Approximation properties of q$Bernoulli polynomials

4 downloads 0 Views 139KB Size Report
Aug 8, 2017 - [6] T. Ernst, the history of q$calculus and a new method, Department of ... [19] W. Rudin, Principles of mathematical analysis, 3rd edition, Mac ...
Approximation properties of q-Bernoulli polynomials M. Momenzadeh and I. Y. Kakangi Near East University Lefkosa, TRNC, Mersiin 10, Turkey Email: [email protected] [email protected] August 8, 2017 Abstract We study the q analogue of Euler-Maclurian formula and by introducing a new q-operator we drive to this form. Moreover, approximation properties of q-Bernoulli polynomials is discussed. We estimate the suitable functions as a combination of truncated series of q-Bernoulli polynomials and the error is calculated. This paper can be helpful in two di¤erent branches, …rst solving di¤erential equations by estimating functions and second we may apply these techniques for operator theory.

1

Introduction

In this paper, we investigate the approximation of suitable function f (x) as a linear combination of q Bernoulli polynomials. We reach to the kind of Euler expansion for f (x) by using q operators and in a parallel way, we expand the function in a terms of q Bernoulli polynomials. This expansion gives the appropriate tool to solve q-di¤erence equations or normal di¤erential equation as well. In fact, several approaches for solving di¤erential equations used a kind of approximation of a function in a terms of Bernoulli polynomials.[20] [21] [7] In this paper, we provide a conditions to approximate capable functions as a linear combination of q-Bernoulli polynomials and some example will be given. We introduce the new q operator to reach the q analogue of Euler-Maclaurian formula. We start by introducing some q-calculus concepts. There are several type of q Bernoulli polynomials and numbers that can be generated by di¤erent q exponential functions.q-analogue of the Bernoulli numbers were …rst studied by Carlitz [2], in the middle of the last century when he introduced a new sequence f m gm 0 : m X

k=0

m k

kq

k+1

m

=

1; 0;

m = 1; m > 1:

(1)

Here and in the remainder of the paper, we make the assumption that jqj < 1:Clearly we recover the ordinary forms when q tends to one from a left side Since Carlitz, there have been many distinct q-analogue of Bernoulli numbers arising from varying motivations. In [3], they used improved q exponential functions to introduce a new class of q Bernoulli polynomials. In addition, they investigate some properties of these q-Bernoulli polynomials. Let us introduce a class of q Bernoulli polynomials in a form of generating function as follows 1 X zeq (zt) = eq (z) 1 n=0

n;q (t)

zn ; [n]q !

jzj < 2 :

(2)

1

where [n]q is q number and q-numbers factorial is de…ned by [a]q =

1 qa 1 q

(q 6= 1) ;

[0]q ! = 1;

[n]q ! = [n]q [n

1]q

n 2 N; a 2 C:

The q-shifted factorial and q-polynomial coe¢ cient are de…ned by (a; q)0 = 1;

(a; q)n =

n Y1

qj a ;

1

n 2 N;

j=0

(a; q)1 = n k

1 Y

qj a ;

1

j=0

= q

jqj < 1; a 2 C:

(q; q)n ; (q; q)n k (q; q)k

respectively.q-standard terminology and notation can be found at [1] and [6]. We call n and n;q = n;q (0). In addition, q-analogue of (x a) is de…ned as

(x

n

a)q =

1; (x

a)(x

aq):::(x

aq n

1

);

n = 0; n 6= 0:

n;q

Bernoulli number

;

In the standard approach to the q calculus, two exponential function are used.These q exponentials are de…ned by

eq (z) =

1 1 Y X zn = [n]q ! (1 n=0

Eq (z) = e1=q (z) =

1 q) q k z)

(1

k=0 1 X

;

0 < jqj < 1; jzj b

(16)

Now in the aid of q Taylor expansion, we may write F (x + h) as follows

F (x + h) =

j

1 X ((x + h)

x)q

[j]q !

j=0

1 X Dqj Hqj

Dqj F (x) =

(F (x)) = eq (Dq Hq ) (F (x))

[j]q !

j=0

(17)

Here, eq is q shifted operator and can be expressed as expansion of Dq and Hq : We may assume that Dh (F (x)) = f (x) and then f (x) =

F (x + h) h

F (x)

