Research On The Multi-parameter Identification Of Permanent Magnet Synchronous Machines Considering VSI Dead-time Effect

Permanent magnet synchronous machines(PMSM)are now widely employed in industrial field due to their high efficiency,high starting torque and high power/torque density,etc.However,in the real application,the motor parameters will be affected by the influence of the operating conditions of PMSM control system.Thus,the accurate identification of the motor parameters is of great significance for the accurate control of the motor.The on-line parameter identification technique of permanent magnet synchronous motor control system is based on the stator voltage equation.Because of the stator voltage can not be measured directly in the actual control system,the stator voltage have to be replaced by the controller output.The distorted voltage between stator voltage and controller output can affects the identification accuracy of motor parameters.The paper research the on-line multi-parameters identification technology of PMSM based on its vector control theory.The paper firstly introduces the practical significance of motor parameter identification in the control system of PMSM and its present research status at home and abroad.Then introduce the mathematical model and the vector control system of PMSM,and design the motor parameters identification strategy using model reference adaptive algorithm based on compensating the voltage-source inverter(VSI)dead time effect.In order to reduce the influence of VSI dead-time effect on motor parameters identification accuracy more effective,the paper establishes a parameter identification model which is not affected by VSI dead time effect to improve identification accuracy of motor parameters,and the process only need to transform voltage equations.Besides,in order to improve convergence speed and reduce steady-state error of the identification results,the proposed method uses Adaline neural network method with a momentum item to identify motor parameters,and the momentum factor can be adjusted due to different convergence stage.Finally,the on-line parameters identification strategy is verified by the experimental control system of PMSM.The experimental results show that the compensation of the VSI dead-time effect can improve the identification accuracy of the motor parameters.The improved Adaline neural network identification method based on the established model can identify the stator resistance,inductance and rotor flux quickly and accurately.