Advances in Applied Mathematics
Vol. 10  No. 10 ( 2021 ), Article ID: 45949 , 6 pages
10.12677/AAM.2021.1010360

带Markov切换Poisson跳的非线性随机时滞微分方程解的矩有界性

林宇璇,李光洁*

广东外语外贸大学数学与统计学院,广东 广州

收稿日期:2021年9月21日;录用日期:2021年10月14日;发布日期:2021年10月22日

摘要

研究了一类带Markov切换Poisson跳的非线性随机时滞微分方程解的矩有界性。首先,证明了该方程解的存在唯一性;其次,利用随机分析和不等式技巧得到了该方程的解是矩有界的。

关键词

随机时滞微分方程,Markov切换Poisson跳,解的矩有界性

Moment Boundedness of Solutions to Nonlinear Stochastic Delay Differential Equations with Markovian Switching and Poisson Jumps

Yuxuan Lin, Guangjie Li*

School of Mathematics and Statistics, Guangdong University of Foreign Studies, Guangzhou Guangdong

Received: Sep. 21st, 2021; accepted: Oct. 14th, 2021; published: Oct. 22nd, 2021

ABSTRACT

This paper investigates the moment boundedness of the solutions to nonlinear stochastic delay differential equations with Markovian switching and Poisson jumps. It is first proved the existence and uniqueness of the solution for such an equation. By using stochastic analysis and inequality techniques, it is then obtained that the solution is moment bounded.

Keywords:Stochastic Delay Differential Equations, Markovian Switching and Poisson Jumps, Moment Boundedness of Solutions

Copyright © 2021 by author(s) and Hans Publishers Inc.

This work is licensed under the Creative Commons Attribution International License (CC BY 4.0).

http://creativecommons.org/licenses/by/4.0/

1. 引言

随机微分系统常用随机微分方程来刻画且已广泛地应用于生物学、经济、金融以及控制工程等诸多领域,见文献 [1] [2] [3] 及其中的参考文献。但实际中,许多随机微分系统的演化过程不仅和当前的状态有关而且和过去的状态也有关,对于这样的情形可用随机时滞微分方程来描述。时滞也是这些实际系统经常发生振荡、不稳定性和性能差的原因。因此,很有必要研究随机时滞微分方程(参阅文献 [4] [5] [6] [7] )。

实际中,许多系统在遭遇突变现象(如分支和内部联系紊乱、参数转移等)时结构易发生随机改变。对于这样的系统,人们常用包含连续系统状态和离散系统状态的Markov切换系统来描述 [8] [9]。据我们所知Brown运动是连续的随机过程,然而许多实际系统会遭受跳跃形式的随机突发扰动,如随机故障、地震、海啸等。在这些情形下,不能用Brown运动来刻画这些系统,因此将带跳过程引入到这些系统中来处理实际情形是合理的 [10] [11] [12]。

王等人在文献 [13] 和文献 [14] 中分别研究了带分布时滞的线性随机时滞微分方程解的矩有界性和线性随机时滞微分方程解的二阶矩有界性。文献 [15] 研究了脉冲随机时滞微分方程解的有界性。毛等人在文献 [16] 中研究了带Markov切换和跳的随机微分方程解的渐近有界性。但这些文献均未研究带Markov切换和Poisson跳的非线性随机时滞微分方程解的矩有界性。基于此,本文研究了带Markov切换和Poisson跳的非线性随机时滞微分方程解的矩有界性。

