A mathematical model of engine throttle as the controlled object is established and then the neural network algorithms and PID control are combined. With the self -learning function of the neural network, self -tunings of PID parameters are realized. The method overcomes disadvantages of PID as parameters which are difficult to determine and embodies better intelligence and robustness of the neural network, the simulation is researched by Matlab and the results show that the PID neural network controller is more accurate and adaptive than conventional PID.
Paper
Full text
Application of PID Neural Network Controller on Engine Throttle System
Semantic Scholar · Engineering · 2012
Abstract
A mathematical model of engine throttle as the controlled object is established and then the neural network algorithms and PID control are combined. With the self -learning function of the neural network, self -tunings of PID parameters are realized. The method overcomes disadvantages of PID as parameters which are difficult to determine and embodies better intelligence and robustness of the neural network, the simulation is researched by Matlab and the results show that the PID neural network controller is more accurate and adaptive than conventional PID.