1 INTRODUCTION
Hopfield neural network(HNN)[1], [2] is introduced in 1982 and 1984. It has been studied extensively owing to its important applications in associative memory and optimization computation. It is well known that stability is prerequisite in the applications and design of neural networks. However, time delays inevitably exist in biological and artificial neural networks due to the finite switching speed of neurons and amplifiers and communication time, and they may lead to oscillation and instability of networks. In recent years, a number of stability conditions for neural networks with time delays have been proposed [3]–[12]. On the other hand, in the design and hardware implementation of neural networks, vital data such as the neurons fire rate, the synaptic interconnection weight and the signal transmission delays, etc., usually need to be measured, acquired and processed by means of statistical estimation which definitely leads to estimation errors. Moreover, parameter fluctuation in neural network implementation on very large scale integration (VLSI) chips is also unavoidable. Thus, a good neural networks should have certain robustness. The robust stability of neural networks with delays has been extensively investigated in recent years. In [5], [7], several sufficient conditions on asymptotic robust stability was derived for interval Hopfield neural networks with constant delays. In [10], the authors investigated the global robust stability of a class of interval Hopfield neural networks with time-varying delays. In [11], the authors studied the global asymptotical robust stability of multi-delayed interval neural networks by use of linear matrix inequality (LMI) technique and Lyapunov-Krasovskii functional method. In [12], a new sufficient condition was presented for the existence, uniqueness, and the global robust stability of equilibria for interval neural networks with time delays by constructing Lyapunov functional and using matrix-norm inequality.