A Compact Difference Scheme for Solving Fractional Neutral ...

2 downloads 0 Views 3MB Size Report
Sep 17, 2017 - 2School of Finance, Zhongnan University of Economics and Law, Wuhan 430073, China. Correspondence should be addressed to Yanli Zhou;ย ...
Hindawi Journal of Function Spaces Volume 2017, Article ID 3679526, 8 pages https://doi.org/10.1155/2017/3679526

Research Article A Compact Difference Scheme for Solving Fractional Neutral Parabolic Differential Equation with Proportional Delay Wei Gu,1 Yanli Zhou,2 and Xiangyu Ge1 1

School of Statistics & Mathematics, Zhongnan University of Economics and Law, Wuhan 430073, China School of Finance, Zhongnan University of Economics and Law, Wuhan 430073, China

2

Correspondence should be addressed to Yanli Zhou; [email protected] and Xiangyu Ge; xiangyu [email protected] Received 5 July 2017; Accepted 17 September 2017; Published 18 October 2017 Academic Editor: Xinguang Zhang Copyright ยฉ 2017 Wei Gu et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. A linearized compact finite difference scheme is constructed for solving the fractional neutral parabolic differential equation with proportional delay. By the energy method, the unconditional stability of the scheme is proved, and the convergence order of the scheme is proved to be ๐‘‚(๐œ2โˆ’๐›ผ + โ„Ž4 ). A numerical test is also conducted to validate the accuracy and efficiency of the numerical algorithm.

1. Introduction

๐‘ค (0, ๐‘ก) = ๐›พ๐‘ค (๐‘ก) ,

In the past few years, more and more scholars have been attracted to the research of delay partial differential equations (DPDEs) [1โ€“3]. However, most DPDEs have no exact solutions. Constructing efficient numerical methods for DPDEs is of great importance [4โ€“7]. For details on numerically solving neutral delay parabolic differential equations (NDPDEs), the reader is referred to [5, 8]. Recently, fractional delay partial differential equations have been of great interest due to their application in automatic control, population dynamics, economics, and so forth [9, 10]. For details on numerical solutions to fractional delay partial differential equations, we refer the reader to [11, 12]. The work in [11] considers the numerical method without theoretical analysis, and the work in [12] considers the numerical method for a type of semilinear fractional partial differential equation with time delay. In this paper, we consider the following fractional neutral parabolic differential equation with proportional delay:

๐‘ค (1, ๐‘ก) = ๐›ฝ๐‘ค (๐‘ก) ,

2 ๐œ•2 ๐‘ค (๐‘ฅ, ๐‘ก) ๐œ• ๐‘ค (๐‘ฅ, ๐‘๐‘ก) ๐œ•๐›ผ ๐‘ค = ๐ท( + ) ๐›ผ 2 ๐œ•๐‘ก ๐œ•๐‘ฅ ๐œ•๐‘ฅ2

๐‘ก โˆˆ (0, ๐‘‡] , (1) where ๐ท > 0 is a constant, 0 < ๐‘ < 1. Let ๐‘ข(๐‘ฅ, ๐‘ก) = ๐‘ค(๐‘ฅ, ๐‘’๐‘ก ) for ๐‘ก โ‰ฅ ๐‘ก0 + ln(๐‘), where ๐‘ก0 โ‰ฅ 0. Then, ๐‘ข(๐‘ฅ, ๐‘ก) satisfies the following equation: ๐œ•๐›ผ ๐‘ข ๐œ•2 ๐‘ข (๐‘ฅ, ๐‘ก) ๐œ•2 ๐‘ข (๐‘ฅ, ๐‘ก โˆ’ ๐‘ ) = ๐ท ( + ) ๐œ•๐‘ก๐›ผ ๐œ•๐‘ฅ2 ๐œ•๐‘ฅ2 + ๐‘“๐‘ค (๐‘ข (๐‘ฅ, ๐‘ก) , ๐‘ข (๐‘ฅ, ๐‘ก โˆ’ ๐‘ ) , ๐‘ฅ, ๐‘’๐‘ก ) , (๐‘ฅ, ๐‘ก) โˆˆ (0, 1) ร— (๐‘ก0 , ๐‘‡] , ๐‘ข (๐‘ฅ, ๐‘ก) = ๐œ™๐‘ค (๐‘ฅ, ๐‘’๐‘ก ) , ๐‘ข (0, ๐‘ก) = ๐›พ๐‘ค (๐‘’๐‘ก ) , ๐‘ข (1, ๐‘ก) = ๐›ฝ๐‘ค (๐‘’๐‘ก ) , ๐‘ก โˆˆ (๐‘ก0 , ๐‘‡] ,

+ ๐‘“๐‘ค (๐‘ค (๐‘ฅ, ๐‘ก) , ๐‘ค (๐‘ฅ, ๐‘๐‘ก) , ๐‘ฅ, ๐‘ก) , (๐‘ฅ, ๐‘ก) โˆˆ (0, 1) ร— [0, ๐‘‡] ,

๐‘ฅ โˆˆ [0, 1] , ๐‘ก โˆˆ [โˆ’๐‘ , ๐‘ก0 ] ,

where ๐‘  = โˆ’ ln(๐‘).

(2)

2

Journal of Function Spaces

For simplicity, we consider the following fractional neutral parabolic differential equation with delay instead of (1): ๐œ•๐›ผ ๐‘ข ๐œ•2 ๐‘ข (๐‘ฅ, ๐‘ก) ๐œ•2 ๐‘ข (๐‘ฅ, ๐‘ก โˆ’ ๐‘ ) = ๐ท ( + ) ๐œ•๐‘ก๐›ผ ๐œ•๐‘ฅ2 ๐œ•๐‘ฅ2

๐‘˜ = ๐›ฟ๐‘ฅ V๐‘–+1/2

(3)

+ ๐‘“ (๐‘ข (๐‘ฅ, ๐‘ก) , ๐‘ข (๐‘ฅ, ๐‘ก โˆ’ ๐‘ ) , ๐‘ฅ, ๐‘ก) , (๐‘ฅ, ๐‘ก) โˆˆ (0, 1) ร— (0, ๐‘‡] , ๐‘ข (๐‘ฅ, ๐‘ก) = ๐œ™ (๐‘ฅ, ๐‘ก) ,

be the grid function space defined on ฮฉโ„Ž๐œ . The following notations are used:

๐‘ฅ โˆˆ [0, 1] , ๐‘ก โˆˆ [โˆ’๐‘ , 0] ,

(4)

๐‘ข (0, ๐‘ก) = ๐›พ (๐‘ก) , ๐‘ข (1, ๐‘ก) = ๐›ฝ (๐‘ก) ,

(5) ๐‘ก โˆˆ (0, ๐‘‡] ,

where ๐‘  > 0 is a constant delay term. Time fractional partial derivative (๐œ•๐›ผ ๐‘ข/๐œ•๐‘ก๐›ผ ) (0 < ๐›ผ < 1) is defined in the Caputo sense by the following: ๐‘ก ๐œ•๐›ผ ๐‘ข 1 โˆ’๐›ผ ๐œ•๐‘ข (๐‘ฅ, ๐œ‰) = ๐‘‘๐œ‰, โˆซ (๐‘ก โˆ’ ๐œ‰) ๐œ•๐‘ก๐›ผ ฮ“ (1 โˆ’ ๐›ผ) 0 ๐œ•๐œ‰

(6)

where ฮ“( ) is the Gamma function. In this paper, a linearized compact finite difference scheme is constructed for solving (3)โ€“(5). By the energy method, the unconditional stability of the scheme is then proved, and the convergence order of the scheme is proved to be ๐‘‚(๐œ2โˆ’๐›ผ +โ„Ž4 ). A numerical test is also conducted to validate the accuracy and efficiency of the numerical algorithm. The rest of the paper is organized as follows. In Section 2, a compact difference scheme is constructed to solve (3)โ€“(5). Section 3 considers the solvability, convergence, and stability of the provided difference scheme. In Section 4, a numerical test is presented to illustrate the validity of the theoretical results. Section 5 gives a brief conclusion of this paper.

