Research on IPMSM Sensorless Control Based on Lattice Kalman Filters

Embedded permanent magnet synchronous motor (IPMSM) sensorless control system. In view of the problems of traditional Kalman filter algorithm, such as the decline of filtering accuracy when the system state changes suddenly, long running time, slow convergence speed, and large amount of data required for calculation, an improved algorithm of lattice Kalman filter is proposed to estimate the motor speed, and it is applied to the research of IPMSM sensorless control. This method uses the Quasi Monte Carlo method of lattice rules, uses the Kropov lattice rules to generate sample points, and uses the Cranley-Patterson rotation shift method to perform random shift. In the case of extracting a small number of sample points, it can quickly ensure the accuracy and strong robustness in the estimation process. Under the environment of MATLAB/Simulink, the simulation experiment of sudden load in the process of given speed is carried out. Lattice Kalman filter algorithm has high filtering accuracy, strong speed tracking and small amount of data required for calculation. It can realize high-precision control of IPMSM and has practical engineering value.