Research on Power Grid Contact Aging Prediction Based on Variable Modified Grey Model

Research of grid contact is a hot topics of electrical power system. Serious aging of grid contacts will bring accidents or even fire. At present, the main research method of contact aging is grey model. It has the characteristics of convenient calculation and high prediction accuracy. However, the locality of factor variables is not considered in Traditional single variable grey model. For the multivariate gray model, it has high arithmetic and its prediction accuracy is too low. For this purpose, a Variable revision gray model is put forward in this paper. Firstly, Key factor variables is extracted for Modified variable. Secondly, Single behavior variable grey model is used to revise multivariate gray model. Finally, the proposed variable modified grey model is used in Prediction of contact aging. The results show that: In a complex environment, prediction accuracy of variable modified grey model much better than Traditional forecasting model. Up to 95.3%. Furthermore, the model can also be used in the occasion that Variable effects have a significant difference in multivariate prediction. Keywords—contact aging; grey model; variable correction