Você está na página 1de 10

Mathematical and Computer Modelling 55 (2012) 791800

Contents lists available at SciVerse ScienceDirect

Mathematical and Computer Modelling


journal homepage: www.elsevier.com/locate/mcm

Numerical solution of stochastic Volterra integral equations by a


stochastic operational matrix based on block pulse functions
K. Maleknejad , M. Khodabin, M. Rostami
Department of Mathematics, Karaj Branch, Islamic Azad University, Karaj, Iran

article

info

Article history:
Received 9 February 2011
Received in revised form 25 August 2011
Accepted 25 August 2011
Keywords:
Block pulse functions
Stochastic operational matrix
Stochastic Volterra integral equations
It integral
Brownian motion process

abstract
This article proposes an efficient method for solving stochastic Volterra integral equations.
By using block pulse functions and their stochastic operational matrix of integration,
a stochastic Volterra integral equation can be reduced to a linear lower triangular
system, which can be directly solved by forward substitution. The results show that the
approximate solutions have a good degree of accuracy.
2011 Elsevier Ltd. All rights reserved.

1. Introduction
Stochastic Volterra integral equations arise in many applications such as mathematical finance, biology, medical, social
sciences, etc. There is an increasing demand for studying the behavior of a number of sophisticated dynamical systems
in physical, medical and social sciences, as well as in engineering and finance. These systems are often dependent on
a noise source, on a Gaussian white noise, for example, governed by certain probability laws, so that modeling such
phenomena naturally requires the use of various stochastic differential equations [17] or, in more complicated cases,
stochastic Volterra integral equations and stochastic integro-differential equations [814]. Because in many problems, such
equations, of course, cannot be solved explicitly, it is important to find their approximate solutions by using some numerical
methods [15,1012].
Many orthogonal functions or polynomials, such as block pulse functions, Hybrid functions, Walsh functions, Fourier
series, Legendre polynomials, Chebyshev polynomials, Bernstein polynomials and Laguerre polynomials, were used to derive
solutions of different integral equations, [1523]. In this paper, we use block pulse functions and a stochastic integration
operational matrix.
We consider the following linear stochastic Volterra integral equation,
X (t ) = f (t ) +

k1 (s, t )X (s)ds +
0

k2 (s, t )X (s)dB(s) t [0, T ),


0

where X (t ), f (t ), k1 (s, t ) and k2 (s, t ), for s, t [0, T ), are the stochastic processes
defined on the same probability space
t
( , z, P ), and X (t ) is unknown. Also B(t ) is a Brownian motion process and 0 k2 (s, t )X (s)dB(s) is the It integral.
This paper is organized as follows.
In Section 2, we describe the basic properties of the block-pulse functions and functions approximation by blockpulse functions and an integration operational matrix. In Section 3, we introduce the concept of the stochastic integration

Corresponding author. Tel.: +98 21 732 254 16; fax: +98 21 732 234 16.
E-mail addresses: maleknejad@iust.ac.ir (K. Maleknejad), m-khodabin@kiau.ac.ir (M. Khodabin), m.rostami@kiau.ac.ir (M. Rostami).

0895-7177/$ see front matter 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mcm.2011.08.053

792

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

operational matrix. In Section 4, we solve stochastic Volterra integral equations by using the stochastic integration
operational matrix. In Section 5, we report our numerical findings and demonstrate the accuracy of the proposed scheme
by considering numerical examples. Section 6 is devoted to applications of stochastic Volterra integral equation in the
mathematical finance. Finally, Section 7 gives some brief conclusion.
2. Block pulse functions (BPFs)
BPFs have studied by many authors and applied for solving different problems; for example, see [1523]. The goal of
this section is to recall notations and definition of the block pulse functions, state some known results, and derive useful
formulas that are important for this paper. These have discussed thoroughly in [24,25].
2.1. Definition
We define the m-set of BPFs as

i (t ) =

(i 1)h t < ih,

1
0

(1)

otherwise

T
with t [0, T ), i = 1, 2, . . . , m and h = m
.
The elementary properties of BPFs are as follows

(1) Disjointness: The BPFs are disjointed with each other in the interval t [0, T )

i (t )j (t ) = ij i (t ),

(2)

where i, j = 1, 2, . . . , m and ij is Kronecker delta.


