  \rho _ {i}  \rightarrow  0,\ \
 +
\textrm{ as }  k \rightarrow \infty .
 +
$$
  
There are analogous methods for solving non-linear functional equations and operator equations, and also methods with increased accuracy of Runge–Kutta type with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070110.png" />-th order accuracy, and other types.
+
There are analogous methods for solving non-linear functional equations and operator equations, and also methods with increased accuracy of Runge–Kutta type with $  s $-
 +
th order accuracy, and other types.
  
 
The solution of a functional equation in the narrow sense, or of systems of such equations, can be a specific function as well as a class of functions depending on arbitrary parameters or arbitrary functions. In the theory of functional equations there are few general methods known for solving such equations. Hence it is necessary, as a rule, to investigate in each particular case the degree of generality of the solution obtained.
 
The solution of a functional equation in the narrow sense, or of systems of such equations, can be a specific function as well as a class of functions depending on arbitrary parameters or arbitrary functions. In the theory of functional equations there are few general methods known for solving such equations. Hence it is necessary, as a rule, to investigate in each particular case the degree of generality of the solution obtained.
  
One of the more-or-less general methods for solving functional equations is by reducing them to equations in finite differences. For example, take a functional equation of order <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070111.png" /> and class 1 of the form
+
One of the more-or-less general methods for solving functional equations is by reducing them to equations in finite differences. For example, take a functional equation of order $  n $
 +
and class 1 of the form
 +
 
 +
$$ \tag{9 }
 +
F ( x, \phi ( x) \dots \phi  ^ {n} ( x))  =  0,
 +
$$
 +
 
 +
where the function  $  F $
 +
is given and  $  \phi ( x) $
 +
is the unknown function. Here
 +
 
 +
$$
 +
\phi  ^ {0} ( x)  = x,\
 +
\phi  ^ {1} ( x)  = \phi ( x),\
 +
\phi  ^ {2} ( x)  = \phi ( \phi ( x)) , . . . .
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070112.png" /></td> <td valign="top" style="width:5%;text-align:right;">(9)</td></tr></table>
+
Further, assume that  $  x = u _ {z} $,
 +
$  \phi ( x) = u _ {z + 1 }  $.  
 +
The transition from  $  x $
 +
to  $  \phi ( x) $
 +
is replaced by introducing a new variable  $  z $
 +
and increasing  $  z $
 +
by 1 in the function  $  u _ {z} $.  
 +
After such a change, equation (9) has the form
  
where the function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070113.png" /> is given and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070114.png" /> is the unknown function. Here
+
$$ \tag{10 }
 +
F ( u _ {z} \dots u _ {z + n }  )  = 0.
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070115.png" /></td> </tr></table>
+
The solution of the  $  n $-
 +
th order finite-difference equation (10) gives an expression for  $  u _ {z} $
 +
in terms of  $  z $
 +
and  $  n $
 +
arbitrary periodic functions  $  C _ {i} $
 +
in  $  z $
 +
with period one. The most general form of the solution to (9) is a system of two compatible equations
  
Further, assume that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070116.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070117.png" />. The transition from <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070118.png" /> to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070119.png" /> is replaced by introducing a new variable <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070120.png" /> and increasing <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070121.png" /> by 1 in the function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070122.png" />. After such a change, equation (9) has the form
+
$$ \tag{11 }
 +
\left .
 +
\begin{array}{c}
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070123.png" /></td> <td valign="top" style="width:5%;text-align:right;">(10)</td></tr></table>
+
u _ {z}  = = f ( z, C _ {1} \dots C _ {n} ),
 +
\\
  
The solution of the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070124.png" />-th order finite-difference equation (10) gives an expression for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070125.png" /> in terms of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070126.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070127.png" /> arbitrary periodic functions <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070128.png" /> in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070129.png" /> with period one. The most general form of the solution to (9) is a system of two compatible equations
+
u _ {z + 1 }  = \phi ( x) = \
 +
f ( z + 1, C _ {1} \dots C _ {n} ).
 +
\end{array}
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070130.png" /></td> <td valign="top" style="width:5%;text-align:right;">(11)</td></tr></table>
+
\right \}
 +
$$
  
By choosing particular values for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070131.png" /> one can eliminate <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070132.png" /> from (11) and thus obtain a particular solution to (9). For example, for the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070133.png" />-nd order functional equation
+
By choosing particular values for $  C _ {i} $
 +
one can eliminate $  z $
 +
from (11) and thus obtain a particular solution to (9). For example, for the $  2 $-
 +
nd order functional equation
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070134.png" /></td> <td valign="top" style="width:5%;text-align:right;">(12)</td></tr></table>
+
$$ \tag{12 }
 +
\phi [ \phi ( x)] + a \phi ( x) + bx  = 0
 +
$$
  
 
the general solution (11) takes the form
 
the general solution (11) takes the form
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070135.png" /></td> <td valign="top" style="width:5%;text-align:right;">(13)</td></tr></table>
+
$$ \tag{13 }
 +
\left .
 +
\begin{array}{c}
 +
 
 +
u _ {z}  = = \
 +
C _ {1} \lambda _ {1}  ^ {z} +
 +
C _ {2} \lambda _ {2}  ^ {z} ,
 +
\\
 +
 
 +
u _ {z + 1 }  = \phi ( x)  = \
 +
C _ {1} \lambda _ {1} ^ {z + 1 } +
 +
C _ {2} \lambda _ {2} ^ {z + 1 } ,
 +
\end{array}
 +
 
 +
\right \}
 +
$$
 +
 
 +
where  $  \lambda _ {1} $
 +
and  $  \lambda _ {2} $
 +
are the roots of the quadratic equation  $  \lambda  ^ {2} + a \lambda + b = 0 $,
 +
and  $  C _ {1} $
 +
and  $  C _ {2} $
 +
are arbitrary fixed periodic functions with period one. If  $  C _ {1} $
 +
and  $  C _ {2} $
 +
are taken to be arbitrary constants and  $  z $
 +
is eliminated from (13), then the complete solution of (12) is obtained:
 +
 
 +
$$
 +
 
 +
\frac{x \lambda _ {2} - \phi ( x) }{\lambda _ {2} - \lambda _ {1} }
 +
  = \
 +
C _ {1} \left [
 +
 
