A new global asymptotic stability criterion of Takagi-Sugeno fuzzy Cohen-Grossberg neural networks with probabilistic time-varying delays was derived, in which the diffusion item can play its role. Owing to deleting the boundedness conditions on amplification functions, the main result is a novelty to some extent. Besides, there is another novelty in methods, for Lyapunov-Krasovskii functional is the positive definite form of p powers, which is different from those of existing literature. Moreover, a numerical example illustrates the effectiveness of the proposed methods.
Department of Science and Technology of Sichuan Province2011JYZ0102012JY010Education Department of Sichuan Province14ZA027412ZB34911ZA17208ZB0021. Introduction
Cohen-Grossberg neural networks (CGNNs) have many practical applications, like artificial intelligence, parallel computing, image processing and recovery, and so on ([1–6]). But the success of these applications largely depends on whether the system has some stability, and so people began to be interested in the stability analysis of the system. In recent decades, reaction-diffusion neural networks have received much attention ([7–13]), including various Laplacian diffusion ([6, 14–20]). Besides, people are paying more and more attention to fuzzy neural network system ([21–34]), due to encountering always some inconveniences such as the complicity, the uncertainty, and vagueness ([27, 35–37]). For example, in [27], Zhu and Li investigated the following fuzzy CGNNs model: (1)dxit=-aixitbixit-⋀j=1naijfjxjt-⋁j=1nbijgjxjt-⋀j=1ncijfjxjt-τ-⋁j=1ndijgjxjt-τdt+∑j=1nσijxjt,xjt-τdwjt,xit=ϕit,-τ⩽t⩽0.
In [36], Muralisankar and Gopalakrishnan studied the following T-S fuzzy neutral type CGNNs with distributed delays: (2)dxit=∑j=1rhjωt-AjxtBjxt-Cjfxt-τt-Mj∫t-ρttfxsds-Djx˙t-rt.Besides, Balasubramaniam and Syed Ali discussed Takagi-Sugeno fuzzy Cohen-Grossberg BAM neural networks with discrete and distributed time-varying delays in [37].
Note that there is the following bounded condition on amplification functions in many literatures (see, e.g., [38, Theorem 4]) related to CGNNs: (3)0<a_i⩽air⩽a¯i,r∈R,i=1,2,…,n.
So, in this paper, we try to delete this bounded condition on amplification functions. This is the main purpose of this paper.
2. Preliminaries
Consider the following fuzzy Takagi-Sugeno p-Laplace partial differential equations with distributed delay.
Fuzzy Rule j. IF ω1(t) is μj1 and⋯ωs(t) is μjs THEN(4)∂u∂t=∇·Dt,x,u∘∇pu-AuBu-Cjfut-τt,x-Mj∫t-ρttfus,xds,uθ,x=ϕθ,x,θ,x∈-∞,0×Ω,ut,x=0∈Rn,t,x∈R×∂Ω,where Ω is an arbitrary open bounded subset in Rm. ωk(t)(k=1,2,…,s) is the premise variable and μjk(j=1,2,…,r;k=1,2,…,s) is the fuzzy set that is characterized by membership function. And r is the number of the IF-THEN rules; s is the number of the premise variables. u(t,x)=(u1(t,x),u2(t,x),…,un(t,x))T∈Rn, where ui(t,x) is the state variable of the ith neuron and the jth neuron at time t and in space variable x. Matrix D(t,x,u)=Dijt,x,un×m with each Dij(t,x,u)⩾0, and Dij(t,x,u) is diffusion operator. D(t,x,u)∘∇pu=Djkt,x,u∇uip-2∂ui/∂xkn×m denotes the Hadamard product of matrix D(t,x,u) and ∇pu (see [39] for details). Matrices A(u)=diag(a1(u1),a2(u2),…,an(un)) and B(u)=diag(b1(u1),b2(u2),…,bn(un)), where ai(ui) and bi(ui) represent an amplification function at time t and an appropriate behavior function at time t. Cj=cikjn×n is the connection matrix. Time delays τ(t)∈[0,+∞). fut-τt,x=(f1(u1(t-τ(t)t,x)),f2(u2(t-τ(t),x)),…,fn(un(t-τ(t),x)))T is the activation function of the neurons. And the second and third equations of (4) imply the initial condition and the Dirichlet boundary condition, respectively.
