【摘要】 神经网络以其可学习的特性和高度的并行结构成为自控界的热门研究课题 它能以任意精度逼近任意连续非线性函数 ,能同时处理定量和定性的信息 ,能很好的协调多种输入信息关系 ,适用于多信息融合和多媒体技术 它为解决不确定性系统的许多控制问题及高度非线性问题提供了一个有力的工具 介绍了网络用于自动控制系统的主要特征 ,重点阐述了神经网络在系统辨识和模糊控制系统中的应用 ,并讨论了发展概况及趋势更多还原

【Abstract】 Neural networks are widely used in solving many highnonlinear control porblems because of their massive parallel structure and leavning capability. It can approximate to random continuous non-liner function with any precision, dispose the quantitative and qualitative information at the same time, and reconcile the relation of manifold input information. It apply to amalgamation of manifold information and multimedia technology. It’s a powerful instrument to resolve many control and highly non-linear problems of uncertainty system. This paper introduces the main features of the neural network, and presents the neural network application in system identification and fuzzy control system. Some trends of neutal network is to be discussed.更多还原