Fault Detection Method in Micro-Grid Multi-Pulse Thyristor Rectifier Circuit

Multi-pulse rectifier unit can effectively limit the number of harmonics injected into the grid frequency in micro-grid, but thyristor open-circuit or short-circuit fault may occur if don´t handle in time, the power side and load side will have serious consequences. Studying fast detection device of the thyristor fault has an important significance for rapidly inhibiting the spread of fault effects, reducing the maintenance and repair difficulty, improving reliability and maintainability of power transformation device. This paper studies the fault classification and diagnosis method of the bridge 12-pulse waveform controlled rectifier circuit based on Support Vector Machines. All the faults in thyristors are analyzed, the fault classification method according to the rectifier aberrant voltage waveforms is put forward, fault voltage waveforms are summarized. By keeping the relations between faults and waveforms in a support vector machines model, the support vector machines model can be trained to detect faults and realize automation of fault diagnosis.12-pulse rectifier circuit fault is presented as an example, coupled with the results of a certain power electronic circuit experiment, the test result proves the system obtained better effect about the fault diagnosis and the method is correct and feasible.