Direct Torque Control For PMSM With Sensorless Based On Model Reference Adaptive System

Nowadays, because permanent magnet synchronous motor (PMSM) has suchadvantages as small size, simple structure, good performance, high reliability, itsapplications are broaden continuously in electric drive field. Direct torque control(DTC) of PMSM has a simple control structure, fast dynamic response, so DTC hasreceived extensive attention.At first, this thesis introduced the background and current technologies of directtorque control and the sensorless control of permanent magnet synchronous motoraround the world. And then the structure and characteristics of permanent magnetsynchronous motor were described briefly, and the mathematical model of thepermanent synchronous motor in the different coordinate systems were analyzed.Secondly, according to the mathematical model of the PMSM, the theoreticalbasis of permanent magnet synchronous motor direct torque control were demonstrat-ed by derivation. Several control strategy for DTC of PMSM were introduced, ofwhich the control strategy adopted space voltage vector modulation (SVM) wasanalyzed in detail. Meantime，the advantages and disadvantages of various of variouscontrol strategies were compared.Thirdly，the installation of speed sensor increased the complexity of the systemand reduced system reliability and robustness. Moreover, it increased the system costand maintenance requirements. In order to overcome these drawbacks of the intro-ducetion of speed sensor, a speed sensorless based on the principle of the modelreference adaptive system (MRAS) was studied to obtain the speed and the location ofthe rotor.At last, a novel direct torque control system based on space vector pulse widthmodulation technique (SVM-DTC) was proposed, of which the stator flux and torquewere regulated by PI controllers to decouple motor space vector in the twocomponents of stator flux and torque. Meanwhile, because the traditional statorvoltage integration method to estimate stator flux was prone to produce offset driftingand integral saturation, a stator flux estimation method based on rotor position andstator current was used to estimate the stator flux. And a sensorless based on MRASwas designed to estimate the rotor speed and position. Simulation and experimentalresults show that the proposed method can effectively estimate the stator flux, andalso reduce the electromagnetic torque and stator flux ripples, so the system has good statistic and dynamic performance.