Online fault detection and post-fault switching strategies to improve the fault-tolerance of matrix converters

Jaya Deepti Dasika, Purdue University

Abstract

A Matrix Converter (MC) is an ac-ac power converter which has been proposed for applications requiring a critical volume/weight ratio. However, those applications often require a high level of reliability/availability. Therefore, a fault-tolerant capability needs to be incorporated within the control/operation of the MC so that a fault occurrence does not lead to serious effects on the safety and system performance. Fault tolerance involves several steps including fault detection/location, isolation, and reconfiguration of the system so the converter can continue to operate as close as possible to its normal conditions. In this way, a partial failure in the converter does not lead to the shutdown of the whole system. The aim of this research is to improve the fault tolerance of the MC under open-switch fault conditions. This research develops two different fault detection methods, one based on monitoring load voltages and one based on monitoring load currents to detect and isolate a faulty switch. A post-fault modulation strategy in which, without inclusion of any additional hardware and based on the remaining healthy switches, the MC continues to operate under an open-switch fault, is proposed. The proposed post-fault strategy is extended to operate the MC under unbalanced input voltage conditions. Performance/effectiveness of the proposed fault detection methods and the post-fault strategies is evaluated based on simulations studies in the MATLAB/SIMULINK software environment and experiments.

Degree

Ph.D.

Advisors

Saeedifard, Purdue University.

Subject Area

Electrical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS