Less conservative results of H∞ filtering for neural networks with time-varying delay

A more effective Lyapunov functional has been constructed to investigate the H∞ filtering problems for a class of neural networks with time-varying delay. By combining with some inequality technic, the delaydependent conditions have been proposed such that the filtering error system is globally asymptotically stable with guaranteed H∞ performance. The time delay is divided into several subintervals, more information about time delay is utilised and less conservative results can be expected. Examples and simulations have been provided to illustrate the less conservatism and effectiveness of the designed filter.