一类三次超越多项式零点的分布及其在时滞生物系统的应用

来源 :南京信息工程大学学报(自然科学版) | 被引量 : 0次 | 上传用户:cqc465330937
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
  摘要本文简要讨论Gronwall不等式的研究进展,并给出关于如下的一类非线性Volterra积分不等式的一个结果:
  w(u(t))≤g(t)+∑ni=1∫αi(t)αi(t0)fi(t,s)∏mj=1Hij(u(s))Gij(maxs-h≤ξ≤su(ξ))ds.
  关键词非线性积分不等式;Gronwall不等式;Gronwall类不等式;Volterra积分微分不等式;Lyapunov第二方法
  中图分类号O178
  文献标志码A
  1美国德克萨斯农工大学康莫斯分校数学系,康莫斯,德州,75428
  0关于Gronwall不等式
  1919年,在研究一个带参变量的微分方程系统时,Gronwall[1]给出如下的著名引理:
  定理1(Gronwall原始不等式)对x0≤x≤x0+h,连续函数z=z(x)满足不等式:
  0≤z≤∫xx0[Mz+A]dx,
  其中常数M和A非负,那么
  0≤z≤AheMh,x0≤x≤x0+h.
  在以后的24年里,Gronwall原始不等式都没引起关注.1943年,Bellman[2]推广了Gronwall原始不等式使得M可以是函数,并且不等式也被陈述得更为简单、明了.这个结果被称之为GronwallBellman不等式,在许多文献中均可查到,如文献[26].
  定理2(GronwallBellman不等式)已知u(t)和f(t)是定义在[a,b]区间上的非负连续函数,c是非负常数,如果
  u(t)≤c+∫taf(s)u(s)ds,a≤t≤b,
  那么
  u(t)≤ce∫taf(s)ds,a≤t≤b.
  1958年,Bellman[7]进一步改进了上述定理,使得c可以是一个非负非增连续函数.
  定理3(Bellman不等式)如果y(t)是正的,且单调增,x(t),z(t)≥0,那么
  x(t)≤y(t)+∫tαx(s)z(s)ds,
  蕴含着
  x(t)≤y(t)e∫tαz(s)ds,α≤t≤β.
  今天看,以上的3个定理都较为粗糙,因为定理的陈述不够完整、条件还可以改进.作为对前面3个定理的统一推广,1966年,Halanay在专著[8]中,给出了下面的定理4.这个定理被广泛引用,如文献[9].其条件比以上3个定理都要弱一些.这里,我们使用HaleLunel[1993,p15]的陈述.
  定理4如果u和α是定义在[a,b]区间上的实值连续函数,
  学报(自然科學版),2017,9(4):391394 Journal of Nanjing University of Information Science and Technology(Natural Science Edition),2017,9(4):391394
  王廷秀.一类非线性Volterra积分不等式.
  WANG Tingxiu.
  A nonlinear Volterratype integral inequality.
  β≥0在[a,b]区间上可积,且满足
  u(t)≤α(t)+∫taβ(s)u(s)ds,a≤t≤b,
  那么
  u(t)≤α(t)+∫taβ(s)α(s)e∫tsβ(u)duds,a≤t≤b.
  此外,如果α非减,那么
  u(t)≤α(t)e∫taβ(s)ds,a≤t≤b.
  不像前面3个定理,定理4只要求β非负.当然,如果u是非负的,α也相应必须非负.Gronwall不等式在微分方程有界性、稳定性、存在性及其他定性性质的研究中有了大量、广泛的应用,对Gronwall不等式的应用、推广、研究爆发性增长,并产生了许多新的研究方向.1998年出版的Pachpatte等的专著[6],收集、总结了在此之前对Gronwall不等式的研究、推广、应用.在众多推广中,本文讨论下面的推广.2000年,Lipovan[10]研究了
  u(t)≤k+∫α(t)α(t0)f(s)w(u(s))ds.