 +
\frac{x \lambda _ {1} - \phi ( x) }{C _ {2} ( \lambda _ {1} - \lambda _ {2} ) }
 +
 
 +
\right ]  ^  \gamma  ,\ \
 +
\gamma  = \
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070136.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070137.png" /> are the roots of the quadratic equation <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070138.png" />, and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070139.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070140.png" /> are arbitrary fixed periodic functions with period one. If <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070141.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070142.png" /> are taken to be arbitrary constants and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070143.png" /> is eliminated from (13), then the complete solution of (12) is obtained:
+
\frac{ \mathop{\rm log}  \lambda _ {1} }{ \mathop{\rm log}  \lambda _ {2} }
 +
.
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070144.png" /></td> </tr></table>
+
The method of reduction to finite-difference equations is also applied in solving direct problems in functional calculus. For example, let the function  $  \phi ( x) = a + bx $
 +
be given and suppose it is required to construct the expression  $  \phi  ^ {n} ( x) $(
 +
where  $  \phi  ^ {i} ( x) = \phi ( \phi ^ {i - 1 } ( x) ) $).
 +
Assume that  $  \phi  ^ {n} ( x) = u _ {n} $
 +
and write the equation in finite differences  $  u _ {n + 1 }  = a + bu _ {n} $,
 +
the solution of which is  $  u _ {n} = Cb  ^ {n} + a/( 1 - b) $.
 +
Thus, for  $  n = 0 $
 +
one obtains  $  u _ {0} = x = c + a/( 1 - b) $,
 +
that is,  $  C = x - a/( 1 - b) $.  
 +
Thus,
  
The method of reduction to finite-difference equations is also applied in solving direct problems in functional calculus. For example, let the function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070145.png" /> be given and suppose it is required to construct the expression <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070146.png" /> (where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070147.png" />). Assume that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070148.png" /> and write the equation in finite differences <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070149.png" />, the solution of which is <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070150.png" />. Thus, for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070151.png" /> one obtains <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070152.png" />, that is, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070153.png" />. Thus,
+
$$
 +
\phi  ^ {n} ( x) = \
 +
\alpha _ {n} + \beta _ {n} x,\ \
 +
\alpha _ {n}  = \
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070154.png" /></td> </tr></table>
+
\frac{a ( b  ^ {n} - 1) }{b - 1 }
 +
,\ \
 +
\beta _ {n}  = b  ^ {n} .
 +
$$
  
Here, if desired, one can write the functional equation <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070155.png" />, which will have a solution <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070156.png" /> for any <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070157.png" />, etc. By solving this functional equation by the same method one can construct other solutions <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070158.png" />. In particular, for odd <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070159.png" /> one gets another real solution <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070160.png" />.
+
Here, if desired, one can write the functional equation $  \phi  ^ {n} ( x) = \beta _ {n - 1 }  \phi ( x) + \alpha _ {n - 1 }  $,  
 +
which will have a solution $  \phi ( x) $
 +
for any $  n $,  
 +
etc. By solving this functional equation by the same method one can construct other solutions $  \phi ( x) $.  
 +
In particular, for odd $  n $
 +
one gets another real solution $  \phi ( x) = - bx + a ( 1 + b)/( 1 - b) $.
  
 
One also applies substitution methods to solve functional equations. For example, suppose one has the functional equation
 
One also applies substitution methods to solve functional equations. For example, suppose one has the functional equation
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070161.png" /></td> <td valign="top" style="width:5%;text-align:right;">(14)</td></tr></table>
+
$$ \tag{14 }
 +
f ( x + y) + f ( x - y)  = 2f ( x) \cos  y.
 +
$$
  
 
Applying successively the substitutions
 
Applying successively the substitutions
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070162.png" /></td> </tr></table>
+
$$
 +
= 0, y  = t; \ \
 +
= {
 +
\frac \pi {2}
 +
} +
 +
t, y  = {
 +
\frac \pi {2}
 +
} ; \ \
 +
= {
 +
\frac \pi {2}
 +
} ,\
 +
= {
 +
\frac \pi {2}
 +
} + t,
 +
$$
  
 
(14) implies, respectively,
 
(14) implies, respectively,
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070163.png" /></td> </tr></table>
+
$$
 +
f ( t) + f (- t)  = 2a  \cos  t,\ \
 +
f ( \pi + t) + f ( t)  = 0
 +
$$
  
 
and
 
and
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070164.png" /></td> </tr></table>
+
$$
 +
f ( \pi + t) + f (- t)  = \
 +
2b  \cos \left ( {
 +
\frac \pi {2}
 +
} + t \right )  = - 2b  \sin  t,
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070165.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070166.png" />. Hence, subtracting the third equation from the sum of the first two equations one obtains <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070167.png" />. The function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070168.png" /> is the general solution of (14). This method can also be applied to other equations of the type
+
where $  f ( 0) = a $,
 +
f ( \pi /2) = b $.  
 +
Hence, subtracting the third equation from the sum of the first two equations one obtains $  2f ( t) = 2a  \cos  t + 2b  \sin  t $.  
 +
The function $  f ( x) = a  \cos  x + b  \sin  x $
 +
is the general solution of (14). This method can also be applied to other equations of the type
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070169.png" /></td> </tr></table>
+
$$
 +
H ( f ( x + y), f ( x- y), f ( x), x, y)  = 0
 +
$$
  
under certain assumptions on the function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070170.png" />. Various other substitutions can be applied to equations of other types.
+
under certain assumptions on the function $  H $.  
 +
Various other substitutions can be applied to equations of other types.
  
