摘要
利用WNN(小波神经网络 )逼近未知函数 ,将未知离散非线性系统转化为一类参数化严格反馈系统 ,进而对变换后的系统给出一个避免过参数化的自适应反推控制器 ,并证明该控制器可保证在存在参数不确定性和函数不确定性的条件下 ,整个自适应系统的状态全局有界 ,同时也可保证系统的跟踪误差落在一个大小与不确定性成比例的紧集中 ,仿真结果表明该控制器具有较强的鲁棒性 。
The strict-feedback discrete-time nonlinear systems with unknown functions are firstly transformed into a class of parametric strict-feedback discrete_time nonlinear systems via a wavelet_based neural network (WNN) function approximator.Then a robust backstepping adaptive controller without overparameterization is proposed,which shows the global boundedness of the whole adaptive system with parametric and functional uncertainties.It can also ensure that the tracking error falls within a compact set whose size is proportional to the magnitude of the uncertainties.The results of simulation represent that the proposed controller has the robustness and can be applied to the different objects.
出处
《自动化技术与应用》
2003年第6期6-9,共4页
Techniques of Automation and Applications
