Because the short-term power load and power consumption are affected by many complicated non-linear factors

the traditional single BP neural network overall prediction method has problems such as low accuracy and slow convergence. In order to improve the convergence speed and prediction accuracy

it is divided into long-term and short-term influencing factors according to the characteristics of the influencing factors

and it is divided into reference and sensitive quantities according to the characteristics of the load and power consumption curves.Finally the required short-term influencing factors are determined by using the determination coefficient method.The genetic algorithm is used to optimize the initial weight and threshold of the BP neural network

and the BP neural prediction error is used as the fitness function of the genetic algorithm. An improved BP neural network short-term power prediction method based on characteristic analysis is established. The data of a winter period during the five-year peak period of from 2015 to 2019 in a central province is selected for verification

and the results show that the accuracy and convergence rate of the prediction method are improved to meet the requirements.