Algorithm Analysis Of Speed Control On Marine Diesel Engine

With the rapid development of modern society, energy saving and environmental protection as an important method of sustainable development is one of the key for the shipping industry. Diesel engine as its main power has a very important role in economy and emission performance, and speed control system design in marine diesel engine is one of the key points of fuel consumption and pollutant emission control. In view of this, doing research on marine diesel engine speed regulation system modeling and controller design is imperative.Firstly, according to the principle of the diesel engine speed control system, and based on the "linearization" method of constructing a mathematical model of speed control system, it is necessary to analyse the effect of parameters on the system control performance. Then, concentrate on the mainstream intelligent algorithm and focus on the neural-network control, fuzzy-control and expert-control through combining them with the traditional PID control to design a new type of intelligent controller and give an analyze of the contract between different design plans. Finally, use the MATLAB simulation software to build GUI graphical interface. Furthermore, measure and compare the intelligent controller and the conventional controller, and try to improve the impact of algorithm on the theory.As the research shows that the intelligent PID controller is superior to the traditional PID controller on the control effect, but different intelligent controllers have their own pros and cons. Among them, the fuzzy-adaptive PID controller has a good performance on response while the RBF-network-adaptive PID controller has the minimal overshoot and the expert-adaptive PID controller has the best tracking performance. So, combine the fuzzy-control and the expert-control in order to make a new intelligent controller which called the fuzzy-expert controller. And the simulation results verify that its performance is better than fuzzy-control and expert-control, and it also prove organic combination between intelligent-algorithms is feasible, which makes some useful support for the development of the comprehensive intelligent controller.