Research On Permanent Magnet Synchronous Motor Drive Techniques Based On Kalman Filtering

The modern production level, product quality and economic efficiency depend greatly on the performance of the automated equipment. The Permanent Magnet Synchronous Motor (PMSM) servo drive system is studied because of its importance in motion control system. An efficient PMSM servo drive system can help to improve the product quality and productivity, and enhance the level of manufacturing in China. Therefore, the research of PMSM servo drive system is advantageous to improve product performance and efficiency, even strengthen enterprise competitive power.To improve the precision of PMSM servo drive system and reduce costs, the speed sensorless technique and the dead-time compensation algorithm are considered based on summarizing the relevant literatures at home and abroad. The main contents and contribution of this thesis are as follows:1. The structure and the mathematical model of PMSM are discussed. A new digital servo drive system model with current loop and speed loop is designed based on rotor flux orientation vector control.2. A speed sensorless technique based on EKF (Extended Kalman Filter) is adopted to estimate the PMSM speed considering the unreliability and high cost of speed sensor. The output characteristics between sensorless PMSM vector control system and the system with speed sensor are compared.3. The influence of the dead-time and on-voltage of power devices on the inverter output voltage is analyzed. An on-line dead-time compensation method using Kalman filtering is presented for vector-controlled PMSM drives. The simulation results illustrate that the zero-current clamping phenomenon is eliminated and the output current ripple is reduced by using the proposed dead-time compensation method.4. A new PMSM servo drive system is designed in hardware and software. The PMSM servo controller was designed on TMS320F2812, and the power driver is based on PS21563 in hardware. An EKF speed estimation method is adopted to accomplish the speed sampling, and Kalman filter was used to compensate the dead-time effect in software.5. The speed estimation of EKF and the dead-time compensation of Kalman filter are verified through the experiment. The experiment results show that the PMSM servo drive system has preferable controlling precision and stability.6. The work accomplished in this thesis is briefly summarized and the key points for further research of the PMSM servo drive system are also presented.