Neural Network Based Decoupling Motion Control

Against a variety of uncertainties and nonlinearities of the neutral buoyancy plant motion control, this paper proposed a multichannel control algorithm based on neural network decoupling and CMAC. Based on the online learning function of neurons, a simple neural network optimization compensation method is designed. Combined with the advantages of CMAC feedforward control and PID feedback control, a new controller is composed. By MATLAB simulation, the proposed algorithm can effectively control the position and attitude of the experimental plant. The adjustment time is short and the steady state precision is high, and the effect is much better than PID control.