2. The Construction of the Compact Difference Scheme

๐›ฟ๐‘ฅ2 V๐‘–๐‘˜ =

๐‘˜ ๐‘˜ โˆ’ 2V๐‘–๐‘˜ + V๐‘–โˆ’1 V๐‘–+1 , โ„Ž2

AV๐‘–๐‘˜ =

1 ๐‘˜ ๐‘˜ ), (V + 10V๐‘–๐‘˜ + V๐‘–+1 12 ๐‘–โˆ’1

๐›ฟ๐‘ก๐›ผ V๐‘˜ =

โˆ’๐›ผ

(9)

๐‘˜โˆ’1

๐œ [๐‘‘0 V๐‘˜ โˆ’ โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) V๐‘— ฮ“ (2 โˆ’ ๐›ผ) ๐‘—=1 [

โˆ’ ๐‘‘๐‘˜โˆ’1 V0 ] , ] where ๐‘‘๐‘— = (๐‘— + 1)1โˆ’๐›ผ โˆ’ ๐‘—1โˆ’๐›ผ , ๐‘— โ‰ฅ 0. For the time fractional derivative, we have the following lemma. Lemma 1 (see [13]). Suppose 0 < ๐›ผ < 1, ๐‘ฆ โˆˆ ๐ถ2 [0, ๐‘ก๐‘˜ ]; it holds that ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ ๐‘ก๐‘˜ ๐‘ฆ๓ธ€  (๐‘ ) ๐‘‘๐‘  1 ๐œโˆ’๐›ผ [ ๓ต„จ๓ต„จ ๐‘‘0 ๐‘ฆ (๐‘ก๐‘˜ ) โˆซ ๓ต„จ๓ต„จ ๐›ผ โˆ’ ๓ต„จ๓ต„จ ฮ“ (1 โˆ’ ๐›ผ) 0 (๐‘ก๐‘˜ โˆ’ ๐‘ ) ฮ“ (2 โˆ’ ๐›ผ) ๓ต„จ๓ต„จ [ ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ ๓ต„จ โˆ’ โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) ๐‘ฆ (๐‘ก๐‘— ) โˆ’ ๐‘‘๐‘˜โˆ’1 ๐‘ฆ (๐‘ก0 )]๓ต„จ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ ๐‘—=1 ]๓ต„จ๓ต„จ ๐‘˜โˆ’1

โ‰ค

(10)

1 โˆ’ ๐›ผ 22โˆ’๐›ผ 1 [ + โˆ’ (1 + 2โˆ’๐›ผ )] ฮ“ (2 โˆ’ ๐›ผ) 12 2โˆ’๐›ผ

๓ต„จ ๓ต„จ โ‹… max ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐‘ฆ๓ธ€ ๓ธ€  (๐‘ก)๓ต„จ๓ต„จ๓ต„จ๓ต„จ ๐œ2โˆ’๐›ผ , 0โ‰ค๐‘กโ‰ค๐‘ก๐‘˜ and ๐‘‘๐‘— satisfies the following lemma.

Throughout this paper, assume ๐‘ข(๐‘ฅ, ๐‘ก) โˆˆ ๐ถ6,2 ([0, 1] ร— (0, ๐‘‡]). Function ๐‘“(๐‘ข(๐‘ฅ, ๐‘ก), ๐‘ข(๐‘ฅ, ๐‘ก โˆ’ ๐‘ ), ๐‘ฅ, ๐‘ก) is sufficiently smooth and satisfies ๓ต„จ๓ต„จ๓ต„จ๐‘“ (๐œ‡ + ๐œ–1 , ] + ๐œ–2 , ๐‘ฅ, ๐‘ก) โˆ’ ๐‘“ (๐œ‡, ], ๐‘ฅ, ๐‘ก)๓ต„จ๓ต„จ๓ต„จ ๓ต„จ ๓ต„จ (7) ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ โ‰ค ๐‘0 ๓ต„จ๓ต„จ๐œ–1 ๓ต„จ๓ต„จ + ๐‘1 ๓ต„จ๓ต„จ๐œ–2 ๓ต„จ๓ต„จ , where ๐œ–1 , ๐œ–2 are arbitrary real numbers and ๐‘0 and ๐‘1 are positive constants. First, let ๐‘€ and ๐‘ be two positive integers; then, we take โ„Ž = 1/๐‘€, ๐œ = ๐‘ /๐‘› (๐‘› > 0 is a positive integer), ๐‘ฅ๐‘– = ๐‘–โ„Ž, ๐‘ก๐‘˜ = ๐‘˜๐œ. Define ฮฉโ„Ž๐œ = ฮฉโ„Ž ร— ฮฉ๐œ , where ฮฉโ„Ž = {๐‘ฅ๐‘– | 0 โ‰ค ๐‘– โ‰ค ๐‘€}, ฮฉ๐œ = {๐‘ก๐‘˜ | โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค ๐‘}, ๐‘ = [๐‘‡/๐œ]. Denote ๐‘ˆ๐‘–๐‘˜ = ๐‘ข(๐‘ฅ๐‘– , ๐‘ก๐‘˜ ), 0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค ๐‘ throughout this paper. Let W = {V๐‘–๐‘˜ | 0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค ๐‘}

๐‘˜ โˆ’ V๐‘–๐‘˜ V๐‘–+1 , โ„Ž

(8)

Lemma 2 (see [14]). Assume 0 < ๐›ผ < 1; then, it holds that (1) ๐‘‘๐‘— decreases monotonically as ๐‘— increases, and 0 < ๐‘‘๐‘— โ‰ค 1; (2) ๐‘‘0 = 1, โˆ‘๐‘˜โˆ’1 ๐‘—=1 (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) = ๐‘‘0 โˆ’ ๐‘‘๐‘˜โˆ’1 . Lemma 3 (see [5]). Suppose ๐‘ฆ(๐‘ฅ) โˆˆ ๐ถ6 [๐‘ฅ๐‘–โˆ’1 , ๐‘ฅ๐‘–+1 ]; then, one has 1 [๐‘ฆ๓ธ€ ๓ธ€  (๐‘ฅ๐‘–โˆ’1 ) + 10๐‘ฆ๓ธ€ ๓ธ€  (๐‘ฅ๐‘– ) + ๐‘ฆ๓ธ€ ๓ธ€  (๐‘ฅ๐‘–+1 )] 12 โˆ’ =

1 [๐‘ฆ (๐‘ฅ๐‘–โˆ’1 ) โˆ’ 2๐‘ฆ (๐‘ฅ๐‘– ) + ๐‘ฆ (๐‘ฅ๐‘–+1 )] โ„Ž2

โ„Ž4 (6) ๐‘ฆ (๐œ„๐‘– ) , 240

where ๐œ„๐‘– โˆˆ (๐‘ฅ๐‘–โˆ’1 , ๐‘ฅ๐‘–+1 ).

(11)

Journal of Function Spaces

3

Considering (3) at the point (๐‘ฅ๐‘– , ๐‘ก๐‘˜ ), we have

where

๐œ•๐›ผ ๐‘ข ๐œ•2 ๐‘ข ๐œ•2 ๐‘ข (๐‘ฅ , ๐‘ก ) = ๐ท ( (๐‘ฅ , ๐‘ก ) + (๐‘ฅ , ๐‘ก โˆ’ ๐‘ )) ๐‘– ๐‘˜ ๐‘– ๐‘˜ ๐œ•๐‘ก๐›ผ ๐œ•๐‘ฅ2 ๐œ•๐‘ฅ2 ๐‘– ๐‘˜ + ๐‘“ (๐‘ข (๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) , ๐‘ข (๐‘ฅ๐‘– , ๐‘ก๐‘˜ โˆ’ ๐‘ ) , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) ,

๐‘˜ + ๐‘…๐‘–๐‘˜ = A๐‘…0๐‘–

From Lemma 1, we obtain ๐œ•๐›ผ ๐‘ข (๐‘ฅ , ๐‘ก ) = ๐›ฟ๐‘ก๐›ผ ๐‘ˆ๐‘–๐‘˜ + ๐‘Ÿ๐‘–๐‘˜ , ๐œ•๐‘ก๐›ผ ๐‘– ๐‘˜

(13)

where ๐‘Ÿ๐‘–๐‘˜ =

1 1 โˆ’ ๐›ผ 22โˆ’๐›ผ [ + โˆ’ (1 + 2โˆ’๐›ผ )] ฮ“ (2 โˆ’ ๐›ผ) 12 2โˆ’๐›ผ ๓ต„จ๓ต„จ ๐œ•2 ๐‘ข (๐‘ฅ , ๐‘ก ) ๓ต„จ๓ต„จ ๓ต„จ ๐‘– ๐‘˜ ๓ต„จ๓ต„จ 2โˆ’๐›ผ ๓ต„จ๓ต„จ ๐œ . โ‹… max ๓ต„จ๓ต„จ๓ต„จ๓ต„จ 2 ๓ต„จ๓ต„จ 0โ‰ค๐‘กโ‰ค๐‘ก๐‘˜ ๓ต„จ ๐œ•๐‘ก ๓ต„จ ๓ต„จ

