基于数据挖掘的地铁车站热湿特征抽取

Aiming at reducing the high proportion of energy consumption for environment control system (ECS) in urban rail transit system (UTS), the energy efficiency improvement strategy of ECS operation is explored from the perspective of thermal and humid environment of subway stations. K-means was adopted to cluster the annual average daily temperature and humidity data in each station of the metro company, and then the influence of physical attributes such as spatiotemporal distribution and burial depth of each clustered station were analysed. The results show as follows. The temperature and humidity curves of the same line and adjacent stations are quite different. The second-class temperature clustering of the stations is colder in winter and hotter in summer, and the comfort is the worst. The southwest monsoon from November to December cause the fourth temperature clustering stations to be relatively low, which can be adjusted strategically. After conducting the standard deviation analysis on the indoor and outdoor temperature and humidity differences, it is found that the curve fluctuation is related to physical characteristics such as burial depth and azimuth, and showed strong correlation. The ECS of different stations should be operated in a differentiated manner due to differences in physical characteristics, and different environmental control energy consumption quota standards should be set for different station categories. This research provides a basis for the subway stations design and the power selection of ECS, it has positive environmental significance for the dark green operation of UTS and the construction of low-carbon cities.