一类非线性系统反馈线性化的遗传神经网络实现

利用Ｈｏｐｆｉｅｌｄ反馈神经网络对一类仿射非线性系统进行反馈线性化，然后利用常规的ＰＩ控制方法设计控制器，同时指出，利用神经网络不仅可以对系统的状态进行辨识，而且可以辨识其相对阶数，并给出了完整的证明，在训练神经网络时，提出了一直直接基于寻优参数的遗传算法ＤＰＧＡ，仿真结果说明了该线性化方法的有效性。

Linearising Feedback of One Class of Nonlinear System with Application of Genetic Evoloved Neural Network

One class of nonlinear affine systems are linearized by Hopfield neural network feedback and then controlled by a standard PI controller.In the meantime,it is proved strictly that not only the system states but also the relative degree can be identified by using the neural network.Furthermore,a direct parameter based genetic algorithem (DPGA) is presented to train the neural network.The simulation result shows that the proposed method is realizable and efficient.