R&D On Sensorless Control Technology For Brushless DC Motor

Brushless DC motor with permanent magnet excitation, in which electrical commutator is used instead of mechanical, has the same good characteristics of speed control as traditional DC motor. Brushless DC motors have found wide application due to their high power density and ease of control. Moreover, the machines have high efficiency over wide speed range. Attachment of sensorless control not only overcomes a lot of disadvantages generated by the sensor, but also expands the application range of brushless DC motor further. At present, sensorless control has become a development trend of control technique for brushless DC motors.This dissertation makes a comprehensive view of the rising, development and present situation of control technique for brushless DC motor, and summarizes the present level and shortages of sensorless control technique. Depending on the Research and development of all DC variable speed air-conditioner, this dissertation researches and discusses the following problems in theory and practice:(1) "Lagging 90 -a commutation" and "leading 60 -ycommutation " methods are bring forward to overcome the shortages existing in the traditional zero back EMF sensing and commutation point sensing respectively. At the same time, the commutation method realized by combining the software and hardware breaks through the limitation of commutation method depending on pure hardware sensing circuit, then the speed-variable range is expandedand the stability and reliability of the control system are increased.(2) The principle of commutation torque ripples is analyzed. According to the analysis and deduction, a compensation method is put forward creatively to reduce the torque ripples generated by commutation. Simulation and experimental results with compensation prove that the compensation method is correct and practical.(3) The influences of different switching mode of PWM are discussed. Then a practical method, "three-state algorithm of invariable duty cycle under forced output mode" is brought out to eliminate the commutation dead band generated by switching among the output interfaces of PWM waveform. Experimental results prove that this method is effective.(4) In the control system of BLDC motors, different PWM mode has different influence on the pulsating current generated by BEMF in the switch-off phase, which may make commutation error in the commutation control and then decline the reliability and the stability of the control system. The principle of BEMF current is analyzed and the best PWM mode in five given types to reduce its influence on the commutation control is selected. Simulation and experimental results prove that the selection is correct and practical.(5) The dynamic model of brushless DC motor is investigated, and the close-loop speed control system with single neural adaptive controller used as speed regulator is established. Analysis shows that single neural adaptive controller can improve the robustness of the control system.