Hybrid load system energy control method and system for low-carbon park

The invention discloses a hybrid load system energy control method and system for a low-carbon park, and the method comprises the steps: carrying out the variable selection through an LASSO method, and obtaining a meteorological variable factor which has the greatest impact on the load of a comprehensive energy system; preprocessing the meteorological variable factors and converting the meteorological variable factors into three-dimensional tensor data, and inputting the three-dimensional tensor data into an LSTM layer for processing to obtain a load prediction result; optimizing the hybrid load prediction model of the LSTM by using a PSO algorithm; and performing optimization control on the integrated energy system by using a mixed integer linear programming framework through analyzing a load prediction result, the actual working condition of internal optimization operation of the integrated energy network and the uncertainty and limitation of overall equipment operation. By means of high-precision predictio