A fault prediction method and system for a power distribution device

Abstract

The invention discloses a fault prediction method and system for power distribution equipment, which can realize the evaluation and prediction of the future health state of the equipment. First, collect the state data of power distribution equipment with various usage periods in a region, and convert them into status data under no-load conditions; calculate the mean value of the status data of power distribution equipment with the same usage period as the status corresponding to the usage period data z (m) , use the model to calculate the health index y corresponding to the usage time m (m) ;z (m) and y (m) Constitute the training samples; use the hidden Markov model to model the state data and health index of the power distribution device, and use the self-circulation neural network to realize it. The input of the self-circulating neural network is the state data z at time m (m) and the health index at time m-1, the output is the health index at time m; the training samples are used to train the self-circulating neural network; finally, the self-circulating neural network can be used to realize the prediction of the health index of power distribution equipment.

Description

technical field [0001] The invention relates to the technical field of fault detection, in particular to a fault prediction method and system of a power distribution device. Background technique [0002] With the increasing dependence of society on electricity, once the power distribution device fails, it will bring huge losses to the production enterprises. Therefore, the traditional fault detection technology lacks in advance, and there are hidden dangers in safety, while the regular fault inspection has the problem of excessive maintenance and wastes manpower and material costs. Therefore, the use of fault prediction technology to replace the traditional fault diagnosis mode will become the trend of technology development. [0003] For the fault prediction of power distribution equipment, the difficulty is that it is difficult to obtain the historical state data of the equipment to be predicted. If the current state data is used to evaluate and predict the future health ...

AI Technical Summary

Problems solved by technology

[0003] For the fault prediction of power distribution equipment, the difficulty is that it is difficult to obtain the historical state data of the equipment to be predicted. If the current state data is used to evaluate and predict the future health state, the data based on it is too single, which is not conducive to guaranteeing the fault. predictive validity

Images