This paper investigates the problem of \(l_{2}\)–\(l_{\infty }\) asynchronous filtering for a class of discrete-time fuzzy neural networks subject to Markov jump parameters and unreliable communication links. Due to the fact that neural networks possess the nonlinear dynamic characteristic, it is difficult to deal with such a nonlinear characteristic directly, so the Takagi–Sugeno fuzzy model is introduced to approximate the system. Directed against the unreliable communication links, the data packet loss depicted by a stochastic variable with Bernoulli distribution and the signal quantization phenomenon occurring in communication channels are taken into consideration simultaneously. The attention of this paper is mainly centered on devising an asynchronous \(l_{2}\)–\(l_{\infty }\) filter for ensuring the \(l_{2}\)–\(l_{\infty }\) performance of the studied system under asynchronous conditions. Some sufficient conditions for the existence of the asynchronous \(l_{2}\)–\(l_{\infty }\) filter are presented. Finally, a numerical example is given to carry out the simulation experiment, which can verify the effectiveness of the obtained results.

本文研究了一类离散马尔可夫跳跃参数和不可靠通信链接的离散时间模糊神经网络的\（l_ {2} \） – \（l _ {\ infty} \）异步过滤问题。由于神经网络具有非线性动态特性，因此很难直接处理这种非线性特性，因此引入了Takagi–Sugeno模糊模型来对该系统进行近似。针对不可靠的通信链路，同时考虑了具有伯努利分布的随机变量描述的数据包丢失和在通信通道中发生的信号量化现象。本文的注意力主要集中在设计异步\（l_ {2} \）上– \（l _ {\ infty} \）过滤器，用于确保所研究系统在异步条件下的\（l_ {2} \） – \（l _ {\ infty} \）性能。给出了存在异步\ （l { {2} \） – \（l _ {\ infty} \）过滤器的一些充分条件。最后给出了算例进行仿真实验，验证了所获得结果的有效性。