In order to improve the power meteorological service and provide human comfort index prediction according to the demand of power load forecasting in Tianjin. We use the power load data and meteorological data in summer from 2002 to 2005 in Tianjin to calculate the human comfort index. The results of various models and the corresponding relationship between mean temperature and power load are analyzed in detail. The results show that some models are more accordant with the feeling of body comfort and has better linear dependence with meteorological power loads, such as UTCI (Universal Thermal Climate Index), Lyu Weilin outdoor model and Li Yuan model. In order to build a better human comfort index model for Tianjin, the least squares method is applied to correct the Lyu Weilin model, which is widely used in China.