国产欧美精品一区二区,中文字幕专区在线亚洲,国产精品美女网站在线观看,艾秋果冻传媒2021精品,在线免费一区二区,久久久久久青草大香综合精品,日韩美aaa特级毛片,欧美成人精品午夜免费影视

The set value of district heating system’s subsystem depends on forecasting heat load, providing accurate heat load prediction is the basis to improve the heating quality. For the traditional heat load prediction, the outdoor temperature was considered as the only one influence factor, the heat load prediction model’s parameters identification depend on traditional mathematical tools, and the model’s accuracy isn’t high. In order to predict the heat load more accurately, the multiple weather factors’ influences on the district heating system heat load were researched in this paper, moreover, the multiple regression method was used in this paper, and then, the district heating system heat load prediction models were established by Levenberg Marquarat-Universal Global Optimization (LM-UGO) algorithm in 1stOpt software. The experimental results shown that the outdoor temperature has a direct influence on the heat load, the wind speed or solar radiation has a direct influence on the outdoor temperature so as to impact the heat load indirectly. The results shown that the multiple regression fitting’s square correlation coefficient were all greater than 0.900 0, model training and testing’s mean absolute percentage error were all less than 4.00%. This application indicated that the training and testing process of the heat load prediction models were reasonable, it proved that this multiple regression method can be promoted into the heat load prediction field.