Research on Energy Efficiency Improvement Strategy of Air Conditioning in Comprehensive Energy System Based on Neural Network Algorithm

To enhance the precision of the integrated energy system's operational approach, the study suggests utilizing a BP neural network for forecasting thermal and cooling demands, thereby minimizing forecast inaccuracies. This paper develops models for ice storage, gas boilers, and ground source heat pump cooling and heating equipment outputs. To enhance system efficiency, an energy efficiency target function is established. Simulation analysis shows that the BP neural network reduces prediction errors more effectively than exponential smoothing and gray prediction methods. The optimized operation model for the integrated energy system boosts the air conditioning system's comprehensive energy efficiency from 3.2 to 3.7.