Ship Autopilot System Design and Testing on Santana Ship Model Based on Neural-Fuzzy Method

Abstract

This paper presents the design of Ship autopilot system based on the output feedback neural - fuzzy (ANFIS) controller in 3-axis coordinate system which controls a heading of the ship under the effect of the waves, wind and water currents. The inputs of system include the feedback heading of the ship, angle of the waves, wind and water currents, and then the outputs are a rudder angle and the heading. In order to increase the adaptability of the system under the effect of the environment, ANFIS controller is designed using of neural-fuzzy with the Takagi – Sugeno method that approximate the input signals and linearized the output signal. The simulation results via Matlab software and the experimental results on Santana ship model showed the effectiveness and advantages of the proposed method.