Sensorless Control for Switched Reluctance Motor Drives

Gallegos Lopez, Gabriel (1998) Sensorless Control for Switched Reluctance Motor Drives. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 13818705.pdf] PDF
Download (11MB)


The thesis presents the work carried out on two methods for indirect sensing of the rotor position in switched reluctance motors. In the first method, PWM voltage control is used to control the motor. The method, called current gradient sensorless method, uses the change of the derivative of the phase current to detect the position where a rotor pole and stator pole start to overlap, giving one position update per energy conversion stroke. As no a priori knowledge of motor parameters is required (except for the numbers of stator and rotor poles), the method is applicable to most switched reluctance motor topologies in a wide power and speed range and for several inverter topologies. A single current sensor and detection stage is enough to drive a multiphase SRM. To start up the motor a feedforward method is used which assures robust start-up from stand-still up to a minimum speed even under load. The method is comparable to the back-EMF position estimation for brushless DC motors in performance and cost. The major contributions to the current gradient sensorless method are its real-time implementation, its experimental test and the reduction of current sensors and detection stages from three to one. The second method proposes a principle of high resolution sensorless position estimation for a switched reluctance motor drive, using either flux linkage or current to correct for errors in rotor position. The estimation algorithm makes full use of the nonlinear magnetic characteristics of the SRM through correlation of current, flux linkage and rotor position. The estimation model is simple, but with no loss in accuracy, leading to few real-time computations. Furthermore, a criterion is proposed to choose the phase most suited for position estimation when more than one phase conducts. The algorithm can also predict flux linkage, which in turn may be used to correct the position estimate further, and the features of this are discussed. The major contributions to this algorithm are the use of a simpler model for the observer, the introduction of a novel method to estimate the rotor position form the best phase from all conducting phases, and the experimental validation of the algorithm. Simulation and experimental results, which include steady-state with and without load, and dynamic transient tests, are presented, and confirm the concept and the robust functionality of both methods. The detailed operation and real-time implementation of both methods are described. However, it should be noted that these methods present advantages and disadvantages as other methods proposed before in the literature. Therefore, it is fair to say that these methods cannot replace in fully the mechanical sensor such as encoder or resolver, however they can offer a simpler and a better way than previous methods to estimate the rotor position. A second topic covered in this thesis is a functionality study of the classic switched reluctance motor inverter utilising a resonant dc-link to provide current regulation with zero-voltage switching. Although this topic is not directly related to the major part of the thesis on sensorless control, it is important in the wider study of power electronic converters for SR machines, and could form the basis for extended research in the combination of soft switching and sensorless control.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Additional Information: Adviser: Tim JE Miller
Keywords: Electrical engineering
Date of Award: 1998
Depositing User: Enlighten Team
Unique ID: glathesis:1998-75950
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 19 Nov 2019 17:13
Last Modified: 19 Nov 2019 17:13

Actions (login required)

View Item View Item


Downloads per month over past year