Park energy consumption optimization method based on power load data prediction

The invention discloses a park energy consumption optimization method based on power load data prediction, and solves the problem that the pressure of a power system is increased due to low park energy consumption efficiency and poor energy supply complementarity in the prior art. Establishing an improved K-means power utilization behavior clustering model of the users in the park; establishing a short-term load prediction model based on an LSTM algorithm; establishing a park energy consumption optimization double-layer decision-making model considering the demand response, the lower layer of the double-layer decision-making model being a park user energy consumption optimization decision-making model, and the upper layer being a park operator demand response pricing decision-making model; and solving the park user energy consumption optimization decision-making model by adopting linear programming, solving the park operator demand response pricing decision-making model by adopting a differential evolution me