| The GBB1004ball-beam system provided by Googol technology Co., Ltd. is one of the most famous laboratory equipments for studying control theory. Many classical and modern control methods can be verified by ball-beam system. In this thesis, the research on control method of ball-beam system based on genetic algorithm-fuzzy technology achieves satisfactory experimental results. The main work is as follows:Firstly, the thesis introduces the feature, composition and working principle of ball-beam system briefly. On this basis, the thesis establishes the math model of ball-beam system.Secondly, PID controller, which is most commonly used in industry, is designed to ball-beam system and simulated. Because PID control has some shortages itself, such as:it is far to meet the people’s requirement for high control quality when applied in nonlinear and uncertain systems; and its parameters can’t do the self-adjusting job on line. Aimed at the above shortages, in the thesis, fuzzy control is introduced to achieve the adjustment of PID parameters on line, that is, the fuzzy PID controller is designed to ball-bean and simulated. The simulated results verify that fuzzy PID controller can improve dynamic performance of the system compared with PID controller.Then, in the fuzzy controller, obtaining the fuzzy rules and adjusting the parameters, which is lack of systemic method to achieve, mainly depends on the experience of the control experts and the repeated experiments. However, with the improvement for the systematic complexity, the intuitive experience, which becomes more and more difficult to obtain, is not accurate in expression and hard to be used directly. Direct at the above questions, genetic algorithm is introduced to the fuzzy PID controller. In this thesis, quantitative factors and scaling factors of the fuzzy PID controller are optimized and simulated respectively by basic genetic algorithm and self-adaptive genetic algorithm. The simulated results verify that fuzzy PID controller based on self-adaptive genetic algorithm can improve not only dynamic but also static performance of the system compared with fuzzy PID controller based on basic genetic algorithm.After that, the experimental platform of ball-beam system is built, where the PID controller and fuzzy PID controller are introduced into the actual debugging and their control effects verify the simulated results.Finaly, the remainder of this dissertation summarizes the work done, and put forward the prospect of further research directions and problems. |
