Sensorless Vector Control Of PMSM Based On Extended Kalman Filters

Permanent magnet synchronous motor(PMSM),sparked by a number of contributing factors such as small loss、high efficiency and excellent speed performance,has currently become an integral part in motor speed controlling system,The PMSM speed controlling system has critically enhanced the development both in industry and agriculture field.As a result,the study in the field of PMSM controlling system has important application value and theoretical significance.In this dissertation,we present an improved Extended Kalman Filter(EKF)algorithm approach capable of estimating motor rotor position while avoiding the limitations caused by mechanical sensor installed in PMSM control system.Those limitations include increasing cost and decreasing reliability.Our approach uses Genetic Algorithm(GA)to optimize complex matrix calculation loop and applies the load torque observer technique to overcome the delay effects in traditional EKF approach,increasing the rotor position estimation accuracy,while providing the motor motion system with the feed-forward compensation.In conclusion,the improved Extended Kalman Filter algorithm could satisfy the high requirements of sensorless PMSM vector control system.First,we discuss the principle of PMSM vector control,so as to deducing the PMSM mathematical model under static α-β coordinates.Second,in addition to providing the theory of traditional EKF algorithm,we integrate it with PMSM model in detail.We demonstrate the specific steps on how to use GA,together with load torque observer,to facilitate the traditional EKF and establish the improved EKF mathematical motor model.The Improved EKF simulation model is set up by integrating the SIMULINK into S Function,and simulation experiments are carried out.The proposed approach is able to replace rotor position sensor and it can achieve targeting accuracy that exceeds tradition EKF while simultaneously enabling avoidance of load disturbances.Finally,the PMSM experimental platform is established based on the TI company’s controlling chip TMS320F28335.The system’s hardware design and programming are given and the experiments are implemented.The experiment results validate the proposed improved EKF algorithm successfully meeting the needs of estimation on motor rotor position and speed,achieving the high performance of PMSM controlling system.