 
The substitution method is also applied to reduce some functional equations to others of the same type; in particular, to functional equations with known solutions. For example, the functional equation
 
The substitution method is also applied to reduce some functional equations to others of the same type; in particular, to functional equations with known solutions. For example, the functional equation
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070171.png" /></td> <td valign="top" style="width:5%;text-align:right;">(15)</td></tr></table>
+
$$ \tag{15 }
 +
f \left ( {
 +
\frac{x + y }{2}
 +
} \right )  = \
 +
 
 +
\frac{f ( x) }{2}
 +
+
 +
 
 +
\frac{f ( y) }{2}
 +
 
 +
$$
  
 
can be reduced to the Cauchy functional equation
 
can be reduced to the Cauchy functional equation
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070172.png" /></td> <td valign="top" style="width:5%;text-align:right;">(16)</td></tr></table>
+
$$ \tag{16 }
 +
f ( x + y)  = f ( x) + f ( y),
 +
$$
 +
 
 +
which has the continuous solution  $  f ( x) = Cx $.  
 +
In order to obtain this solution,  $  x + y $
 +
is substituted in (15) for  $  x $
 +
and 0 for  $  y $:
  
which has the continuous solution <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070173.png" />. In order to obtain this solution, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070174.png" /> is substituted in (15) for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070175.png" /> and 0 for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070176.png" />:
+
$$
 +
f \left ( {
 +
\frac{x + y }{2}
 +
} \right ) = \
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070177.png" /></td> </tr></table>
+
\frac{f ( x + y) }{2}
 +
+
 +
{
 +
\frac{a}{2}
 +
} ,\ \
 +
= f ( 0).
 +
$$
  
Comparing this with (15) one obtains a functional equation of the form <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070178.png" />, whence <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070179.png" />, where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070180.png" /> and thus <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070181.png" />. The function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070182.png" /> is the solution.
+
Comparing this with (15) one obtains a functional equation of the form $  f ( x + y) = f ( x) + f ( y) - a $,  
 +
whence $  \phi ( x + y) = \phi ( x) + \phi ( y) $,  
 +
where $  \phi ( x) = f ( x) - a $
 +
and thus $  \phi ( x) = Cx $.  
 +
The function $  f ( x) = Cx + a $
 +
is the solution.
  
 
Taking logarithms and other methods can also be used to reduce a functional equation to another of the same type. For example, solving the functional equation
 
Taking logarithms and other methods can also be used to reduce a functional equation to another of the same type. For example, solving the functional equation
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070183.png" /></td> <td valign="top" style="width:5%;text-align:right;">(17)</td></tr></table>
+
$$ \tag{17 }
 +
f ( x + y)  = f ( x) f ( y)
 +
$$
  
can be reduced to solving the functional equation (16) by taking logarithms. A continuous function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070184.png" /> satisfying (17) is always strictly positive and the function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070185.png" /> is continuous (as a composition of continuous functions) and satisfies the condition <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070186.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070187.png" />. The solution thus is <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070188.png" />.
+
can be reduced to solving the functional equation (16) by taking logarithms. A continuous function f ( x) $
 +
satisfying (17) is always strictly positive and the function $  \phi ( x) = \mathop{\rm log}  f ( x) $
 +
is continuous (as a composition of continuous functions) and satisfies the condition $  \phi ( x + y) = \phi ( x) + \phi ( y) $,  
 +
$  \phi ( x) = Cx $.  
 +
The solution thus is $  f ( x) = e  ^ {Cx} = a  ^ {x} $.
  
In many cases the method of reduction to differential equations can also be applied to solve functional equations. For example, the functional equation (14) can be reduced to an equation of the type <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070189.png" />. There are other functional equations which reduce to this equation. This method only gives solutions in the class of differentiable functions. For example, the solution to the Cauchy functional equation (16) with the condition that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070190.png" /> is differentiable can be found in the following way. Differentiating (16) with respect to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070191.png" /> one obtains <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070192.png" />, hence, since <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070193.png" /> here is arbitrary, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070194.png" />. Then integration gives <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070195.png" />, where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070196.png" /> is a new constant. Substituting again the expression found for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070197.png" /> in (16) one ascertains that for all values of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070198.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070199.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070200.png" />. The function <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/f/f042/f042070/f042070201.png" /> is the solution.
+
In many cases the method of reduction to differential equations can also be applied to solve functional equations. For example, the functional equation (14) can be reduced to an equation of the type $  f ^ { \prime\prime } ( x) = - f ( x) $.  
 +
There are other functional equations which reduce to this equation. This method only gives solutions in the class of differentiable functions. For example, the solution to the Cauchy functional equation (16) with the condition that f ( x) $
 +
is differentiable can be found in the following way. Differentiating (16) with respect to $  x $
 +
one obtains $  f ^ { \prime } ( x + y) = f ^ { \prime } ( x) $,  
 +
hence, since $  y $
 +
here is arbitrary, $  f ^ { \prime } ( x) = C $.  
 +
Then integration gives $  f ( x) = Cx + C _ {1} $,  
 +
where $  C _ {1} $
 +
is a new constant. Substituting again the expression found for f ( x) $
 +
in (16) one ascertains that for all values of $  x $
 +
and $  y $,  
 +
$  C _ {1} = 0 $.  
 +
The function $  f ( x) = Cx $
 +
is the solution.
  
 
Iteration methods are also applied in a number of cases to solve functional equations.
 
Iteration methods are also applied in a number of cases to solve functional equations.
Line 189: Line 583:
 
