Intelligent Maintenance Decision Model of Distribution Transformer Based on Multivariate Data Fusion

Abstract

In order to effectively reduce the failure rate of distribution transformer and save maintenance cost, an intelligent maintenance decision-making model of distribution transformer based on multivariate data fusion is proposed. By effectively integrating deep belief network and D-S evidence theory, a multi-level decisionmaking model for intelligent maintenance of distribution transformer is constructed. The fault location and specific causes of the fault are analyzed layer by layer. The multi-dimensional data of distribution transformer fault are extracted and classified by deep belief network. The uncertainty problem in fault maintenance is solved by combining D-S evidence theory, The intelligent maintenance decision of transformer is realized. The experimental results show that: after the application of the model, the total positive judgment rate of distribution transformer maintenance is high, up to 96.60%; the maintenance cost is low; and the failure rate of transformer can be significantly reduced, the maintenance times are reduced, and the overhaul of equipment is avoided, and the service life of distribution transformer is extended. It can be proved that the model has good maintenance decision-making effect.