Sensorless control of switch reluctance motors

Submission note: A thesis submitted in total fulfillment for the degree of Doctor of Philosophy to the Department of Electronic Engineering, Faculty of Science, Technology and Engineering, La Trobe University.

This thesis presents a novel sensorless position technique for a Switched Reluctance Motor (SRM). The widespread use of SRMs has been relatively limited when compared to other popular motor technologies. Renewed interest has occurred in SRMs due to low cost, environmentally friendly materials and its wide torque operating region. One of the disadvantages affecting the wider use of an SRM, is the requirement for a rotor position sensor. This increases the cost and reduces the durability of the system. As a result, much research has been dedicated to making the motor `sensorless'. The proposed sensorless technique seeks to improve the operating range of currently available methods, with the advantage that there is no need for an offline calibration technique. To achieve sensorless operation, the proposed technique evaluates the inductance waveform on a per phase basis, to find the minimum and maximum inductance positions. These locations correspond to the unaligned and aligned rotor positions respectively. A sensorless rotor position is created via a novel digital phase locked loop (DPLL), which generates intermediate positions between these detected positions. To enhance sensorless tracking and accuracy a multiple model adaptive estimator Kalman filter is introduced. This improves the position estimates of the un-aligned and aligned rotor positions and improves the sensorless tracking output of the DPLL. Lastly, to ensure sensorless operation from zero speed to normal sensorless operation, a novel hesitation-free starting technique is presented. The performance of the proposed sensorless technique is evaluated using simulations and real-world testing utilising a three-phase test SRM. Novel methods have been developed to remove the effects of mutual coupling, flux-linkage term errors and magnetic hysteresis. These techniques ensure that accurate sensorless operation is achievable over a wide range of speeds and loads, without the need for an offline calibration technique. This results in improved durability and reduced cost of the SRM.