(14)

Replacing ๐‘ˆ๐‘–๐‘˜ by ๐‘ข๐‘–๐‘˜ in (19), (22), and (23) and omitting ๐‘…๐‘–๐‘˜ , we can obtain the following compact difference scheme: (15)

2

where ๐œ‚๐‘˜ โˆˆ (๐‘ก๐‘˜โˆ’2 , ๐‘ก๐‘˜ ), ๓ฐœš๐‘–๐‘˜ in between ๐‘ข(๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) and 2๐‘ˆ๐‘–๐‘˜โˆ’1 โˆ’๐‘ˆ๐‘–๐‘˜โˆ’2 . Substituting (13) and (15) into (12) and applying the operator A on both sides of (12), we obtain ๐œ•2 ๐‘ข ๐œ•2 ๐‘ข A๐›ฟ๐‘ก๐›ผ ๐‘ˆ๐‘–๐‘˜ = ๐ท (A 2 (๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) + A 2 (๐‘ฅ๐‘– , ๐‘ก๐‘˜ โˆ’ ๐‘ )) ๐œ•๐‘ฅ ๐œ•๐‘ฅ

(17)

๐œ•2 ๐‘ข โ„Ž4 ๐œ•6 ๐‘ข ๐‘˜ 2 ๐‘˜ (๐‘ฅ , ๐‘ก ) = ๐›ฟ ๐‘ˆ + (๐œƒ , ๐‘ก ) , ๐‘ฅ ๐‘– ๐œ•๐‘ฅ2 ๐‘– ๐‘˜ 240 ๐œ•๐‘ฅ6 ๐‘– ๐‘˜

๐‘ข0๐‘˜ = ๐›พ (๐‘ก๐‘˜ ) , ๐‘˜ = ๐›ฝ (๐‘ก๐‘˜ ) , ๐‘ข๐‘€

(26) 1 โ‰ค ๐‘˜ โ‰ค ๐‘.

Define the following grid function space on ฮฉโ„Ž : Vโ„Ž,0 = {๐‘ข | ๐‘ข = (๐‘ข0 , ๐‘ข1 , . . . , ๐‘ข๐‘€) , ๐‘ข0 = ๐‘ข๐‘€ = 0} .

(27)

๐‘€โˆ’1

๐œƒ๐‘–๐‘˜ โˆˆ (๐‘ฅ๐‘–โˆ’1 , ๐‘ฅ๐‘–+1 ) , โ„Ž4 ๐œ•6 ๐‘ข ๐‘˜โˆ’๐‘› ๐œ•2 ๐‘ข (๐œƒ , ๐‘ก๐‘˜โˆ’๐‘› ) , A 2 (๐‘ฅ๐‘– , ๐‘ก๐‘˜ โˆ’ ๐‘ ) = ๐›ฟ๐‘ฅ2 ๐‘ˆ๐‘–๐‘˜โˆ’๐‘› + ๐œ•๐‘ฅ 240 ๐œ•๐‘ฅ6 ๐‘–

(๐‘ข, V) = โ„Ž โˆ‘ ๐‘ข๐‘– V๐‘– , ๐‘–=1

(18)

๓ต„จ ๓ต„จ โ€–๐‘ขโ€– = โˆš(๐‘ข, ๐‘ข), โ€–๐‘ขโ€–โˆž = max ๓ต„จ๓ต„จ๓ต„จ๐‘ข๐‘– ๓ต„จ๓ต„จ๓ต„จ , 1โ‰ค๐‘–โ‰ค๐‘€โˆ’1 ๐‘€โˆ’1

โŸจ๐›ฟ๐‘ฅ ๐‘ข, ๐›ฟ๐‘ฅ VโŸฉ = โ„Ž โˆ‘ (๐›ฟ๐‘ฅ ๐‘ข๐‘–+1/2 ) (๐›ฟ๐‘ฅ V๐‘–+1/2 ) ,

๐œƒ๐‘–๐‘˜โˆ’๐‘› โˆˆ (๐‘ฅ๐‘–โˆ’1 , ๐‘ฅ๐‘–+1 ) .

๐‘–=0

Substituting (18) into (16), we have

๓ต„จ๓ต„จ ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐‘ข๓ต„จ๓ต„จ1 = โˆšโŸจ๐›ฟ๐‘ฅ ๐‘ข, ๐›ฟ๐‘ฅ ๐‘ขโŸฉ.

A๐›ฟ๐‘ก๐›ผ ๐‘ˆ๐‘–๐‘˜ = ๐ท (๐›ฟ๐‘ฅ2 ๐‘ˆ๐‘–๐‘˜ + ๐›ฟ๐‘ฅ2 ๐‘ˆ๐‘–๐‘˜โˆ’๐‘› ) โˆ’

(25)

If ๐‘ข, V โˆˆ Vโ„Ž,0 , we introduce the following inner products and corresponding norms:

From Lemma 3 and Taylor expansion, we have

+

(24)

3. The Solvability, Convergence, and Stability of the Difference Scheme

where

๐‘ˆ๐‘–๐‘˜โˆ’2 , ๐‘ˆ๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ )

๐‘ข๐‘–๐‘˜ = ๐œ™ (๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) , 0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค 0,

(16)

๐‘˜ + A๐‘“ (2๐‘ˆ๐‘–๐‘˜โˆ’1 โˆ’ ๐‘ˆ๐‘–๐‘˜โˆ’2 , ๐‘ˆ๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) + A๐‘…0๐‘– ,

๐œ•2 ๐‘ข (๐‘ฅ , ๐œ‚๐‘˜ ) ๐‘“๐œ‡ (๓ฐœš๐‘–๐‘˜ , ๐‘ˆ๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) . ๐œ•๐‘ก2 ๐‘–

A๐›ฟ๐‘ก๐›ผ ๐‘ข๐‘–๐‘˜ = ๐ท (๐›ฟ๐‘ฅ2 ๐‘ข๐‘–๐‘˜ + ๐›ฟ๐‘ฅ2 ๐‘ข๐‘–๐‘˜โˆ’๐‘› ) + A๐‘“ (2๐‘ข๐‘–๐‘˜โˆ’1 โˆ’ ๐‘ข๐‘–๐‘˜โˆ’2 , ๐‘ข๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) ,

๐‘ข + ๐œ 2 (๐‘ฅ๐‘– , ๐œ‚๐‘˜ ) ๐‘“๐œ‡ (๓ฐœš๐‘–๐‘˜ , ๐‘ˆ๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) , ๐œ•๐‘ก

A๐‘“ (2๐‘ˆ๐‘–๐‘˜โˆ’1

(23) 1 โ‰ค ๐‘˜ โ‰ค ๐‘.

= ๐‘“ (2๐‘ˆ๐‘–๐‘˜โˆ’1 โˆ’ ๐‘ˆ๐‘–๐‘˜โˆ’2 , ๐‘ˆ๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ )

A

(22)

๐‘ˆ0๐‘˜ = ๐›พ (๐‘ก๐‘˜ ) , ๐‘˜ = ๐›ฝ (๐‘ก๐‘˜ ) , ๐‘ˆ๐‘€

๐‘“ (๐‘ข (๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) , ๐‘ข (๐‘ฅ๐‘– , ๐‘ก๐‘˜ โˆ’ ๐‘ ) , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ )

๐‘˜ = ๐‘Ÿ๐‘–๐‘˜ + ๐œ2 ๐‘…0๐‘–

(20)

Noticing ๐‘ข(๐‘ฅ, ๐‘ก) โˆˆ ๐ถ6,2 ([0, 1] ร— (0, ๐‘‡]) and (7), we can easily obtain ๓ต„จ๓ต„จ๓ต„จ๐‘…๐‘˜ ๓ต„จ๓ต„จ๓ต„จ โ‰ค ๐ถ (๐œ2โˆ’๐›ผ + โ„Ž4 ) , 0 โ‰ค ๐‘– โ‰ค ๐‘€, 1 โ‰ค ๐‘˜ โ‰ค ๐‘. (21) ๓ต„จ๓ต„จ ๐‘– ๓ต„จ๓ต„จ ๐‘… Discretizing the initial and boundary conditions of (4) and (5), we obtain ๐‘ˆ๐‘–๐‘˜ = ๐œ™ (๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) , 0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค 0,

From Taylor expansion, we have

2๐œ•

โ„Ž4 ๐œ•6 ๐‘ข ๐‘˜โˆ’๐‘› (๐œƒ , ๐‘ก๐‘˜โˆ’๐‘› ) . + 240 ๐œ•๐‘ฅ6 ๐‘–

(12)

0 โ‰ค ๐‘– โ‰ค ๐‘€, 0 โ‰ค ๐‘˜ โ‰ค ๐‘.

