Optimization of Engine Speed Neural Network PID Controller Based on Genetic Algorithm

Engine-Dynamometer system is a two-input, dual output system with nonlinear, time-varying characteristics of large inertia, and exists coupling within the system input and output. Using the traditional PID controller, the control is often difficult to achieve the desired effect. In addition, at the production site, because of being cumbersome and precision tuning effects, the traditional PID parameter tuning methods can lead to poor control of engine speed. In this paper, an engine speed neural network PID controller based on genetic algorithm which use genetic algorithm to optimizate three control parameters of neural network PID and achieving the system input and output decoupling control is studied. Simulation results show that the genetic algorithm optimizating engine speed neural network PID control system can effectively improve the accuracy, enhance stability and fast of the system, and also have increased the engine speed control effect.