By way of a standard fuzzy inference method, (4) can be inferred as follows. (5)∂u∂t=∇·Dt,x,u∘∇pu-Aut,xBut,x-∑j=1rhjωtCjfut-τt,x+Mj∫t-ρttfus,xds,uθ,x=ϕθ,x,θ,x∈-∞,0×Ω,ut,x=0∈Rn,t,x∈R×∂Ω,where ω(t)=[ω1(t),ω2(t),…,ωs(t)]T and hj(ω(t))=wj(ω(t))/∑k=1rwk(ω(t)),wj(ω(t)):Rs→[0,1](j=1,2,…,r) is the membership function of the system with respect to the fuzzy rule j. hj can be regarded as the normalized weight of each IF-THEN rule, satisfying hj(ω(t))⩾0 and ∑j=1rhj(ω(t))=1.
Next, we consider the following information for probability distribution of time delays τ(t): (6)P0⩽τt⩽τ1=c0,Pτ1<τt⩽τ2=1-c0.Here the nonnegative scalar c0⩽1. Define a random variable as follows: (7)Ct=1,0⩽τt⩽τ1;0,τ1<τt⩽τ2.So, in this paper, we consider the following Takagi-Sugeno (T-S) fuzzy system with probabilistic time-varying delays:(8)∂u∂t=∇·Dt,x,u∘∇pu-Aut,xBut,x-∑j=1rhjωtc0Cjfut-τ1t,x+1-c0Cjfut-τ2t,x+C-c0Cjfut-τ1t,x-Cjfut-τ2t,x+Mj∫t-ρttfus,xds,uθ,x=ϕθ,x,θ,x∈-∞,0×Ω,ut,x=0∈Rn,t,x∈R×∂Ω.
System (8) includes the following integrodifferential equations: (9)dxtdt=-AxtBxt-∑j=1rhjωtc0Cjfxt-τ1t+1-c0Cjfxt-τ2t+C-c0Cjfxt-τ1t-Cjfxt-τ2t+Mj∫t-ρttfxsds,t⩾0,xθ=ϕθ,θ∈-∞,0.
Particularly when p=2, system (8) degenerates into the so-called reaction-diffusion CGNNs:(10)∂u∂t=∇·Dt,x,u∘∇u-Aut,xBut,x-∑j=1rhjωtc0Cjfut-τ1t,x+1-c0Cjfut-τ2t,x+C-c0Cjfut-τ1t,x-Cjfut-τ2t,x+Mj∫t-ρttfus,xds,uθ,x=ϕθ,x,θ,x∈-∞,0×Ω,ut,x=0∈Rn,t,x∈R×∂Ω.
Throughout this paper, we assume p=p1/p2>1 with p1 being even number and p2 being odd number. Besides, suppose that the following conditions hold:
There exist positive definite matrices A_=diag(a_1,a_2,…,a_n) and A¯=diag(a¯1,a¯2,…,a¯n) such that (11)0<a_i⩽aissp-2⩽a¯i,0≠s∈R,i=1,2,…,n,
where A(u)=diag(a1(u1),a2(u2),…,an(un)) and u=(u1,u2,…,un)T∈Rn.
There exists a positive definite matrix B=diag(b1,b2,…,bn) such that bi(0)=0 and (12)biss⩾bi,0≠s∈R,i=1,2,…,n.
There is a positive definite matrix F=diag(F1,F2,…,Fn) such that (13)fis⩽Fis,s∈R,i=1,2,…,n.
From (H1)–(H3), we know that bi(0)=fi(0)=0 and u=0 is an equilibrium of fuzzy system (8).
Remark 1.
There are numerous functions satisfying (H1). For example, if p=4/3, we may set (14)ais=0.11+e-s2s23,∀0≠s∈R,ai0=0.2.It is obvious that (15)lims→0ais=lims→00.11+e-s2s23=+∞.So the function ai(s) is unbounded for s∈R. Moreover, (16)0.1⩽aissp-2=0.11+e-s2⩽0.2.One can know from (16) that 0.1⩽ai(s)/sp-2⩽0.2 with a_i=0.1 and a¯i=0.2.
Remark 2.
The amplification function ai(s) defined as (7) is actually unbounded for s∈R. However, various bounded conditions always imposed restrictions on the amplification functions of existing literature ([3–6, 9, 10, 24, 27, 28]). Hence, our condition (H1) is weaker, which will make a corollary with regard to ordinary integrodifferential equations (9) become novel.
