Analysis on AUV Heading Control System Based on RS - Chaos - LSSVM

Aimed at the submarine uncertainties and model’s impreciseness, an integrated RS-Chaos-LSSVM model is put forward because of nonlinearity and time-variability of AUV heading control system. The LSSVM is a typical nonlinear regression modeling based on small sample. Its parameters are optimized by Chaos algorithm. It is to gain the optimal model and get the higher accurate. By using Rough theory the monitor data attribute of AUV is reduced to eliminate the redundant information to improve efficiency. Taken a certain typed AUV as an example, the two controller compare results .The results show the RS-Chaos-LSSVM controller markedly reduce the time. And the heading error and control rudder are simulated in the ideal and ocean current environment. In the ideal environment the two controllers all have good performances. But in the ocean current environment the simulation results show the integrated controller improve the model’s stability and accuracy. The overshoot, the rise and adjust time reduced greatly. In a word the integrated RS-Chaos-LSSVM heading predictive controller has better performances especially strong anti-interfere capability. Copyright © 2013 IFSA.