Research On Position Detection Of Switched Reluctance Motor Without Position Sensor

The traditional position detection method of Switched Reluctance Motor(SRM)uses the position sensor to detect the position of the rotor.However,the presence of the position sensor in different situations brings many disadvantages.For example,the switch of the photoelectric sensor will be out of control when the amount of dust is too high in the coal mine environment.The position sensor will deviate in the severe vibration environment.In the high electromagnetic interference environment,the electromagnetic position sensor will be interfered,so that the detection of the rotor position is not accurate high.For the above problems,based on the traditional position sensorless detection method,this paper studied the rotor position detection method of SRM at different speed stages.The main research contents are as follows:First of all,in view of the problem that the inductor intersection will be offset when the inductor is saturated in the SRM low-speed stage,which will affect the accuracy of rotor position estimation,the error compensation function based on current is established in this paper to compensate the error of the inductor intersection which is greatly affected by saturation.In addition,to solve the problem that the rotational speed fluctuates greatly due to the influence of voltage and current detection errors,Kalman filter is used to smooth the estimated value,and the accuracy of the inductance intersection compensation method after filtering is verified by simulation.The simulation results show that,compared with the traditional method,the improved position detection method can effectively improve the detection accuracy at low speed.Then,in order to solve the problem of single turn-off angle in the simplified flux method adopted by SRM in the middle and high speed operation stage,the Back Propagation(BP)neural network SRM flux model based on the preprocessing function was established.Particle Swarm Optimization(PSO)was introduced to optimize the convergence rate of BP network,and then the optimized flux model was applied to the simplified flux method to detect the rotor position of SRM.In addition,Kalman filter is used to smooth the estimated speed,and the accuracy of the optimized simplified flux method is verified by simulation.The simulation results show that the improved position prediction model can not only realize the variable turnoff Angle,but also estimate the rotor position information more accurately,and has higher detection accuracy.Finally,the sensorless position detection technology used in this paper is verified experimentally with the three-phase 12/8 SRM on the sensorless position detection experimental platform of switched reluctance motor.In order to avoid SRM rotor inversion,high-frequency pulse injection method was used to determine the initial conduction phase.Then,the rotor position was detected by the optimized inductance intersection point compensation method and the improved simplified flux flux method,so as to verify the accuracy and feasibility of SRM detection method in the low-speed and medium-speed operation stages.