Real-time wavelet-based distribution systems disturbances detection

Distribution feeders are prone to different types of events, such as disturbances and faults. Since these events are both unwanted and unexpected, their detection is essential for prompt system restoration, ensuring its reliability. This paper proposes the use of a power component, defined for non-sinusoidal conditions, as a disturbance detection index, which is estimated using the real-time stationary discrete wavelet packet transform. The proposed method analyzes the voltage and current of each phase separately, which are acquired at a single upstream measurement point, the substation. The performance and effectiveness of the proposed method are evaluated using events simulated in a test system based on a real distribution network. The results attest that the proposed method is suitable for detecting non-stationary disturbances.