Ship Steering Based On Immune Genetic Algorithm

The most widely used control method of modern ship autopilot is PID control. It is difficult to adjust the PID parameter, because the ship motion is a great inertia, large delay and nonlinear control system. The adaptive immune genetic algorithm is an improved immune genetic algorithm, which through improving coding method and selection, crossover and mutation operators. The adaptive immune genetic algorithm can avoid local optimum and accelerate convergence speed, and it has a reference significance to optimizing PID parameter for ship course control.The adaptive immune genetic algorithm is based on the immune genetic algorithm, which through improving coding method and selection, crossover and mutation operators. This adaptive immune genetic algorithm implements the selection operation and individual concentration adjusting through selection operator which is based on similarity and vector distance. This allows the diversity of the population to maintain a high level, thus avoiding local optimum. In order to get better results, the scope of the next generation is based on the individual fitness of previous generation. This makes the offspring individual evolution for better and accelerate the convergence rate. The mutation operator uses an adaptive mutation strategy, which can avoid late stage of evolution have a greater oscillation and make the population tends to converge. A series of experiments results demonstrate that the immune genetic algorithm proposed has better global convergence effect and speed.Finally, using adaptive immune genetic algorithm to optimize PID controller parameter, and compare with manual tuning and genetic algorithm with elitist. Simulation result shows that PID control for ship’s course steering based on the adaptive immune genetic algorithm has shorter response time, no overshoot, more robust and better control effect than the standard genetic algorithm PID controller.