RESEARCH ON THE HEADING CONTROL METHOD AND EXPERIMENT OF WAVE GLIDER

The wave glider is a typical nonlinear, strongly coupled, underdriven system. Conventional proportional integral derivative (PID) control algorithm, which has the disadvantages such as difficulty in parameter adjustment and inability to adjust online, hardly meets the requirements of high-precision heading control in complex and time-varying marine environment. In this paper, a back propagation (BP) neural network incremental PID control method based on the improved particle swarm optimization (IPSO) optimization is proposed for this problem. Firstly, a mathematical model for the wave glider is developed. Secondly, the self-learning ability of BP neural network is utilized to adaptively adjust the PID parameters. Meanwhile, to overcome the shortcomings of BP neural network such as high sensitivity to initial weights and easyness to fall into local extremes by back propagation, an IPSO algorithm is introduced to optimally select the initial weights of the network, which can ensure the BP-PID network to obtain the global optimal solution. Finally, simulations and sea trials are conducted and the results show that the proposed algorithm is feasible, and can significantly improve the heading control performance of the wave glider, which verifies the effectiveness and realism of the algorithm.