Research on PID control algorithm based on RBF neural network optimization
In order to better solve the problems of poor precision,no adaptability and poor following performance of conventional PID control,the combination of RBF neural network and conventional PID control algorithm can realize dynamic identification.Using the learning ability of neural network,the proportional,integral and differential parameters of PID control can be modified online ac-cording to the control environment to make it more in line with the adjustment requirements.In this way,the real-time performance and adaptability of the system can be improved.By adding two different signals,step signal and sine signal,the control system is sim-ulated based on Simulink environment in Matlab software,and the control performance of PID control algorithm based on RBF neural network is verified.Through the simulation results of the control system,it is concluded that the PID control algorithm based on RBF neural network has the advantages of fast response speed,small overshoot,good following performance and no static deviation,and its control effect is obviously superior to the conventional PID control algorithm.
conventional PID controlsimulink simulationPID control based on RBF neural network
万方数据
维普
