计算机应用与软件2011，Vol.28Issue(12)：206-209,4.
基于ARIMA和神经网络模型的城市燃气负荷预测
ARIMA AND NEURAL NETWORK MODEL BASED CITY GAS LOAD FORECASTING
摘要
Abstract
City gas load forecasting is an important stage in city gas deployment. On the basis of wavelet periodic analysis over gas load time series, an ARIMA-based neural network temperature error correction model for gas load is built. ARIMA model smoothens annually periodic data to effectively eliminate the short-term impact in the past; it regards the atomosphere temperature as neural network input to rectify ARIMA model predicted values. It is proven by testings that the model well reveals the characteristics of gas load time series. It performs fine on forecasting.关键词
自回归移动平均模型/神经网络/小波/燃气负荷/时间序列/预测模型Key words
ARIMA/Neural network/Wavele/ Gas load/Time series/Forecasting model分类
信息技术与安全科学引用本文复制引用
黄岳嵘,徐晓钟,张益铭,王劲松..基于ARIMA和神经网络模型的城市燃气负荷预测[J].计算机应用与软件,2011,28(12):206-209,4.基金项目
上海师范大学产学研究项目(DCL200801). （DCL200801）