=

eq (Dq Hq ) Hq

1

F (x)

(18)

Since the Jackson integral is q antiderivative, we have Hq Dq eq (Dq Hq )

1

Z

f (x)dq x = F (x)

(19)

And in the aid of (2), the left hand side of the equation can be expressed as a q-Bernoulli numbers, therefore

F (x) =

1 X

n

(Hq Dq ) n;q [n]q !

n=0

= =

Z

Z

Z

f (x)dq x

(20)

1 X h n;q f (x) + Dqn [2]q [n] ! q n=1

f (x)dq x

1 n X X h f (x) + [2]q n=1

f (x)dq x

1

f (x) Hqn (x)

n k

k=0

q

k(k 1) 2

(21) n k

(h + x)

k

( x)

q

!

Dqn

1

f (x) [n]q !

n;q

(22)

Therefore we can state the following q analogue of Euler-Maclaurian formula Theorem 6 If the function f (x) is capable of expansion as a convergent power series and f (x) decrease so rapidly with x such that all normal derivatives approach zero as x ! 1, then we can express the series of function f (x) as follow 1 X

f (n) =

Z

1

f (x)dq x +

a

n=a

lim

b!1

1 n X X

n=1

1 (f (a)) [2]q

n k

k=0

q

(23)

k(k 1) 2

n k

(1 + b)

k

( b)

n k

(1 + a)

f (n) =

n=a

Zb

( a)

q

!

hn Dqn

1

f (a) [n]q !

n;q

(24)

Proof. In the aid of theorem (4) and (18), If we suppose that h = 1 and b b 1 X

k

f (x)d1 x = F (b)

a 2 N; we have

F (a)

(25)

a

=

Z

b

1 (f (b) [2]q

f (x)dq x

a

+

1 n X X

n=1

k=0

n k

q

k(k 1) 2

f (a))

(26) n k

(1 + b)

k

( b)

q

n k

(1 + a)

k

( a)

!

Dqn

1

f (b)

Dqn [n]q !

1

f (a) (27)

5

n;q

f (x) x(q 1) ;

Let f (x) decrease so rapidly with x then Dq f (x) can be estimated by

tend x to in…nity and use

0

f (x) (q 1) :Now

L’Hopital to reach If normal derivatives of f (x) is tend to zero, then q derivatives is also tending zero. The same discussion make Dqn 1 f (b) = 0 for big enough b: Therefore 1 X

f (n) =

Z

1

f (x)dq x +

a

n=a

lim

b!1

1 n X X

n=1

1 (f (a)) [2]q

n k

k=0

q

(28)

k(k 1) 2

n k

k

(1 + b)

n k

( b)

(1 + a)

!

k

( a)

q

hn Dqn

1

f (a) [n]q !

n;q

(29)

Example 7 Let f (x) = xs where s is a positive integer, then we can apply this function at (20) where a = 0 and b 1 2 N b 1 X

bs+1 n = [s + 1]q n=0 s

+

s+1 X n X

n=2

k=0

bs [2]q n k

! q

(30) k(k 1) 2

n k

(1 + b)

k

( b)

q

!

bs

s n

1

n+1

n;q

(31)

[n]q

q

This relation shows the sum of power in the combination of q Bernoulli numbers. when q ! 1 from the right side, we have an ordinary form of this relation. For the another forms of sum of power, see [22] Example 8 Let f (x) = eq x = E1x then this function decreases so rapidly with x such that all normal q derivatives approach zero as x ! 1: For com…rming this, let us mention that Eqx ; for some …xed 0 < jqj < 1 1 1 P Q d and j xj < j1 1 qj is increasing rapidly, since dx Eqx = (1 q)q j 1 + (1 q)q k x > 0: j=1

1 X

eq

n

n=0

1 =1+ [2]q

lim

b!1

1 n X X

n=1

k=0

P1 Moreover, this is q-analogue of n=0 e numbers that is generated by e 11 1 :

3

n k n

q

k(k 1) 2

k=0 k6=j

n k

(1 + b)

k

( b)

q

=

1 e

1

1

=

3 2

+

P1

n=1 (2n)! ; 2n

!

n;q

(32)

[n]q !

