Reinforcement Learning Control for Ship Steering using Recursive Least-Squares Algorithm

Authors:	Shen Zhipeng, Dalian Maritime University,China, China Guo Chen, Dalian Maritime University,China, China Yuan Shichun, Dalian Maritime University,China, China
Topic:	1.2 Adaptive and Learning Systems
Session:	Applications of Adaptive and Learning Control
Keywords:	Recursive squares methods; Learning algorithm; Action network; Ship control ;Simulation

Abstract

Recursive least-squares temporal difference algorithm (RLS-TD) is deduced, which can use data more efficiently with fast convergence and less computational burden. Reinforcement learning based on recursive least-squares methods is applied to ship steering control, as provides an efficient way for the improvement of ship steering control performance. It removes the defect that the conventional intelligent algorithm learning must be provided with some sample data. The parameters of controller are on-line learned and adjusted. Simulation results show that the ship course can be properly controlled in case of the disturbances of wave, wind, current. It is demonstrated that the proposed algorithm is a promising alternative to conventional autopilots.