(2) Orthogonality: The BPFs are orthogonal with each other in the interval t [0, T )

i (t )j (t )dt = hij ,

i, j = 1, 2, . . . , m.

(3)

(3) Completeness: If m , then the BPFs set is complete; i.e. For every f L2 ([0, T )), Parsevals identity holds,

f 2 (t )dt =

fi2 i (t )2 ,

(4)

i=1

where
fi =

1
h

f (t )i (t )dt .

(5)

Vector form: Consider the first m terms of BPFs and write them concisely as m-vector,

T
(t ) = 1 (t ), 2 (t ), . . . , m (t ) ,

t [0, T ).

The above representation and disjointness property follows

1 (t )

0
(t ) T (t ) =
..
.

2 (t )
..
.

..
.

0
0

,
..

.
m (t ) mm

(6)

furthermore, we have

T (t ) (t ) = 1,
and,

(t ) T (t )F T = DF (t ),

(7)

where DF usually denotes a diagonal matrix whose diagonal entries are related to a constant vector F = f1 , f2 , . . . , fm

2.2. Functions approximation


An arbitrary real bounded function f (t ), which is square integrable in the interval t [0, T ), can be expanded into a
block pulse series in the sense of minimizing the mean square error between f (t ) and its approximation
f (t ) fm (t ) =

i =1

fi i (t ),

(8)

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

793

where fi is the block pulse coefficient with respect to the ith BPF i (t ). In the vector form we have,
f (t ) fm (t ) = F T (t ) = T (t )F ,

(9)

where
F = f1 , f2 , . . . , fm

Let k(s, t ) L2 [0, T1 ) [0, T2 ) . It can be similarly expanded with respect to BPFs such as

k(s, t ) k m (s, t ) = T (s)K (t ) = T (t )K T (s),

(10)


where (s) and (t ) are m1 and m2 dimensional BPFs vectors respectively, and K = kij , i = 1, 2, . . . , m1 , j =
1, 2, . . . , m2 is the m1 m2 block pulse coefficient matrix with
T1 T2
1
k(s, t )i (s)j (t )dtds,
kij =
h1 h2

where h1 =

T1
m1

, h2 =

T2
.
m2

For convenience, we put m1 = m2 = m.

2.3. Integration operational matrix


Computing
t

t
0

i (s)ds follows

i (s)ds = t (i 1)h
h

0 t < (i 1)h,
(i 1)h t < ih,
ih t < T .

(11)

Since t (i 1)h, equals to 2h , at mid-point of [(i 1)h, ih), we can approximate t (i 1)h, for (i 1)h t < ih, by 2h .
Now expressing

t
0

i (s)ds, in terms of the BPFs follows

i (s)ds 0, . . . , 0, , h, . . . , h (t ),
2

(12)

where 2h is the ith component of vector.


Therefore, [24],
t

(s)ds P (t ),

(13)

where operational matrix of integration is given by


1
0
h 0
P =
2
..

2
1
0

..
.

2
2
1

...
...
...
..
.

..
.

...

2
2
2

..
.

(14)

mm

So, the integral of every function f (t ) can be approximated as follows


t

f (s)ds
0

F T (s)ds F T P (t ).

(15)

3. Stochastic integration operational matrix


The It integral of each single BPFs i (t ) can be computed as follows,
t

i (s)dB(s) = B(t ) B((i 1)h)


B(ih) B((i 1)h)

0 t < (i 1)h,
(i 1)h t < ih,
ih t < T .

(16)

Since B(t )B((i1)h), equals to B((i0.5)h)B((i1)h), at mid-point of [(i1)h, ih), we can approximate B(t )B((i1)h),
for (i 1)h t < ih, by B((i 0.5)h) B((i 1)h).

