Genetic-based Hybrid Intelligent Control For Ship Steering

This thesis has systematically researched several control strategies, such as fuzzy control, genetic algorithm, and intelligent integral control, and their applications to ship's course steering control.A novel efficient fuzzy controller has been proposed, after the presentation of fuzzy control theory. The controller employs simplified fuzzy inference method, to lessen the computational times. Not conventional storing rules but on-line rule computing saves the computer's memory. The increase of the fuzzy rules has little effect on computational load and memory expenditures. Meanwhile, highly precise and efficient fuzzy control scheme can be realized. The controller also overcomes the drawbacks of table-query method, i.e.. the existence of dead zone. Based on inverted pendulum on a cart, simulation research shows the effectiveness of the novel fuzzy controller. A course steering autopilot is designed, based on the above novel control approach. Simulation research shows that, the autopilot is featured by fast response and good robustness. However, similar to the conventional autopilot, there exists final offset, when external disturbances appear. Then, integral control, i.e. integrating course error and rudder angle, respectively, is fused into the above controller, and two kinds of fuzzy-PID autopilot are designed. Simulation researches show that both of them can effectively eliminate final offset. Furthermore, the fuzzy-PID autopilot, with integral item of rudder angle, shows smaller overshoot, and better performance.Genetic-based fuzzy-PID autopilot is proposed. Genetic algorithm is charactered by biological evolution and global optimization. Then, the shortcomings in fuzzy controller is successfully overcome, i.e. difficulties in manually regulation, arising from the complicate coupling in quantification factor, proportional factor, integral factor and fuzzy rules. Simulation research shows that, after genetic optimization, the performanceof the fuzzy-PID autopilot is greatly improved, and the control system is characterized by little overshoot, short rising time, and excellent robustness. A genetic-based hybrid intelligent control for ship steering is proposed. Traditional integral control has the drawback of blind integral, which contains unfavorable information to course steering. At the different stages, different control strategies are utilized, and the autopilot's parameters are optimized via genetic algorithm. Simulation research shows that such an autopilot is of good performance, small overshoot.To verify the performance of the autopilot design in this thesis, ship's mathematical model is researched, in consideration of the forces and moments from the environmental disturbances, such as winds, waves, currents, etc. By using Matlab's Simulink toolbox, simulation experiments are carried out on ship steering autopilot.