โ„Ž4 ๐œ•6 ๐‘ข ๐‘˜ (๐œƒ , ๐‘ก ) 240 ๐œ•๐‘ฅ6 ๐‘– ๐‘˜

+

๐‘…๐‘–๐‘˜ ,

(19)

It is easy to obtain the following lemma. Lemma 4 (see [12]). โˆ€๐‘ข โˆˆ Vโ„Ž,0 , one has โ€–A๐‘ขโ€–2 โ‰ค โ€–๐‘ขโ€–2 .

(28)

4

Journal of Function Spaces

Lemma 5 (see [15]). โˆ€๐‘ข โˆˆ Vโ„Ž,0 , one has

where ๐œ“(๐‘ฅ, ๐‘ก) is the perturbation caused by ๐œ™(๐‘ฅ, ๐‘ก). The following difference scheme can be obtained for solving (33):

๓ต„จ๓ต„จ ๓ต„จ๓ต„จ ๓ต„จ๐›ฟ๐‘ฅ ๐‘ข๓ต„จ โ€–๐‘ขโ€–โˆž โ‰ค ๓ต„จ ๓ต„จ1 , 2 ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ ๓ต„จ๐›ฟ๐‘ฅ ๐‘ข๓ต„จ โ€–๐‘ขโ€– โ‰ค ๓ต„จ ๓ต„จ1 . โˆš6

A๐›ฟ๐‘ก๐›ผ V๐‘–๐‘˜ = ๐ท (๐›ฟ๐‘ฅ2 V๐‘–๐‘˜ + ๐›ฟ๐‘ฅ2 V๐‘–๐‘˜โˆ’๐‘› ) + A๐‘“ (2V๐‘–๐‘˜โˆ’1 โˆ’ V๐‘–๐‘˜โˆ’2 , V๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) ,

(29)

V๐‘–๐‘˜ = ๐œ™ (๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) + ๐œ“๐‘–๐‘˜ ,

The following lemma will be used in the proof of the stability and convergence analysis. ๐‘˜

Lemma 6 (see [15]). Assume that {๐น | ๐‘˜ โ‰ฅ 0} is a nonnegative sequence and satisfies

0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค 0,

(34) (35)

V0๐‘˜ = ๐›พ (๐‘ก๐‘˜ ) , ๐‘˜ = ๐›ฝ (๐‘ก๐‘˜ ) , V๐‘€

(36) 1 โ‰ค ๐‘˜ โ‰ค ๐‘.

Denote ๐‘˜

๐น๐‘˜+1 โ‰ค ๐ด + ๐ต๐œโˆ‘๐น๐‘™ ,

๐‘˜ = 0, 1, . . . ;

๐‘–=1

then, ๐น๐‘˜+1 โ‰ค ๐ด exp (๐ต๐‘˜๐œ) , ๐‘˜ = 0, 1, 2, . . . ,

๐œ”๐‘–๐‘˜ = V๐‘–๐‘˜ โˆ’ ๐‘ข๐‘–๐‘˜ ,

(30)

(31)

0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค ๐‘.

(37)

Definition 8. Assume that ๐‘ข๐‘–๐‘˜ satisfies (24)โ€“(26) and V๐‘–๐‘˜ satisfies (34)โ€“(36); then, a numerical scheme for (3)โ€“(5) is stable if one has ๓ต„ฉ๓ต„ฉ ๐‘˜ ๓ต„ฉ๓ต„ฉ ๓ต„จ ๓ต„จ ๓ต„ฉ๓ต„ฉ๐œ” ๓ต„ฉ๓ต„ฉ โ‰ค ๐ถ max ๓ต„จ๓ต„จ๓ต„จ๐œ“๐‘— ๓ต„จ๓ต„จ๓ต„จ , (38) ๓ต„ฉ ๓ต„ฉโˆž โˆ’๐‘›โ‰ค๐‘—โ‰ค0 ๓ต„จ ๓ต„จ1 where ๐ถ is a bounded constant independent of โ„Ž and ๐œ.

where ๐ด and ๐ต are nonnegative constants. Theorem 7. The difference scheme (24)โ€“(26) has a unique solution. ๐‘˜ Proof. Denote ๐‘ข๐‘˜ = [๐‘ข1๐‘˜ , ๐‘ข2๐‘˜ , . . . , ๐‘ข๐‘€โˆ’1 ]๐‘‡ ; the difference scheme (24)โ€“(26) is a linear tridiagonal system M๐‘ข๐‘˜ = ๐‘‘, ๐‘˜ and is where ๐‘‘ only depends on ๐‘ข๐‘— (๐‘— โ‰ค ๐‘˜ โˆ’ 1), ๐‘ข0๐‘˜ , and ๐‘ข๐‘€ ๐‘˜ independent of ๐‘ข .

1 ๐ท 10 2๐ท 1 ๐ท M = tridiag ( , โˆ’ , + โˆ’ ), 12๐œ† โ„Ž2 12๐œ† โ„Ž2 12๐œ† โ„Ž2

(32)

where ๐œ† = ๐œ๐›ผ ฮ“(2 โˆ’ ๐›ผ). We can see that M is a strictly diagonally dominant coefficient matrix. Thus, scheme (24)โ€“(26) has a unique solution.

Theorem 9. Assume ๐‘ข(๐‘ฅ, ๐‘ก) โˆˆ ๐ถ6,2 ([0, 1] ร— (โˆ’๐‘ , ๐‘‡]) is the solution of (3)โ€“(5); the difference scheme (24)โ€“(26) is stable with respect to the initial perturbation of ๐œ“๐‘–๐‘˜ ; that is, โ€–๐œ”๐‘˜ โ€–โˆž โ‰ค ฬ‚ maxโˆ’๐‘›โ‰ค๐‘—โ‰ค0 |๐œ“๐‘— |1 , where ๐ถ ฬ‚ is a positive constant independent ๐ถ of โ„Ž and ๐œ. Proof. Subtracting (34)โ€“(36) from (24)โ€“(26), respectively, we can obtain the following equations: A๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ = ๐ท (๐›ฟ๐‘ฅ2 ๐œ”๐‘–๐‘˜ + ๐›ฟ๐‘ฅ2 ๐œ”๐‘–๐‘˜โˆ’๐‘› ) + A๐‘ž๐‘–๐‘˜ ,

(39)

๐œ”๐‘–๐‘˜ = ๐œ“๐‘–๐‘˜ , 0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค 0,

(40)

๐œ”0๐‘˜ = 0, ๐‘˜ = 0, ๐œ”๐‘€

(41) 1 โ‰ค ๐‘˜ โ‰ค ๐‘,

To discuss the stability of the difference scheme (24)โ€“(26), we consider the following problem: A

๐œ•๐›ผ V ๐œ•2 V (๐‘ฅ, ๐‘ก) ๐œ•2 V (๐‘ฅ, ๐‘ก โˆ’ ๐‘ ) = ๐ท ( + ) ๐œ•๐‘ก๐›ผ ๐œ•๐‘ฅ2 ๐œ•๐‘ฅ2

where ๐‘ž๐‘–๐‘˜ = ๐‘“(2V๐‘–๐‘˜โˆ’1 โˆ’ V๐‘–๐‘˜โˆ’2 , V๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) โˆ’ ๐‘“(2๐‘ข๐‘–๐‘˜โˆ’1 โˆ’ ๐‘ข๐‘–๐‘˜โˆ’2 , ๐‘ข๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ). Multiplying โ„Ž๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ on both sides of (39) and summing up for ๐‘– from 1 to ๐‘€ โˆ’ 1, we obtain ๐‘€โˆ’1

โ„Ž โˆ‘ (A๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ ) (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ )

+ A๐‘“ (V (๐‘ฅ, ๐‘ก) , V (๐‘ฅ, ๐‘ก โˆ’ ๐‘ ) , ๐‘ฅ, ๐‘ก) ,

๐‘–=1

(๐‘ฅ, ๐‘ก) โˆˆ (0, 1) ร— (0, ๐‘‡] , V (๐‘ฅ, ๐‘ก) = ๐œ™ (๐‘ฅ, ๐‘ก) + ๐œ“ (๐‘ฅ, ๐‘ก) , ๐‘ฅ โˆˆ [0, 1] , ๐‘ก โˆˆ [โˆ’๐‘ , 0] , V (0, ๐‘ก) = ๐›พ (๐‘ก) ,

๐‘€โˆ’1

(33)

= ๐ทโ„Ž โˆ‘ (๐›ฟ๐‘ฅ2 ๐œ”๐‘–๐‘˜ + ๐›ฟ๐‘ฅ2 ๐œ”๐‘–๐‘˜โˆ’๐‘› ) (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ ) ๐‘–=1

๐‘€โˆ’1

+ โ„Ž โˆ‘ (A๐‘ž๐‘–๐‘˜ ) (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ ) ,

๐‘ข (1, ๐‘ก) = ๐›ฝ (๐‘ก) ,

๐‘–=1

๐‘ก โˆˆ (0, ๐‘‡] ,

1 โ‰ค ๐‘– โ‰ค ๐‘€ โˆ’ 1, 0 โ‰ค ๐‘˜ โ‰ค ๐‘ โˆ’ 1.

