Modeling and Optimization Control for Aircraft AC Generator Brushless Excitation System Based on Improved Adaptive PSO

ABSTRACT

The brushless excitation system of aircraft AC generator is a strong coupled and nonlinearity dynamic system which is often being subjected to disturbances. Therefore, the conventional PID controller is unable to meet the brushless excitation system control requirements of More Electric Aircraft (MEA) or All Electric Aircraft (AEA). A new brushless excitation compound control controller (RBFPID) is proposed in this paper based on radical basis function (RBF) neural networks and the conventional PID control. Because the new brushless excitation compound controller (RBFPID) has a number of mutually coupled parameters that needs to be set, the improved adaptive particle swarm optimization (APSO) algorithm is used to optimize mutually coupled PID parameters Kp , Ki , Kd and RBF parameters η ,α , m, n on line. In order to validate performances of the new brushless excitation compound controller based on multi-parameter optimization by the improved APSO, the simulation model of the aircraft brushless excitation system is implemented in MATLAB/SIMULINK according to differential equations of each component of brushless excitation system. The simulation results show that the optimized adaptive compound excitation controller (APSORBFPID) exhibits quick response speed, short adjustment time and high steady state accuracy.

Keywords:

Aircraft generator, brushless excitation system, modeling and optimiztion control, PSO, RBF neural network.