Prediction of the Heat Load in Central Heating Systems Using GA-BP Algorithm

Abstract

This paper presented the research on heat load prediction method of central heating system. The combined simulation data at Xi'an in January was used as the samples for training and predicting. This paper selected the daily average outdoor wind speed, the daily average outdoor temperature, date type, sunshine duration as input variables and the heating load value as output variable. After preprocessing of the historical data, the BP neural network algorithm and the GA-BP algorithm were employed to predict and verify heat load respectively. Based on the analysis of prediction results, it showed that the error between the predicted data and the actual value using the BP algorithm is large (maximum:-39.8%) and not suitable for heating load prediction while the error between the predicted data and the actual value using the GA-BP algorithm is small (maximum:-16.6%) and within the acceptable range. This paper provided a feasible method for heating load prediction.