====References====
 
====References====
 
<table><TR><TD valign="top">[1]</TD> <TD valign="top">  L.V. Kantorovich,  "On Newton's method"  ''Trudy Mat. Inst. Steklov.'' , '''28'''  (1949)  pp. 104–144  (In Russian)</TD></TR><TR><TD valign="top">[2]</TD> <TD valign="top">  M.A. Krasnosel'skii,  G.M. Vainikko,  P.P. Zabreiko,  et al.,  "Approximate solution of operator equations" , Wolters-Noordhoff  (1972)  (Translated from Russian)</TD></TR><TR><TD valign="top">[3]</TD> <TD valign="top">  A.A. Samarskii,  "Theorie der Differenzverfahren" , Akad. Verlagsgesell. Geest u. Portig K.-D.  (1984)  (Translated from Russian)</TD></TR><TR><TD valign="top">[4]</TD> <TD valign="top">  N.S. Bakhvalov,  "Numerical methods: analysis, algebra, ordinary differential equations" , MIR  (1977)  (Translated from Russian)</TD></TR><TR><TD valign="top">[5]</TD> <TD valign="top">  D.F. Davidenko,  , ''Approximate methods for solving operator equations and their applications'' , Irkutsk  (1982)  pp. 71–83  (In Russian)</TD></TR><TR><TD valign="top">[6]</TD> <TD valign="top">  M.A. Mertvetsova,  "Analogue of the process of tangent hyperbolas for general functional equations"  ''Dokl. Akad. Nauk SSSR'' , '''88''' :  4  (1953)  pp. 611–614  (In Russian)</TD></TR><TR><TD valign="top">[7]</TD> <TD valign="top">  M.I. Nechepurenko,  "On Chebychev's method for functional equations"  ''Uspekhi Mat. Nauk'' , '''9''' :  2  (1954)  pp. 163–170  (In Russian)</TD></TR><TR><TD valign="top">[8]</TD> <TD valign="top">  E.E. Tamme,  "On approximately solving functional equations by the method of series expansion in the inverse operator"  ''Dokl. Akad. Nauk SSSR'' , '''103''' :  5  (1955)  pp. 769–772  (In Russian)</TD></TR><TR><TD valign="top">[9]</TD> <TD valign="top">  I.P. Mysovskikh,  "On convergence of L.V. Kantorovich's method for solving non-linear functional equations, and its applications"  ''Vestn. Leningrad. Univ.'' , '''11'''  (1953)  pp. 25–48  (In Russian)</TD></TR><TR><TD valign="top">[10]</TD> <TD valign="top">  D.F. Davidenko,  , ''Conf. programming and mathematical methods for solving physical problems''  (1974)  pp. 542–548  (In Russian)</TD></TR><TR><TD valign="top">[11]</TD> <TD valign="top">  B.A. Bel'tyukov,  "On a certain method of solution of non-linear functional equations"  ''Zh. Vychisl. Mat. i Mat. Fiz.'' , '''5''' :  5  (1965)  pp. 927–931  (In Russian)</TD></TR><TR><TD valign="top">[12]</TD> <TD valign="top">  O. Vaarmann,  "Approximations of pseudo-inverse operators as applied to the solution of non-linear equations"  ''Izv. Akad. Nauk EstSSR'' , '''20''' :  4  (1971)  pp. 386–393  (In Russian)  (Estonian and English summaries)</TD></TR><TR><TD valign="top">[13]</TD> <TD valign="top">  D.F. Davidenko,  , ''Theoretical and applied problems in computing'' , Moscow  (1981)  pp. 61–63  (In Russian)</TD></TR><TR><TD valign="top">[14]</TD> <TD valign="top">  A.I. Liventsov,  ''Mat. Sb.'' , '''8''' :  1  (1876)  pp. 80–160</TD></TR><TR><TD valign="top">[15]</TD> <TD valign="top">  D.M. Sintsov,  "Notes on functional calculus"  ''Izv. Fiz.-Mat. Obshch. Kazan. Univ. (2)'' , '''13''' :  2  (1903)  pp. 46–72  (In Russian)</TD></TR></table>
 
<table><TR><TD valign="top">[1]</TD> <TD valign="top">  L.V. Kantorovich,  "On Newton's method"  ''Trudy Mat. Inst. Steklov.'' , '''28'''  (1949)  pp. 104–144  (In Russian)</TD></TR><TR><TD valign="top">[2]</TD> <TD valign="top">  M.A. Krasnosel'skii,  G.M. Vainikko,  P.P. Zabreiko,  et al.,  "Approximate solution of operator equations" , Wolters-Noordhoff  (1972)  (Translated from Russian)</TD></TR><TR><TD valign="top">[3]</TD> <TD valign="top">  A.A. Samarskii,  "Theorie der Differenzverfahren" , Akad. Verlagsgesell. Geest u. Portig K.-D.  (1984)  (Translated from Russian)</TD></TR><TR><TD valign="top">[4]</TD> <TD valign="top">  N.S. Bakhvalov,  "Numerical methods: analysis, algebra, ordinary differential equations" , MIR  (1977)  (Translated from Russian)</TD></TR><TR><TD valign="top">[5]</TD> <TD valign="top">  D.F. Davidenko,  , ''Approximate methods for solving operator equations and their applications'' , Irkutsk  (1982)  pp. 71–83  (In Russian)</TD></TR><TR><TD valign="top">[6]</TD> <TD valign="top">  M.A. Mertvetsova,  "Analogue of the process of tangent hyperbolas for general functional equations"  ''Dokl. Akad. Nauk SSSR'' , '''88''' :  4  (1953)  pp. 611–614  (In Russian)</TD></TR><TR><TD valign="top">[7]</TD> <TD valign="top">  M.I. Nechepurenko,  "On Chebychev's method for functional equations"  ''Uspekhi Mat. Nauk'' , '''9''' :  2  (1954)  pp. 163–170  (In Russian)</TD></TR><TR><TD valign="top">[8]</TD> <TD valign="top">  E.E. Tamme,  "On approximately solving functional equations by the method of series expansion in the inverse operator"  ''Dokl. Akad. Nauk SSSR'' , '''103''' :  5  (1955)  pp. 769–772  (In Russian)</TD></TR><TR><TD valign="top">[9]</TD> <TD valign="top">  I.P. Mysovskikh,  "On convergence of L.V. Kantorovich's method for solving non-linear functional equations, and its applications"  ''Vestn. Leningrad. Univ.'' , '''11'''  (1953)  pp. 25–48  (In Russian)</TD></TR><TR><TD valign="top">[10]</TD> <TD valign="top">  D.F. Davidenko,  , ''Conf. programming and mathematical methods for solving physical problems''  (1974)  pp. 542–548  (In Russian)</TD></TR><TR><TD valign="top">[11]</TD> <TD valign="top">  B.A. Bel'tyukov,  "On a certain method of solution of non-linear functional equations"  ''Zh. Vychisl. Mat. i Mat. Fiz.'' , '''5''' :  5  (1965)  pp. 927–931  (In Russian)</TD></TR><TR><TD valign="top">[12]</TD> <TD valign="top">  O. Vaarmann,  "Approximations of pseudo-inverse operators as applied to the solution of non-linear equations"  ''Izv. Akad. Nauk EstSSR'' , '''20''' :  4  (1971)  pp. 386–393  (In Russian)  (Estonian and English summaries)</TD></TR><TR><TD valign="top">[13]</TD> <TD valign="top">  D.F. Davidenko,  , ''Theoretical and applied problems in computing'' , Moscow  (1981)  pp. 61–63  (In Russian)</TD></TR><TR><TD valign="top">[14]</TD> <TD valign="top">  A.I. Liventsov,  ''Mat. Sb.'' , '''8''' :  1  (1876)  pp. 80–160</TD></TR><TR><TD valign="top">[15]</TD> <TD valign="top">  D.M. Sintsov,  "Notes on functional calculus"  ''Izv. Fiz.-Mat. Obshch. Kazan. Univ. (2)'' , '''13''' :  2  (1903)  pp. 46–72  (In Russian)</TD></TR></table>
 
