Study on application of neural network structural PID in electro-hydraulic injection timing actuator

To meet electronic control demand of electro-hydraulic injection timing actuator for heavy duty diesel engine, application of neural network structural PID (N-PID) in the actuator was developed. Work character and control demand of the actuator were analyzed, method of neural network structural PID was introduced to perform coefficient tuning of PID automatically. Closed-loop control for piston displacement in the actuator was performed, an un-sensitive angle of the actuator was presented to compensate errors of mechanical hardware and fulfill control for injection timing. Simulation comparison between N-PID and PID under Matlab/Simulink was made, control effect of N-PID was prior to that of general PID. The bench control test of the actuator was made, its control precision and maximum response time were 0.3 (°) CA and within 2 s respectively and control effect of N-PID was superior to that of PID. The simulated and tested results show that application of N-PID method in the actuator is available.