Permanent Magnet Synchronous Motor Without Position Sensor Control Technology

Permanent Magnet Synchronous Motors (PMSM) have been widely used in high performance servo systems in the last two decades for its advantages, such as compactness, high efficiency, reliability and adaptability to the environment. To achieve effective control of PMSM, information of the rotor position is essential to achieve magnetic filed orientation. This can be done with different types of mechanical sensors, such as optical encoders and resolvers. However, such solutions would introduce additional cost, and lower the reliability of the system. Therefore, sensorless control of PMSM has become a topic of interest in the filed of motor control. Within the last two decades, many researchers and scientists have published papers on this topic, which fall into two broad categories, namely, those that only work at high speed and those that only work at low speed. In fact, there are very few sensorless control algorithms that single-handedly permit operation over entire speed range. To solve this problem, a novel fusion algorithm is proposed in this thesis which is valid over the whole speed range. The following research works have been done.Firstly, the formation of the salient poles in Interior Permanent Magnet Synchronous Motors(IPM) is analyzed. The vector control system based on IPM is introduced, and the mathematical model of IPM is given.Secondly, after analyzing the mathematical model of IPM, the High Frequency Injection method is introduced to estimate the rotor position. The High Frequency Revolving Voltage Injection method is given which relies on the tracking of the position of the salient pole. The simulation result proves its performance.Thirdly, through transforming the mathematical model of IPM, the traditional Back-EMF tracking observer estimation method, which only works for Surface Mounted Permanent Synchronous Motors(SPM), is adapted for the control of IPM. Using this method, the rotor position is effectively estimated by tracking the extended back-EMF, which is part of the transformed mathematical model.Lastly, a novel fusion algorithm is proposed, which uses the information generated by High Frequency Injection method at low speed and that generated by extended Back-EMF tracking observer at high speed, and functions seamlessly over the entire speed spectrum.