文献类型：期刊文献

中文题名：基于人工神经网络的石油化工工程建设项目管理绩效评价

英文题名：MANAGEMENT PERFORMANCE EVALUATION IN PETROCHEMICAL ENGINEERING CONSTRUCTION PROJECT BY USING ARTIFICIAL NEURAL NETWORK

作者：韩志国[1];王基铭[1,2];陈智高[1]

机构：[1]华东理工大学,上海200237;[2]中国石油化工集团公司,北京100728

年份：2010

卷号：26

期号：3

起止页码：317

中文期刊名：石油学报（石油加工）

外文期刊名：Acta Petrolei Sinica(Petroleum Processing Section)

收录：CSTPCD;;EI(收录号：20102813077080);Scopus;北大核心:【北大核心2008】；CSCD:【CSCD_E2011_2012】；

语种：中文

中文关键词：石油化工工程建设项目;项目管理;绩效评价;人工神经网络(ANN)

外文关键词：petrochemical engineering construction project; project management; performance evaluation; artificial neural network （ANN）

摘要：针对非线性多输入多输出的石油化工工程建设项目管理绩效评价问题,应用人工神经网络(ANN)构建评价模型。使用50个项目的287个学习案例数据,以10个影响因素为输入,6个指标为输出,对BP神经网络、基于遗传算法的BP神经网络、径向基函数神经网络与广义回归神经网络4类网络模型进行训练和测试。通过均方误差的比较,发现基于遗传算法的BP神经网络优于一般的BP神经网络,广义回归神经网络的测试结果优于BP神经网络,径向基函数神经网络具有最好的误差精度。2个应用示例表明,人工神经网络应用于石油化工工程建设项目管理绩效的评价是可行和有效的。
An artificial neural network model （ANN） was developed for the management performance evaluation of petrochemical engineering construction project.By using the data from 287 samples of 50 projects,in which 10 factors were as inputs and 6 indicators as outputs,training and test were given to four kinds of ANN models,BP-NN,GA-based BP-NN,radial basis function neural network （RBF-NN） and generalized regression neural network （GR-NN）.By comparing the mean square errors,it is found that GA-based BP-NN is prior to BP-NN,GR-NN is prior to former,and RBF-NN has best accuracy.It is verified by the illustrations of two cases that ANN model is feasible and valid to the management performance evaluation of petrochemical engineering construction project.