| The Permanent Magnet Synchronous Motor is an important type of synchronous motor. For using rare earth materials as magnetic pole, the volume and weight of PMSM is reduced, and it has simple structure, reliable operation and convenient maintenance. One of the trends on modern electronic drive control is developing new type alternating current (AC) speed regulation and servo system. Whether vector control or scalar control, speed and position close -loop control need a speed sensor in the motor shaft .However, the extra speed sensor cause many problems, for example, the rising of cost, reducing of the stability, even the problem of installation. So the high performance sensorlesss control has become a research hotspot of motor control.This paper applies the Kalman filter in the sensorless control of PMSM based on the introduction of it. The PMSM is a several-orders, strong-coupling and nonlinear system, the way of dealing with nonlinear system in this paper is applying to the truncation of Taylor series expansion, which is the most used method in industry application, the system noises and measurement noises are introduced to reflect the uncertainty of modeling and disturbance of environment including the variation of loads and wheel inertia, the kalman filter algorithm is extended to nonlinear system, which means Extended kalman filter (EKF).The rotor position and speed are considered as system states to be estimated by EKF,and then feedback to the sensorless control system of PMSM. Simulation results show that the EKF has good speed observation capability, well dynamic response and immunity to the noises.For the defects of the EKF in sensorless control, this paper proposes Reduced -order Linear Kalman Filter (RLKF). RLKF re-selects system states and the motor system equations are completely linearized, square -variance of system equations are the invariable in every step, this type Kalman filter is Linear Kalman Filter (LKF).From the simulation, compared with traditional EKF, RLKF algorithm is conciser, task of adjusting parameters is simplified, the algorithm is easier to implement digitally. RLKF not only inherit all the benefits from classic EKF but also overcome almost the flows.By analyzing Kalman filter, for the reflection of the uncertainty of modeling and disturbance of environment including the variation of loads and wheel inertia, it has good performance on states estimating. This paper has designed a program of states observation system for PMSM based RLKF and digital signal processor (DSP).The system estimates the rotor position and speed of motor in real time from the phase current based the abundant resources and great computing power of DSP.This system can provide information of rotor position and speed by taking the place of the traditional sensor. The main work of next step in this paper is to put the program into practice, continue the research of Kalman fiter and its performance in PMSM sensorless control. |