For convenience’s sake, we need to introduce the following standard notations similarly as [38]:(17)Q=qijn×n>0<0,Q=qijn×n⩾0⩽0,Q1⩾Q2Q1⩽Q2,Q1>Q2Q1<Q2,λmaxΦ,λminΦ,C=cijn×n,ut,x,andtheidentitymatrixwithcompatibleI.
The Sobolev space =u∈Lp:Du∈Lp (see [40] for details).
Denote by λ1 the lowest positive eigenvalue of the boundary value problem (see [40] for details). (18)-Δpφt,x=λφt,x,x∈Ωφt,x=0,x∈∂Ω.
Lemma 3.
One has(19)aq-1b⩽q-1qaq+bqq,∀a,b∈0,+∞,q>1.
Note that Lemma 3 is the particular case of the famous Young inequality.
3. Results and DiscussionLemma 4.
Let P=diag(p1,p2,…,pn) be a positive definite matrix and u be a solution of the fuzzy system (8). Then one has (20)∫ΩuTP∇·Dt,x,u∘∇pudx⩽-λ1p_Dupp, where D=minik(inft,x,u∣Dij(t,x,u)), upp=∑i=1n∫Ωuipdx, and p_ is a positive scalar, satisfying P>p_I.
Proof.
Since u is a solution of system (8), it follows by Gauss formula and the Dirichlet zero-boundary condition that (21)∫Ω∑j=1npjuj∑k=1m∂∂xkDjk∇ujp-2∂uj∂xkdx=-∑k=1m∑j=1n∫ΩpjDjk∇ujp-2∂uj∂xk2dx⩽-λ1Dp_∑j=1n∫Ωujpdx=-λ1p_Dupp.
Remark 5.
Lemma 4 extends the conclusion of [2, Lemma 2.1] and [10, Lemma 2.4] from Hilbert space H01(Ω) to Banach space W01,p(Ω). Particularly, in the case of Ω=(0,T)⊂R1 or W01,p(0,T), the first eigenvalue λ1=(2/T∫0p-11/pdt/1-tp/p-11/p)p (see, e.g., [40]).
Theorem 6.
If there exists a positive definite matrix P=diag(p1,p2,…,pn) and two positive scalars p_, p¯ such that the following inequalities hold: (22)λ1Dp_+p_λminA_B>np¯p∑j=1rp-1cj+ρp-1mj+c0cj1-τ1+1-c0cj1-τ2+ρmjλmaxA¯λmaxF,(23)P>p_I,P<p¯I,then the null solution of fuzzy system (8) is globally asymptotically stable, where matrices Cj=(cik(j))n×n,Mj=(mik(j))n×n, cj=maxikcikj, mj=maxikmikj, and τ1′(t)⩽τ1<1,τ2′(t)⩽τ2<1,0⩽ρ(t)⩽ρ.
Proof.
Firstly, we can conclude from (H1)–(H3) that u=0 is an equilibrium point for system (8).
Next, consider the Lyapunov-Krasovskii functional: (24)Vt=V1t+11-τ1V2t+11-τ2V3t+V4t,where (25)V1t=∫ΩuTt,xPut,xdx=∑i=1n∫Ωpiui2dx,V2t=2c0np¯p∑j=1rcjλmaxA¯λmaxF∑k=1n∫t-τ1tt∫Ωuks,xpdxds,V3t=21-c0np¯p∑j=1rcjλmaxA¯λmaxF∑k=1n∫t-τ2tt∫Ωuks,xpdxds,V4t=2p¯p∑j=1rmjλmaxA¯λmaxF∑i=1n∑k=1n∫Ω∫-ρ0dζ∫t+ζtuks,xpdsdx.
Here, u(t,x)=(u1(t,x),u2(t,x),…,un(t,x))T is a solution for stochastic fuzzy system (8). Below, we may denote u(t,x) by u and ui(t,x) by ui for simplicity.
Remark 7.
It is obvious that our Lyapunov-Krasovskii functional is the positive definite form of p powers, which is different from those of existing literature ([41–43]). For example, in [41], the model is also neural networks with discrete time delay and distributed delays: (26)dxt=-Cixt+Aifyt-τt,i+Bi∫t-ρttfys,rsdt+σtdw1t,dyt=-C~iyt+A~igxt-τ~t,i+B~i∫t-ρ~ttgxs,rsdt+σ~tdw2t.In [42, Theorem 1], the corresponding Lyapunov-Krasovskii functional is as follows: (27)V~2=∫-ρ0dθ∫t+θtfTys,rsLfys,rsds+∫-ρ~0dθ∫t+θtgTxs,rsLgxs,rsds, which is the positive definite form of 2 powers. And the conclusion of [42, Theorem 1] is the asymptotical stability in the mean square, which is also similar to that of our Theorem 6. However, by means of our Lyapunov-Krasovskii functional with the positive definite form of p powers, we shall derive the asymptotical stability in the mean square for nonlinear p-Laplacian diffusion system (8).
