A Hybrid Prediction Model for Air Conditioning Load based on VMD-GA-SVM

Aiming at the problem that it is difficult to realize the accurate prediction of air-conditioning electric loads due to the fixed office hours, users' behaviors and the influence of meteorological parameters of certain buildings, a hybrid prediction model of air-conditioning loads based on variational modal decomposition-genetic algorithm-support vector machine is proposed, which fully exploits the non-smoothness characteristics of loads through VMD decomposition and effectively combines the advantages of genetic algorithms and support vector machine, and the results of the arithmetic examples show that the method can improve the prediction accuracy effectively.