基于ARIMA-ES混合模型的电网工程造价指数预测  

Study on Grid Engineering Cost Index Forecasting Based on ARIMA-ES Hybrid Model

作　　者：刘卫东[1] 石华军[2] 路妍[3] 滑福宇[3] 邱金鹏[3] 

机构地区：[1]西安交通大学经济与金融学院,西安710061 [2]浙江大学电力系统自动化系,杭州310007 [3]华北电力大学经济与管理学院,北京102206

出　　处：《管理评论》2016年第3期45-53,105,共10页Management Review

摘　　要：电网工程造价指数是反映了市场价格变动对电网工程造价影响的综合性指标,准确有效的对电网工程造价指数进行预测可以提高建设管理部门的管理水平。本文以电网工程中的线路工程为例,以浙江省2008-2014年线路工程造价数据为基础,在统一工程技术、规模条件下,创新的构建出一种适用于电网工程的综合造价指数,并且合理测算历史造价指数。在电网工程造价指数预测模型中引入了GDP平减指数,分析造价指数与GDP平减指数之间的弹性系数,并基于ARIMA-ES混合模型构建造价指数的分析预测进行了实例计算,通过与GM(1,1)、线性回归模型的对比分析结果表明,ARIMA-ES混合模型预测结果更加符合地区经济水平发展趋势,提高了电网工程造价指数的预测精度,在预测方法的应用上具有一定参考价值,对提高电网工程管理水平,降低工程造价投资具有一定的实际意义。The power grid project cost index is a comprehensive index which reflects the impact of market price fluctuation on power grid project cost. The accurate and effective forecasting of power grid project cost index can improve the management level of construction management department. This paper takes the engineering of power grid as an example,based on the line project cost data of Zhejiang province from 2008 to 2014,under the unified engineering technology and scale conditions,constructs a comprehensive cost index for power grid project innovatively,and estimates the historical cost index reasonably. In the power grid project cost index forecasting model,the GDP index is introduced,and the elasticity coefficient between the cost index and GDP index is analyzed. The ARIMA-ES mixed model is used to calculate the cost index. Through the comparison with GM（ 1,1） and linear regression model,the results show that the forecasting results of ARIMA-ES mixed model are more consistent with the development trend of regional economy and improve the forecasting precision of the power grid project cost index. In the application of forecasting method,the reference value is of certain practical significance to improving the management level and reducing the investment of project cost.

关 键 词：电网工程 ARIMA-ES混合模型 造价指数 预测 

分 类 号：TU723.3[建筑科学—建筑技术科学] TM7[电气工程—电力系统及自动化]