Evaluating the time derivation of V1(t) along the trajectory of the fuzzy system (8), we can get by [38, Lemma 6] and Lemma 4(28)V1′t=2∫ΩuTP∇·Dt,x,u∘∇pu-uTPAuBudx+2∑j=1rhjωt∫ΩuTPAuc0Cjfut-τ1t,xdx+∫ΩuTPAu1-c0Cjfut-τ2t,xdx+∫ΩuTPAuMj∫t-ρttfus,xdsdx⩽-2λ1p_Dupp-2∫ΩuTPAuBudx+2c0∑j=1r∫ΩuTPAuCjfut-τ1t,xdx+21-c0∑j=1r∫ΩuTPAuCjfut-τ2t,xdx+2∑j=1r∫ΩuTPAuMj∫t-ρttfus,xdsdx.Besides, gathering (H1) and (H2) gives (29)∫ΩuTPAuBudx⩾p_λminA_Bupp.
It follows by (H1), (H3), and Lemma 3 that (30)c0∫ΩuTPAuCjfut-τ1t,xdx=c0∑k=1n∑i=1n∫Ωpiuiaiuicikjfkukt-τ1t,xdx⩽c0p¯∑k=1n∑i=1n∫Ωa¯icjuip-1Fkukt-τ1t,xdx⩽c0p¯cjλmaxA¯λmaxF∑k=1n∑i=1n∫Ωp-1puip+ukt-τ1t,xppdx=c0p-1p¯pncjλmaxA¯λmaxFupp+c0p¯pncjλmaxA¯λmaxF∑k=1n∫Ωukt-τ1t,xpdx.Similarly, (31)1-c0∫ΩuTPAuCjfut-τ2t,xdx⩽1-c0p-1p¯pncjλmaxA¯λmaxFupp+1-c0p¯pncjλmaxA¯λmaxF∑k=1n∫Ωukt-τ2t,xpdx,∫ΩuTPAuMj∫t-ρttfus,xdsdx=∑k=1n∑i=1n∫Ωpiuiaiuimikj∫t-ρttfkuks,xdsdx⩽p¯∑k=1n∑i=1n∫Ω∫t-ρtta¯imjuit,xp-1Fkuks,xdsdx⩽p¯mjλmaxA¯λmaxF∑k=1n∑i=1n∫Ω∫t-ρttp-1puit,xp+uks,xppdsdx⩽ρp-1p¯pnmjλmaxA¯λmaxFupp+p¯pmjλmaxA¯λmaxF∑i=1n∑k=1n∫Ω∫t-ρttuks,xpdsdx.
On the other hand,(32)V2′t=2c0np¯p∑j=1rcjλmaxA¯λmaxF∑k=1n∫Ωukt,xpdx-∑k=1n1-τ1′t∫Ωukt-τ1t,xpdx⩽2c0np¯p∑j=1rcjλmaxA¯λmaxFupp-2c0np¯p∑j=1rcjλmaxA¯λmaxF1-τ1∑k=1n∫Ωukt-τ1t,xpdx.Similarly, (33)V3′t⩽21-c0np¯p∑j=1rcjλmaxA¯λmaxFupp-21-c0np¯p∑j=1rcjλmaxA¯λmaxF1-τ2∑k=1n∫Ωukt-τ2t,xpdx.
Next, we need to recall some facts derived by mathematical analysis. Assume that η(t,s) is continuous on variables t and s, and ∂η/∂t exists, utilizing the integral middle value theorem reaches (34)ddt∫ξtϖtηt,sds=ϖ′tηt,ϖt-ξ′tηt,ξt+∫ξtϖt∂ηt,s∂tds,where both ξ(·) and ϖ(·) are differentiable.
