Transformer state risk assessment method and device based on ensemble learning

Abstract

The invention provides a transformer state risk assessment method and device based on ensemble learning. The transformer state risk assessment method comprises the steps that S1, original data of thetransformer state are obtained, and a transformer state equilibrium data set is obtained through a mixed sampling equalization algorithm based on probability distribution estimation; S2, by combiningwith a cost-sensitive learning algorithm, a decision tree standard algorithm is optimized, and a cost-sensitive decision tree optimization algorithm is obtained; S3, based on the state equilibrium data set and the cost-sensitive decision tree optimization algorithm, cost-sensitive decision tree optimization classifiers are obtained, and a transformer state risk assessment model is constructed through an ensemble learning algorithm combined by AdaBoost and the multiple cost-sensitive decision tree optimization classifiers; and S4, whether a transformer has an urgent and serious defect in the next cycle or not is assessed through the transformer state risk assessment model. The technical problem that the model generalization capability is low due to low quality and low value density of existing sample data is solved.

Description

technical field [0001] The present application relates to the field of power grid risk assessment, in particular to a transformer state risk assessment method and device based on integrated learning. Background technique [0002] Transformer condition risk assessment is an important topic for the safe and stable operation of power equipment and an important basis for transformer condition-based maintenance, which directly affects the quality and effect of condition-based maintenance. At the same time, with the gradual expansion of the scale of power grid construction, more and more transformers have been put into use. The risk assessment of transformers is directly related to the fundamental interests of power grid construction and public electricity consumption. [0003] At present, the transformer status evaluation business is carried out under the guidance of the power grid company enterprise standard "Guidelines for Oil-immersed Transformer (Reactor) Status Evaluation", ...

Images