A Study Of Permanent Magnet BLDCM System Based On Square-wave Drive

Brushless dc motor has been widely used in various aspects of the national economy due to its superior performance. However, the existence of the position sensor limits its application in complex environments. For the above mentioned reason, the position sensorless control system has aroused more and more concerns. Combining with square-wave drive technology, namely three-phase six state 120 o conduction mode, three key problems,including motor start, rotor position estimation and controller design, will be further research.In the traditional "three-stage" startup method, pre-positioning and outer synchrotron acceleration is an open loop operation mode which can not deal with the startup problem in case of different load. This paper proposes a new closed-loop startup method without position sensor based on a short pulse injection strategy. The bus current difference has been used as feedback signal to regulate voltage vector size at the pre-positioning stage. During the acceleration, two pulses will be injected each times to finish the acceleration and detection simultaneously. The commutation conditions are specified by the corresponding bus current difference less than a given threshold. The motor speed calculated before the communication is used as feedback signal to adjust the voltage vector size. The whole startup process is in form of close-loop.As far the rotor position estimation problem, a simplified method of rotor position estimation is proposed by real-time calculation of the terminal voltage through analyzing the traditional BEMF detection principle. The method simplifies the hardware structure and avoids BEMF zero-crossing phase shift caused by the terminal voltage filter circuit. As a result, a higher precision of rotor position estimation can be obtained.In order to improve the control performance of BLDCM speed regulating system, a fuzzy self-tuning anti-disturbance IMC is designed in this paper. This controller regulates the IMC filter parameters real-time using fuzzy controller and disturbance observer is added to compensate the uncertain dynamics. The simulation results show that the controller has better dynamic performance and robustness than the traditional IMC-PID algorithm.Finally, hardware and software platform are designed using Active-Semi PAC5220 as the core of the controller. The motor startup and rotor position estimations are tested in this platform and results demonstrate the effectiveness of proposed methods.