794

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

Now expressing
t

t
0

i (s)dB(s), in terms of the BPFs follows

i (s)dB(s) B((i 0.5)h) B((i 1)h) i (t ) + B(ih) B((i 1)h)


j (t ),

(17)

j=i+1

and it has the vector form,


t

i (s)dB(s)
0

0, . . . , 0, B((i 0.5)h) B((i 1)h), B(ih) B((i 1)h), . . . , B(ih) B((i 1)h) (t ),

(18)

in which the ith component is B((i 0.5)h) B((i 1)h).


Therefore
t

(s)dB(s) PS (t ),

(19)

where stochastic operational matrix of integration is given by

h
B
2

PS =
0

.
..

B(h)

3h

B(h)

B(2h) B(h)

B(2h) B(h)

5h
B(2h)
B

B(3h) B(2h)

..
.

..
.

..

..
.

B(h)

B(h)

(2m 1)h
2

B((m 1)h)

(20)

mm

So, the It integral of every function f (t ) can be approximated as follows


t

f (s)dB(s)

F T (s)dB(s) F T PS (t ).

(21)

4. Solving stochastic Volterra integral equations by using a stochastic operational matrix


We consider following linear stochastic Volterra integral equation,
X (t ) = f (t ) +

k1 (s, t )X (s)ds +
0

k2 (s, t )X (s)dB(s),

t [0, T ),

(22)

where X (t ), f (t ), k1 (s, t ) and k2 (s, t ), for s, t [0, T ), are the stochastic processes
defined on the same probability space
t
( , z, P ), and X (t ) is unknown. Also B(t ) is a Brownian motion process and 0 k2 (s, t )X (s)dB(s) is the It integral.
We approximate X (t ), f (t ), k1 (s, t ) and k2 (s, t ) by relations (9), (10) as follows
X (t ) X T (t ) = T (t )X ,
f (t ) F T (t ) = T (t )F ,
k1 (s, t ) T (s)K1 (t ) = T (t )K1T (s),
k2 (s, t ) T (s)K2 (t ) = T (t )K2T (s).
In the above approximates, X and F are stochastic block pulse coefficients vector, and K1 and K2 are stochastic block pulse
coefficient matrix.
With substituting above approximation in Eq. (22), we get
X (t ) F (t ) + X
T

(s) (s)ds K1 (t ) + X
T

(s) (s)dB(s) K2 (t ).
T

(23)

Let K1i be the ith row of the constant matrix K1 , and K2i be the ith row of the constant matrix K2 , and Ri be the ith row of
the integration operational matrix P , RiS be the ith row of the stochastic integration operational matrix PS , DK i be a diagonal
1

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

795

matrix with K1i as its diagonal entries, and DK i be a diagonal matrix with K2i as its diagonal entries. By the previous relations
2

and assuming m1 = m2 , we have,

(s) (s)ds K1 (t ) =
T

(s) (s)ds K1 (t )
T

R1 (t )K11 (t )
R2 (t )K 2 (t )
1

R1 DK 1

R D 2
K

= . 1 (t ) = B1 (t ),
..

.
.
.
m
m
m
R (t )K1 (t )
R DK m

(24)

where
k111
0

h
B1 = 0
2 .

2k112
k122
0

..

..
.

2k113
2k123
k133

..
.

..
.

2k11m
2k12m

2k13m

k1mm

..
.

(25)

mm

also,

(s) (s)dB(s) K2 (t ) =
T

(s) (s)dB(s) K2 (t )
T

R1S (t )K21 (t )
R2 (t )K 2 (t )
2
S

R1S DK 1

R2S D 2
K

=
. 2 (t ) = B2 (t ),

..

.
.
.
m
m
Rm

(
t
)
K

(
t
)
RS DK m
S
2

(26)

where

k211 B

B2 =

k212 B(h)

k222 B

k213 B(h)

3h

B(h)

.
.
.

.
.
.

k223 B(2h) B(h)

k233 B

k21m B(h)

5h

B(2h)

.
.
.

k22m B(2h) B(h)

k23m B(3h)

..

