Adaptive Iterative Learning Controller Design for a Class of Strict-feedback Time-Varying Nonlinear Systems

This paper elaborates the design of a new iterative learning control scheme for a class of strict-feedback high-order uncertain time-varying nonlinear system. A novel iterative learning neural network approximator (ILNNA) is firstly proposed to eliminate the time-varying uncertainties. Then, combining composite energy function (CEF), robust adaptive control and backstepping techniques, a new iterative learning control mechanism with both differential and difference updating laws is constructed. The learning control scheme can warrant a $L_{pe}$ boundedness of all state variables and a $L_{T}^{2}$ convergence of the output along the iteration axis in the presence of unknown time-varying parametric nonlinearities, needless of Lipschitz continuous assumption. Simulation studies are undertaken to illustrate the effectiveness of the proposed scheme.