(42)

Journal of Function Spaces

5 ๐‘€โˆ’1

Then, each term of (42) will be estimated. From the discrete Green formula and inequality โˆ’(๐‘Ž+๐‘)2 โ‰ฅ โˆ’2(๐‘Ž2 +๐‘2 ), we have

๐‘˜โˆ’๐‘› ๐‘˜ ) [๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 โ‹… โ„Ž โˆ‘ (๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 ๐‘–=0 [

๐‘€โˆ’1

โ„Ž โˆ‘ (A๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ ) (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ )

๐‘˜โˆ’1

๓ต„ฉ ๓ต„ฉ2 โ„Ž2 ๐‘€โˆ’1 = ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ๐›ฟ๐‘ก๐›ผ ๐œ”๐‘˜ ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ + โ„Ž โˆ‘ (๐›ฟ๐‘ฅ2 (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ )) (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ ) 12 ๐‘–=1

(43)

โ‰ค

๓ต„ฉ2 โ„Ž2 ๓ต„ฉ๓ต„ฉ ๐›ผ ๐‘˜ ๓ต„ฉ๓ต„ฉ2 2 ๓ต„ฉ๓ต„ฉ ๐›ผ ๐‘˜ ๓ต„ฉ๓ต„ฉ2 ๓ต„ฉ ๓ต„ฉ๓ต„ฉ๐›ฟ ๐›ฟ ๐œ” ๓ต„ฉ๓ต„ฉ โ‰ฅ ๓ต„ฉ๓ต„ฉ๐›ฟ๐‘ก ๐œ” ๓ต„ฉ๓ต„ฉ . = ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ๐›ฟ๐‘ก๐›ผ ๐œ”๐‘˜ ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ โˆ’ ๓ต„ฉ 12 ๓ต„ฉ ๐‘ฅ ๐‘ก ๓ต„ฉ 3๓ต„ฉ

๓ต„จ2 ๓ต„จ ๓ต„จ2 ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ + ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘— ๓ต„จ๓ต„จ๓ต„จ ๓ต„จ ๓ต„จ1 ๓ต„จ ๓ต„จ 1 + โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) 2 ๐‘—=1

From the discrete Green formula and inequality โˆ’๐‘Ž๐‘ โ‰ฅ โˆ’(๐‘Ž + ๐‘2 )/2, we have ๐ทโ„Ž โˆ‘ ๐‘–=1

๓ต„จ๓ต„จ ๐‘˜ ๓ต„จ๓ต„จ2 ๓ต„จ๓ต„จ๓ต„จ๐›ฟ ๐œ”๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ2 + ๓ต„จ๓ต„จ๓ต„จ๐›ฟ ๐œ”0 ๓ต„จ๓ต„จ๓ต„จ2 } ๓ต„จ ๐‘ฅ ๓ต„จ๓ต„จ1 ๓ต„จ๓ต„จ ๐‘ฅ ๓ต„จ๓ต„จ1 ๐ท { ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ” ๓ต„จ๓ต„จ๓ต„จ1 + ๐‘‘๐‘˜โˆ’1 ๓ต„จ = } ๐œ†{ 4 2 } {

(๐›ฟ๐‘ฅ2 ๐œ”๐‘–๐‘˜ ) (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ ) ๐‘€โˆ’1

๐‘˜ ๐‘˜ = โˆ’๐ทโ„Ž โˆ‘ (๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 ) (๐›ฟ๐‘ก๐›ผ ๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 )=โˆ’ ๐‘–=0

๐ท ๐œ†

+

๐‘€โˆ’1

๐‘˜ ๐‘˜ โ‹… โ„Ž โˆ‘ (๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 ) [๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 ๐‘–=0 [ ๐‘˜โˆ’1

๐ท {๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ { ๓ต„จ ๓ต„จ1 ๐œ† {

From the Cauchy-Schwarz inequality, Lemma 4, and (7), we have ๐‘€โˆ’1

โ„Ž โˆ‘ (A๐‘ž๐‘–๐‘˜ ) (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ ) ๐‘–=1

โˆ’ โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) โŸจ๐›ฟ๐‘ฅ ๐œ”๐‘˜ , ๐›ฟ๐‘ฅ ๐œ”๐‘— โŸฉ

โ‰ค

๐‘—=1

0

โ‰ค

๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ ๓ต„จ2 ๓ต„จ๓ต„จ ๐‘˜ ๓ต„จ๓ต„จ2 ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ + ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”0 ๓ต„จ๓ต„จ๓ต„จ } ๐ท { ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ” ๓ต„จ๓ต„จ๓ต„จ1 ๓ต„จ ๓ต„จ ๓ต„จ ๓ต„จ 1 1 โˆ’ ๐‘‘๐‘˜โˆ’1 }=โˆ’๐œ† { 2 2 } { 2 ๓ต„จ ๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ ๓ต„จ๓ต„จ๐›ฟ ๐œ”๐‘— ๓ต„จ๓ต„จ ๐‘˜โˆ’1 ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”0 ๓ต„จ๓ต„จ๓ต„จ } ๓ต„จ๓ต„จ ๐‘ฅ ๓ต„จ๓ต„จ1 ๓ต„จ ๓ต„จ1 , โˆ’ โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) โˆ’ ๐‘‘๐‘˜โˆ’1 } 2 2 ๐‘—=1 }

โ‰ค

๐‘€โˆ’1 2 1 ๓ต„ฉ๓ต„ฉ ๐›ผ ๐‘˜ ๓ต„ฉ๓ต„ฉ2 ๓ต„ฉ๓ต„ฉ๐›ฟ๐‘ก ๐œ” ๓ต„ฉ๓ต„ฉ + ๐œ€๐‘02 โ„Ž โˆ‘ (2๐œ”๐‘–๐‘˜โˆ’1 โˆ’ ๐œ”๐‘–๐‘˜โˆ’2 ) ๓ต„ฉ ๓ต„ฉ 2๐œ€ ๐‘–=1 ๐‘€โˆ’1

+ ๐œ€๐‘12 โ„Ž โˆ‘ (๐œ”๐‘–๐‘˜โˆ’๐‘› )

๐‘€โˆ’1

2

๐‘–=1

๐‘–=1

๐‘–=0

1 ๓ต„ฉ๓ต„ฉ ๐›ผ ๐‘˜ ๓ต„ฉ๓ต„ฉ2 ๓ต„ฉ๓ต„ฉ๐›ฟ ๐œ” ๓ต„ฉ๓ต„ฉ 2๐œ€ ๓ต„ฉ ๐‘ก ๓ต„ฉ ๐œ€ ๐‘€โˆ’1 ๓ต„จ ๓ต„จ ๓ต„จ ๓ต„จ 2 + โ„Ž โˆ‘ (๐‘0 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ2๐œ”๐‘–๐‘˜โˆ’1 โˆ’ ๐œ”๐‘–๐‘˜โˆ’2 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ + ๐‘1 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐œ”๐‘–๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ๓ต„จ) 2 ๐‘–=1

๐ทโ„Ž โˆ‘ (๐›ฟ๐‘ฅ2 ๐œ”๐‘–๐‘˜โˆ’๐‘› ) (๐›ฟ๐‘ก๐›ผ ๐œ”๐‘–๐‘˜ ) ๐‘˜โˆ’๐‘› ๐‘˜ = โˆ’๐ทโ„Ž โˆ‘ (๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 ) (๐›ฟ๐‘ก๐›ผ ๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 )=โˆ’