where

2n

is a normal Bernoulli

approximation by q-Bernoulli polynomial

We know that the class of q-Bernoulli polynomials are not orthogonal polynomials. In addition, we may apply Gram-Schmit algorithm to make these polynomials orthonormal then according to Theorem 8.11 [19] we have the best approximation in the form of Fourier series. Now, Instead of using that algorithm, we apply properR1 ties of lemma (1) to achieve an approximation. Let H = L2q [0; 1] = fq : [0; 1] ! [0; 1] j fq2 (t) dq t < 1 0

be the Hilbert Space[18], then n;q (x) 2 H for n = 0; :::; N and Y = Span f 0;q (x); 1;q (x); :::; N;q (x)g is …nite dimensional vector subspace of H. Unique best approximation for any arbitrary elements of H like h; is b h 2 Y such that for any y 2 Y the inequality h b h kh yk2;q ; where the norm is de…ned 2;q

by kf k2;q :=

R1

1 2

fq2 (t) dq t

: Following proposition determine the coe¢ cient of q Bernoulli polynomials,

0

6

when we estimate any f 2 L2q [0; 1] by truncated q Bernoulli series. In fact, in order to see how well a certain partial sum approximates the actual value, we would like to drive a formula similar to (19), but with the in…nite sum on the right hand side replaced by the N th partial sum SN , plus an additional term RN: Sippose a 2 Z and b = a + 1:Consider the Nth partial sum SN =

N X

k=0

k;q

Let g(x; y) =

SN =

N X

N;q (x;y)

[N ]q !

[k]q !

N X

DqN

k

Hqk Dqk (f ) (a)

; then in the aid of lemma (1) we have Dqn

k;q (0; 0)

k=0

=

Hqk Dqk (f ) (a + 1)

[k]q !

Hqk Dqk

(f ) (a + 1)

k;q (0; 1)

[k]q !

g(0; 0)Hqk Dqk (f ) (a + 1)

DqN

k

Hqk Dqk

k

g(x; y) =

!

(f ) (a)

k;q (x;y)

[k]q !

and we have

Dq (f ) (a)

g(0; 1)Hqk Dqk (f ) (a)

Dq (f ) (a)

k=0

= Dq (f ) (a)

N X

DqN

k

g(0; x)Hqk Dqk (f ) (a + 1

x)jx=1 x=0

k=0

By taking q derivative of the summation, we may rewrite this partial sum by integral presentation, Moreover this relation give a boundary for RN (x): PN Proposition 9 Let f 2 L2q [0; 1] and be estimated by truncated q Bernoulli series n=0 Cn n;q (x): Then R1 (n) Cn coe¢ cients for n = 0; 1; :::; N can be calculated as an integral form Cn = [n]1 ! Dq f (x)dq x. In addition q

0

we can write f (x) in the following form

f (x) =

Z

N X

1

0

h @ 1 f (x) + [2]q [n]q ! n=1

f (x)dq x

0

Z1 0

1

Dq(n) f (x)dq xA

Moreover Rq;n (x) as a reminder part is bounded by

n;q (x)

+ Rq;n (x)

(n) 2n [N ]q ! Supx2[0;1] j n;q (x)jSupx2[0;1] jDq f (x)j

Proof. In fact we approximate f (x) as a linear combinations of q Bernoulli polynomials and we assume PN that f (x) ' n=0 Cn n;q (x); so take the Jackson integral from both sides lead us to Z

1

0

f (x)dq x '

N X

n=0

Cn

Z

1

n;q (x)dq x

0

= C0

Z

1

0;q (x)dq x

+ C1

0

Z

1

1;q (x)dq x

0

+ ::: + CN

Z

1

N;q (x)dq x

0

In the aid of Lemma 1 part (d) all the terms at the right side except the …rst one has to be zero. we R1 calculate 0;q (x) = 1; so C0 = [0]1 ! f (x)dq x:Using part (a) of that lemma for q derivative of q Bernoulli q

0

polynomial gather by taking q-derivative of f (x) leads to Z 1 Z 1 Z N X Dq (f (x)) dq x ' Cn [n]q ! (x)d x = C n 1;q q 1 0

n=0

0

1 0;q (x)dq x

= C1

0

Repeating this procedure n-times, yields the form of Cn as we mention it at theorem. We mention that if x 2 [0; 1] and 0 < jqj < 1, then Hqn (x) for h = 1 is bounded by 2n :In addition, the reminder part can be presented by integral forms and this boundary can be found easily.

