Brushless DC motor sensor-less control method based on extreme learning machine classification

Abstract

The invention provides a brushless DC motor sensor-less control method based on extreme learning machine (ELM) classification, and aims at the problem of brushless DC motor rotor position detection. According to the sensor-less control based on the extreme learning machine, the brushless DC motor stator voltage and current act as the input of the extreme learning machine network, the rotor position information acts as the output, the DC motor rotor position is divided into six areas and each area is corresponding to conduction of the corresponding switching tube, i.e. the switching logic signal. The network parameters are determined through training of the ELM network and then the trained network model is applied to motor operation so that the rotor position information can be solved by using the motor stator voltage and current. The sensor-less control method has the advantages of great dynamic performance and high robustness, and the accuracy of the controller can be enhanced throughapplication of the extreme learning machine so that the learning rate of the algorithm is fast and the reaction speed of the controller can be enhanced.

Description

technical field [0001] The invention relates to a control method in the field of brushless direct current motors, specifically a method for controlling a brushless direct current motor without a position sensor based on extreme learning machine classification. Background technique [0002] Since the brushless DC motor has no brushes, it needs an electronic commutation circuit for current commutation, and a position signal is necessary to realize electronic commutation. The electronic commutation circuit is controlled by the rotor position signal so that the windings of the stator armature are continuously commutated and energized, so that the stator magnetic field changes continuously with the position of the rotor, so that the stator magnetic field and the rotor permanent magnetic field always maintain a left and right space Angle, generate torque to drive the rotor to rotate. [0003] Because the electronic commutation circuit needs the rotor position signal control, it i...

AI Technical Summary

Problems solved by technology

[0005] Technical problem: A variety of position sensorless rotor signal detection methods have their limitations, and the accuracy is not very high, so it is difficult to apply to occasions that require relatively high motor operation

Images