该文结构如下:第2节给出了本文所需的预备知识;第3节证明了该类方程解的存在唯一性以及解的矩有界性。

2. 预备知识

R n 表示n-维的欧式空间。若 x R n | x | 表示欧式范数。记 R = ( , + ) R + = [ 0 , + ) 。记 ( Ω , F , P ) 为一个完备的概率空间,其滤子 F t 满足通常的条件,即 { F t } t 0 是右连续的且 F 0 包含所有的零测集。令 τ > 0 C ( [ τ , 0 ] ; R n ) 表示所有定义在 [ τ , 0 ] 上的 R n -值连续函数 φ 的全体。 C F 0 b ( Ω ; R n ) 表示所有 F 0 -可测的且定义在 C ( [ τ , 0 ] ; R n ) 上的有界函数全体,其上的范数为 φ = sup τ θ 0 | φ ( θ ) | 。对 x , y R n x , y x T y 均表示其内积。

{ r ( t ) } t 0 代表的是定义在完备概率空间 ( Ω , F , P ) 上的一个取值为 S = { 1 , 2 , , N } 的右连续时齐Markov链,其生成元(密度矩阵) Γ = ( γ i j ) N × N 由转移概率矩阵确定,即

P { r ( t + Δ ) = j | r ( t ) = i } = { γ i j Δ + o ( Δ ) , i j , 1 + γ i i Δ + o ( Δ ) , i = j ,

这里, Δ > 0 lim Δ 0 o ( Δ ) / Δ = 0 ,对于 i j γ i j 0 表示从状态i到状态j的转移速率且 γ i i = j i γ i j 。作

为一个事实, r ( ) 的每个样本轨道是一个右连续的阶梯函数,且在 R + = [ 0 , + ) 上的任何一个有限区间里至多存在有限个跳跃点(见文献 [17] )。

考虑如下形式的带Markov切换Poisson跳的非线性随机时滞微分方程:

d x ( t ) = f ( x ( t ) , x ( t τ ) , r ( t ) , t ) d t + g ( x ( t ) , x ( t τ ) , r ( t ) , t ) d B ( t ) + h ( x ( t ) , x ( t τ ) , r ( t ) , t ) d N ( t ) , t 0 (1)

初始值

x 0 = φ = { x ( t ) : τ t 0 } C F 0 b ( Ω ; R n ) r ( 0 ) = i 0 S (2)

其中 f , g , h : R n × R n × S × R + R n B ( t ) N ( t ) 分别是定义在该概率空间上的1-维Brown运动和强度为 λ 的Poisson过程,这里假设Markov链 r ( t ) 与Brown运动 B ( t ) 和Poisson过程 N ( t ) 是相互独立的。 N ˜ ( t ) = N ( t ) λ t 表示 N ( t ) 的补偿Poisson过程。

为了证明方程(1)全局解的存在唯一性,假设如下的局部Lipschitz条件成立。

(A1) 对任意的 h > 0 ,存在一个常数 L h > 0 使得对 ( i , t ) S × R + x 1 , x 2 , y 1 , y 2 R n ,有

| f ( x 1 , y 1 , i , t ) f ( x 2 , y 2 , i , t ) | | g ( x 1 , y 1 , i , t ) g ( x 2 , y 2 , i , t ) | | h ( x 1 , y 1 , i , t ) h ( x 2 , y 2 , i , t ) | L h ( | x 1 x 2 | + | y 1 y 2 | )

成立,其中, | x 1 | | x 2 | | y 1 | | y 2 | h

C 2 , 1 ( R n × S × R + ; R + ) 表示关于变量x二阶连续可导且关于变量t一阶连续可导的全体非负函数 V ( x , i , t ) 的集合。给定任意的 V ( x , i , t ) C 2 , 1 ( R n × S × R + ; R + ) ,定义算子

L V ( x , y , i , t ) = V t ( x , i , t ) + V x ( x , i , t ) f ( x , y , i , t ) + 1 2 trace [ g T ( x , y , i , t ) V x x ( x , i , t ) g ( x , y , i , t ) ] + λ [ V ( x + h ( x , y , i , t ) , i , t ) V ( x , i , t ) ] + j = 1 N γ i j V ( x , j , t ) ,

其中, V t ( x , i , t ) = V ( x , i , t ) t V x x ( x , i , t ) = ( 2 V ( x , i , t ) x i x j ) n × n

