基于灰色神经网络模型的区域供热负荷预测研究

通过灰色关联分析法对区域供热负荷影响因素进行了评价,并将灰色预测与BP神经网络算法相结合,建立了灰色神经网络结构,能够对影响供热负荷的因素进行筛选,并对供热负荷进行预测。对某区域供热负荷进行了供热负荷预测与验证,通过对比筛选不同影响因素灰色神经网络的预测结果与误差,表明灰色神经网络模型在热负荷预测中能够选择合适的影响因素,排除关联度低的影响因素,可提高供热负荷预测的准确性,为区域供热负荷的预测提供理论依据。

Prediction of district heating load based on grey neural network model

Evaluates the influence factors of district heating load by grey relativity analysis,and establishes a structure of grey neural network by combining grey prediction with BP neural network algorithm,which can screen the factors affecting the heating load and predict the heating load.Predicts and verifies the heating load of a district heating system.By comparing and screening the prediction results and errors of grey neural networks with different influence factors,the results show that the grey neural network model can select the appropriate influence factors and exclude the influence factors with low relativity degree in heating load prediction,improve the accuracy of heating load prediction,and provide a theoretical basis for district heating load prediction.