Brushless Dc Motor Without Position Sensor Location Detection Technology

The brushless direct current motor (BLDCM) with its superior performance has been broadly applied in each field. It's important to check the rotor position in the controlling method of BLDCM, while the defects is becoming more and more obvious because of the usage of the position sensor of the traditional BLDCM. So the study of sensorless control has a better application prospect.First, this dissertation discusses the development of BLDCM and compares some general methods of the rotor position detection. And then it analyzes the structure and the work principle of BLDCM and deduces a mathematic model of BLDCM, then Matlab/Simulink is used to construct the simulation of BLDCM and its control system. The simulation model is built though the organic integration of the independent function model, also for the follow-up study laying foundation.Second, one of the methods of rotor position detection - BEMF zero crossing point detection is explained in detail. Based on lagging 30°—(?) commutation, the lagging 90°—(?) commutation method is brought forward to improve the traditional detection method, and designs a zero crossing detection circuit. Once the circuit parameters of detection circuit selected, the phase shift angle size of BEMF signal only depends on the frequency of BEMF, that is, depends on the speed of motor. We calculate the phase shift compensation angle in different speed, based on the analysis error of rotor position, and a correlation between motor speed and compensation angle is proposed, so the error can be amended simply in software.Finally, in this dissertation, the principle of position sensorless control for BLDCM is analyzed and a new position detection method for BLDCM is proposed. A wavelet neural network (WNN) model is built whose parameters are trained based on genetic algorithm. The input of WNN is phase current and flux linkage, the output of WNN is rotor position. By using Matlab/Simulink, the simulation results demonstrate that the model is feasible. It is unnecessary to consider the motor speed, and has a wider range of speed than the BEMF zero crossing point detection method.