1 ๓ต„ฉ๓ต„ฉ ๐›ผ ๐‘˜ ๓ต„ฉ๓ต„ฉ2 ๓ต„ฉ๓ต„ฉ๐›ฟ ๐œ” ๓ต„ฉ๓ต„ฉ 2๐œ€ ๓ต„ฉ ๐‘ก ๓ต„ฉ ๐œ€ ๐‘€โˆ’1 ๓ต„จ ๓ต„จ ๓ต„จ ๓ต„จ 2 + โ„Ž โˆ‘ (A [๐‘0 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ2๐œ”๐‘–๐‘˜โˆ’1 โˆ’ ๐œ”๐‘–๐‘˜โˆ’2 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ + ๐‘1 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐œ”๐‘–๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ๓ต„จ]) 2 ๐‘–=1

} ๐ท {๓ต„จ ๓ต„จ2 โˆ’ ๐‘‘๐‘˜โˆ’1 โŸจ๐›ฟ๐‘ฅ ๐œ” , ๐›ฟ๐‘ฅ ๐œ” โŸฉ} โ‰ค โˆ’ {๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 ๐œ† } { ๓ต„จ๓ต„จ๓ต„จ๐›ฟ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ2 + ๓ต„จ๓ต„จ๓ต„จ๐›ฟ ๐œ”๐‘— ๓ต„จ๓ต„จ๓ต„จ2 ๐‘˜โˆ’1 ๓ต„จ ๐‘ฅ ๓ต„จ๓ต„จ1 ๓ต„จ๓ต„จ ๐‘ฅ ๓ต„จ๓ต„จ1 โˆ’ โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) ๓ต„จ 2 ๐‘—=1

๐‘€โˆ’1

2

๓ต„จ2 ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘— ๓ต„จ๓ต„จ๓ต„จ ๓ต„จ1 ๓ต„จ + โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) 2 ๐‘—=1 ๐‘˜โˆ’1

(44)

๐‘˜โˆ’1

๐‘˜

๓ต„จ2 ๓ต„จ 3 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1

๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”0 ๓ต„จ๓ต„จ๓ต„จ } ๓ต„จ ๓ต„จ1 . + ๐‘‘๐‘˜โˆ’1 2 } }

๐‘—

0 ] โˆ’ โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) ๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 โˆ’ ๐‘‘๐‘˜โˆ’1 ๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 ๐‘—=1 ]

=โˆ’

๐ท { 1 ๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ ๓ต„จ2 ๓ต„จ๓ต„จ๐›ฟ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ + ๓ต„จ๓ต„จ๓ต„จ๐›ฟ ๐œ”๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ ๓ต„จ1 ๐œ† { 4 ๓ต„จ ๐‘ฅ ๓ต„จ1 ๓ต„จ ๐‘ฅ { ๐‘˜โˆ’1

2

๐‘€โˆ’1

๐‘—

0 ] โˆ’ โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) ๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 โˆ’ ๐‘‘๐‘˜โˆ’1 ๐›ฟ๐‘ฅ ๐œ”๐‘–+1/2 ๐‘—=1 ]

๐‘–=1

โ‰ค ๐ท ๐œ†

1 ๓ต„ฉ๓ต„ฉ ๐›ผ ๐‘˜ ๓ต„ฉ๓ต„ฉ2 ๓ต„ฉ ๓ต„ฉ2 ๓ต„ฉ ๓ต„ฉ2 ๓ต„ฉ๓ต„ฉ๐›ฟ ๐œ” ๓ต„ฉ๓ต„ฉ + 8๐œ€๐‘02 (๓ต„ฉ๓ต„ฉ๓ต„ฉ๐œ”๐‘˜โˆ’1 ๓ต„ฉ๓ต„ฉ๓ต„ฉ + ๓ต„ฉ๓ต„ฉ๓ต„ฉ๐œ”๐‘˜โˆ’2 ๓ต„ฉ๓ต„ฉ๓ต„ฉ ) ๓ต„ฉ ๓ต„ฉ ๓ต„ฉ ๓ต„ฉ 2๐œ€ ๓ต„ฉ ๐‘ก ๓ต„ฉ ๓ต„ฉ ๓ต„ฉ2 + ๐œ€๐‘12 ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ๐œ”๐‘˜โˆ’๐‘› ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ .

(45)

6

Journal of Function Spaces

Substituting (43)โ€“(45) into (42) and taking ๐œ€ = 3/4 in (45), we obtain

From Lemma 5, we have ๓ต„ฉ๓ต„ฉ ๐‘˜ ๓ต„ฉ๓ต„ฉ ฬ‚ max ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐œ“๐‘— ๓ต„จ๓ต„จ๓ต„จ๓ต„จ . ๓ต„ฉ๓ต„ฉ๐œ” ๓ต„ฉ๓ต„ฉ โ‰ค ๐ถ ๓ต„ฉ ๓ต„ฉโˆž โˆ’๐‘›โ‰ค๐‘—โ‰ค0 ๓ต„จ ๓ต„จ1

๐ท ๓ต„จ๓ต„จ ๓ต„จ2 3๐ท ๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ โ‰ค ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ ๓ต„จ ๓ต„จ ๓ต„จ ๓ต„จ1 1 4๐œ† 2๐œ† +

๐ท ๐‘˜โˆ’1 ๓ต„จ2 ๓ต„จ โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘— ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 โˆ‘ (๐‘‘ ๐œ† ๐‘—=1 ๐‘˜โˆ’๐‘—โˆ’1

The proof is completed.

๐ท ๓ต„จ ๓ต„จ2 ๐‘‘ ๓ต„จ๓ต„จ๓ต„จ๐›ฟ ๐œ”0 ๓ต„จ๓ต„จ๓ต„จ ๐œ† ๐‘˜โˆ’1 ๓ต„จ ๐‘ฅ ๓ต„จ1 ๓ต„ฉ2 ๓ต„ฉ ๓ต„ฉ2 ๓ต„ฉ + 6๐‘02 (๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ๐œ”๐‘˜โˆ’1 ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ + ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ๐œ”๐‘˜โˆ’2 ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ )

Denote ๐‘’๐‘–๐‘˜ = ๐‘ˆ๐‘–๐‘˜ โˆ’ ๐‘ข๐‘–๐‘˜ , 0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค ๐‘; by subtracting (24)โ€“(26) from (19), (22), and (23), respectively, the following error equations can be obtained:

+

3 ๓ต„ฉ ๓ต„ฉ2 + ๐‘12 ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ๐œ”๐‘˜โˆ’๐‘› ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ 4 โ‰ค

(46)

๐ท ๓ต„จ2 ๓ต„จ โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘— ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 โˆ‘ (๐‘‘ ๐œ† ๐‘—=1 ๐‘˜โˆ’๐‘—โˆ’1

๐‘’0๐‘˜ = 0,

1 โ‰ค ๐‘˜ โ‰ค ๐‘, where ๐‘๐‘–๐‘˜ = ๐‘“(2๐‘ˆ๐‘–๐‘˜โˆ’1 โˆ’ ๐‘ˆ๐‘–๐‘˜โˆ’2 , ๐‘ˆ๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ) โˆ’ ๐‘“(2๐‘ข๐‘–๐‘˜โˆ’1 โˆ’ ๐‘ข๐‘–๐‘˜โˆ’2 , ๐‘ข๐‘–๐‘˜โˆ’๐‘› , ๐‘ฅ๐‘– , ๐‘ก๐‘˜ ). Similar to the proof of Theorem 9, the following convergence result can be obtained.

๐‘˜โˆ’๐‘› ๓ต„จ2

๓ต„จ๓ต„จ , ๓ต„จ๓ต„จ1

where Lemma 5 has been used. Multiplying (46) by 4๐œ†/๐ท, we have

Theorem 10. Assume ๐‘ข(๐‘ฅ, ๐‘ก) โˆˆ ๐ถ6,2 ([0, 1] ร— (โˆ’๐‘ , ๐‘‡]) is the solution of (3)โ€“(5) and {๐‘ข๐‘–๐‘˜ | 0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค ๐‘} is the solution of (24)โ€“(26). Then, one has

๐‘˜โˆ’1 ๓ต„จ ๓ต„จ ๓ต„จ2 ๓ต„จ2 ๓ต„จ2 ๓ต„จ๓ต„จ ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ โ‰ค 4 โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘— ๓ต„จ๓ต„จ๓ต„จ + 4 ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”0 ๓ต„จ๓ต„จ๓ต„จ ๓ต„จ ๓ต„จ ๓ต„จ1 ๓ต„จ1 ๓ต„จ1 ๓ต„จ ๐‘—=1

+

4๐œ† 2 ๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ ๓ต„จ2 ๐‘0 (๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’1 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 + ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’2 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 ) ๐ท

+ (6 +

(52)