Moreover, we can derive by employing (32) time and again (35)V4′t=2p¯p∑j=1rmjλmaxA¯λmaxF∑i=1n∑k=1n∫Ω∫-ρ0ukt,xpds-∫-ρ0ukt+s,xpdsdx=2p¯p∑j=1rmjλmaxA¯λmaxF∑i=1n∑k=1n∫Ωρukt,xp-∫t-ρttuks,xpdsdx=2p¯p∑j=1rmjλmaxA¯λmaxFnρupp-∑i=1n∑k=1n∫Ω∫t-ρttuks,xpdsdx.
Combining (28)–(35) results in (36)V′t⩽-2λ1Dp_+p_λminA_B-np∑j=1rp¯p-1cj+ρp-1p¯mj+c0p¯1-τ1cj+1-c0p¯1-τ2cj+ρp¯mjλmaxA¯λmaxFupp⩽0.
Now the standard Lyapunov functional theory derives that the null solution of the fuzzy system (8) is globally asymptotically stable.
Remark 8.
In the case of Takagi-Sugeno fuzzy model, our Theorem 6 is better than [38, Theorem 4] because the condition (H1) is weaker than the bounded assumption (2).
Remark 9.
In Theorem 6, (22) illustrates the influence of nonlinear diffusion on the stability of system (8) while its role was always ignored in existing results (see, e.g., [5, 39, 44]).
Theorem 6 derives the following corollary.
Corollary 10.
If there exists a positive definite matrix P=diag(p1,p2,…,pn) and two positive scalars p_, p¯ such that the following inequalities hold: (37)p_λminA_B>np¯p∑j=1rp-1cj+ρp-1mj+c0cj1-τ1+1-c0cj1-τ2+ρmjλmaxA¯λmaxF,P>p_I,P<p¯I,then the null solution of the ordinary integrodifferential equations (9) is globally asymptotically stable.
Furthermore, if both diffusion behaviors and distributed delay are ignored, we derive from Corollary 10.
Corollary 11.
If there exists a positive definite matrix P=diag(p1,p2,…,pn) and two positive scalars p_, p¯ such that the following inequalities hold: (38)p_λminA_B>np¯p∑j=1rp-1cj+c0cj1-τ1+1-c0cj1-τ2λmaxA¯λmaxF,P>p_I,P<p¯I,then the null solution of the following fuzzy system (39)dxtdt=-AxtBxt-∑j=1rhjωtc0Cjfxt-τ1t+1-c0Cjfxt-τ2t+C-c0Cjfxt-τ1t-Cjfxt-τ2t,t⩾0,xθ=ϕθ,θ∈-∞,0is globally asymptotically stable.
Remark 12.
Condition (H1) is weaker than the bounded conditions on amplification functions of existing literature ([3–6, 9, 10, 24, 27, 28]).
Discussion 1.
In recent related literature ([27, 45–51]), some new conditions and methods were presented, and their results were very good. However, some of the methods and conditions are not applicable for system (8) with nonlinear p-Laplacian diffusion. How to apply the new conditions and methods of [45–49] to our system (8) is an interesting problem.
4. Methods and Numerical Example4.1. Methods
In this paper, Lyapunov functional method is employed to derive the stability criterion. In this process, the integral middle value theorem together with the derivation formula on integral upper limit functions plays the important roles.
Example 1.
Consider the following Takagi-Sugeno p-Laplace fuzzy T-S dynamic equations.
Fuzzy Rule 1. IF ω1(t) is μ11, and ω2(t) is μ12, THEN(40)∂u∂t=∇·Dt,x,u∘∇pu-AuBu-c0C1fut-τ1t,x-1-c0C1fut-τ2t,x-C-c0C1fut-τ1t,x-C1fut-τ2t,x-M1∫t-ρttfus,xds,uθ,x=ϕθ,x,θ,x∈-∞,0×Ω,ut,x=0∈R2,t,x∈R×∂Ω.
