The Internal Model Control Theory Based On Wavelet Networks And Implementation On Reheat Steam Temperature Control System

This paper points out merits of frequency domain analysis method based on transfer function. The architecture and classification of wavelet networks are summarized. Its function approximation property is analyzed. Then, the design problem of wavelet networks structure is studied deeply. An intelligent control system based on wavelets neural networks is used in the reheat steam temperature control. Based on inverse system method for nonlinear continuous system, internal model system is proposed.The steady performance of the control system is analyzed. Some simulation examples are given to demonstrate the effectiveness of the method. For nonlinear SISO internal model control system, the sufficient conditions of the plant inverse model existence is given, the design methods is studied. A kind of adaptive inverse control arithmetic presented, which overcomes the intrinsic defect of a nonminimum-phase inverse plant mmodeling. The stability of system and convergence of arithmetic were analysed. Simulation results verify the effectiveness of the proposed method.