Sensorless Control Of PMSM Based On Kalman Filter

The electric motor is a mechanical energy conversion device. It is clean,efficient, and has already played a very important role in people’s lives. Permanentmagnet synchronous motor has many advantages, such as simple structure, small size,high efficiency and wide speed range. It can meet the high requirements in mostcases, and is very important in the motor system in recent years.Using of basic mathematical models and coordinate motor transformationmethods to get models and formulas in the coordinate system. Then introduces avariety of control methods used in permanent magnet synchronous motor control.Vector control technology is currently widely used control method. In order toachieve precise control of PMSM the SVPWM (Space Vector Pulse WidthModulation) is used. This control system is currently used as the primary controlscheme of PMSM by scientists in many fields.In PMSM vector control theory, adaptive feedback system is needed to know thelocation and speed of the motor rotor. Some motors contain position sensors. Thecontrol system can obtain the speed and angle information from these sensors. But inmany cases there is no motor position sensor, then we need to estimate the speed andangle of software-level algorithms. This paper describes some strategis of SensorlessControl of PMSM which are often used. Among these observers, the principles andapplications of the Kalman filter algorithm is mainly introduced. Since the motorsystem is nonlinear, therefore it needs to be linearized. So that it becomes theExtended Kalman filter (EKF). Then presenting the mathematical formula of EKF inα-β coordinate system, and listing the error parameter matrix in this experiment.In this paper, the MATLAB-Simulink is used to build the motor systemimplementation and system simulation model validation. EKF mathematical formulasin the α-β coordinate system used to build the model, also including vector control(FOC), space vector pulse width modulation (SVPWM) and so on. We need to payattention to some details of the model maybe affect the simulation results, which aremetioned in this paper. We found the Kalman filter has excellent performance bysimulating conditions at low speeds and different motor parameters within a certainrange. In the end of this paper, some places which need to be further improved areexplained, and looking forward to the next phase of the work.