Large-scale experimental device power device fault diagnosis method based on deep convolutional neural network

Abstract

The invention discloses a large-scale experimental device power device fault diagnosis method based on a deep convolutional neural network. The method comprises the steps of collecting the historicalfault data according to online monitoring data of a power device; forming an initial sample set, carrying out data preprocessing; obtaining a normalized sample set; mining the hidden fault informationdeeply by using the deep convolutional neural network; and then adjusting the internal weight parameters of a fault diagnosis model according to the deviation between the predicted fault type and thereal fault type of the model, and finally carrying out performance testing on the power device fault diagnosis model, so that the fault diagnosis accuracy of the power device fault diagnosis model based on the deep convolutional neural network is further improved. Whether a fault occurs or not can be accurately judged according to the power device monitoring data, the fault type is output, and the corresponding fault solving method is obtained according to the fault type, so that the power device system can rapidly and effectively recover to a normal working state.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to a method for diagnosing faults of electric equipment of large-scale experimental devices based on a deep convolutional neural network. Background technique [0002] At present, the inspection and maintenance of the power equipment of large-scale experimental installations is carried out by means of on-site patrol by the staff. Relying on the staff to find work faults on the spot has a large workload, and the probability of finding faults is low. Therefore, for this type of maintenance The plan is not comprehensive, which makes it difficult to detect small faults in time and cause more serious consequences. [0003] Diagnosis methods for power equipment faults in the prior art are mainly divided into three categories, including three types based on mathematical models, based on digital signals and based on machine learning. Among them, the fault diagnosis method b...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide a method for diagnosing faults of large-scale experimental equipment power equipment based on deep convolutional neural networks, thereby solving the aforementioned problems in the prior art

Images