Wavelet based fault tolerant control of induction motor / Khalaf Salloum Gaeid

Gaeid, Khalaf Salloum (2012) Wavelet based fault tolerant control of induction motor / Khalaf Salloum Gaeid. PhD thesis, University of Malaya.

Preview

PDF (Full Text) Download (4Mb) | Preview

Abstract

IMs are an important part of the industry as they provide convenience through their simple construction and a minimum maintenance. In addition they are highly rigid and reliable. They are mostly used in compressors, fans and pumps, but their usage has progressed to a much advanced level such as, aircrafts, space shuttles and military applications. These operations are crucial and cannot tolerate compromise in reliability and safety. This thesis aims at exploring the fault tolerant control of the IMs. This involves sensor based closed loop vector control for a healthy IM, Variable frequency with constant flux(V/F) closed loop for both stator open winding and stator short winding faults , V/F open loop to control the drive in case of minimum voltage fault and using the sensorless vector control in case of encoder faults. The model that has been suggested for the fault tolerant control has been verified through a 0.5 hp IM with the inclusion of condition monitoring and protection against the above faults. A platform of more than 13 trips has been included in this work. The faults will be classified according to the location, severity and the time. The speed for the sensorless vector control and closed loop V/F controllers is evaluated through the model reference adaptive control estimator. There are two steps in the fault tolerant control process; in the first step the fault is identified with the feature extraction module, a fault decision module and a feature cluster module. After that the controller is redesigned accordingly. In this work however, the fault protection is an additional feature of the control system, where the wavelet based fault tolerant system has been tested and simulated using a 1kW IM, which has a short stator winding and sensor faults, while the primary faults are open stator winding. The additional fault presented to the system is under voltage. The wavelet packed decomposition covers the transient and steady state regions of the IM operations with respect to both time and frequency domain. The motor features are extracted through the proposed Discrete Wavelet Transform (DWT) based analysis method, while the wavelet inlet acts as an expert tool to adapt the right controller in accordance with the fault type. The fault detection algorithm identifies the time and location of each fault. The optimal levels of decomposition of the stator current error signal and mother wavelet function are selected with the help of the maximum entropy and description length data. The effect of faults and the effectiveness of the fault tolerant algorithm is demonstrated by observing the speed response of the induction machine, which is the strategy adopted by the majority of researchers in this area. The reliability and the effectiveness of the proposed DWT based fault tolerant controller will be confirmed by the simulation and experimental results on the 1kW IM. This will further ascertain the effectiveness of the controller for high performance motor drive applications in the industry.

Actions (For repository staff only : Login required)