城市燃气管网长期规划用气量的预测  

Gas Consumption Forecast for Long-term Planning of City Gas Pipeline Network

作　　者：袁树明[1] 赖建波[1] 张德坤 马俊峰[1] 

机构地区：[1]中国市政工程华北设计研究总院,天津300074 [2]广州市煤气公司,广东广州510060

出　　处：《煤气与热力》2009年第5期24-27,共4页Gas & Heat

摘　　要：天然气用量与城市国民经济发展情况、产业结构、能源结构、城市规划人口等宏观因素有关,采用传统的回归分析法预测天然气用量会产生较大偏差,而偏最小二乘回归分析法可以解决各影响因素之间的相关性问题。采用组合预测法能够避免单一预测方法自身的缺陷,使得预测结果更加合理。给出了组合预测法的应用实例。The natural gas consumption is related to many macro factors, such as national economic development situation, industrial structure, energy source structure, population planning and so on. The traditional regression analysis to forecast would lead to obvious deviation. However, the partial leastsquares regression analysis can solve relativity problems among different influence factors. The combined forecast method can avoid the defect of unique forecast method itself, thus making the forecast result even more rational. The application example of the combined forecast method is given.

关 键 词：负荷预测 偏最小二乘回归分析法 时间序列法 组合预测法 

分 类 号：TU996[建筑科学—供热、供燃气、通风及空调工程]