Fuzzy Rule 2. IF ω1(t) is μ21, and ω2(t) is μ22, THEN(41)∂u∂t=∇·Dt,x,u∘∇pu-AuBu-c0C2fut-τ1t,x-1-c0C2fut-τ2t,x-C-c0C2fut-τ1t,x-C2fut-τ2t,x-M2∫t-ρttfus,xds,uθ,x=ϕθ,x,θ,x∈-∞,0×Ω,ut,x=0∈R2,t,x∈R×∂Ω,where u(t,x)=(u1(t,x),u2(t,x))T,Ω=(0,π),p=4/3, and then Remark 1 gives (42)λ1=2π∫0p-11/pdt1-tp/p-11/pp=0.7915. Let τ1(t)=t/3,τ2(t)=t/2, and then τ1=1/3,τ2=1/2. Let ai(ui)=0.1ui-2/3(1+e-iui2),i=1,2,b1(u1)=2u1(1+sin2u1),b2(u2)=1.95u2, f1(u1(t-τ(t)))=0.16u1t-τtsinu1(t-τ(t)),f2(u2(t-τ(t)))=0.166u2(t-τ(t)),τ=0.5, D=0.003, c0=0.75,c1=0.2,c2=0.3,m1=0.02,m2=0.03, and (43)A_=0.01000.02,A¯=0.1000.2,B=2001.95,Dt,x,u=0.0030.0050.0040.006,M1=0.020.0100.01,F=0.16000.166,C1=0.20.100.15,C2=0.20.100.3,M2=0.010.0100.03. Now we use MATLAB to solve LMIs (22)-(23), obtaining the feasibility data (44)P=0.9381001.013,p_=0.9103,p¯=1.023.
Now Theorem 6 derives that the null solution of this Takagi-Sugeno fuzzy equations is globally asymptotically stable (see Figures 1 and 2).
Computer simulations of the state u1(t,x).
Computer simulations of the state u2(t,x).
5. Conclusions
By constructing a novel Lyapunov function, we employed Young inequality and LMI technique to derive the asymptotic stability criteria for CGNNs with distributed delays and nonlinear diffusion. Since the stability of nonlinear p-Laplacian diffusion neural networks was originally investigated in [2], various p-Laplacian diffusion neural networks have attracted a lot of interest ([6, 17, 34, 39, 44]). As pointed out in Discussion 1, some new conditions and methods may not be applicable to CGNNs model with nonlinear p-Laplacian diffusion. So our results are a novelty to some extent.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Authors’ Contributions
Xiongrui Wang wrote the original manuscript, carrying out the main part of this paper. Shouming Zhong checked it, and Ruofeng Rao is in charge of correspondence. All authors read and approved the final manuscript.
Acknowledgments
This work is supported by Scientific Research Fund of Science Technology Department of Sichuan Province (2011JYZ010, 2012JY010) and Scientific Research Fund of Sichuan Provincial Education Department (14ZA0274, 12ZB349, 11ZA172, and 08ZB002).
ZhangX.WuS.LiK.Delay-dependent exponential stability for impulsive Cohen-Grossberg neural networks with time-varying delays and reaction-diffusion terms201116315241532MR273682910.1016/j.cnsns.2010.06.0232-s2.0-77957333769RaoR.ZhongS.WangX.Stochastic stability criteria with LMI conditions for Markovian jumping impulsive BAM neural networks with mode-dependent time-varying delays and nonlinear reaction-diffusion201419125827310.1016/j.cnsns.2013.05.024MR3142464LiuQ.XuR.Periodic solutions of a cohen-grossberg-type BAM neural networks with distributed delays and impulses201220121764341810.1155/2012/6434182-s2.0-84861038985SongH.ChenD.LiW.QuY.Graph-theoretic approach to exponential synchronization of stochastic reaction-diffusion Cohen-Grossberg neural networks with time-varying delays20161771791872-s2.0-8495931346410.1016/j.neucom.2015.11.036XiangH.CaoJ.Periodic oscillation of fuzzy cohen-grossberg neural networks with distributed delay and variable coefficients200820081845362710.1155/2008/4536272-s2.0-41949125937QingfeiP.ZifangZ.JingchangH.Stability of the stochastic reaction-diffusion neural network with time-varying delays and p-laplacian201220121040593910.1155/2012/4059392-s2.0-84871843179SongQ.YanH.ZhaoZ.LiuY.Global exponential stability of impulsive complex-valued neural networks with both asynchronous time-varying and continuously distributed delays2016811102-s2.0-8496993087010.1016/j.neunet.2016.04.012LiK.ZhangL.ZhangX.LiZ.Stability in impulsive Cohen-Grossberg-type BAM neural networks with distributed delays2010215113970398410.1016/j.amc.2009.12.001MR2578862LiK.SongQ.Exponential stability of impulsive Cohen-Grossberg neural networks with time-varying delays and reaction-diffusion