Abstract: By using the daily icing data of Erlangshan Mountains in Sichuan Province during three winter seasons provided by Southwest Electric Power Design Institute and related meteorological data, methods of correlation analysis, regression analysis are used to study the meteorological factors and icing index associated with the icing over Erlangshan Mountains. From the view of physical concept, the main meteorological factors influencing the strength of icing conductor and the microscopic process of conductor icing are analyzed by the fluid mechanics and thermodynamic principles. The theoretical framework of the rime icing conductor model and the glaze icing conductor model are built. By analyzing icing data and conventional meteorological data, it is found that the icing density is not well correlated with other meteorological elements except for temperature. An icing density model is established by the method of nonlinear regression analysis whose fitting correlation ratio is 0.5652. This density model can reflect the real icing density. According to the relationship between the visibility, extinction coefficient and liquid water, the liquid water content and its transportation can be estimated by the visibility.To choose the most proper scheme for Sichuan mountain climate characteristics, liquid water content estimated by different model is compared using observed data. Combined with the observed icing data, Kunkel scheme is selected for fitting test. As a result, it is found there is a good correlation between the growth of icing thickness and the liquid water content. The conductor icing thickness is decomposed into the rime icing which is formed by horizontal droplet and the glaze icing formed by precipitation. An icing process efficiency index is defined for further transformation of the theoretical models of conductor icing.From physical significance analysis, temperature and wind are main influencing factors of the icing process efficiency index. But the icing process efficiency index is not significantly correlated with temperature and it is related to the wind speed with exponential relationship. Based on these results, an icing conductor model is established to achieve the purpose of engineering application. From the analysis of the model fitting results, this icing conductor model can reflect the actual icing conductor on the whole, especially reflecting the trend well. The correlation coefficient between the icing thickness and fitting thickness is 0.8340, and the mean square error of icing thickness is 28.61 mm.