基于低频信号注入法的无轴承异步电机转速自检测控制

针对无轴承异步电机运行中悬浮转子转速检测问题,提出了一种基于低频信号注入法的无速度传感器控制新策略。该策略在无轴承异步电机基波模型基础上,通过注入低频信号引起的响应来构造转子位置偏差角,进一步通过PI控制器对偏差角进行调节,得到电机气隙磁场旋转速度,进而估计电机转速。运用该转速自检测方法,在Matlab/Simulink平台中搭建了无轴承异步电机无速度传感器矢量控制系统仿真模型,并进行了仿真研究。仿真结果表明,该方法能够在0.15 s内快速跟踪转子转速,并且具有优良的悬浮和转矩特性。试验结果同样表明,该方法不仅具有良好的转速在线自检测能力,而且能在无速度传感器方式下实现转子稳定悬浮运行,验证了所提方法的有效性和实用性。

Revolving speed self-detecting control based on low-frequency signal injection for bearingless induction motor

Abstract: A bearingless induction motor has the advantages of no friction, no wear, high speed, ultra-high-speed operation, and so on, so it is widely used in the field of life science with the difficulties of periodic maintenance, the field of chemical industry, the semiconductor industry and other fields. However, the installation of mechanical speed sensor not only leads to the increase of axial length of the motor and the cost issue, but also limits the high-speed, ultra-high-speed development of bearingless induction motors. In order to eliminate the adverse effect of the mechanical speed sensor on the high speed running of the bearingless induction motor, to reduce the axial dimension of the bearingless induction motor, to promote the development of bearingless induction motors towards being small, low-cost and practical, exploring a new kind of speed self-detecting strategy is particularly important. A new speed sensorless control strategy based on low-frequency signal injection method was proposed to solve the problem of rotor speed identification in the operation of bearingless induction motor. This strategy has many advantages such as non ideal characteristics of the motor, not easy to introduce other high-frequency harmonic signal and simple structure, so it has strong applicability. With the bearingless induction motor fundamental model, through the response caused by low-frequency signal injection, rotor position deviation angle was constructed, which was adjusted further through the PI (proportion integration) controller, and then the rotational speed of the motor''s air gap magnetic field was obtained. Then, the rotor speed was estimated. Using this speed self-detecting method, the simulation model of bearingless induction motor''s speed sensorless vector control system was built in MATLAB/Simulink platform. The simulation included the rotor speed response, the radial offset in x and y axis, the torque response and the self-tracking ability to detect rotor speed when the rotor speed mutated. Simulation results showed that this method could fast track the rotor speed, besides, the rotor speed curve from the self-detecting and the actual speed curve could be fully consistent in a short time. In addition, the radial displacement obtained by the low-frequency signal injection method was compared with that obtained by the high-frequency signal injection method. The comparison results showed that the proposed method not only could reduce the maximum radial deviation of the rotor, but also enabled it at the center position in a shorter time. At the same time the starting torque of the motor was large. After the speed mutation, the control system also had a good tracking ability for a given speed, and a fast response, besides, stable error was very small. Finally, in the bearingless induction motor''s control system experimental platform, the experiment was carried out using the proposed strategy. We selected the DSP TMS320F2812 as experiment control chip; a bearingless induction motor was used as a prototype, and the prototype parameters and simulation parameters were consistent. In order to more accurately compare the actual speed with the self-testing speed, the prototype was equipped with optical encoder disk. Test results showed that the self-detecting speed using the low-frequency signal injection method was more accurate than that using the high-frequency signal injection method, and the rotor center of mass offset distance using low-frequency signal injection method was smaller than that using high-frequency signal injection method. The results verify that the method has not only a good capability of speed online self-testing, but also a stable suspension operation of the rotor, and therefore the proposed method is effective and practical.