B(2h)

.
.
.

k2mm B

(2m 1)h
2

(27)

B((m 1)h)

With substituting relations (24) and (26) in (23), we get,


X T (t ) F T (t ) + X T B1 (t ) + X T B2 (t ).
Then,
X T (I B1 B2 ) F T .

(28)

So, by setting M = (I B1 B2 ) and replacing by =, we will have,


T

MX = F .

(29)

Which is a linear system of equations with lower triangular coefficients matrix that gives the approximate block pulse
coefficient of the unknown stochastic processes X (t ).
5. Numerical examples
To illustrate the method stated in Section 4, we consider following examples. The computations associated with the
examples were performed using Matlab 7. Let Xi denote the block pulse coefficient of exact solution of the given examples,

796

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

Table 1
Mean, standard deviation and mean confidence interval for error in Example 1 with m = 8.
n

xE

30
50
75
100
125
150
200

sE

0.03689954
0.03540561
0.03928085
0.03958671
0.03954587
0.03658486
0.03871999

0.01062056
0.01251955
0.01598618
0.01580345
0.01537519
0.01445615
0.01409427

0.95 confidence interval for mean of E


Lower

Upper

0.03309902
0.03193537
0.03566283
0.03648923
0.03685048
0.03427139
0.03676662

0.04070006
0.03887586
0.04289886
0.04268418
0.04224126
0.03889832
0.04067336

Table 2
Mean, standard deviation and mean confidence interval for error in Example 1 with m = 32.
n

xE

30
50
75
100
125
150
200

sE

0.13182518
0.13332280
0.13507344
0.12575306
0.12785005
0.12548772
0.12876281

0.02498830
0.02686517
0.02526732
0.02557079
0.02564942
0.02434792
0.02755984

0.95 confidence interval for mean of E


Lower

Upper

0.12288323
0.12587615
0.12935491
0.12074118
0.12335350
0.12159124
0.12494321

0.14076713
0.14076944
0.14079198
0.13076494
0.13234659
0.12938420
0.13258241

Fig. 1. The trajectory of the approximate solution and exact solution of Example 1 for m = 32, m = 64, n = 30.

and let Yi be the block pulse coefficient of computed solutions by the presented method. The error is defined as

E = max |Xi Yi |.
1im

Example 1. Consider the following linear stochastic Volterra integral equation,


X (t ) = 1 +

s X (s)ds +
2

sX (s)dB(s) s, t [0, 0.5),

(30)

0
t3

with the exact solution X (t ) = e 6 + 0 sdB(s) , for 0 t < 0.5, where X (t ) is an unknown stochastic process defined on the
probability space ( , z, P ), and B(t ) is a Brownian motion process. The numerical results are shown in Tables 1 and 2. In
Tables 1 and 2, n is the number of iterations, xE is mean of error, and sE is standard deviation of error. The curves in Fig. 1
represents a trajectory of the approximate solution computed by the presented method with a trajectory of exact solution.
The curves in Fig. 2 represents variation process of error.
Example 2. Consider the following linear stochastic Volterra integral equation,
X (t ) =

1
12

cos(s)X (s)ds +

+
0

sin(s)X (s)dB(s) s, t [0, 0.5),


0

(31)

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

797

Fig. 2. Variation trend of error in Example 1 for m = 32, n = 30, n = 100.


Table 3
Mean, standard deviation and mean confidence interval for error in Example 2 with m = 8.
n

xE

30
50
75
100
125
150
200

sE

0.00788410
0.00793958
0.00848855
0.00838714
0.00837599
0.00858557
0.00833360

0.00297687
0.00340213
0.00331070
0.00341973
0.00333239
0.00350444
0.00363129

0.95 confidence interval for mean of E


Lower

Upper

0.00681884
0.00699656
0.00773927
0.00771687
0.00779179
0.00802474
0.00783033

0.00894936
0.00888261
0.00923784
0.00905740
0.00896018
0.00914639
0.00883687

Table 4
Mean, standard deviation and mean confidence interval for error in Example 2 with m = 32.
n

xE

30
50
75
100
125
150
200

sE

0.02308947
0.02341165
0.02345945
0.02364843
0.02345691
0.02377126
0.02322547

0.00442835
0.00511389
0.00459037
0.00524000
0.00477156
0.00462729
0.00470545
sin(2t )