7

References [1] V. kac, P. Cheung, Quantum Calculus ISBN 0-387-95341-8 (2000) [2] L. Carlitz, q-Bernoulli numbers and polynomials, Duke Math. J. 15 (1948) 987–1000. [3] N. I. Mahmudov and M. Momenzadeh, On a Class of q-Bernoulli, q-Euler, and q-Genocchi Polynomials, Abstract and Applied AnalysisVolume 2014 (2014), Article ID 696454, 10 pages [4] W. Van Assche, Little q Legendre polynomials and irrationality of certain Lambert series, The Ramanujan Journal 5 (2001), 295-310 [5] Jan L. Cie´sli´nski, Improved q-exponential and q-trigonometric functions,Applied Mathematics Letters Volume 24, Issue 12, December 2011, Pages 2110–2114. [6] T. Ernst, the history of q-calculus and a new method, Department of Mathematics, Uppsala University (2000). [7] E. Tohidi, A. Klicman, A Collocation Method Based on the Bernoulli Operational Matrix for Solving Nonlinear BVPs Which Arise from the Problems in Calculus of Variatio, Vol 2013, Article ID 757206, Hindawi(2013). [8] D.J. Evans, K.R. Raslan, The adomain decomposition method for solving di¤erential equation, int. J. Comput. Math. 82 (2005) 49-54 [9] F.H. Jackson, On q-functions and a certain di¤erence operator. Trans. Roy. Soc. Edinb. 46, 64–72 (1908) [10] R.D. Carmichael, On the theory of linear di¤erence equations. Am. J. Math. 35(2), 163–182 (1913) [11] B.Q. Chen, X.Z. Chen, S. Li, Properties on solutions of some q-di¤erence equations. Acta Math. Sin. (Engl. Ser.) 26(10), 1877–1886 (2010) [12] A. Dobrogowska, A. Odzijewicz, Second order q-di¤erence equations solvable by factorization method. J. Comput. Appl. Math. 193(1), 319–346 (2006) [13] H.I. Abdel-Gawad, A.A. Aldailami, On q-dynamic equations modelling and complexity. Appl. Math. Model. 34(3), 697–709 (2010) [14] A. Algin, A comparative study on q-deformed fermion oscillators. Int. J. Theor. Phys. 50(5), 1554–1568 (2011) [15] A. Dobrogowska, A. Odzijewicz, Solutions of the q-deformed Schr¨ odinger equation for special potentials. J. Phys. A Math. Theor. 40(9), 2023–2036 (2007) [16] H.Mabrouk, q-heat operatorand q-Poisson’s operator. Fract.Calc.Appl.Anal.9(3), 265–286 (2006) [17] A. Nemri, A. Fitouhi, Polynomial expansions for solution of wave equation in quantum calculus. Matematiche 65(1), 73–82 (2010) [18] M. H. Annaby, Z. Mansour, q-fractional calculus and equations, Springer ISBN 978-3-642-30897-0 (2012) [19] W. Rudin, Principles of mathematical analysis, 3rd edition, Mac Graw-Hill (1964) [20] E. Tohidi, M. zak, A new matrix approach for solving second-order linear matrix partial di¤erential equations, Mediterranean journal of mathematics (2016), vol 13, pp 1353-1376 [21] B. Zogheib, E. Tohidi, A new matrix method for solving two-dimentional time dependent di¤usion equation with Dirichlet boundary conditions (2016), vol 291, pp 1 8

[22] O. Chan, D. Manna, A new q-analogue for Bernoulli numbers (2010), Ams clas. no. 11B68, 30c15, 05A30 [23] R. Carmichael, A. M., A treatise on the calculus of operations, Longman, brown, green, and longmans pp 92-103(1855) [24] P. Ricci, I. Tavkelidze, An introduction to opertional techniques and special polynomials, Journal of mathematical sciences, Vol 157, No.1 (2009) [25] P. Ricci, I. Tavkelidze, Operational methods and solutions of boundary-value problems with periodic data, Journal of mathematical sciences, Vol 157, No.1 (2009) [26] H. Davis, The thory of linear operators, Principia press (1936) [27] G. Boole, A tretise on the calculus of …nite di¤erences, Macmillan and co. (1880)

9