Fast detection of short circuits in low-voltage AC systems based on wavelet packet transform and current change rate

Abstract

The safe and stable operation of a new low-voltage power system requires the research and application of fast detection technology of short-circuit faults. However, the existing methods are difficult to effectively realize the rapidity and reliability in case of short circuits with unclear features and various interference conditions. Therefore, the fault feature of short-circuit current in the full phase angle range is analyzed, and a new method of short-circuit detection based on wavelet packet transform and current change rate is proposed. Among them, the fast-action detection method of wavelet packet transform is used for faults with obvious features to achieve detection rapidity, and the double criteria backup detection method of wavelet packet transform and current change rate is used for faults with unclear features to ensure detection reliability. On this basis, the hardware technology realization of short-circuit fault fast detection algorithm is studied. For short circuit faults, load startup, load switching, and other situations, the real low-voltage AC system experiment is carried out to verify. Finally, the test shows that the detection method can effectively and rapidly detect short-circuit faults within 1ms in the full phase angle range, and has robustness under interference conditions.