  2012年,Bohner等[11] 研究了下面的不等式:
  ψ(u(t))≤k+∑ni=1∫αi(t)αi(t0)fi(s)up(s)ωi(u(s))ds+∑mj=1∫βj(t)βj(t0)gj(s)up(s)j(maxξ∈[s-h,s] u(ξ))ds. (1)
  2015年,Wang[12]推广了不等式(1),用更一般的复合函数Hij(u(s))取代up(s),用
  fi(s)up(s)ωi(u(s))j(maxξ∈[s-h,s] u(ξ))
  合并了两个级数,研究了下面的不等式:
  w(u(t))≤K+
  ∑ni=1∫αi(t)αi(t0)fi(s)∏mj=1Hij(u(s))Gij(maxs-h≤ξ≤su(ξ))ds. (2)
  本文对不等式(2)进一步加以推广,使得K可以是函数,fi(s)可以是fi(t,s).从而,我们研究Volterra不等式:
  w(u(t))≤g(t)+∑ni=1∫αi(t)αi(t0)fi(t,s)∏mj=1Hij(u(s))Gij(maxs-h≤ξ≤su(ξ))ds. (3)
  1一类非线性Volterra积分不等式
  我们研究不等式(3),并得到一个结果.由于不等式(3)涉及7类函数,为此,我们需要如下6个条件和记号:已知h>0,t0,T为常数,0≤t0  (A2) αi∈C1([t0,T),R+)非减,并且αi(t)≤t,t∈[t0,T),i=1,2,…,n;
  (A3) fi(t,s)∈C([t0,T)×[t0,T),R)对t为连续非减函数,i=1,2,…,n;
  (A4) Hij,Gij∈C(R+,R+)非减,且当x>0,Hij(x)>0,Gij(x)>0;
  (A5) w∈C(R+,R+)为严格递增函数,w(0)=0,limt→∞w(t)=∞;
  (A6) u∈C([-h,T),R+).
  定理5如果u(t)满足不等式(3)以及以上6个条件,从(A1)—(A6).那么不等式(3)的解是:
  u(t)≤w-1W-1W(g(t))+∑ni=1∫αi(t)αi(t0)fi(t,s)ds,t∈[t0,T).
  其中
  W(r)=∫rr01Hm(w-1(s))Gm(w-1(s))ds,0≤r<∞,
  r0是一个合适的非负常数,使得W(r)有定义.
  H(r)=max1≤i≤n,1≤j≤m{Hij(r)},
  G(r)=max1≤i≤n,1≤j≤m{Gij(r)}.
  证明取η为一任意常数满足t0≤t≤η  w(u(t))≤g(η)+
  ∑ni=1∫αi(t)αi(t0)fi(η,s)∏mj=1Hij(u(s))Gijmaxs-h≤ξ≤su(ξ)ds.
  对t0≤t≤η,定义上面的不等式的右边为Z(t),即
  Z(t)=g(η)+
  ∑ni=1∫αi(t)αi(t0)fi(η,s)∏mj=1Hij(u(s))Gij(maxs-h≤ξ≤su(ξ))ds.
  不难看出,Z(t)非减,且0≤w(u(t))≤Z(t),t∈[t0,η].
  由(A5),w-1存在且具有和w相同的性质.因此,
  u(t)≤w-1(Z(t)),t0≤t≤η.
  此外,
  maxs-h≤ξ≤su(ξ)≤maxs-h≤ξ≤sw-1(Z(ξ))=w-1(Z(s)),t0≤s≤η.
  因此,对t∈[t0,η],
  Z(t)≤g(η)+∑ni=1∫αi(t)αi(t0)fi(η,s)∏mj=1
  Hij(w-1(Z(s)))Gij(w-1(Z(s)))ds≤
  g(η)+∑ni=1∫αi(t)αi(t0)fi(η,s)∏mj=1
  H(w-1(Z(s)))G(w-1(Z(s)))ds≤
  g(η)+∑ni=1∫αi(t)αi(t0)fi(η,s)
  Hm(w-1(Z(s)))Gm(w-1(Z(s)))ds.