 
  
 
====Comments====
 
====Comments====

Latest revision as of 19:16, 16 January 2024


Methods for finding exact or approximate solutions of specific or abstract functional equations, that is, equations of the form

$$ \tag{1 } P ( x) = y $$

where $ P ( x) $ is a certain, generally speaking non-linear, operator transforming the elements of a $ B $- space $ X $( or a space of some other type) into elements of a space $ Y $ of the same type (see Functional equation). Exact solutions in the form of analytic expressions have been obtained only for a few types of functional equations, therefore approximate methods of solution are especially valuable.

A number of methods have evolved to find solutions of general functional equations of the form (1). For example, the method of infinite power series, the method of successive approximations, the Galerkin method (the method of moments), the method of tangent hyperbolas, Chebyshev's method (of tangent parabolas), the Newton–Kantorovich method and its modifications, the method of steepest descent (cf. Steepest descent, method of), etc., as well as the method of variation of the parameter (direct, iterative and combined, cf. Parameter, method of variation of the) of particular types and its various modifications, including successive approximations to the inverse operator. General methods have been applied to solve various specific functional equations in mathematical analysis. In addition there are special methods for solving specific functional equations, including numerical methods, for example, the grid method, etc. The method of variation of the parameter, the Newton–Kantorovich method and certain other methods indicated also have theoretical significance, since they can be used to draw conclusions about the existence, uniqueness and the domain of location of a solution of a functional equation without finding the solution itself, which is sometimes no less important then the actual value of the solution. The direct method of variation of the parameter is considered below as an example.

In the Banach space $ [ X] = [ X \rightarrow X] $ of linear bounded operators, let a non-linear operator ordinary differential equation be given on the infinite interval:

$$ \tag{2 } \frac{dx }{d \lambda } = \ x ( I - Ax) \ \ (= ( I - xA) x),\ \ x ( 0) = x _ {0} , $$

where $ A, x _ {0} , I \in [ X] $( $ A $, $ x _ {0} $ are given, $ I $ is the identity operator), and $ x ( \lambda ) $ is an abstract function with values in $ [ X] $( $ X $ is a $ B $- space). The problem of constructing the inverse operator $ A ^ {-1} $, for an invertible operator $ A $, the solution of a linear operator equation of the form $ I - Ax = 0 $, etc., are all reduced to solving equation (2) by the direct method of variation of the parameter. If the spectrum of the operator $ Ax _ {0} $( respectively, $ x _ {0} A $) is situated in the right half-plane, then problem (2) has a unique solution $ x ( \lambda ) $, $ x ( 0) = x _ {0} $, $ \lambda \in [ 0, \infty ) $.

An analytic application to the Cauchy problem (2) of methods for the numerical integration of ordinary differential equations leads to a whole class of direct methods of variation of the parameter for solving (2), and consequently also to the solutions of problems that reduce to (2). For example, the Euler method with irregular step $ h _ {k} $ leads to the following method for constructing $ A ^ {-1} $:

$$ \tag{3 } x _ {k + 1 } = \ x _ {k} + h _ {k} x _ {k} ( I - Ax _ {k} ) \ \ ( = x _ {k} + h _ {k} ( I - x _ {k} A) x _ {k} ), $$

$$ k = 0, 1 , . . . . $$

The steps $ h _ {k} $ in (3) are chosen in various ways. When the spectrum of $ I - Ax _ {0} $( respectively, $ I - x _ {0} A $) is on the real axis in an interval $ [- \rho _ {0} , 0] $, $ 0 < \rho _ {0} < \infty $, then the following method for choosing $ h _ {k} $ is very effective:

$$ \tag{4 } h _ {k} = \ { \frac{1}{1 + \rho _ {k} } } ,\ \ \rho _ {k + 1 } = \ \rho _ {k} L _ {k} ,\ \ 4L _ {k} = \ \frac{\rho _ {k} }{1 + \rho _ {k} } , $$

$$ k = 0, 1 \dots \ h _ {k + 1 } = \frac{4h _ {k} }{( 1 + h _ {k} ) ^ {2} } . $$

Here, the rate of convergence in (3) is of order higher than $ 2 ^ {k} $, and the norm of the discrepancy decreases according to the law $ \rho _ {k + 1 } = \rho _ {k} ^ {2} /4 ( 1 + \rho _ {k} ) $. The case of an interval $ [ 0, \overline \rho \; _ {0} ] $, $ 0 < \overline \rho \; _ {0} < 1 $, can be reduced, with advantage, to the one-step method (3) considered, where

$$ h = \ { \frac{1}{1 - \overline \rho \; _ {0} } } ,\ \ \rho _ {0} = \ \frac{\overline \rho \; {} _ {0} ^ {2} }{4 ( 1 - \overline \rho \; _ {0} ) } . $$

Regarding the convergence of (3), (4), there are some results, stated for specific operator equations, based on general facts and depending on which space is taken as a base.

The problem of constructing projections of the form

$$ \left . P ( Ax _ {0} ) \ \ ( \textrm{ or } P ( x _ {0} A)),\ \ P ( K) = \ e ^ {- K \lambda } \right | _ {\lambda = \infty } ,\ \ K \in [ X], $$

also reduces to the solution of problem (2) (the case when the spectrum of $ I - Ax _ {0} $( respectively, $ I - x _ {0} A $) is in $ [- \rho _ {0} , 1] $) and so does the problem of constructing pseudo-inverse operators $ x ^ {+} $( respectively, $ {} ^ {+} x $) for the operator $ A $ such that

$$ I - Ax ^ {+} = \ P ( Ax _ {0} ),\ \ I - {} ^ {+} xA = \ P ( x _ {0} A),\ \ x ^ {+} = {} ^ {+} x. $$

For a direct construction of the projections $ P ( Ax _ {0} ) $( respectively, $ P ( x _ {0} A) $), the method (3), (4) can be rewritten in the form

$$ P _ {k + 1 } = \ P _ {k} + h _ {k} P _ {k} ( P _ {k} - I),\ \ k = 0, 1 \dots $$

$$ P _ {0} = I - Ax _ {0} \ (= I - x _ {0} A). $$

Depending on the location of the spectrum and the properties of $ A $ one can take, for example, the operators $ I $, $ A $, $ A ^ {*} $, etc., in place of $ x _ {0} $.

