基于BP神经网络的城市燃气月度负荷预测研究

The Study on the Monthly City Gas Load Predicting Based on BP Neural Network

应用MATLAB科学计算软件编制了对浙江省某城市燃气月度负荷的预测程序,建立了燃气月度负荷预测的BP神经网络模型,对模型进行训练后,利用它对该市燃气月度负荷进行了预测,预测结果的相对误差最在±6.34%以内,由预测结果可知该模型预测精度较高,完全可以达到工程实际应用的目标。

This paper gave the procedure using which to predict monthly gas load for a city in Zhejiang province with utilizing software of MATLAB,and built a BP neural network model to predict city gas load. After training this model, it was applied to predict monthly gas load for this city, the results showed that the relative error is under ±6.34%, so this model have a high accuracy and absolutely it can be applied to engineering and accomplish the goal.