  再定義上面的不等式的右边为R(t),即:
  R(t)=g(η)+
  ∑ni=1∫αi(t)αi(t0)fi(η,s)Hm(w-1(Z(s)))Gm(w-1(Z(s)))ds,t0≤t≤η  显然,R(t0)=g(η),0≤Z(t)≤R(t),t0≤t≤η  R′(t)=∑ni=1fi(η,αi(t))Hm(w-1(Z(αi(t))))·
  Gm(w-1(Z(αi(t))))αi′(t)≤
  ∑ni=1fi(η,αi(t))Hm(w-1(Z(t)))·
  Gm(w-1(Z(t)))αi′(t)≤
  ∑ni=1fi(η,αi(t))Hm(w-1(R(t)))·
  Gm(w-1(R(t)))αi′(t)≤
  Hm(w-1(R(t)))Gm(w-1(R(t)))∑ni=1fi(η,αi(t))αi′(t).
  所以,
  R′(t)Hm(w-1(R(t)))Gm(w-1(R(t)))≤∑ni=1fi(η,αi(t))αi′(t).
  即:
  d(W(R(t)))dt≤∑ni=1fi(η,αi(t))αi′(t).
  对上式从t0到t积分,t∈[t0,η],我们得到:
  W(R(t))-W(R(t0))≤∫tt0∑ni=1fi(η,αi(s))αi′(s)ds=∫α(t)α(t0)∑ni=1fi(η,s)ds,
  W(R(t))≤W(g(η))+∑ni=1∫α(t)α(t0)fi(η,s)ds.
  因为W连续且严格增,所以W-1存在、连续、严格增.对上式使用W-1,我们得到:
  R(t)≤W-1W(g(η))+∑ni=1∫αi(t)αi(t0)fi(η,s)ds,t∈[t0,η],
  w(u(t))≤Z(t)≤R(t)≤W-1W(g(η))+∑ni=1∫αi(t)αi(t0)fi(η,s)ds,t∈[t0,η],
  u(t)≤w-1W-1W(g(η))+
  ∑ni=1∫αi(t)αi(t0)fi(η,s)ds,t∈[t0,η].
  在上式,取t=η,可得:
  u(η)≤w-1W-1W(g(η))+
  ∑ni=1∫αi(η)αi(t0)fi(η,s)ds,t∈[t0,η].
  由于η是区间t0≤η  u(t)≤w-1W-1W(g(t))+
  ∑ni=1∫αi(t)αi(t0)fi(t,s)ds,t∈[t0,T).   证毕.
  2应用
  我们的结果可以很容易地应用到微分方程的稳定性理论中.例如,我们考虑滞后型泛函微分方程
  dudt=F(t,ut), (4)
  在用Lyapunov第二方法研究微分方程稳定性时,如下条件经常可见:
  1) W1(|(0)|)≤V(t,)  2) W1(|(0)|)≤V(t,)  3) W1(|u(t)|X)≤V(t,ut)≤W2(D(t,ut))+∫tt-hL(s)W1(|u(s)|X)ds;
  4) V′(4)(t,)≤0;
  5) V′(4)(t,)≤-W4(|(0)|);
  6) V′(4)(t,)≤-W4(∫tt-h|(s)|ds);
  7) V(4)(t,ut)≤-η(t)W2(D(t,ut))+P(t).
  關于与这些条件相关的假设,定理可在众多的文献中查到,如文献[45,1220].这些条件加上定理5,可以得到对Lyapunov函数和解的上界,文献[13]有一个很相近的例子.再如,定理5可以很容易地应用到Volterra积分方程:
  X(t)=a(t)+∫tt-αg(t,s)X(s)ds,X∈Rn. (5)
  关于方程(5),很多论文研究过,如文献[5,2122].
