Research on Fault Diagnosis of Electric Power Collection Equipment Based on Big Data Analysis

As the application of power consumption information collection system continues to deepen, the business scope continues to expand, and the functional applications continue to increase, the existing abnormality monitoring and diagnostic model can no longer meet the data analysis application needs of metering device status detection, efficient operation and maintenance of the collection system, and equipment quality evaluation. Existing operation and maintenance model, due to the lack of fault intelligent diagnosis and screening technology support, only when the equipment failure, can find the problem, due to the huge number of equipment, complex operating environment, fault characterisation is difficult, in the discovery of individual equipment problems, can not be related to the operation of the hidden danger of the equipment to do in advance screening and prediction, and can not effectively guard against possible systematic, continuity of the equipment operation problems. In this paper, by collecting the equipment attributes, operation attributes and external attribute influence factors of the collection equipment, using data dimensionality reduction, statistical average method and gradient boosting tree (GBDT) algorithm, combined with remote fault intelligent diagnosis and manual diagnosis, comprehensively carry out the research on the collection fault intelligent diagnosis model, and achieve the practicality of a number of contents such as the precise research and judgement of the collection faults, the qualitative analysis of the equipment faults, and the early warning, so as to provide guidance for the front-line staff and to reduce the operation problems. It provides guidance for on-site operation and maintenance, reduces the difficulty of operation and maintenance, and improves the quality and efficiency of operation and maintenance.