Transformer hotspot temperature predicting method for multi-working-condition-parameter recognition and optimization

Abstract

The invention discloses a transformer hotspot temperature predicting method for multi-working condition-parameter recognition and optimization. The method includes the following steps of collecting test data of a transformer in real time, dividing test data into a training set and a prediction set according to running working conditions of the transformer, selecting obviously-changing parameters of an IEEE model under different working conditions as to-be-recognized and optimized parameters, establishing a parameter recognizing and optimizing target function of the IEEE model under various working conditions, searching for the optimal solution of the parameter recognizing and optimizing target function under various working condition through a search method on the basis of the training set of various running working conditions, substituting the recognizing and optimizing parameters under various working conditions, and predicting test data of the prediction set under each working condition through the IEEE model to obtain a transformer hotspot temperature prediction sequence. High-precision hotspot temperature prediction value can be obtained, the dynamic change tendency of hotspot temperature can be reliably analyzed, and the precision of the transformer on the aspect of thermal characteristic measurement is improved.

Description

technical field [0001] The invention relates to a transformer hot spot temperature prediction method with multi-working condition parameter identification and optimization. Background technique [0002] As one of the most important equipments in the power system, the power transformer deserves attention for its operation safety and economy. Transformers allow short-time overload operation. Under the premise of accurately and reliably predicting the temperature of the top oil temperature of the transformer, increasing the temperature rise limit of the transformer by 2% to 3% will bring huge economic benefits. The hot spot temperature of the transformer is the biggest limiting factor of the dynamic operating capacity of the transformer. When the hot spot temperature exceeds the safety limit, the insulation will be damaged and the safe operation of the transformer will be affected. Therefore, it is very important to accurately and reliably predict the hot spot temperature of th...

AI Technical Summary

Problems solved by technology

The thermal circuit model requires more transformer heat transfer parameters; the neural network method lacks clear physical meaning; the IEEEStd C57.91 recommended model is a traditional method, and the model parameters are relatively easy to calculate, but there is a problem of low accuracy, especially in Under load conditions, the prediction accuracy is severely reduced, and the predicted value under overload conditions is lower than the actual measured value, which increases the risk of transformer operation beyond the nameplate

[0004] Therefore, none of the existing models considered the predicted safety margin of the transformer under overload conditions

Images