Microgrid fault detection based on wavelet transformation and Park's vector approach

This paper presents a new algorithm for fault detection in microgrids application. This algorithm is used in real time by applying the dq0 and wavelet transformation over local measurements. This approach consists of transforming three-phase voltage or current signals into dq0 components to analyze its behavior during faults in order to then recognize patterns that indicate the start of a fault. By filtering one of the dq0 components through the wavelet transformation, and by isolating band frequencies of interest, faults can be detected using the finite difference between samples of the filtered signal. In order to demonstrate the algorithm's performance, we have performed an experiment on a scaled microgrid and worked with the data obtained from this simulation. This algorithm will likely impact the development of new microgrid architectures with fewer resources directed to power monitoring.