The LM algorithm is used to improve the BP neural network to enhance the convergence speed of the model,reduce the number of iterations.The LM-BP neural network is used to train and test samples of air-conditioning energy consumption simulation data in summer and winter respectively to construct a time-by-time air-conditioning energy consumption prediction model for office buildings.The results show that the MAPE error between the predicted value and the simulated value is 5.74％on average in summer model,and 5％in winter model,both of which are within 10％,which indicates the feasibility of using LM-BP neural network in predicting building air conditioning energy consumption.The model is verified on the basis of the actual air-conditioning energy consumption data of an office building in Chengdu area.The average relative error between the predicted value and the actual value is 6.6％,indicating that the established energy consumption prediction model can meet the needs of practical projects,and the accurate electricity prediction data can provide scientific decision-making basis for the actual electricity market transactions of office buildings.
air-conditioning energy consumptionLM algorithmneural networkoffice building
维普
万方数据