Using the direct method of variation of the parameter one can reduce to the abstract linear functional ordinary differential equation $ ( 0 \leq \lambda \leq \infty ) $

$$ \tag{5 } \frac{dy }{d \lambda } = \ x _ {0} ( b - Ay),\ \ y ( 0) = y _ {0} , $$

where $ b, y _ {0} \in X $, $ x _ {0} , A \in [ X] $ and $ y ( \lambda ) $ is an abstract function with values in $ X $, the problem of constructing directly the pseudo-solutions $ y ^ {+} $ of a linear functional equation of the form

$$ \tag{6 } b - Ay ^ {+} = \ P ( Ax _ {0} ) ( b - Ay _ {0} ) $$

or of the functional equations $ b - Ay ^ {+} = 0 $ if $ P ( Ax _ {0} ) ( b - Ay _ {0} ) = 0 $. Other problems also reduce to (5), including problems for ordinary and partial differential equations, integral equations, etc. The formula

$$ y ( \lambda ) = \ e ^ {- x _ {0} A \lambda } y _ {0} + r ( \lambda ) x _ {0} b,\ \ r ( \lambda ) = \ \int\limits _ { 0 } ^ \lambda e ^ {- x _ {0} A ( \lambda - s) } ds, $$

gives the unique solution to equation (5) satisfying $ y ( 0) = y _ {0} $.

For example, applying Euler's method with forward step $ h _ {k + 1 } $ to problem (5) leads to the following method for constructing pseudo-solutions $ y ^ {+} $ to (6):

$$ \tag{7 } y _ {k + 1 } = \ y _ {k} + h _ {k + 1 } x _ {0} ( b - Ay _ {k} ),\ \ k = 0, 1 , . . . . $$

The steps $ h _ {k + 1 } $ are chosen, for example, in terms of the roots of the Chebyshev polynomials $ T _ {N} ( t) $:

$$ \tag{8 } h _ {j} = \ { \frac{2}{[ M + m - ( M - m) t _ {j} ] } } ,\ \ t _ {j} = \cos \left [ { \frac{( 2 \kappa _ {j} - 1) \pi }{2N} } \right ] , $$

$$ j = 1 \dots N, $$

when $ x _ {0} A $ is a self-adjoint operator with non-zero part of the spectrum in $ [ m, M] $, $ 0 < m \leq M $. Here, $ P _ {N} = ( \kappa _ {1} \dots \kappa _ {N} ) $ is a certain permutation of $ 1 \dots N $ in order that the computation be stable and $ N $ is the degree of the required polynomial. The method (7), (8) gives an optimum estimate of convergence only at the $ N $- th step. When $ N = 2 ^ {i} $( or $ \approx 2 ^ {i} $) the choice of steps in (7) is very effective if they depend successively on the roots of the Chebyshev polynomials

$$ T _ {2} ( t) - T _ {0} ( t),\ T _ {1} ( t), T _ {1} ( t) \dots T _ {2 ^ {i - 2 } } ( t), T _ {2 ^ {i - 2 } } ( t), $$

instead of on $ T _ {N} ( t) $; this substantially simplifies the problem of ordering the steps $ h _ {j} $ and increases the effectiveness of the computation, especially for large $ N $. In this case the error decreases optimally after using all the ordered roots of each polynomial, which is important for a simplification of the control of the computation. If $ P ( Ax _ {0} ) ( b - Ay _ {0} ) = 0 $, then $ P ( x _ {0} A) ( y ^ {+} - y _ {0} ) = 0 $; moreover, if $ P ( x _ {0} A) $ is an orthogonal projection operator, then $ y ^ {+} - y _ {0} $ is the normal solution of the functional equation $ A ( y ^ {+} - y _ {0} ) = b - Ay _ {0} $. On the other hand (7) implies (for $ k = 0, 1 , . . . $)

$$ b - Ay _ {k + 1 } = \ U _ {k + 1 } ( b - Ay _ {0} ),\ \ U _ {k + 1 } = \ \prod _ {i = 1 } ^ { {k } + 1 } ( I - h _ {i} Ax _ {0} ). $$

If $ P ( Ax _ {0} ) $ is a projection operator, then

$$ \| U _ {k} - P ( Ax _ {0} ) \| \leq \ \frac{2q ^ {k} }{( 1 + q ^ {2k} ) } \rightarrow 0, $$

as $ k \rightarrow \infty $ $ ( q = ( \sqrt M - \sqrt m )/( \sqrt M + \sqrt m )) $.

There are also direct methods of variation of the parameter of Euler type using the recurrence relations for Chebyshev, and closely related, polynomials (without using the roots of these polynomials explicitly); for these the error decreases at each step. These are multi-step methods and have an increased rate of convergence. Applying Euler's method with backward step $ h _ {k + 1 } $ to problem (5) gives the following effective class of methods:

$$ y _ {k + 1 } = \ y _ {k} + h _ {k + 1 } ( I + h _ {k + 1 } x _ {0} A) ^ {-1} x _ {0} ( b - Ay _ {0} ),\ \ k = 0, 1 \dots $$

whence

$$ b - Ay _ {k + 1 } = \ W _ {k + 1 } ( b - Ay _ {0} ), $$

$$ W _ {k + 1 } = \prod _ {i = 1 } ^ {k + 1 } ( I + h _ {i} Ax _ {0} ) ^ {-1} ,\ k = 0, 1 , . . . . $$

Moreover, if $ P ( Ax _ {0} ) $ is a projection operator and $ h _ {i} > 0 $ is such that for any $ i $,

$$ \| ( I - P ( Ax _ {0} )) ( I + h _ {i} Ax _ {0} ) ^ {-1} \| \leq \rho _ {i} < 1, $$

then

$$ \| W _ {k} - P ( Ax _ {0} ) \| \leq \ \prod_{i=1}^ { k } \rho _ {i} \rightarrow 0,\ \ \textrm{ as } k \rightarrow \infty . $$

There are analogous methods for solving non-linear functional equations and operator equations, and also methods with increased accuracy of Runge–Kutta type with $ s $- th order accuracy, and other types.

