The Research On Position And Speed Sensorless Control Method Of Brushless DC Motor

A novel mathematical model for Brushless DC Motor (BLDCM) is proposed to meet the requirements of practical analysis and control application. Based on classical phase voltage model, a line voltage model is derived and then the phase current model is obtained. Beside, the solution for this novel mathematical model is also given. A novel simulation model of BLDCM is built, by using MATLAB/Simulink PSB block, to help the characteristics analysis and promote the research of new control strategies.A novel method for speed and rotor position estimation of BLDCM,which applies extend Kalman filter (EKF), is presented in this paper. The rotor velocity and position can be taken as two states variable of the system by using the EKF on the basis of the theory of the BLDCM. Based on the measured stator currents and stator voltages, we can estimate the rotor velocity and position. It is very useful for us to study the application of the velocity and position sensorless operation of brushless DC motor drives. The simulation results have confirmed the feasibility of the presented technique.This paper analyses the producing reason of commutation torque ripple, introduces the reason of commutation torque ripple suppression techniques that are practically effective in low speed and high speed. A method for reducing commutation torque ripple generated in BLDCM using a single DC current sensor, which combines deadbeat current control scheme with commutation compensation techniques, is presented in this paper. Effectiveness of the proposed control method is verified through simulation.A speed-sensorless control method for BLDCM, which applies recurrent fuzzy neural network (RFNN), is presented in this paper based on the dynamic model of BLDCM. The RFNN controller is used as a speed controller to mimic the optimized output of the system. The simulation results show the good performance for the system by using network to adjust the parameters and the recurrent weight of neural network on-line dynamically on the condition of variety of system parameter and the impact of outside uncertainty factors.