Indirect Rotor Position Estimation For 12/10 Switched Reluctance Motor

The research of Switched Reluctance Drives system (SRD) have gained more attention due to its flexible control methods,compactness, robustness and good electric characteristics. Position information is essential for a high performance Switched Reluctance Motor(SRM). Mechanical position sensors are often used to provide rotor position information for the SRD.However,these mechanical position sensors add to the cost and dimension and lower the reliability of the SRD.As a result,the technique of indirect rotor "position estimation has become one major direction in the study of SRD.First,this dissertation build the linear model and non-linear model of SRM,and give the three operation models of SRM.And then proposes a sensorless control methods for the SRD based on the linear model of SRM.That is,we inject narrow voltage pulses into two non-conducting phases of SRM and detect the corresponding currents. So we can get the position information from the relationship of the rotor position,stator winding inductance and the corresponding currents.In this paper we also studied the frequency of the injected voltage pulses and the methods of measuring the motor speed.This methods got the position information by detecting the variation trends of the corresponding current instead of many a complicated calculation, which not only eliminates the detection errors,but also decreases the effects of valtage fluctuation and load variation,and then strengthens the stability and robustness.However, when the motor speed is high, the flux will be saturated and the inductance between phases can not be neglected, which leed to the invalidation of this methods. So, we can only use this sensorless methods to detect the rotor position in start-up stage. According to this methods a design of software and hardware is presented in this paper. The corresponding currents excited by the voltage pulses was detected too. The torque-angle trait of the 12/10 SRM was measured, and the position signals from the photoelectricity sensor and from the sensorless methods was compared. Experimental results proved that the proposed sensorless methods has high reliability and utility in start-up stage.