๐‘˜ = 0, ๐‘’๐‘€

๐ท ๓ต„จ ๓ต„จ2 ๐‘‘ ๓ต„จ๓ต„จ๓ต„จ๐›ฟ ๐œ”0 ๓ต„จ๓ต„จ๓ต„จ ๐œ† ๐‘˜โˆ’1 ๓ต„จ ๐‘ฅ ๓ต„จ1 ๓ต„จ2 ๓ต„จ ๓ต„จ2 ๓ต„จ + ๐‘02 (๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’1 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 + ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’2 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 ) +

3๐ท 1 2 ๓ต„จ๓ต„จ + ๐‘ ) ๓ต„จ๓ต„จ๐›ฟ ๐œ” 2๐œ† 8 1 ๓ต„จ ๐‘ฅ

A๐›ฟ๐‘ก๐›ผ ๐‘’๐‘–๐‘˜ = ๐ท (๐›ฟ๐‘ฅ2 ๐‘’๐‘–๐‘˜ + ๐›ฟ๐‘ฅ2 ๐‘’๐‘–๐‘˜โˆ’๐‘› ) + A๐‘๐‘–๐‘˜ + ๐‘…๐‘–๐‘˜ , ๐‘’๐‘–๐‘˜ = 0, 0 โ‰ค ๐‘– โ‰ค ๐‘€, โˆ’๐‘› โ‰ค ๐‘˜ โ‰ค 0,

๐‘˜โˆ’1

+(

(51)

(47)

๓ต„ฉ๓ต„ฉ ๐‘˜ ๓ต„ฉ๓ต„ฉ ๓ต„ฉ๓ต„ฉ๐‘’ ๓ต„ฉ๓ต„ฉ โ‰ค ๐ถ (๐œ2โˆ’๐›ผ + โ„Ž4 ) , ๓ต„ฉ ๓ต„ฉโˆž

1 โ‰ค ๐‘˜ โ‰ค ๐‘,

(53)

where ๐ถ is a positive constant independent of โ„Ž and ๐œ.

๐œ† 2 ๓ต„จ๓ต„จ ๓ต„จ2 ๐‘ ) ๓ต„จ๓ต„จ๐›ฟ ๐œ”๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 . 2๐ท 1 ๓ต„จ ๐‘ฅ

4. Numerical Test

Denote ๐ถ๐‘˜ = max {

4ฮ“ (2 โˆ’ ๐›ผ) 2 ฮ“ (2 โˆ’ ๐›ผ) 2 ๐‘0 , ๐‘1 + 6} > 0. ๐ท 2๐ท

(48)

Noticing ๐œ† = ๐œ๐›ผ ฮ“(2 โˆ’ ๐›ผ) for 0 < ๐œ < 1, we have ๐œ๐›ผ < 1. Then, from (47), we obtain ๐‘˜โˆ’1

๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ2 ๓ต„จ2 ๓ต„จ ๓ต„จ ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ โ‰ค 4 โˆ‘ (๐‘‘๐‘˜โˆ’๐‘—โˆ’1 โˆ’ ๐‘‘๐‘˜โˆ’๐‘— ) ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘— ๓ต„จ๓ต„จ๓ต„จ + 4 ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”0 ๓ต„จ๓ต„จ๓ต„จ ๓ต„จ ๓ต„จ1 ๓ต„จ1 ๓ต„จ1 ๓ต„จ ๓ต„จ ๐‘—=1

(49)

๓ต„จ2 ๓ต„จ ๓ต„จ2 ๓ต„จ ๓ต„จ2 ๓ต„จ + ๐ถ๐‘˜ (๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’1 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 + ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’2 ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 + ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜โˆ’๐‘› ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 ) . From Lemmas 2 and 6, we have ๓ต„จ๓ต„จ ๓ต„จ2 ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘˜ ๓ต„จ๓ต„จ๓ต„จ ๓ต„จ ๓ต„จ1 ๓ต„จ ๓ต„จ2 โ‰ค (3๐ถ๐‘˜ + 4) exp (3๐ถ๐‘˜ + 4 (1 โˆ’ ๐‘‘๐‘˜โˆ’1 )) max ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ”๐‘— ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 (50) โˆ’๐‘›โ‰ค๐‘—โ‰ค0 ๓ต„จ ๓ต„จ2 โ‰ค (3๐ถ๐‘˜ + 4) exp (3๐ถ๐‘˜ + 4) max ๓ต„จ๓ต„จ๓ต„จ๓ต„จ๐›ฟ๐‘ฅ ๐œ“๐‘— ๓ต„จ๓ต„จ๓ต„จ๓ต„จ1 . โˆ’๐‘›โ‰ค๐‘—โ‰ค0

In this section, a numerical test is used to validate the performance of scheme (24)โ€“(26). Denote the maximum error at all grid points as ๓ต„ฉ ๓ต„ฉ ๐‘’ (โ„Ž, ๐œ) = max ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉ๐‘ˆ๐‘˜ โˆ’ ๐‘ข๐‘˜ ๓ต„ฉ๓ต„ฉ๓ต„ฉ๓ต„ฉโˆž ; 1โ‰ค๐‘˜โ‰ค๐‘

(54)

the convergence order in time and space is defined, respectively, as Rate๐œ =

log (๐‘’ (โ„Ž, ๐œ1 ) /๐‘’ (โ„Ž, ๐œ2 )) , log (๐œ1 /๐œ2 )

log (๐‘’ (โ„Ž1 , ๐œ) /๐‘’ (โ„Ž2 , ๐œ)) . Rateโ„Ž = log (โ„Ž1 /โ„Ž2 )

(55)

For Rate๐œ , we require โ„Ž to be fixed and small enough, while for Rateโ„Ž , ๐œ should be fixed and small enough.

Journal of Function Spaces

7 ๎‹ฃ = 0.2

๎‹ฃ = 0.4 ร—10โˆ’6

ร—10โˆ’7 5

2.5

4

2

3

1.5

2

1

1

0.5

0 1 t

1

0.5

0.5 0

0

0 1 t

x

1

0.5

0.5 0

0

๎‹ฃ = 0.6

x

๎‹ฃ = 0.8

ร—10โˆ’5 1

ร—10โˆ’5 4

0.8

3

0.6

2

0.4 0.2

1

0 1

0 1

t

1

0.5

0.5 0

0

t

x

1

0.5

0.5 0

0

x

Figure 1: Error planes for ๐œ = โ„Ž = 1/100 for different ๐›ผ in Example 1.

โˆ’ ๐‘’๐‘ฅ (2 โˆ’ 8๐‘ฅ + ๐‘ฅ2 + 6๐‘ฅ3 + ๐‘ฅ4 )

Example 1. Consider the following problem:

โ‹… (๐‘ก2+๐›ผ + (๐‘ก โˆ’ 0.1)2+๐›ผ ) .

๐œ•๐›ผ ๐‘ข ๐œ•2 ๐‘ข (๐‘ฅ, ๐‘ก) ๐œ•2 ๐‘ข (๐‘ฅ, ๐‘ก โˆ’ 0.1) = ( + ) โˆ’ ๐‘ข (๐‘ฅ, ๐‘ก)2 ๐œ•๐‘ก๐›ผ ๐œ•๐‘ฅ2 ๐œ•๐‘ฅ2

(57)

+ ๐‘ข (๐‘ฅ, ๐‘ก โˆ’ 0.1) + ๐บ (๐‘ฅ, ๐‘ก) , (๐‘ฅ, ๐‘ก) โˆˆ (0, 1) ร— (0, 1] , ๐‘ข (๐‘ฅ, ๐‘ก) = ๐‘’๐‘ฅ ๐‘ฅ2 (1 โˆ’ ๐‘ฅ)2 ๐‘ก2+๐›ผ ,

(56) ๐‘ฅ โˆˆ [0, 1] , ๐‘ก โˆˆ [โˆ’0.1, 0] ,

๐‘ข (0, ๐‘ก) = 0, ๐‘ข (1, ๐‘ก) = 0, ๐‘ก โˆˆ (0, 1] ;

the exact solution of (56) is ๐‘ข(๐‘ฅ, ๐‘ก) = ๐‘’๐‘ฅ ๐‘ฅ2 (1 โˆ’ ๐‘ฅ)2 ๐‘ก2+๐›ผ , and ๐บ (๐‘ฅ, ๐‘ก) = ๐‘’2๐‘ฅ ๐‘ฅ4 (1 โˆ’ ๐‘ฅ)4 ๐‘ก4+2๐›ผ โˆ’ ๐‘’๐‘ฅ ๐‘ฅ2 (1 โˆ’ ๐‘ฅ)2 ฮ“ (3 + ๐›ผ) 2 ๐‘ฅ 2 โ‹… (๐‘ก โˆ’ 0.1)2+๐›ผ + ๐‘ก ๐‘’ ๐‘ฅ (1 โˆ’ ๐‘ฅ)2 2

From Table 1, we can see the maximum errors between the numerical solution and the exact solution in the temporal directions for ๐›ผ = 0.3, 0.5, 0.8, respectively, where the spatial step is fixed to be โ„Ž = 1/400. The results show that the temporal convergence order matches well the theoretical convergence order of 2 โˆ’ ๐›ผ. Table 2 shows the maximum errors in the spatial directions for ๐›ผ = 0.2 when the temporal step is fixed at ๐œ = 1/2000. From the results, we can see that the spatial convergence order is 4, which coincides with the theoretical result. Figure 1 gives the error plane for ๐›ผ = 0.2, 0.4, 0.6, 0.8, respectively. From this figure, we can see that the error becomes larger when a larger ๐›ผ is taken.

