The Study On Intelligent Control And Simulation Of Marine Systems

In this thesis, the intelligent control and simulation for some marine systems have been discussed in detail as follows. The approaches include the modeling, intelligent control and simulation for main diesel marine engine, the designing and development of a single-chip computer based synchronous-meter utilized in ship power station simulation of marine engine room simulator DMS-2000, the software interface of simulation to the ship main engine remote control system AotoChief-4 and the hardware and software development for the ship autopilot control system.The ship main engine remote control system is one of the most important automatic systems in the modern ships, and the main engine speed regulating system is a key part in it. The control algorithms of both traditional and intelligent main-engine speed regulator are emphatically discussed in the thesis. The author established an simplified integrated mathematical model for marine main diesel engine, screw propeller and axis series with the mother type of large low speed diesel MAN B&W. The digital simulations for PI speed regulator are completed. Aim at overcoming the disadvantages of conventional PID control algorithm, which adopts fixed-parameters, author introduces a variable parameters PID speed regulator which gives much improvement on anti-disturbance. This control system has fast respondence when increasing or decreasing speed at emergency work conditions.Referencing previous research work, author also proposed three intelligent control algorithms for marine controlled plants. Based on improvement of standard fuzzy controller, an integrated intelligent algorithm combining fuzzy logic with integral function and adaptive parameters is presented. The simulation results show it has better performance than conventional PID controller. Another fuzzy-neural hybrid intelligent controller, which can make better use of fuzzy logic and the neural networks is given. During developing the controller, the author uses an alterable-slope BP algorithm for overcoming the shortcomings of potential local minimum-value, the BP algorithm adopted can properly finish the learning process.The thesis also introduces the author's research work including implementation of a single-chip system based synchronous-meter successfully utilized in marine engine room simulator DMS-2000 which belongs to one key "211 Project" of DLMU: "Engine Room Simulator Adopted Virtual Reality"; the realization of software interface on "AutoChief-4 Ship Main Engine Remote Control System"; and the designing and realization of the hardware and software of autopilot for "Ship Hybrid Intelligent Control System".