Parameter Tuning Of PID Controller Based On Hybrid Genetic Algorithm

PID control technology has developed for nearly eighty years. It played an important role in the past industrial production and now still plays a dominant role in the industrial production. The performance of PID controllers determines whether industrial equipments run efficiently and stably. It can also affect the quality of products. In the past industrial production process, the parameters' tuning of the PID controller mainly depended on the manual mode. This approach requires not only specialized technical personnel, and it is more time-consuming. The effect of this approach is also not ideal. With the development of modern industry, the traditional turning method has clearly lagged behind. Therefore, in this context, a large number of parameters' auto-tuning method of PID controller has begun to appear. This paper is to design a fast and reliable parameters' tuning method of the PID controller.Genetic algorithm is a new optimizing searching method based on biology evolutionary theory. The basic principle of genetic algorithm and its basic operation process are firstly studied. Then the disadvantages of the basic genetic algorithm are analyzed. For example, it is easy to premature in the early times and has a slow evolution speed in the later times. Several methods are used to improve the performance of the basic genetic algorithm in ths paper. Those methods are protecting the best individual, employting adaptive crossver and mutation operators and using coordinate rotation algorithm to strengthen the local searching ability in the later times. By applying the basic genetic algorithm and the improved genetic algorithm to the optimizing of parameters of several kinds of system's PID controller, better performances of control system are achieved. So the results prove that the feasibility and availability of genetic algorithm in the optimizing of PID controllers' parameters. Then by applyting this basic algorithm and improved genetic algorithm to AC servo motor, comparision result shows that the modified algorithm has good convergence.In the end, a conclusion is made based on summarizing all the work done by the author. Also, we have some problems need to settle in the future is indicated.