Transformer hot spot temperature time sequence prediction method based on data mining algorithm

Abstract

The invention relates to a transformer hot spot temperature time sequence prediction method based on a data mining algorithm, and belongs to the technical field of transformers. According to the technical scheme, the method comprises the steps: establishing a local area load prediction model based on support vector regression according to the fact that the characteristic quantity of the transformer hot-spot temperature and the load rate has the maximum correlation, and taking the local area load prediction model as front input of the transformer hot-spot temperature prediction model; predicting the hot spot temperature of the transformer at the next moment by using a support vector regression-based time sequence model with external input. The invention has the beneficial effects that the hot spot temperature of the transformer at the next moment can be accurately predicted, and the method has the advantages of being high in learning speed and good in generalization performance, so theprediction precision of the hot spot temperature of the transformer is improved; the operation efficiency is high, the prediction precision is high, and a good data basis is provided for dynamic capacity increase of the transformer.

Description

technical field [0001] The invention relates to a time series prediction method for transformer hot spot temperature based on a data mining algorithm, and belongs to the technical field of transformers. Background technique [0002] During the operation of power transformers, the internal temperature environment is complex and affected by many factors. The hot spot temperature of the transformer is usually generated on the low-voltage winding side. The hot spot temperature limits the running time and service life of the transformer to a certain extent. Its accurate calculation or Prediction becomes a key factor affecting the decision-making of transformer dynamic capacity increase. [0003] At present, the calculation methods of hot spot temperature can be generally divided into the following two categories: one is the thermal circuit model method; the other is direct calculation according to the "Guidelines for Oil-immersed Power Transformer Load" given by the national stan...

AI Technical Summary

Problems solved by technology

These prediction methods are usually similar to more accurate calculation results, and have little significance for the hot spot temperature trend of the transformer at the next moment

Images