Data-driven random forest forecasting method of monthly electricity consumption

Accurate forecast of monthly electricity consumption has guiding significance for the economic dispatch of the power system, and it is also a prerequisite for the power company to formulate a reasonable sales plan. The traditional forecasting method of monthly electricity consumption performs poorly in processing the sequence of monthly electricity consumption with a dual trend, and it cannot consider multiple influencing factors at the same time and cannot screen the influencing factors of monthly electricity consumption. This paper proposes a random forest prediction method of monthly electricity consumption based on the maximum mutual information coefficient. First, the maximum mutual information coefficient between monthly electricity consumption and its influencing factors is calculated; second, high-relevance factors are filtered out based on the maximum mutual information coefficient value; third, the data of high-relevance factors are combined, and random forest is used to predict monthly electricity consumption; finally, the program of the abovementioned method is compiled in Python language with the electricity consumption data of the whole society in Shenyang, Liaoning Province as an actual calculation example, and the method is compared with the method that does not use correlation factor identification. Simulation results show that the proposed method has high prediction accuracy and can provide a basis for making reasonable grid operation plans and making power decisions correctly.