Research On Intelligent PID Control Based On GA & FNN And Its Application

In numerical control machine, the position servo system links the computer numerical controller and the mechanical transmission equipment. So its performance is very important to the performance of the numerical control machine. Such as it can affect the fast ratio of the machine, position precission, the quality of surface of the manufacture, productivity, reliability and etc. At the present,we usually used conventional PID controller because its algorithm is every simple and it has strong robustness and fine reliability. And especially, it is suitable to the system that we can make its accurate mathematical model. However, with the development of industry, the manufacturing process is becoming more and more complicated. Fist, it is difficult to make the precise mathematical model of object. And the conventional PID controller can't adjust itself to the variation of the surroundings and the object. So we must find a new method to design the parameters of the PID controller.Fuzzy logic control and neural network have many advantages which conventional PID control doesn't have. They don't need the accurate mathematical model of object, and have fine ability of self-studying and adapting. Genetic algorithm is a kind of random searching optimal algorithm based on biological evolution. It can find the best result by reproduction, crossover and mutation. So we come up with a new method to design the parameters of PID controller. This kind of controller has the advantages of conventional PID controller, fuzzy logic controller and neural controller. The simulation curve on position servo system of numerical control machine shows that it improves the ability of self-studying, adapting and anti-jamming of the system, and betters the robustness.