Windowed Lift Wave Packet Transform Algorithm And Application On Fault Location

Transmission lines are essential parts of the power systems, attaching power plants and end-users. Meanwhile power lines are the places where the most faults occur in power system, and the fault seeking is very difficult because these lines are mostly working in harsh circumstance. When power line faults occur, correctly locating the fault point and quickly removing the fault would be critical for repairing power lines and restoring power supply in time. Moreover, it would also play an important role in safety stability and economical operation of the whole power systems.For traditional double-ended fault location algorithm based on wavelet transform, it is found that calculating quantity is large and error is obvious if wavelet transform is done on all of the data directly, through analyzing the Mallat wavelet decomposition algorithm. Thus, windowed wavelet transform algorithm is proposed firstly. In this method sampling data was multiplied with window function, filtering the low-frequency effectively and reducing the number of decomposition layers, could locate the accurate time when singularity happens quickly. This method has the features of less calculating quantity and quicker speed, comparing with the classical Mallat algorithm. However, the algorithm above only decomposes the low frequency element in each layer and precision is affected, so windowed wavelet packet is presented to promote the locating precision, which decomposes both low-frequency and high- frequency elements. The simulation result is effective. But because its calculation quantity is increased too much and the speed is decreased, windowed lift wave packet transform is designed in the end, in which the lift scheme could improve the speed obviously. The simulation result shows that it could find the singularity point on the time domain of signal, and locate the accurate time when singularity happens quickly. The method could not only reduce calculating quantity and improve locating speed, but also have high precision.