Research On Model Predictive Control System For Switched Reluctance Motor

Switched Reluctance Motor(SRM)is a key research object in the field of traction drive due to its low manufacturing cost,wide range of speed regulation,good starting performance,and fault tolerance.In recent years,SRM drive systems have developed rapidly and have a broad application prospect in the field of electric vehicles.In order to improve the comprehensive performance of the electric vehicle drive system based on SRM,this thesis improves the control strategy of the SRM based on the model prediction principle,and makes an indepth study on the improvement of the current tracking ability of the motor and the suppression of the torque ripple.Firstly,in order to overcome the shortcomings of the traditional finite set predictive control,which has multiple switch states to be selected for each sector and large computational complexity,a new sector partitioning method based on the SRM inductance and torque current ratio characteristics is designed.This method selects different switch combinations in the non-linear region and the linear rising region of inducance,which reduces the number of switch states to be selected in each subdivided sector.Based on this sector division method,an improved finite set predictive control strategy is designed.The simulation and experimental results have shown that this method can effectively reduce the computational burden in the control process,lower the system’s demand for processor computing power,and improve the real-time control performance.Secondly,the principle and limitations of conventional hysteresis control is analyzed.Then,to overcome the shortage of hysteresis control,a deadbeat control method based on the current intersection is proposed.By calculating the approximate trajectory of the current,this method predicts the ideal switching moment of the switch state in the next cycle,and then gives the Pulse width modulation duty cycle(PWM)accurately to track the reference current.It has the characteristics of high tracking accuracy,small computation and low sampling frequency requirement.Simulation and experimental results prove that the proposed method has a better ability to track the reference current,which can effectively reduce the fluctuation of the current waveform and improve the system’s ability of the torque ripple suppression.Finally,in order to reduce torque ripple during braking process of SRM,A two-step torque prediction method is proposed.The method eliminates the need for a torque inverse model and allows the control signal of the system to be optimized for the next two steps with one sampling,reducing the system’s requirement for sampling frequency.At the same time,in order to improve the unreasonable torque distribution caused by the TSF during commutation region,this method also combines an improved optimization compensation for the torque sharing function to better distribute the reference torque in the two-phase overlap region.Simulation and experimental results show that the proposed method can precisely control the braking torque,significantly suppress the torque ripple in the braking process of the motor,and effectively improve the braking performance of the SRM drive system.This thesis has 55 figures,5 tables and 85 references.