  参考文献
  References
  [1]Gronwall T H.Note on the derivatives with respect to a parameter of the solutions of a system of differential equations[J].Annals of Mathematics,1919,20(4):292296
  [2]Bellman R.The stability of solutions of linear differential equations[J].Duke Mathenatical Journal,1943,10(4):643647
  [3]Bellman R.Stability theory of differential equations[M].New York:McGrawHill,1953
  [4]Burton T A.Stability and periodic solutions of ordinary and functionaldifferential equations[M].Orlando,Florida:Academic Press,1985
  [5]Burton T A.Volterra integral and differential equations[M].2nd ed.New York:Elsevier,2005
  [6]Pachpatte B G.Inequalities for differential and integral equations[M].London:Academic Press,1998
  [7]Bellman R.Asymptotic series for the solutions of linear differentialdifference equations[J].Rendiconti del Circolo Matematico di Palermo Series 2,1958,7(3):261269
  [8]Halanay A.Differential equations:Stability,oscillations,time lags[M].New York and London:Academic Press,1966
  [9]Hale J,Lunel S.Theory of functional differential equations[M].New York:SpringerVerlag,1993
  [10]Lipovan O.A retarded Gronwalllike inequality and its applications[J].Journal of Mathematical Analysis and Applications,2000,252(1):389401
  [11]Bohner M,Hristova S,Stefanova K.Nonlinear integral inequalities involving maxima of the unknown scalar functions[J].Mathematical Inequalities & Applications,2012,15(4):811825
  [12]Wang T X.Generalization of Grownwalls inequality and its applications in functional differential equations[J].Communications in Applied Analysis,2015,19:679688
  [13]Wang T X.Stability in abstract functional differential equations.Part I:General theorems[J].Journal of Mathematical Analysis and Applications,1994,186(2):534558   [14]Wang T X.Stability in abstract functional differential equations.Part II:Applications[J].Journal of Mathematical Analysis and Applications,1994,186(3):835861
  [15]Wang T X.Lower and upper bounds of solutions of functional differential equations[J].Dynamics of Continuous,Discrete and Impulsive Systems,Series A:Mathematical Analysis,2013,20(1):131141
  [16]Wang T X.Inequalities of Solutions of Volterra integral and differential equations[J].Electronic Journal of Qualitative Theory of Differential Equations,2009(28):538543
  [17]Wang T X.Exponential stability and inequalities of abstract functional differential equations[J].Journal of Mathematical Analysis and Applications,2006,342(2):982991
  [18]Wang T X.Inequalities and stability in a linear scalar functional differential equation[J].Journal of Mathematical Analysis and Applications,2004,298(1):3344
  [19]Wang T X.Wazewskis inequality in a linear Volterra integrodifferential equations[M]//Corduneanu C,Sandberg I W.Volterra equations and applications.Amsterdam:Gordon and Breach Science Publishers,2000:483492
  [20]廖曉昕.动力系统的稳定性理论和应用[M].北京:国防工业出版社,2000
  LIAO Xiaoxin.Theory and application of stability for dynamic systems[M].Beijing:National Defense Industry Press,2000
  [21]Burton T A.Lyapunov functional for integral equations[M].Bloomington,Indiana:Trafford Publishing,2008
  [22]Wang T X.Bounded solutions and periodic solutions in integral equations[J].Dynamics of Continuous,Discrete,and Impulsive Systems,2000,7(1):1931
其他文献
摘要  网络虚拟化技术通过对物理资源的抽象,可以有效解决现有互联网架构中存在的网络结构僵化、可扩展性差等问题。虚拟网络映射问题是指将用户发送的所有虚网请求映射到底层物理网络中,同时还要满足虚网请求中对各个资源的限制要求(如节点计算能力、链路带宽等)。从节点负载平衡的角度出发,在基于就近原则的虚网映射算法基础上,引入节点负载平衡的反馈机制,引导各个虚网请求更均匀地映射到底层物理网络中。另外,在k短路
摘要本文考虑了具有混合时变延时和不同时标的混沌忆阻竞争神经网络的自适应同步问题.使用Lyapunov泛函方法和不等式分析技术,设计了一类新的具有反馈控制律的自适应控制器以取得网络同步及指数同步目的,提出了不用过多计算,如求解线性矩阵不等式或复杂代数计算的保证网络同步条件;同时,所获条件也可以应用到已有文献里关于忆阻器网络不同数学模型中.最后,通过实例验证了本文获得的理论结果的有效和正确性.关键词自