In April 2016 Manchester eScholar was replaced by the University of Manchester’s new Research Information Management System, Pure. In the autumn the University’s research outputs will be available to search and browse via a new Research Portal. Until then the University’s full publication record can be accessed via a temporary portal and the old eScholar content is available to search and browse via this archive.

Related resources

Search for item elsewhere

University researcher(s)

Academic department(s)

Sensorless Speed Estimation in Wound Rotor Induction Machines Drives

Tshiloz, Kavul

[Thesis]. Manchester, UK: The University of Manchester; 2016.

Access to files

FULL-TEXT.PDF (pdf)

Abstract

Sensorless speed estimation in induction machines (IMs) presents an attractive proposition for eliminating the need for physical speed measurement sensors and thus avoiding the associated reliability and cost issues, such as the requirement of extra wiring, careful mounting, maintenance and adjustment. In this thesis, the feasibility of utilising the stator current and power signals to establish spectral search based (SSB) sensorless speed estimation schemes in wound rotor induction machines (WRIMs) operating in extended slip and open-loop controlled conditions is investigated. The research is performed on three different industrial WRIM designs.The thesis first investigates the spectral content of WRIM electrical and mechanical signals with the principal aim of identifying spectral patterns that can facilitate the development of real-time sensorless speed estimation. The examination is based on detailed harmonic models of the considered machine designs as well as experimental results obtained from tests performed on laboratory test rigs. A generalised theoretical analysis of the possible spectral content of machine signals that enables the derivation of closed form analytical expressions linking individual spectral frequencies to rotor speed is also undertaken. The results demonstrate that it is possible to clearly identify speed dependent components in the stator current and power signals and map the boundaries of the narrowbands maximised by these for extended slip and open-loop operating conditions. To enable improvement in attainable real time SSB estimation rates a dichotomous search algorithm real-time spectral processing method was employed for frequency tracking in this research. The algorithm performance is evaluated in real-time tests performed on a measured steady-state laboratory machine stator current and power signals. The results demonstrate that the dichotomous routine provides an inherent advantage in the frequency estimation rate without compromising the estimation accuracy and can therefore enable significant estimation rate improvement in SSB speed estimation algorithms.Novel sensorless speed SSB estimation techniques are then proposed for WRIM operation in extended slip and voltage/frequency controlled conditions. The algorithms utilise the reported analysis of electrical signals and are separately defined for each assessed operation mode and the stator current, phase power and three-phase power signals. It is shown that, in principle, power signal based estimation algorithms can offer an inherent capability of estimation rate reduction. A novel adaptive sliding window algorithm is defined for open-loop operating conditions that enable estimation in a wide operating speed range while minimising the potential for undesirable overlap with PWM harmonics. The proposed algorithms have been verified and their performance limitations assessed in real-time experiments on three different industrial WRIM designs. It is shown that reliable real-time speed estimation in steady-state and transient operating conditions is possible at an improved estimation rate while maintaining a low estimation error.