Forecast of Power Grid Material Demand Based on Random Forest

With the higher requirements of material management put forward by power grid companies, it is particularly important to carry out scientific and effective forecasting of electric power material demand, which is helpful for managers to make more effective decisions in the process of material management. However, the characteristics of various kinds of electric power materials, long purchasing period and great influence by policies increase the difficulty of forecasting the demand of electric power materials, and how to predict the demand of electric power materials more accurately becomes a big problem. In view of this, this paper constructs an electric power material demand prediction model based on random forest, and uses the electric power material using record of a power grid enterprise from 2011 to 2019 as a data set to carry out an empirical study. The results show that the model has high prediction accuracy and can provide model reference for material demand forecasting in power enterprises.