Research On Model Predictive Control Optimization Strategy Of Permanent Magnet Synchronous Motor

Due to the advantages of simple structure,high torque and power density and wide speed range,permanent magnet synchronous motor(PMSM)has been widely studied and applied.model predictive control(MPC)has received more attention and been more widely studied in permanent magnet synchronous motor drives due to its advantages of fast response,multi-variable control and easy implementation.However,there are still some problems of the MPC in steady performance,algorithm complexity and parameter robustness.Therefore,this thesis takes the PMSM system driven by two-level voltage source inverter as the research object,and studies the MPC for the PMSM.Based on the mathematical model of the PMSM driven by two-level voltage source inverter,this thesis introduces the basic principle and realization method of mpc in detail.Analyze and compare the transient-state and steady-state performance,algorithm complexity and parameter sensitivity of the different model predictive control method,and then clarify the limitation and problem to be solved of the MPC.This thesis reduces the torque ripple and flux ripple of by optimizing the voltage vector sequence,and proposes an MPC strategy based on the flux linkage ripple optimization for the surface-PMSM.Firstly,the stator flux vector ripples are quantitatively analyzed when the multiple voltage vectors are applied in different sequences in a control period,and the voltage vector sequences(VVSs)which can reduce the stator flux vector ripple are selected as the candidate vector sequences.The cost function is designed to evaluate different VVSs by calculating the stator flux vector errors at different switching instants in one control period,and then the optimal VVS is selected.The experimental results indicate that the proposed model predictive control strategy based on the flux linkage fluctuation optimization can improves the steady-state performance of the PMSM while guaranteeing good tracking performance is also achieved.In order to improve the steady-state performance of the interior-PMSM in the low carrier wave ratio operation condition,this thesis derives a current augmenter prediction model(CAPM)to improve the prediction accuracy under high-speed condition and to suppress the current tracking error caused by parameter mismatch and deadtime effect;and then an adaptive observer(AO)based on the proposed CAPM is designed to estimate the model error of the CAPM caused by parameter mismatch.Moreover,a variable integral gains(VIGs)calculation method is designed to guarantee good performance of the AO under different motor operation conditions.The experimental results indicate that the proposed strategy reduces the current ripple of the PMSM under high-speed operation conditions,and the proposed AO with VIGs ensures that the proposed PMC has good dynamic,steady-state performance and anti-motor parameter disturbance ability under different motor operation conditions.In order to reduce the algorithm complexity of MPC to achieve higher control frequency,a predictive torque control(PTC)based on the discrete three-phase switching duty ratios(3P-SDRs)is proposed.This thesis analyzes the relation between the 3P-SDRs and the prediction torque and prediction stator flux,discretizes the 3P-SDRs,and selects the optimal discrete 3P-SDRs.The complexity of the proposed strategy is reduced by eliminating the processes of the synthesis of the optimal voltage vector and the additional calculation of the corresponding 3P-SDRs.Finally,only the discrete results of one-phase SDR need to be stored,which greatly reduces the storage burden of the digital controller.