A Study on Indirect Rotor Position Estimation using Phase Inductance Nonlinearity of 4/2 Switched Reluctance Motor

Abstract

In this dissertation, we propose a new indirect rotor position estimation method for driving without using position sensor by using the inductance nonlinearity of Switched Reluctance Motor (SRM). SRM is a motor with a salient pole type in both stator and rotor, and rotor position information is needed to generate continuous reluctance torque. This is because the stator winding is excited according to the rotor position information to generate the reluctance torque. The rotor position information is generally obtained using a position sensor but it causes problems such as cost, volume increase, complexity and reliability reduction of the drive system. To solve this problem, various methods for indirectly detecting the rotor position information have been studied. The conventional methods estimate the rotor position information using look-up table and Fourier series inductance modeling, which require prior knowledge of magnetic characteristics. However, a large amount of data or higher order modeling is required to increase the resolution of rotor position estimation. For this reason, there are problems that external hardware is required and the computation quantity increases. In this dissertation, we propose a new indirect rotor position estimation method using the inflection point of inductance generated in 4/2 SRM used in vacuum cleaner application. The proposed method indirectly obtains the rotor position information by using the characteristics of each of the two inflection point generated for one phase. In particular, the method using the inflection point at the alignment position indirectly estimates the rotor position using the change in slope of the inductance at the alignment position, the method using the inflection point at the unaligned position indirectly estimates the rotor position using the change in slope of the inductance and the current at the unaligned position. By using the estimated rotor position information, the switching point for phase commutation can be calculated. As a result, unlike methods such as look-up table and higher order inductance modeling, which require prior knowledge of magnetic characteristics, inductance inflection points are detected using the physical amount from the terminals of the motor, it does not require prior knowledge of the magnetic characteristics and complicated modeling. To verify the validity of the proposed method, various experiments using a prototype set with 400 W 4/2 SRM were performed and the proposed indirect rotor position estimation method was verified.