基于BP神经网络的核电工程人工单价预测研究  

Research on dynamic adjustment of labor unit price in nuclear power engineering based on BP neural network

作　　者：徐婧 黄骏 裴丽倩 卞逢源 Xu Jing;Huang Jun;Pei Liqian;Bian Fengyuan(China Institute of Nuclear Industry Strategy,Beijing 100048,China;Hebei Branch of China Nuclear Power Engineering Co.,Ltd.,Shijiazhuang Hebei 050021,China)

机构地区：[1]中核战略规划研究总院有限公司,北京100048 [2]中国核电工程有限公司河北分公司,河北石家庄050021

出　　处：《山西建筑》2023年第16期184-187,198,共5页Shanxi Architecture

摘　　要：核电工程人工单价是计算核电项目直接人工费的重要依据,是造价工作中建安成本测算的核心要素。为科学预测核电工程人工单价,首先分析影响核电工程人工单价的因素指标,通过主成分分析法提取关键影响因素,在此基础上,构建BP神经网络预测模型,并以优化的模型对2021年人工单价进行预测,预测结果为160元/工日。结果表明,该模型计算过程便捷高效,结论科学合理,可用于后续定额人工单价的测算及调整工作,具有较好的推广应用意义。The unit price of labor in nuclear power engineering is an important basis for calculating the direct labor cost of nuclear power projects,and is the core element of construction and installation cost calculation in cost work.To scientifically predict the labor unit price of nuclear power engineering,the factors and indicators that affect the labor unit price of nuclear power engineering are first analyzed.Key influencing factors are extracted through principal component analysis.On this basis,a BP neural network prediction model is constructed,and the optimized model is used to predict the labor unit price in 2021,with a prediction result of 160 Yuan/workday.The prediction results indicate that the calculation process of this model is convenient and efficient,and the conclusions are scientific and reasonable.It can be used for the calculation and adjustment of the fixed labor unit price in the future,and has good promotion and application significance.

关 键 词：核电工程 人工单价 主成分分析法 BP神经网络 

分 类 号：TU723.3[建筑科学—建筑技术科学]