0.95 confidence interval for mean of E


Lower

Upper

0.02150480
0.02199415
0.02242056
0.02262139
0.02262042
0.02303074
0.02257333

0.02467413
0.02482915
0.02449835
0.02467548
0.02429340
0.02451178
0.02387762

1 4 +sin(t )+ 8 + 0 sin(s)dB(s)
with the exact solution X (t ) = 12
e
, for 0 t < 0.5, where X (t ) is an unknown stochastic process
defined on the probability space ( , z, P ), and B(t ) is a Brownian motion process. The numerical results are shown in
Tables 3 and 4. In Tables 3 and 4, n is the number of iterations, xE is mean of error, and sE is standard deviation of error. The
curves in Fig. 3 represents a trajectory of the approximate solution computed by the presented method with a trajectory of
exact solution. The curves in Fig. 4 represents variation process of error.

6. Applications in the mathematical finance


6.1. The basic BlackScholes model
In this subsection we provide a rigorous derivation of the BlackScholes pricing formula, our market consists of just two
securities. The first is our old friend the cash bond, {B(t )}t 0 . We retain (for now) our assumption that the risk-free interest
rate is constant, so that if B0 = 1 then B(t ) = ert . The second security in our market is a risky asset whose price at time t,
we denote by S (t ). In this, our basic reference model, we suppose that {S (t )}t 0 is a geometric Brownian motion, that is it
solves,
dS (t ) = S (t )dt + S (t )dW (t ),
or,
S ( t ) = S0 +

S (s)ds +
0

S (s)dW (s),
0

798

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

Fig. 3. The trajectory of the approximate solution and exact solution of Example 2 for m = 32, m = 64, n = 30.

Fig. 4. Variation trend of error in Example 2 for m = 32, n = 30, n = 100.

for some constants and , where {W (t )}t 0 is a P-Brownian motion. We call P the market measure. Exactly as in the
discrete world, the market measure tells us which market events have positive probability, but we shall reweight the
probabilities for the purposes of pricing and hedging, [27].
6.2. General stock model
In our classical BlackScholes framework we assume that the riskless borrowing rate is constant and that the returns of
the stock follow a Brownian motion with constant drift. In this subsection we consider much more general models to which
we can apply the BlackScholes analysis although, in practice, even for vanilla options the prices that we obtain must now
be evaluated numerically. The key assumption that we retain is that there is only one source of randomness in the market,
the Brownian motion that drives the stock price, [27].
Writing {z(t )}t 0 for the filtration generating the driving Brownian motion, we replace the riskless borrowing rate, r,
the drift and the volatility in our basic BlackScholes model by {z(t )}t 0 -predictable processes {r (t )}t 0 , {(t )}t 0 and
{ (t )}t 0 . In particular, r (t ), (t ) and (t ) can depend on the whole history of the market before time t. Our market model
is then as follows, [27].
General stock model: The market consists of a riskless cash bond, {B(t )}t 0 , and a single risky asset with price process
{S (t )}t 0 governed by

dB(t ) = r (t )B(t )dt , B0 = 1,


dS (t ) = (t )S (t )dt + (t )S (t )dW (t )

(32)

or,

dB(t ) = r (t )B(t )dt , B0 = 1,


t
t
S (t ) = S0 +
(s)S (s)ds +
(s)S (s)dW (s)
0

(33)

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

799

Table 5
Mean, standard deviation and mean confidence interval for error in above example with n = 20.
m

xE

4
8
16
32
48
64
80

sE

0.01495613558
0.02607134096
0.03559123141
0.05520441602
0.07947141115
0.09662823642
0.11213804896

0.95 confidence interval for mean of E

0.00803435227
0.01053117415
0.00982903575
0.01451123788
0.01377181160
0.01970281731
0.01784149356

