Research On Control Strategy For PM Brushless DC Motors

In allusion to the key problem on the control of BLDCM, the thesis discuss some methods about control of it in aspects of commutation of motor, closed-loop control and the disturbance restraint.The thesis first expounds the mathematic model and the basic running principle of BLDCM. Then one of the methods of rotor position detection-BEMF zero crossing point detection is explained in detail. In succession, the chapter dissertates the method of measuring phase voltage and supply voltage, and analyses the error of rotor position using this method. Based on "lagging 30 - a commutation ", the "lagging 90 - a commutation "method is brought forward to overcome the shortages existing in the traditional zero BEMF, then the rang of a is expanded and the reliability of commutation is increased.Secondly, in order to make the motor commutate exactly, a fuzzy neural network is used- to estimate the rotor position. Based on the relationship of voltage, current, flux linkage and corresponding position (), a FNN is proposed whose structure and parameters are trained based on improved genetic algorithm. Through measurement of the phase voltages and phase currents, the flux linkage can be figure out. The input of FNN is flux linkage and phase current, the output of FNN is rotor position. The FNN is trained off line with the genetic algorithm. Simulation results demonstrate that the method is feasible.When running in closed-loop system, the motor will be affect inevitably by load disturbance ?and disturbance due to variety of motor parameter, then the dynamic performance of closed-loop is affected. The precise mathematic model is not brought forward in this thesis. Because of disturbance uncertainty, a disturbance observer is used to estimate the equivalent disturbance of whole control system, and then the estimation of disturbance is compensated to system to counteract the disturbance of system. Simulation results indicate that system disturbance is restrained effectively.Finally, a DSP-based BLDCM control system is introduced, and the experimental results are presented.