Reliability Improvement of Matrix Converter Drive Systems

碩士 === 國立臺灣科技大學 === 電機工程系 === 99 === The matrix converter system is becoming a very promising candidate to replace the conventional two-stage ac/dc/ac converter. Since its commercialization, modulation techniques and commutation issues for the matrix converter have been thoroughly investigated and h...

Full description

Bibliographic Details
Main Authors: Khiem Nguyen-Duy, Nguyen Duy Khiem
Other Authors: none
Format: Others
Language:en_US
Published: 2011
Online Access:http://ndltd.ncl.edu.tw/handle/me2nsx
Description
Summary:碩士 === 國立臺灣科技大學 === 電機工程系 === 99 === The matrix converter system is becoming a very promising candidate to replace the conventional two-stage ac/dc/ac converter. Since its commercialization, modulation techniques and commutation issues for the matrix converter have been thoroughly investigated and have reached their maturity. This is because this direct ac/ac converter possesses several key advantages such as: inherent four-quadrant operation, nonexistence of bulky dc link electrolytic capacitors, high-quality input power factor, and improved power density. However, system reliability remains an open issue and has not received much attention. The most common reliability problem is that a bi-directional switch has an open-switch fault during operation. In this thesis, a matrix converter driving an adjustable speed permanent magnet synchronous motor is examined under both single open-switch fault and two open-switch faults. First, a new fault detection method is proposed for the situation of single open-switch fault using only the motor currents. Second, novel fault-tolerant switching strategies are presented. The proposed methods utilize the concept of modeling the matrix converter as a two-stage rectifier/inverter. With the proposed methods, existing modulation techniques for the inverter stage can be reused, whereas the rectifier stage is modified by control to counteract the fault. The principal advantage is that the proposed techniques require no additional hardware devices or circuit modifications to the matrix converter. Experimental results show that the proposed method can maintain the motor speed with a maximum ripple of 2% - a five-fold improvement over the uncompensated system. The proposed method, therefore, offers a very economical and effective solution for the matrix converter fault tolerance problem.