Lower

Upper

0.01143492545
0.02145585039
0.03128346656
0.04884458561
0.07343564859
0.08799309601
0.10431866964

0.01847734572
0.03068683153
0.03989899626
0.06156424644
0.08550717370
0.10526337684
0.11995742828

Fig. 5. The trajectory of the approximate solution and exact solution of the above example for m = 32, m = 80, n = 20.

where {W (t )}t 0 is a P-Brownian motion generating the filtration {z(t )}t 0 and r (t ), (t ) and (t ) are {z(t )}t 0 predictable processes, [27].
Evidently a solution to these equations should take the form,

B(t ) = exp

r (s)ds ,

(34)

S (t ) = S0 exp

(s)
2

2
s

(s)dW (s) .

ds +

(35)

For example, Consider the following General stock model,

B0 = 1 t [0, 0.8),
dB(t ) = sin(t )B(t )dt ,
t
t
1
S (t ) =
+
ln(1 + s)S (s)ds +
sS (s)dW (s) s, t [0, 0.8)
10

(36)

t3

1 (1+t ) ln(1+t )t 6 + 0 sdW (s)


with the exact solution B(t ) = e1cos(t ) , and S (t ) = 10
e
, for 0 t < 0.8. The numerical results
are shown in Table 5. In Table 5, n is the number of iterations, xE is mean of error, and sE is standard deviation of error. The
curves in Fig. 5 represents a trajectory of the approximate solution computed by the presented method with a trajectory of
exact solution. The curves in Fig. 6 represents variation process of error.

7. Conclusion
Because for some SDEs that can be write as Volterra integral equations, it is impossible to find the exact solution of
Eq. (22), it would be convenient to determine its numerical solution based on stochastic numerical analysis. Using block
pulse functions as basis functions to solve the linear stochastic Volterra integral equations is very simple and effective in
comparison with other methods. Here, the applicability and accuracy of this method discussed in two examples. The results
of numerical solution compared with analytical solution. So, if we encounter with the SDEs similar to Volterra integral
equations as we cannot solve them analytically, we can use this method. Moreover, one could also apply the ItoTaylor
expansion described by Kloeden and Platen [3], or those from article [26], for example. Certainly, it could be the topic of
some future work.

800

K. Maleknejad et al. / Mathematical and Computer Modelling 55 (2012) 791800

Fig. 6. Variation trend of error in the above example for m = 48, n = 20, n = 50.

