Model-Based Fault Diagnosis Applied to 6-pulse Cycloconverter

The motivation for this work is the need of study and reduction of current problems and faults affecting high- power AC drive systems. Availability and reliability of power electronics systems has a tremendous importance in mining applications, which is the reason why the knowledge on the fault mode behaviour of an AC drive system is extremely important from the idea of improving system design, operation, protection, and fault tolerance control. This work presents a model-based fault diagnosis approach for a 6-pulse cycloconverter drive system based on output voltages surveillance and fault model process. The proposed strategy is mainly focused to detect, isolate and identify faults originated in the semiconductors of the converter. To probe the fault diagnosis strategy and its behaviour, the algorithm was implemented on fixed-point DSP for a 5KVA laboratory prototype of a cycloconverter system.