Ship roll stabilization using supervision control based on inverse model wavelet neural network

Abstract

An intelligent supervision control for ship roll motion stabilization is presented in this paper. The control system consists of a linear PID controller based on ship fin stabilizer three parameters, an inverse model wavelet neural network (IMWNN) controller and a wavelet neural network identification (WNNI). The linear PID controller is adopted to carry out feedback control. The WNNI is used to learn the dynamics characteristics of the nonlinear system, and obtain a fast dynamics internal model. The IMWNN controller is employed to implement intelligent supervision control, and obtain the inverse dynamics model of the process. In simulation, the roll mode input-output signals to train the wavelet neural network off-line are obtained by introducing PRBS sequence. The simulation results illustrate the efficiency of the proposed method, and prove that the intelligent supervision control can make sure the stability and robustness of control system and effectively improve the system adaptive ability. This method can also be used in other complexity, non-linearity system control.