V x ( x , i , t ) = ( V ( x , i , t ) x 1 , V ( x , i , t ) x 2 , , V ( x , i , t ) x n )

在条件(A1)下,给定初始条件(2),方程(1)有唯一的局部解。本文假设方程(1)的系数不满足线性增长条件,为了防止解的爆破,本文假设另一个条件成立。

(A2) 令 H ( ) C ( R n × [ τ , ) ; R + ) 。假设存在一个函数 V C 2 , 1 ( R n × S × R + ; R + ) 和正常数 c 1 , c 2 , c 3 满足

c 3 < c 2 | x | q V ( x , i , t ) H ( x , t ) ( x , i , t ) R n × S × R +

L V ( x , y , i , t ) c 1 c 2 H ( x , t ) + c 3 H ( y , t τ ) ( x , y , i , t ) R n × R n × S × R +

3. 主要结果

定理1给定初始值(2)。若条件(A1)-(A2)成立,则方程(1)存在唯一的全局解 x ( t ) ( t τ )

证明给定初始值(2)。由条件(A1)知方程(1)在 [ τ , τ e ] 上有唯一的最大局部解 x ( t ) ,其中 τ e 是爆破时间。若要证 x ( t ) 是全局的,则只需证 τ e = a.s.。令 m 0 是一个充分大的正数且使得 x 0 = φ = sup τ s 0 | φ ( s ) | m 0 。对任意的整数 m > m 0 ,定义停时 τ m = inf { t [ 0 , τ e ) : | x ( t ) | m } 。通常规定 inf = ,这里 是一个空集。由 τ m 的定义可知 τ m 是一个单调递增的序列且 τ = lim m τ m τ e 。利用Itô公式,对 t > 0 ,有

E V ( x ( t τ m ) , r ( t τ m ) , t τ m ) = V ( x ( 0 ) , r ( 0 ) , 0 ) + E 0 t τ m L V ( x ( s ) , x ( s τ ) , r ( s ) , s ) d s

由条件(A2)得

E V ( x ( t τ m ) , r ( t τ m ) , t τ m ) V ( x ( 0 ) , r ( 0 ) , 0 ) + c 1 t c 2 E 0 t τ m H ( x ( s ) , s ) d s + c 3 E 0 t τ m H ( x ( s τ ) , s τ ) d s

注意,经计算可得

0 t τ m H ( x ( s τ ) , s τ ) d s = τ t τ m τ H ( x ( s ) , s ) d s τ 0 H ( x ( s ) , s ) d s + 0 t τ m H ( x ( s ) , s ) d s

因此

E V ( x ( t τ m ) , r ( t τ m ) , t τ m ) V ( x ( 0 ) , r ( 0 ) , 0 ) + c 1 t + c 3 τ 0 H ( x ( s ) , s ) d s ( c 2 c 3 ) E 0 t τ m H ( x ( s ) , s ) d s

因为 c 2 < c 3 ,所以

E V ( x ( t τ m ) , r ( t τ m ) , t τ m ) V ( x ( 0 ) , r ( 0 ) , 0 ) + c 1 t + c 3 τ 0 H ( x ( s ) , s ) d s

K 1 = V ( x ( 0 ) , r ( 0 ) , 0 ) + c 3 τ 0 H ( x ( s ) , s ) d s

又因 | x | q V ( x , i , t ) ,故

E | x ( t τ m ) | q K 1 + c 1 t

进一步可得 E | x ( τ m ) I { τ m t } | q K 1 + c 1 t ,从而可得

m q P ( τ m t ) K 1 + c 1 t

m ,可得 P ( τ t ) 0 ,即 τ > t a.s.由于t的任意性可得

τ = a.s.