5. Conclusion This paper presents a compact finite difference scheme for solving the fractional neutral parabolic differential equation with proportional delay. The unconditional stability and the

8

Journal of Function Spaces Table 1: Maximum errors and convergence order in temporal direction with โ„Ž = 1/400 and ๐›ผ = 0.3, 0.5, 0.8 for Example 1.

๐œ 1/100 1/200 1/300 1/400

๐›ผ = 0.3 ๐‘’(โ„Ž, ๐œ) 1.083๐‘’ โˆ’ 006 3.234๐‘’ โˆ’ 007 1.598๐‘’ โˆ’ 007 69.694๐‘’ โˆ’ 008

Rate๐œ โˆ— 1.744 1.739 1.737

๐›ผ = 0.5 ๐‘’(โ„Ž, ๐œ) 4.502๐‘’ โˆ’ 006 1.571๐‘’ โˆ’ 006 8.501๐‘’ โˆ’ 007 5.501๐‘’ โˆ’ 007

Table 2: Maximum errors and convergence order in spatial direction with ๐œ = 1/2000 and ๐›ผ = 0.2 for Example 1. โ„Ž 1/8 1/12 1/16 1/20

๐‘’(โ„Ž, ๐œ) 7.949๐‘’ โˆ’ 005 1.578๐‘’ โˆ’ 005 4.976๐‘’ โˆ’ 006 2.045๐‘’ โˆ’ 006

Rateโ„Ž โˆ— 3.988 4.011 3.985

global convergence of the scheme in the maximum norm are proved. The convergence order of the considered scheme is ๐‘‚(๐œ2โˆ’๐›ผ + โ„Ž4 ). A numerical experiment is presented to support the theoretical results and validate the efficiency of the difference scheme.

[5]

[6]

[7]

[8]

Conflicts of Interest The authors declare that there are no conflicts of interest regarding the publication of this paper.

Acknowledgments This research work is supported by the National Natural Science Foundation of China (11401591) and the Humanities and Social Science Foundation of the Ministry of Education of China (17YJC630236). The first author acknowledges the Project of the Scientific Research Foundation for the Returned Overseas Chinese Scholars, State Education Ministry (2013693).

References [1] A. V. Rezounenko and J. Wu, โ€œA non-local PDE model for population dynamics with state-selective delay: local theory and global attractors,โ€ Journal of Computational and Applied Mathematics, vol. 190, no. 1-2, pp. 99โ€“113, 2006. [2] B. Zubik-Kowal, โ€œSolutions for the cell cycle in cell lines derived from human tumors,โ€ Computational and Mathematical Methods in Medicine. An Interdisciplinary Journal of Mathematical, Theoretical and Clinical Aspects of Medicine, vol. 7, no. 4, pp. 215โ€“ 228, 2006. [3] D. Li, C. Zhang, and W. Wang, โ€œLong time behavior of nonFickian delay reaction-diffusion equations,โ€ Nonlinear Analysis: Real World Applications, vol. 13, no. 3, pp. 1401โ€“1415, 2012. [4] Z.-X. Sun and Z. B. Zhang, โ€œA linearized compact difference scheme for a class of nonlinear delay partial differential equations,โ€ Applied Mathematical Modelling: Simulation and

[9]

[10]

[11]

[12]

[13]

[14]

[15]

Rate๐œ โˆ— 1.519 1.515 1.513

๐›ผ = 0.8 ๐‘’(โ„Ž, ๐œ) 3.800๐‘’ โˆ’ 005 1.651๐‘’ โˆ’ 005 1.014๐‘’ โˆ’ 005 7.176๐‘’ โˆ’ 006

Rate๐œ โˆ— 1.203 1.202 1.202

Computation for Engineering and Environmental Systems, vol. 37, no. 3, pp. 742โ€“752, 2013. Q. Zhang and C. Zhang, โ€œA compact difference scheme combined with extrapolation techniques for solving a class of neutral delay parabolic differential equations,โ€ Applied Mathematics Letters, vol. 26, no. 2, pp. 306โ€“312, 2013. W. Gu, โ€œA compact difference scheme for a class of variable coefficient quasilinear parabolic equations with delay,โ€ Abstract and Applied Analysis, Article ID 810352, Art. ID 810352, 8 pages, 2014. D. Li, C. Zhang, and J. Wen, โ€œA note on compact finite difference method for reaction-diffusion equations with delay,โ€ Applied Mathematical Modelling: Simulation and Computation for Engineering and Environmental Systems, vol. 39, no. 5-6, pp. 1749โ€“1754, 2015. C. R. Jin, Z. H. Yu, and R. N. Qu, โ€œAn implicit difference scheme for solving a neutral delay parabolic differential equation,โ€ Journal of Shandong University. Natural Science. Shandong Daxue Xuebao. Lixue Ban, vol. 46, no. 8, pp. 13โ€“16, 2011. R. Metzler and J. Klafter, โ€œThe random walkโ€™s guide to anomalous diffusion: a fractional dynamics approach,โ€ Physics Reports, vol. 339, pp. 1โ€“77, 2000. S. Chen, F. Liu, I. Turner, and V. Anh, โ€œAn implicit numerical method for the two-dimensional fractional percolation equation,โ€ Applied Mathematics and Computation, vol. 219, no. 9, pp. 4322โ€“4331, 2013. F. A. Rihan, โ€œComputational methods for delay parabolic and time-fractional partial differential equations,โ€ Numerical Methods for Partial Differential Equations, vol. 26, no. 6, pp. 1556โ€“1571, 2010. Q. Zhang, M. Ran, and D. Xu, โ€œAnalysis of the compact difference scheme for the semilinear fractional partial differential equation with time delay,โ€ Applicable Analysis: An International Journal, vol. 96, no. 11, pp. 1867โ€“1884, 2017. Z. Sun and X. Wu, โ€œA fully discrete difference scheme for a diffusion-wave system,โ€ Appl. Numer. Math, vol. 56, pp. 193โ€“209, 2006. S. Chen, F. Liu, P. Zhuang, and V. Anh, โ€œFinite difference approximations for the fractional Fokker-Planck equation,โ€ Applied Mathematical Modelling: Simulation and Computation for Engineering and Environmental Systems, vol. 33, no. 1, pp. 256โ€“273, 2009. Z. Z. Sun, The numerical methods for partial equations, Science Press, Beijing, China, 2005.

Advances in

Operations Research Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Advances in

Decision Sciences Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Journal of

Applied Mathematics

Algebra

Hindawi Publishing Corporation http://www.hindawi.com

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Journal of

Probability and Statistics Volume 2014

The Scientific World Journal Hindawi Publishing Corporation http://www.hindawi.com

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

International Journal of

Differential Equations Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Volume 2014

Submit your manuscripts at https://www.hindawi.com International Journal of

Advances in

Combinatorics Hindawi Publishing Corporation http://www.hindawi.com

Mathematical Physics Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Journal of

Complex Analysis Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

International Journal of Mathematics and Mathematical Sciences

Mathematical Problems in Engineering

Journal of

Mathematics Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Volume 2014

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

#HRBQDSDฤฎ,@SGDL@SHBR

Journal of

Volume 201

Hindawi Publishing Corporation http://www.hindawi.com

Discrete Dynamics in Nature and Society

Journal of

Function Spaces Hindawi Publishing Corporation http://www.hindawi.com

Abstract and Applied Analysis

Volume 2014

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

International Journal of

Journal of

Stochastic Analysis

Optimization

Hindawi Publishing Corporation http://www.hindawi.com

Hindawi Publishing Corporation http://www.hindawi.com

Volume 2014

Volume 2014