基于神经网络的非线性系统内模控制的渐近完全控制

The Asymptotic Full Control to Internal Model Control of Nonlinear Systems Using Neural Networks

对一类非线性系统首先讨论了系统模型的逆的结构，然后讨论了在神经网络固有逼近误差存在的情况下，系统的稳定性问题，最后研究了系统完全控制的渐近特性，得出了给定输入情况下系统稳态输出的上界，仿真结果验证了关于系统渐近完全控制性质的结论．

In this paper,some qualitative properties of internal model control of nonlinear systems are discussed.First,it is dealt with structure of inversion of the systems model to some kinds of nonlinear systems,then the stability of the systems is discussed under the condition of the neural networks existing intrinsic approximation error.Lastly,it is researched the asymptotic properties of the systems full control and the upper bound of steady stae output of the systems is obtained,the conclusion on properties of asmptotic full control of systems are verified through results of simulation .