Acknowledgements
The authors would like to thank Iran National Science Foundation (INSF) for partially financially supporting this research
and also thank the Islamic Azad University of Karaj Branch for providing facilities and encouraging this work.
References
[1] M. Khodabin, K. Maleknejad, M. Rostami, M. Nouri, Numerical solution of stochastic differential equations by second order RungeKutta methods,
Mathematical and Computer Modelling 53 (2011) 19101920.
[2] M. Khodabin, K. Maleknejad, M. Rostami, M. Nouri, Interpolation solution in generalized stochastic exponential population growth model, Applied
Mathematical Modelling, Manuscript (2011) (in press), Corrected Proof, Available online 23 July 2011.
[3] P.E. Kloeden, E. Platen, Numerical Solution of Stochastic Differential Equations, in: Applications of Mathematics, Springer-Verlag, Berlin, 1999.
[4] J.C. Cortes, L. Jodar, L. Villafuerte, Numerical solution of random differential equations: a mean square approach, Mathematical and Computer
Modelling 45 (2007) 757765.
[5] J.C. Cortes, L. Jodar, L. Villafuerte, Mean square numerical solution of random differential equations: facts and possibilities, Computers and Mathematics
with Applications 53 (2007) 10981106.
[6] B. Oksendal, Stochastic Differential Equations, fifth ed., in: An Introduction with Applications, Springer-Verlag, New York, 1998.
[7] H. Holden, B. Oksendal, J. Uboe, T. Zhang, Stochastic Partial Differential Equations, second ed., Springer-Verlag, New York, 2009.
[8] M.A. Berger, V.J. Mizel, Volterra equations with Ito integrals I, Journal of Integral Equations 2 (1980) 187245.
[9] M.G. Murge, B.G. Pachpatte, On second order Ito type stochastic integrodifferential equations, Analele stiintifice ale Universitatii. I. Cuzadin Iasi,
Mathematica (1984) 2534. Tomul xxx, s.I a.
[10] M.G. Murge, B.G. Pachpatte, Succesive approximations for solutions of second order stochastic integrodifferential equations of Ito type, Indian Journal
of Pure and Applied Mathematics 21 (3) (1990) 260274.
[11] X. Zhang, Euler schemes and large deviations for stochastic Volterra equations with singular kernels, Journal of Differential Equations 244 (2008)
22262250.
[12] S. Jankovic, D. Ilic, One linear analytic approximation for stochastic integro-differential eauations, Acta Mathematica Scientia 30B (4) (2010)
10731085.
[13] X. Zhang, Stochastic Volterra equations in Banach spaces and stochastic partial differential equation, Acta Journal of Functional Analysis 258 (2010)
13611425.
[14] J. Yong, Backward stochastic Volterra integral equations and some related problems, Stochastic Processes and their Applications 116 (2006) 779795.
[15] K. Maleknejad, M. Tavassoli Kajani, Solving second kind integral equations by Galerkin methods with hybrid Legendre and block-Pulse functions,
Applied Mathematics and Computation 145 (2003) 623629.
[16] K. Maleknejad, Y. Mahmoudi, Numerical solution of linear Fredholm integral equation by using hybrid Taylor and block-Pulse functions, Applied
Mathematics and Computation 149 (2004) 799806.
[17] K. Maleknejad, S. Sohrabi, Y. Rostami, Numerical solution of nonlinear Volterra integral equations of the second kind by using Chebyshev polynomials,
Applied Mathematics and Computation 188 (2007) 123128.
[18] K. Maleknejad, M. Shahrezaee, H. Khatami, Numerical solution of integral equations system of the second kind by block Pulse functions, Applied
Mathematics and Computation 166 (2005) 1524.
[19] K. Maleknejad, B. Rahimi, Modification of block Pulse Functions and their application to solve numerically Volterra integral equation of the first kind,
Communications in Nonlinear Science and Numerical Simulation 16 (2011) 24692477.
[20] E. Babolian, K. Maleknejad, M. Mordad, B. Rahimi, A numerical method to solve Fredholm-Volterra integral equations in two dimensional spaces using
block Pulse Functions and operational matrix, Journal of Computational and Applied Mathematics 235 (14) (2011) 39653971.
[21] K. Maleknejad, K. Mahdiani, Solving nonlinear mixed VolterraFredholm integral equations with two dimensional block-pulse functions using direct
method, Communications in Nonlinear Science and Numerical Simulation (2011).
[22] K. Maleknejad, B. Basirat, E. Hashemizadeh, Hybrid Legendre polynomials and Block-Pulse functions approach for nonlinear VolterraFredholm
integro-differential equations, Computers & Mathematics with Applications 61 (9) (2011) 28212828.
[23] K. Maleknejad, E. Hashemizadeh, B. Basirat, Computational method based on Bernestein operational matrices for nonlinear VolterraFredholmHammerstein integral equations, Communications in Nonlinear Science and Numerical Simulation 17 (1) (2012) 5261.
[24] Z.H. Jiang, W. Schaufelberger, block Pulse Functions and Their Applications in Control Systems, Springer-Verlag, 1992.
[25] G. Prasada Rao, Piecewise Constant Orthogonal Functions and Their Application to Systems and Control, Springer, Berlin, 1983.
[26] C. Tudor, M. Tudor, Approximation schemes for ItoVolterra stochastic equations, Boletin Sociedad Matemtica Mexicana 3 (1) (1995) 7385.
[27] A. Etheridge, A Course in Financial Calculus, Cambridge University Press, 2002.

Você também pode gostar