Advanced Monitoring and Parameter Estimation Techniques for Doubly Fed Induction Generator Drives

Abstract

This thesis investigated the development of condition monitoring techniques for generator mechanical fault diagnosis and sensorless speed estimation in a doubly fed induction generator (DFIG) utilizing the DFIG closed loop controller signals. The DFIG harmonic model was first implemented in the MATLAB/Simulink platform to investigate shaft misalignment fault signatures for fault diagnosis as well as harmonic content selection for sensorless speed estimation. Conductor distribution function approach was employed to predict the frequency content of the electrical signals including terminal signals and control loop signals of a DFIG. By analysing the characteristics of DFIG signals, the fault-specific spectral signatures and speed related harmonic contents could be effectively recognized thus further employed for DFIG condition monitoring. To validate the findings in the DFIG harmonic model and evaluate the experimental performance in research, a DFIG test-rig facility using industrial converters along with associated sensors and real-time platform was developed. The frequency contents of the DFIG terminal signals and control loop signals in different operating load and speed conditions acquired from experimental tests were examined and cross-correlated with the DFIG harmonic model. After identifying the spectral nature of the DFIG control loop signals, their potential to be used as a signature for shaft misalignment recognition and sensorless speed estimation in DFIG system were investigated. Fault identification capability of the control loop signals were examined in a range of different load and speed operating points and indicated better performance compared to typically used machine terminal signal. The spectral based searching algorithm based on control loop signal MMF harmonic identification was validated with real-time DFIG constant and variable speed scenarios, which manifested a considerable potential for actual speed estimation for wind turbine applications. Moreover, a novel DFIG shaft misalignment diagnostic technique utilizing Fibre Bragg Grating (FBG) to monitor machine frame strain was developed. Practical strain sensing tests in different load, speed and severities were undertaken and cross-correlated with vibration sensing. It was shown to have the capability to provide a sufficient means of monitoring misalignment induced effects.

Date of Award	31 Dec 2021
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Alexander Smith (Supervisor) & Sinisa Durovic (Supervisor)