Marine Diesel Engine Host Real-time Simulation System Modeling And Speed Control

In the past, three methods were generally used for modeling the marine diesel engines: linear, volume, and quasi-stationary state, which are all on the basis of the mathematic modeling of the engines. But the main diesel engine systems are so complex that a lot of phenomena (such as the oil supporting system, mechanical efficiency of diesel engine) can not be described by the mathematic models. The mathematic model can't recognize the large amount of fuzzy messages in the phenomena courses. So, it is necessary to set up a kind of intelligent model of engines to deal with these fuzzy messages, and make the engine get more effective control.The neural network is an essentially non-linear system, and has a highly parallel processing structure and some networks can be built by using the computer hardware. They have an ability to study and adapt themselves and can deal with the qualitative and quantitative data and the multi-variable systems, etc. The main content of this text is to set up a kind of modeling system for main diesel engine based on RBF neural networks.It is quite difficult for the traditional PID to control and adjust parameters (Kp, Ki, Kd) exactly online simultaneously. The usual way of adjusting parameters is to calculate out the initial values of Kp, Ki and Kd through the actual parameters of real ship's engine, then debug and revise the three parameters to achieve the goal of the controlling. This method is difficult to rectify parameters automatically, and could only choose 1-2 groups ofrelatively good parameters from the history, and isn't possible to completely fulfill the optimum control in all conditions.How to adjust the PID parameters exactly online simultaneously? The way of using Fuzzy Controlling with PID is the main searching direction. Following the dynamic changes of the parameters (Kp, Ki, Kd) by using the Fuzzy Control, we can get a kind of new fuzzy PID controller. In this article the ideal of fuzzy PID parameter controller is based on this.This text is under the research of the Project about the intelligent control and simulation of the marine main diesel engine system, which mainly concerns about the following: The close study about the algorithms of the RBF neural network, and the model establishing of the main diesel engine based on them. The designing of the structure of such an intellectual model, and the verifying to its imitated results. The research of the principles of the general PID controller, and the method of improving on it's shortcomings at its parameter adaptive capacity and bad robust quality. The study of the principles and the algorithms about the fuzzy controller, the designing and testing of a new controller based on the Combining of the fuzzy control and the PID.