一类多时滞随机神经网络的自适应指数同步

Adaptive Exponential Synchronization for A Class of Stochastic Neural Network with Multiple Time-Delay

利用多时滞随机神经网络的动力学特性,以驱动系统与响应系统同步结构为框架,设计了误差系统的均方指数稳定的准则.采用随机理论和优化方法得到参数自适应更新律.研究表明:当时间趋于无穷大时,可得误差系统的状态轨迹趋于零,即实现同步.多时滞神经网络得到的自适应同步准则,与线性矩阵不等式方法得到的条件非常不同,仿真算例验证了该算法的可行性与有效性.

Taking advantage of dynamic characteristics of stochastic neural network with multiple time-delay,based on the synchronization frame of driving system and response system,the criterion of mean square exponential stability for error system was designed. According to the stochastic theory and optimization method,adaptive update was conducted by obtained parameters. When time tends to infinity,state trajectory of the error system tends to zero,thus the synchronization was realized. There is a big difference in the criterions of adaptive synchronization between neural network with multiple time- delay and linear matrix inequality（ LMI）. A simulation example finally was provided to demonstrate the feasibility and effectiveness of the proposed design method.