液压伺服系统的直接自适应神经网络控制

针对液压伺服系统中的非线性和不确定特性，研究了一种基于神经网络的直接自适应控制方法。引入的神经网络模型可以通过学习从而跟踪对象的动力学特性，控制器的设计较少的依赖于对象的先验知识，控制器参数的调整是基于被控系统的测量信号，利用在线辨识的神经网络参数来实现的。仿真结果证明该系统有较好的控制效果。

Directly self-adaptive control using neural network and its application in hydraulic servo system

A hydraulic servo system often has nonlinearity and incertitude. Based on neural network, we developed a directly self-adaptive control method in this paper. The NN applied in the system can identify the dynamic characteristic of hydraulic system. The parameter regulation of controller is inplemented based on the measured signal of the controlled system, and using the parameter of on-line identification NN. The emulate result proves that this method has better control effect.