Early Failure Detection Method for Power Distribution Network Equipment Based On Human-Level Concept Learning

The invention discloses a method for early fault detection of power distribution network based on humanoid concept learning. The method comprises the following steps: decomposing a waveform into an approximate part and a detail part by utilizing wavelet transform, wherein the approximate part is called an approximate shape primitive, and the detail part is called a distortion primitive (S1); splitting, according to an extreme point, the distortion primitive into three primitives, i.e., harmonic wave, pulse and other distortion (S2); extracting the features of the primitives and the time relationship between the primitives (S3); constructing the probability distribution of the waveform according to the features of the primitives and the time relationship between the primitives (S4); and obtaining a determination result of the waveform according to the probability distributions of different types of waveforms (S5). According to the method, a voltage or current waveform is considered as one of visual concepts and decomposed into an approximate shape and various distortions, and the overall probability distribution of the waveform can be obtained by calculating the probability distribution of each component, so as to determine the type of the waveform. The method is greatly superior to the traditional detection in the aspects of required data volume and accuracy, and is of great significance to early fault detection and handling of the power distribution network.