Research On Pmsm Control Of Sensorless Based On Sliding Mode Observer

Permanent magnet synchronous motor (PMSM) which has the advantage of small size,high efficiency and good speed performance has been increasingly widely used in the field ofindustrial control in recent years. The position sensors are usually installed on the PMSM toprovide rotor position information, then the self-controlled system is builded. The positionsensors not only increase the cost, but also increase the failure factors of the PMSM. So thesensorless control system of PMSM has been rapidly developed.In-depth analysis of the position prediction method of the sensorless system of PMSM,sliding mode observer (SMO) which has the advantage of rapid response and good robustnessis used for the rotor position prediction method.The paper first describes the control methods of PMSM and two main variable voltagevariable frequency control strategies of AC motor, building the Mathematicalmodel of PMSM and vector control system of PMSM. Secondly, designing the SMO basedon sliding mode variable structure control, achieving the purpose of the prediction of the rotorposition angle. Compared with the traditional vector control system of PMSM with sensor,SMO estimates the rotor position angle based on the motor back electromotiveforce(back-EMF), the back-EMF is very small when the speed is very low especially atstandstill, so it is difficult to meet the control system requirements. This paper presents a newcontrol strategy that other-control and self-controlled are combined with, the method ofother-control is adopted at standstill and low speed, the PMSM is running with constantvolts/hertz(V/F) control and switches over to self-controlled when the SMO can observe therotor position to realize the space vector pulse width modulation.On this basis, completing the hardware design of the control circuit and the softwaremodular design of this system. The digital filter for the back-EMF causes the phase delay ofthe rotor angle, so the sectionalized linearization is adopted to compensate the phase delay onthe basis of the experiment. In order to verify the practical feasibility of this sensor controlalgorithm, designing the experiments in the control systems platform. The result ofexperiments including position prediction experiment, velocity modulation experiment, loaddisturbance experiment verifies the proposed algorithm of this paper.