The solution of a functional equation in the narrow sense, or of systems of such equations, can be a specific function as well as a class of functions depending on arbitrary parameters or arbitrary functions. In the theory of functional equations there are few general methods known for solving such equations. Hence it is necessary, as a rule, to investigate in each particular case the degree of generality of the solution obtained.

One of the more-or-less general methods for solving functional equations is by reducing them to equations in finite differences. For example, take a functional equation of order $ n $ and class 1 of the form

$$ \tag{9 } F ( x, \phi ( x) \dots \phi ^ {n} ( x)) = 0, $$

where the function $ F $ is given and $ \phi ( x) $ is the unknown function. Here

$$ \phi ^ {0} ( x) = x,\ \phi ^ {1} ( x) = \phi ( x),\ \phi ^ {2} ( x) = \phi ( \phi ( x)) , . . . . $$

Further, assume that $ x = u _ {z} $, $ \phi ( x) = u _ {z + 1 } $. The transition from $ x $ to $ \phi ( x) $ is replaced by introducing a new variable $ z $ and increasing $ z $ by 1 in the function $ u _ {z} $. After such a change, equation (9) has the form

$$ \tag{10 } F ( u _ {z} \dots u _ {z + n } ) = 0. $$

The solution of the $ n $- th order finite-difference equation (10) gives an expression for $ u _ {z} $ in terms of $ z $ and $ n $ arbitrary periodic functions $ C _ {i} $ in $ z $ with period one. The most general form of the solution to (9) is a system of two compatible equations

$$ \tag{11 } \left . \begin{array}{c} u _ {z} = x = f ( z, C _ {1} \dots C _ {n} ), \\ u _ {z + 1 } = \phi ( x) = \ f ( z + 1, C _ {1} \dots C _ {n} ). \end{array} \right \} $$

By choosing particular values for $ C _ {i} $ one can eliminate $ z $ from (11) and thus obtain a particular solution to (9). For example, for the $ 2 $- nd order functional equation

$$ \tag{12 } \phi [ \phi ( x)] + a \phi ( x) + bx = 0 $$

the general solution (11) takes the form

$$ \tag{13 } \left . \begin{array}{c} u _ {z} = x = \ C _ {1} \lambda _ {1} ^ {z} + C _ {2} \lambda _ {2} ^ {z} , \\ u _ {z + 1 } = \phi ( x) = \ C _ {1} \lambda _ {1} ^ {z + 1 } + C _ {2} \lambda _ {2} ^ {z + 1 } , \end{array} \right \} $$

where $ \lambda _ {1} $ and $ \lambda _ {2} $ are the roots of the quadratic equation $ \lambda ^ {2} + a \lambda + b = 0 $, and $ C _ {1} $ and $ C _ {2} $ are arbitrary fixed periodic functions with period one. If $ C _ {1} $ and $ C _ {2} $ are taken to be arbitrary constants and $ z $ is eliminated from (13), then the complete solution of (12) is obtained:

$$ \frac{x \lambda _ {2} - \phi ( x) }{\lambda _ {2} - \lambda _ {1} } = \ C _ {1} \left [ \frac{x \lambda _ {1} - \phi ( x) }{C _ {2} ( \lambda _ {1} - \lambda _ {2} ) } \right ] ^ \gamma ,\ \ \gamma = \ \frac{ \mathop{\rm log} \lambda _ {1} }{ \mathop{\rm log} \lambda _ {2} } . $$

The method of reduction to finite-difference equations is also applied in solving direct problems in functional calculus. For example, let the function $ \phi ( x) = a + bx $ be given and suppose it is required to construct the expression $ \phi ^ {n} ( x) $( where $ \phi ^ {i} ( x) = \phi ( \phi ^ {i - 1 } ( x) ) $). Assume that $ \phi ^ {n} ( x) = u _ {n} $ and write the equation in finite differences $ u _ {n + 1 } = a + bu _ {n} $, the solution of which is $ u _ {n} = Cb ^ {n} + a/( 1 - b) $. Thus, for $ n = 0 $ one obtains $ u _ {0} = x = c + a/( 1 - b) $, that is, $ C = x - a/( 1 - b) $. Thus,

$$ \phi ^ {n} ( x) = \ \alpha _ {n} + \beta _ {n} x,\ \ \alpha _ {n} = \ \frac{a ( b ^ {n} - 1) }{b - 1 } ,\ \ \beta _ {n} = b ^ {n} . $$

Here, if desired, one can write the functional equation $ \phi ^ {n} ( x) = \beta _ {n - 1 } \phi ( x) + \alpha _ {n - 1 } $, which will have a solution $ \phi ( x) $ for any $ n $, etc. By solving this functional equation by the same method one can construct other solutions $ \phi ( x) $. In particular, for odd $ n $ one gets another real solution $ \phi ( x) = - bx + a ( 1 + b)/( 1 - b) $.

One also applies substitution methods to solve functional equations. For example, suppose one has the functional equation

$$ \tag{14 } f ( x + y) + f ( x - y) = 2f ( x) \cos y. $$

Applying successively the substitutions

$$ x = 0, y = t; \ \ x = { \frac \pi {2} } + t, y = { \frac \pi {2} } ; \ \ x = { \frac \pi {2} } ,\ y = { \frac \pi {2} } + t, $$

(14) implies, respectively,

$$ f ( t) + f (- t) = 2a \cos t,\ \ f ( \pi + t) + f ( t) = 0 $$

and

$$ f ( \pi + t) + f (- t) = \ 2b \cos \left ( { \frac \pi {2} } + t \right ) = - 2b \sin t, $$

where $ f ( 0) = a $, $ f ( \pi /2) = b $. Hence, subtracting the third equation from the sum of the first two equations one obtains $ 2f ( t) = 2a \cos t + 2b \sin t $. The function $ f ( x) = a \cos x + b \sin x $ is the general solution of (14). This method can also be applied to other equations of the type

$$ H ( f ( x + y), f ( x- y), f ( x), x, y) = 0 $$

under certain assumptions on the function $ H $. Various other substitutions can be applied to equations of other types.

