Research On Low-Complexity Model Predictive Control Method For Permanent Magnet Synchronous Motor

In recent years,model predictive control has emerged as a research hotspot in the field of control for permanent magnet synchronous motors,owing to its advantages such as control flexibility and fast response.Among various model predictive control approaches,finite control set model predictive control has gained significant attention due to its excellent dynamic performance.However,it still suffers from poor steady-state performance.To improve the control accuracy of torque and flux,a more precise vector synthesis method has been proposed,but it often involves complex selection of reference voltage vectors and duty cycle calculations.To address the trade-off between control performance and calculation complexity,this thesis conducts the following research:To address the unclear relationship between electromagnetic torque,flux,and twophase duty cycles,a two-phase direct duty-cycle prediction model is established based on the mathematical model of permanent magnet synchronous motors,voltage vector generation principles of two-level inverters,and duty cycle synthesis principles.This model reveals the mapping relationship between electromagnetic torque,flux,and twophase duty cycles,without introducing additional vector synthesis steps.This approach directly calculates the duty cycles from control objectives,ensuring high dynamic adjustment capability while reducing calculation costs.Experimental results of a twostage control strategy based on the proposed model validate its effectiveness.To overcome the high calculation cost associated with the optimization process of reference voltage vectors,a virtual reference voltage vector preselection strategy based on flux deviation is proposed.Leveraging the polarities of flux deviation components on the complex plane,this strategy establishes a four-sector virtual reference voltage vector distribution model and allows for rapid determination of the optimal virtual reference voltage vector through simple algebraic calculations.Compared to traditional polling methods,this strategy reduces calculation complexity.Considering the high calculation cost posed by multi-stage and multi-iteration duty cycle calculations in scenarios requiring low torque and flux ripple control,a novel control strategy based on flux constraint is developed.This strategy,based on the twophase direct duty-cycle prediction model,virtual reference voltage vector preselection strategy using flux deviation,and optimization theory,achieves optimal torque control with low flux error while reducing control system order and calculation iteration.Comparative experiments with seven classical control methods under different flux,load,and speed conditions demonstrate that the proposed control strategy keeps the flux and torque within a low range of ripple.The analysis of calculation complexity shows a significant reduction in calculation costs in terms of parameter computation and mathematical operation statistics,compared to high-performance control methods.In summary,the proposed control strategy exhibits improved performance in terms of steady-state and dynamic,as well as calculation complexity.