terms2008721–323124010.1016/j.neucom.2008.01.0092-s2.0-55949127155PanJ.ZhongS.Dynamic analysis of stochastic reaction-diffusion Cohen-Grossberg neural networks with delays2009200911841082310.1155/2009/4108232-s2.0-70449717143RaoR.WangX.ZhongS.LMI-based stability criterion for impulsive delays Markovian jumping time-delays reaction-diffusion BAM neural networks via Gronwall-Bellman-type impulsive integral inequality201520151118585410.1155/2015/1858542-s2.0-84939474266SongQ.CaoJ.ZhaoZ.Periodic solutions and its exponential stability of reaction-diffusion recurrent neural networks with continuously distributed delays2006716580MR218379810.1016/j.nonrwa.2005.01.0042-s2.0-24944581606RaoR.PuZ.ZhongS.LiX.Fixed points and exponential stability for impulsive time-delays BAM neural networks via LMI approach and contraction mapping principle201620168315468310.1155/2016/31546832-s2.0-84994716297LiY.MiY.MuC.Properties of positive solutions for a nonlocal nonlinear diffusion equation with nonlocal nonlinear boundary condition20143437487582-s2.0-8489965744010.1016/S0252-9602(14)60046-1RaoR.ZhongS.Stability analysis of impulsive stochastic reaction-diffusion cellular neural network with distributed delay via fixed point theory201796292597MR3709595GalaktionovV. A.On asymptotic self-similar behaviour for a quasilinear heat equation: single point blow-up199526367569310.1137/s0036141093223419MR1325909Zbl0828.35067RaoR.ZhongS.Existence of exponential p-stability nonconstant equilibrium of markovian jumping nonlinear diffusion equations via ekeland variational principle201520151081215010.1155/2015/8121502-s2.0-84936806569RaoR.ZhongS.PuZ.On the role of diffusion factors in stability analysis for p-Laplace dynamical equations involved to BAM Cohen-Grossberg neural network2017223546210.1016/j.neucom.2016.10.036DingJ.GuoB.-Z.Global existence and blow-up solutions for quasilinear reactiondiffusion equations with a gradient term20112469369422-s2.0-7995245396110.1016/j.aml.2010.12.052RaoR.PuZ.Stability analysis for impulsive stochastic fuzzy p-Laplace dynamic equations under Neumann or Dirichlet boundary condition20132013, article no. 1332-s2.0-8488415707410.1186/1687-2770-2013-133Zbl1294.35178XiaY. H.YangZ.HanM.Lag synchronization of unknown chaotic delayed yang-yang-type fuzzy neural networks with noise perturbation based on adaptive control and parameter identification20092071165118010.1109/tnn.2009.20168422-s2.0-67949121856XiaY.YangZ.HanM.Synchronization schemes for coupled identical YANg-YANg type fuzzy cellular neural networks2009149-103645365910.1016/j.cnsns.2009.01.028MR2509952HeD.XuD.Attracting and invariant sets of fuzzy Cohen-Grossberg neural networks with time-varying delays20083724770577062MR246898610.1016/j.physleta.2008.10.035Zbl1227.920012-s2.0-55549126781LiuY.TangW.Exponential stability of fuzzy cellular neural networks with constant and time-varying delays20043233-4224233MR205048610.1016/j.physleta.2004.01.064Zbl1118.814002-s2.0-1542345660AhnC. K.Delay-dependent state estimation for T-S fuzzy delayed Hopfield neural networks201061348348910.1007/s11071-010-9664-z2-s2.0-77956620204SongQ.CaoJ.Impulsive effects on stability of fuzzy Cohen-Grossberg neural networks with time-varying delays20073737337412-s2.0-3424905350210.1109/TSMCB.2006.887951ZhuQ.LiX.Exponential and almost sure exponential stability of stochastic fuzzy delayed Cohen-Grossberg neural networks2012203749410.1016/j.fss.2012.01.0052-s2.0-84862868271CohenM. A.GrossbergS.Absolute stability of global pattern formation and parallel memory storage by competitive neural networks198313581582610.1109/TSMC.1983.63130752-s2.0-0020970741NguyenA.-T.LaurainT.PalharesR.LauberJ.SentouhC.PopieulJ.