证毕。

定理2给定初始值(2)。若条件(A1)-(A2)成立,则方程(1)的解满足

sup E τ t < | x ( t ) | q <

证明给定初始值(2)。由定理1可知,方程(1)存在唯一的全局解 x ( t ) 。设函数 f ( u ) = c 2 c 3 e u τ u u > 0 。显然 f ( u ) 关于u是一个连续函数且 f ( 0 ) = c 2 c 3 > 0 ,利用连续函数的局部保号性可知存在一个充分小的正常数 ε 使得 f ( ε ) = c 2 c 3 e ε τ ε > 0 。对 e ε t V ( x ( t ) , r ( t ) , t ) ( t > 0 ) 运用Itô公式得

E e ε ( t τ m ) V ( x ( t τ m ) , r ( t τ m ) , t τ m ) = V ( x ( 0 ) , r ( 0 ) , 0 ) + E 0 t τ m e ε s L V ( x ( s ) , x ( s τ ) , r ( s ) , s ) d s + E 0 t τ m ε e ε s V ( x ( s ) , r ( s ) , s ) d s ,

这里 τ m 如定理1中定义。根据条件(A2)可得

E e ε ( t τ m ) V ( x ( t τ m ) , r ( t τ m ) , t τ m ) V ( x ( 0 ) , r ( 0 ) , 0 ) + E 0 t τ m ε e ε s H ( x ( s ) , s ) d s + c 1 E 0 t τ m e ε s d s c 2 E 0 t τ m e ε s H ( x ( s ) , s ) d s + c 3 E 0 t τ m e ε τ e ε ( s τ ) H ( x ( s τ ) , s τ ) d s

从而

E e ε ( t τ m ) V ( x ( t τ m ) , r ( t τ m ) , t τ m ) V ( x ( 0 ) , r ( 0 ) , 0 ) + c 1 ε e ε t + E 0 t τ m ε e ε s H ( x ( s ) , s ) d s c 2 E 0 t τ m e ε s H ( x ( s ) , s ) d s + c 3 e ε τ E τ t τ m τ e ε u H ( x ( u ) , u ) d u

进一步计算得

E e ε ( t τ m ) V ( x ( t τ m ) , r ( t τ m ) , t τ m ) V ( x ( 0 ) , r ( 0 ) , 0 ) + c 1 ε e ε t + c 3 e ε τ τ 0 H ( x ( s ) , s ) d s ( c 2 c 3 e ε τ ε ) E 0 t τ m e ε s H ( x ( s ) , s ) d s

因为 f ( ε ) = c 2 c 3 e ε τ ε > 0 ,所以

E e ε ( t τ m ) V ( x ( t τ m ) , r ( t τ m ) , t τ m ) V ( x ( 0 ) , r ( 0 ) , 0 ) + c 1 ε e ε t + c 3 e ε τ τ 0 H ( x ( s ) , s ) d s

m ,则

E e ε t V ( x ( t ) , r ( t ) , t ) V ( x ( 0 ) , r ( 0 ) , 0 ) + c 1 ε e ε t + c 3 e ε τ τ 0 H ( x ( s ) , s ) d s = K 2 + c 1 ε e ε t ,

其中, K 2 = V ( x ( 0 ) , r ( 0 ) , 0 ) + c 3 e ε τ τ 0 H ( x ( s ) , s ) d s

再次利用 | x ( t ) | q V ( x ( t ) , r ( t ) , t )

E | x ( t ) | q K 2 e ε t + c 1 ε K 2 + c 1 ε <

证毕。

文章引用

林宇璇,李光洁. 带Markov切换Poisson跳的非线性随机时滞微分方程解的矩有界性
Moment Boundedness of Solutions to Nonlinear Stochastic Delay Differential Equations with Markovian Switching and Poisson Jumps[J]. 应用数学进展, 2021, 10(10): 3421-3426. https://doi.org/10.12677/AAM.2021.1010360

参考文献

  1. 1. Mao, X. (1997) Stochastic Differential Equations and Application. Horwood Publication, Chichester.

