In order to solve the problems like the large dynamic overshoot in the control of transient air-fuel ratio of aero piston engine
the model predictive control strategy based on improved particle swarm optimization algorithm and Elman(VPSO-Elman) neural network was proposed. The engine model used in the simulation was established based on a real engine in AMESim. The model predictive control system was established in MATLAB/Simulink. Coupling simulation was carried out to test the effect of the model prediction control. The simulation result shows that
compared with proportional integral derivative (PID) control
using VPSO-Elman network model prediction control the air-fuel overshoot can be reduced by 20%
callback time could be reduced by about 75%; for different throttle opening rate
VPSO-Elman controller also has a good control effect.