Particle Swarm Optimization of Fuzzy Model Reference Learning Controller for Tanker Ship Steering

This paper discussed the implementation of Particle Swarm Optimization (PSO) to optimize a Fuzzy Model Reference Learning Controller (FMRLC) for tanker ship. FMRLC is developed by synthesizing several basic ideas from fuzzy set and control theory. It can achieve the heading regulation of tanker ship exposed to plant changes and disturbances by adjusting the rules in a direct fuzzy controller so that the overall system behaves like a “reference model”. However, the tuning of the fuzzy inverse model scaling gains is considered to be difficult and tedious due to the high nonlinearity of the ship dynamic model and the external disturbances. It is shown that PSO can provide a very promising technique for the design of FMRLC for its simplicity and ease of use. Moreover, Centroidal Voronoi Tessellation (CVT) is implemented to select the starting positions of the particles strategically. The promising results from the experiment provide direct evidence for the feasibility and effectiveness of PSO for the optimization of FMRL controller for tanker ship heading regulation. Key-Words: Particle Swarm optimization, autopilots, nonlinear optimization, tanker ship steering, fuzzy model reference learning controller.