  2. 2. Øksendal, B. (2003) Stochastic Differential Equations. 6th Ed, Springer-Verlag, Berlin.

  3. 3. Arnold, L. (2007) Stochastic Differential Equations: Theory and Applications. World Scientific, Singapore.

  4. 4. Kim, Y.H., Park, C.H. and Bae, M.J. (2016) A Note on the Approximate Solutions to Stochastic Differential Delay Equation. Journal of Applied Mathematics & Informatics, 34, 421-434.

  5. 5. Mao, X. and Rassias, M. (2007) Almost Sure Asymptotic Estimations for Solutions of Stochastic Differential Delay Equations. International Journal of Applied Mathematics & Statistics, 9, 95-109. https://doi.org/10.14317/jami.2016.421

  6. 6. Buckwar, E. (2000) Introduction to the Numerical Analysis of Stochastic Delay Differential Equations. Journal of Computational and Applied Mathematics, 125, 297-307. https://doi.org/10.1016/S0377-0427(00)00475-1

  7. 7. Rodkina, A. and Basin, M. (2006) On Delay-dependent Stability for A Class of Nonlinear Stochastic Delay-Differen- tial Equations. Mathematics of Control, Signals and Systems, 18, 187-197. https://doi.org/10.1007/s00498-006-0163-1

  8. 8. Mao, X. and Yuan, C. (2006) Stochastic Differential Equations with Markovian Switching. Imperial College Press, London. https://doi.org/10.1142/p473

  9. 9. Dieu, N.T. (2016) Some Results on Almost Sure Stability of Non-Autonomous Stochastic Differential Equations with Markovian Switching. Vietnam Journal of Mathematics, 44, 665-677. https://doi.org/10.1007/s10013-015-0181-8

  10. 10. Rong, S. (2006) Theory of Stochastic Differential Equations with Jumps and Applications: Mathematical and Analytical Techniques with Applications to Engineering. Springer Science & Business Media, Berlin.

  11. 11. Wei, Y. and Yang, Q. (2018) Dynamics of the Stochastic Low Concentration Trimolecular Oscillatory Chemical System with Jumps. Communications in Nonlinear Science and Numerical Simulation, 59, 396-408. https://doi.org/10.1016/j.cnsns.2017.11.019

  12. 12. Liu, D., Yang, G. and Zhang, W. (2011) The Stability of Neutral Stochastic Delay Differential Equations with Poisson Jumps by Fixed Points. Journal of Computational and Applied Mathematics, 235, 3115-3120. https://doi.org/10.1016/j.cam.2008.10.030

  13. 13. Wang, Z., Li, X. and Lei, J. (2014) Moment Boundedness of Linear Stochastic Delay Differential Equations with Distributed Delay. Stochastic Process and Application, 124, 586-612. https://doi.org/10.1016/j.spa.2013.09.002

  14. 14. Wang, Z., Li, X. and Lei, J. (2017) Second Moment Boundedness of Linear Stochastic Delay Differential Equations. Discrete and Continuous Dynamical Systems, 19, 2963-2991. https://doi.org/10.3934/dcdsb.2014.19.2963

  15. 15. Xu, L., Dai, Z. and He, D. (2018) Exponential Ultimate Boundedness of Impulsive Stochastic Delay Differential Equations. Applied Mathematics Letters, 85, 70-76. https://doi.org/10.1016/j.aml.2018.05.019

  16. 16. Mao, W., Hu, L. and Mao, X. (2019) Asymptotic Boundedness and Stability of Solutions to Hybrid Stochastic Differential Equations with Jumps and the Euler-Maruyama Approximation. Discrete and Continuous Dynamical Systems, 24, 587-613. https://doi.org/10.3934/dcdsb.2018198

  17. 17. Anderson, W. (1991) Continuous-Time Markov Chains. Springer, New York. https://doi.org/10.1007/978-1-4612-3038-0

  18. NOTES

    *通讯作者。

期刊菜单