Robust passivity analysis of uncertain neutral-type neural networks with distributed interval time-varying delay under the effects of leakage delay

Abstract

This paper deals with the problem of delay-range-dependent robust passivity analysis of uncertain neutral-type neural networks with distributed interval time-varying delay under the effects of leakage delay. The uncertainties under consideration are norm-bounded uncertainties and the restriction on the derivative of the discrete and distributed interval time-varying delays is removed, which means that a fast interval time-varying delay is allowed. By applying a novel Lyapunov-Krasovskii functional approach, improved integral inequalities, Leibniz-Newton formula and utilization of zero equation, then a new delay-rangedependent passivity criterion of neutral-type neural networks with distributed interval time-varying delay under the effects of leakage delay is established in terms of linear matrix inequalities (LMIs). Furthermore, some less conservative delay-dependent passivity criteria are obtained. Moreover, we derived a robust passivity criterion for uncertain neutral-type neural networks with distributed interval time-varying delay under the effects of leakage delay. Besides, a less conservative delay-dependent robust passivity criterion is obtained. Finally, five numerical examples are given to show the effectiveness and less conservativeness of the proposed methods.

Novel H ∞ performance and delay-dependent exponential passivity for neural networks in response to leakage delay

“…It can be seen that our results outperform those found in [9,18,[25][26][27]. Method h d = 0.8 h d = 0.9 Unknown h d [45] 0.8643 0.8344 0.8169 [14] 0.8696 0.8354 0.8169 [15] 0.8784 0.8484 0.8217 [11] 0.8841 0.8570 0.8260 [39] 1.0214 Method α = 0.5 α = 1 α = 1.5 [26] 2.5900 0.9700 0.3500 [9] 2.8200 1.1800 0.5400 [25] 2.9000 1.3200 0.7200 [18] 2.9400 1.3500 0.7200 [27] 6.6021 3.6051 2.5198 Theorem 3.3 19.6010 8.9010 6.0020…”

“…A comparison of the exponential convergence rates of system (4.1) utilizing different approaches is shown in Table 2. It is evident that our results surpass those from [11,14,15,27,39,45]. Table 3 shows a comparison of the allowable upper bound of h M for various α and h d = 0 in Example 4.4.…”

An Adaptive Zeroing Neural Network with Non-Convex Activation for Time-Varying Quadratic Minimization

“…This not only improves the robustness of the model but also results in faster convergence and higher accuracy than previously proposed zeroing neural network models. The paper proposes four theorems and provides corresponding proofs based on the Lyapunov stability theory [35,36] to analyze the global convergence and robustness of the AZNNNA model under different noise interferences. Numerical experiments are also conducted to demonstrate the advantages of the proposed AZNNNA model.…”

Genetic Approach for Joint Transmission Grouping in Next-Generation Cellular Networks

“…Notably, we consider equal power allocation which can be replaced by any power allocation algorithm. Some studies investigated adaptive feedback schemes in neural networks [ 26 , 27 , 28 ].…”