Abstract：The predicting strategy based on fuzzy system is developed for the rise of temperature of connection fitting in value hall. Training data and testing data are obtained from the experiment of the rise of temperature of connection fitting in value hall. By training data,fuzzy system is trained by gradient descent algorithm,the model of rise of temperature is found,then the testing data is used to carrying out the test of fuzzy system and obtained reasonable result,the mean value of relative error is 8. 04%. Regression analysis is used to predict the rise of temperature of connection fitting,the predicting effect of fuzzy system is compared with that of regression analysis,the mean value of relative error of fuzzy system is 5. 56% lower than that of regression analysis. The result of prediction and comparison shows that fuzzy system possesses superiority on the prediction for the rise of temperature of connection fitting in value hall.

Keyword：connection fitting in value hall;fuzzy system;training data;testing data;