-C.LMI-based control synthesis of constrained Takagi-Sugeno fuzzy systems subject to L2 or L∞ disturbances201620779380410.1016/j.neucom.2016.05.0632-s2.0-85006354868MrquezR.GuerraT. M.BernalM.KruszewskiA.Asymptotically necessary and sufficient conditions for Takagi–Sugeno models using generalized non-quadratic parameter-dependent controller design2017306486210.1016/j.fss.2015.12.0122-s2.0-84952058904XuS.SunG.SunW.Takagi–Sugeno fuzzy model based robust dissipative control for uncertain flexible spacecraft with saturated time-delay input20176610512110.1016/j.isatra.2016.10.0092-s2.0-85005847957DingX. M.XuZ. K.CheungN. J.LiuX. H.Parameter estimation of Takagi-Sugeno fuzzy system using heterogeneous cuckoo search algorithm201515131332134210.1016/j.neucom.2014.10.0632-s2.0-84918557978ChoiH. D.AhnC. K.ShiP.LimM. T.SongM. K.L2-L∞ Filtering for takagi-sugeno fuzzy neural networks based on wirtinger-type inequalities20151531171252-s2.0-8492226694910.1016/j.neucom.2014.11.046WangX.RaoR.ZhongS.LMI approach to stability analysis of Cohen-Grossberg neural networks with p-Laplace diffusion201220121252381210.1155/2012/5238122-s2.0-84872808639LiX.RakkiyappanR.BalasubramaniamP.Existence and global stability analysis of equilibrium of fuzzy cellular neural networks with time delay in the leakage term under impulsive perturbations20113482135155MR277183310.1016/j.jfranklin.2010.10.009Zbl1241.920062-s2.0-79251598420MuralisankarS.GopalakrishnanN.Robust stability criteria for Takagi-Sugeno fuzzy Cohen-Grossberg neural networks of neutral type20141445165252-s2.0-8490605877910.1016/j.neucom.2014.04.019BalasubramaniamP.Syed AliM.Stability analysis of Takagi-Sugeno fuzzy Cohen-Grossberg BAM neural networks with discrete and distributed time-varying delays2011531-215116010.1016/j.mcm.2010.07.028MR2739252LindqvistP.On the equation div(∣∇u∣p − 2) + λ∣u∣p − 2u = 01990109115916410.2307/2048375MR1007505RaoR.Delay-dependent exponential stability for nonlinear reaction-diffusion uncertain Cohen-Grossberg neural networks with partially known transition rates via Hardy-Poincare inequality201435457559810.1007/s11401-014-0839-7MR3227746RaoR. F.WangX. R.Infinitely many solutions for the resonant quasi-linear equation without landesman-lazer conditions2012324744752MR3026670ZhuQ.HuangC.YangX.Exponential stability for stochastic jumping BAM neural networks with time-varying and distributed delays201151527710.1016/j.nahs.2010.08.005MR2746644ChenZ.WangX.ZhongS.YangJ.Improved delay-dependent robust passivity criteria for uncertain neural networks with discrete and distributed delays20171032332YangR.GuoR.Adaptive finite-time robust control of nonlinear delay hamiltonian systems via lyapunov-krasovskii method2008202, article 1C1110.1002/asjc.1556RaoR.PuZ.ZhongS.HuangJ.On the role of diffusion behaviors in stability criterion for p-laplace dynamical equations with infinite delay and partial fuzzy parameters under dirichlet boundary value20132013894084510.1155/2013/9408452-s2.0-84893640702AhnC. K.Passive and exponential filter design for fuzzy neural networks2013238126137MR304844710.1016/j.ins.2013.03.004Zbl1320.930562-s2.0-84876830627LiX.WuJ.Stability of nonlinear differential systems with state-dependent delayed impulses20166463692-s2.0-8495179350410.1016/j.automatica.2015.10.002Zbl1329.93108ChengJ.ZhuH.ZhongS.ZhongQ.ZengY.Finite-time H∞ estimation for discrete-time Markov jump systems with time-varying transition probabilities subject to average dwell time switching201520257158210.1016/j.cnsns.2014.06.006MR3251516SongQ. K.ZhaoZ. J.Stability criterion of complex-valued neural networks with both leakage delay and time-varying delays on time scales201617117918410.1016/j.neucom.2015.06.0322-s2.0-84944515067LiX.DengF.Razumikhin method for impulsive functional differential equations of neutral type20171014149TianJ.XiongW.XuF.Improved delay-partitioning method to stability analysis for neural networks with discrete and distributed time-varying delays2014233152164MR321497110.1016/j.amc.2014.01.1292-s2.0-84894570244LiX.CaoJ.An impulsive delay inequality involving unbounded time-varying delay and applications20176273618362510.1109/TAC.2017.2669580MR3669488Zbl06777935