The substitution method is also applied to reduce some functional equations to others of the same type; in particular, to functional equations with known solutions. For example, the functional equation

$$ \tag{15 } f \left ( { \frac{x + y }{2} } \right ) = \ \frac{f ( x) }{2} + \frac{f ( y) }{2} $$

can be reduced to the Cauchy functional equation

$$ \tag{16 } f ( x + y) = f ( x) + f ( y), $$

which has the continuous solution $ f ( x) = Cx $. In order to obtain this solution, $ x + y $ is substituted in (15) for $ x $ and 0 for $ y $:

$$ f \left ( { \frac{x + y }{2} } \right ) = \ \frac{f ( x + y) }{2} + { \frac{a}{2} } ,\ \ a = f ( 0). $$

Comparing this with (15) one obtains a functional equation of the form $ f ( x + y) = f ( x) + f ( y) - a $, whence $ \phi ( x + y) = \phi ( x) + \phi ( y) $, where $ \phi ( x) = f ( x) - a $ and thus $ \phi ( x) = Cx $. The function $ f ( x) = Cx + a $ is the solution.

Taking logarithms and other methods can also be used to reduce a functional equation to another of the same type. For example, solving the functional equation

$$ \tag{17 } f ( x + y) = f ( x) f ( y) $$

can be reduced to solving the functional equation (16) by taking logarithms. A continuous function $ f ( x) $ satisfying (17) is always strictly positive and the function $ \phi ( x) = \mathop{\rm log} f ( x) $ is continuous (as a composition of continuous functions) and satisfies the condition $ \phi ( x + y) = \phi ( x) + \phi ( y) $, $ \phi ( x) = Cx $. The solution thus is $ f ( x) = e ^ {Cx} = a ^ {x} $.

In many cases the method of reduction to differential equations can also be applied to solve functional equations. For example, the functional equation (14) can be reduced to an equation of the type $ f ^ { \prime\prime } ( x) = - f ( x) $. There are other functional equations which reduce to this equation. This method only gives solutions in the class of differentiable functions. For example, the solution to the Cauchy functional equation (16) with the condition that $ f ( x) $ is differentiable can be found in the following way. Differentiating (16) with respect to $ x $ one obtains $ f ^ { \prime } ( x + y) = f ^ { \prime } ( x) $, hence, since $ y $ here is arbitrary, $ f ^ { \prime } ( x) = C $. Then integration gives $ f ( x) = Cx + C _ {1} $, where $ C _ {1} $ is a new constant. Substituting again the expression found for $ f ( x) $ in (16) one ascertains that for all values of $ x $ and $ y $, $ C _ {1} = 0 $. The function $ f ( x) = Cx $ is the solution.

Iteration methods are also applied in a number of cases to solve functional equations.

References

[1] L.V. Kantorovich, "On Newton's method" Trudy Mat. Inst. Steklov. , 28 (1949) pp. 104–144 (In Russian)
[2] M.A. Krasnosel'skii, G.M. Vainikko, P.P. Zabreiko, et al., "Approximate solution of operator equations" , Wolters-Noordhoff (1972) (Translated from Russian)
[3] A.A. Samarskii, "Theorie der Differenzverfahren" , Akad. Verlagsgesell. Geest u. Portig K.-D. (1984) (Translated from Russian)
[4] N.S. Bakhvalov, "Numerical methods: analysis, algebra, ordinary differential equations" , MIR (1977) (Translated from Russian)
[5] D.F. Davidenko, , Approximate methods for solving operator equations and their applications , Irkutsk (1982) pp. 71–83 (In Russian)
[6] M.A. Mertvetsova, "Analogue of the process of tangent hyperbolas for general functional equations" Dokl. Akad. Nauk SSSR , 88 : 4 (1953) pp. 611–614 (In Russian)
[7] M.I. Nechepurenko, "On Chebychev's method for functional equations" Uspekhi Mat. Nauk , 9 : 2 (1954) pp. 163–170 (In Russian)
[8] E.E. Tamme, "On approximately solving functional equations by the method of series expansion in the inverse operator" Dokl. Akad. Nauk SSSR , 103 : 5 (1955) pp. 769–772 (In Russian)
[9] I.P. Mysovskikh, "On convergence of L.V. Kantorovich's method for solving non-linear functional equations, and its applications" Vestn. Leningrad. Univ. , 11 (1953) pp. 25–48 (In Russian)
[10] D.F. Davidenko, , Conf. programming and mathematical methods for solving physical problems (1974) pp. 542–548 (In Russian)
[11] B.A. Bel'tyukov, "On a certain method of solution of non-linear functional equations" Zh. Vychisl. Mat. i Mat. Fiz. , 5 : 5 (1965) pp. 927–931 (In Russian)
[12] O. Vaarmann, "Approximations of pseudo-inverse operators as applied to the solution of non-linear equations" Izv. Akad. Nauk EstSSR , 20 : 4 (1971) pp. 386–393 (In Russian) (Estonian and English summaries)
[13] D.F. Davidenko, , Theoretical and applied problems in computing , Moscow (1981) pp. 61–63 (In Russian)
[14] A.I. Liventsov, Mat. Sb. , 8 : 1 (1876) pp. 80–160
[15] D.M. Sintsov, "Notes on functional calculus" Izv. Fiz.-Mat. Obshch. Kazan. Univ. (2) , 13 : 2 (1903) pp. 46–72 (In Russian)

Comments

Reference [a3] has a very complete bibliography for the period 1747–1964.

References

[a1] J.K. Hale, "Functional differential equations" , Springer (1971)
[a2] J. Aczél (ed.) , Functional equations: history, applications and theory , Reidel (1984)
[a3] J. Aczél, "Functional equations and their applications" , Acad. Press (1966)
How to Cite This Entry:
Functional equation, methods of solution of a. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Functional_equation,_methods_of_solution_of_a&oldid=12978
This article was adapted from an original article by D.F. Davidenko (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article