This paper addresses the global exponential dissipativity (GED) of impulsive recurrent neural networks (IRNNs) with proportional delays. By introducing some adjustable parameters, skillfully designing several Lyapunov functionals and utilizing matrix norm properties, serval delay-dependent GED criteria are developed, and global attractive sets (GAS) and global exponential attractive sets (GEAS) of the proposed system are given. These adjustable parameters are related to the exponential decay rate and contribute greatly to expand the attractive sets of this paper. Here the criteria proposed improve and extend the earlier global dissipativity criteria. Several numerical examples are used to verify the obtained results and show that the obtained results are independent of each other.

本文研究具有比例延迟的脉冲递归神经网络（IRNN）的全局指数耗散性（GED）。通过引入一些可调整的参数，熟练地设计几个Lyapunov函数，并利用矩阵范数性质，开发了依赖于服务时延的GED准则，并给出了该系统的全局吸引集（GAS）和全局指数吸引集（GEAS）。这些可调参数与指数衰减率有关，并极大地扩展了本文的吸引力。这里提出的标准改进和扩展了早期的全球耗散性标准。几个数值示例用于验证所获得的结果，并表明所获得的结果彼此独立。