Ship Course Intelligent Fuzzy Control System And Its Simulation

The control of ship course is the basic control of ship control,which is the symbol of the property of ship steering. But the conventional PID autopilot systems are sensitive to the high-frequency disturbances,easy to lead up operating rudder frequently,and lack of adapting to the ship dynamic property and the changes of sea environment. In addition,the mathematical model of ship is very hard to build up because of the complexity of ship course control system and the randomicity of working environment. On the contrary,fuzzy control don't depend on the mathematical model of controlled objects and have the ability to adapt to the changes of system parameters,so it has the potential to overcome the drawbacks of conventional control schemes. The paper is developed under such engineering background.Fuzzy control is a kind of computer numerical control based on fuzzy set theory,fuzzy linguistic variable and fuzzy logical reasoning,in which the experience of skilled operators and the knowledge of experts are summed up into the control rules in the form of "IF...THEN...",that is fuzzy logical sentence,and then the control results can be attained by fuzzy logical reasoning. This procedure is similar to the ideation of human being.In this thesis,according to nonlinear,time-varying and time-delayed parameters of ship course,the simple fuzzy controller was first designed based on the analysis of the fuzzy control theory,which is the basis of other high-level controllers. This kind of controller is easy to design and operate,and has improved convergence rates and less overshoot than PID controller,but has stable error. In order to improve the properties of the fuzzy controllers,a fuzzy-Pi controller and a neural-fuzzy controller were designed respectively based on the simple fuzzy controller. The former tunes the PI parameters on-line by using fuzzy set theory to improve the self-adaptive ability and to realize non-error control. The latter remembers the control rules by nonlinear mapping and self-learning of the neural network. By BP network's off-line training and on-line self-learning,the controller has the self-tuning and self-adaptive ability,that is,has fine robustness and real-time control effect.At last,the simulation curve was given,and which show that the desired results were attained.