This paper aims to explore the construction and application of the mathematical model of energy economy and to make the most accurate prediction of energy demand. By analyzing various factors affecting energy demand, it constructs a predictive index system and improves it with neural network model and particle swarm algorithm to predict future energy demand. It finds that accurate predictions can be made with this approach, and that the construction and application of the mathematical model of energy economy can provide an important theoretical basis for reasonable energy development strategies.
