Research On Speed Observer Of PMSM Servo System

With the development of power electronic and the improvement of microprocessor performance, the control performance of PMSM servo system continues to improve. It is widely used in industrial production and defence research. The traditional digital speed detection by optical encoder has big errors when the rotate speed is low. It seriously impacts the control performance of the servo system. In order to solve the speed measurement problem, speed observer is discussed in this paper.The traditional open-loop instantaneous speed observer has a poor stability and is easier to be influenced by the motor parameters. A new closed-loop instantaneous speed observer is designed using rotor position angle observation error as feedback variables. Position feedback loop improved the anti-disturbance of the observer.Full order observer and reduced order observer are designed based on the state observer theory. And then the paper makes a comparative analysis, verifying their observation performance and problems.The traditional speed observer based on kalman filter has a low speed in detecting the load torque. In this paper a new kalman filter is designed by combining with disturbance observer. The state variable of the kalman filter is updated by rotor position angle and load torque which are known variables. The new design raises the response speed and detecting precision of the observer.The matrix operation and parameters update of kalman filter need a long time, which would influence the response speed of the servo system. In order to reduce the calculated amount and improve the response speed of the observer, a second order kalman filter is designed based on a disturbance observer.Matlab/Simulink is used to simulate and analysis these design schemes of speed observer. They all can track the real speed well in low speed. The development of speed observer theory will greatly enhance the performance of the PMSM servo system.