混沌最小二乘支持向量机的AUV航向预测控制

摘要

Aim to the uncertain factors of the AUV heading control, the heading control based on the Chaos-LSS-VM was presented. The Chaos was used on-line to optimize LSSVM parameters because the AUV is a nonlinear system. The LSSVM identifier was studied to build the model and the predictive controller has the high control accuracy. The predictive result of the Fuzzy-PID predictive controller is compared with the Chaos-LSSVM. The simulations show that the precision of model is increased, and it has better self梐dapting ability, and more control accuracy especially during the ocean current.%针对AUV航向控制中存在不确定因素,尝试提出一种混沌最小二乘支持向量机(Chaos-LSSVM)的航向预测控制方案.由于AUV是典型的非线性系统,利用LSSVM解决航向非线性建模问题,采用Chaos算法在线优化LSSVM模型参数,预测控制AUV的航向,保证了航向预测控制的精度.最后Chaos-LSSVM与Fuzzy-PID预测控制器仿真结果对比表明,文中方法有效地提高了模型预测控制的精确性,且对于有海流海浪干扰及模型参数摄动具有较好的自适应抗干扰能力.