基于神经网络的工程造价估算研究  

Study on Cost Estimation Based on Artificial Neural Network

作　　者：林骁元[1] 

机构地区：[1]上海核工程研究设计院,上海200233

出　　处：《核电工程与技术》2015年第1期57-61,共5页Nuclear Power Engineering and Technology

摘　　要：通过研究工程造价的构成及其影响因素，利用神经网络有效处理多输入多输出非线性映射的优点，将工程主要特征量作为网络的输入量，工程造价作为输出量，建立了基于神经网络的工程估价模型。收集某核电厂6个BOP电气工程作为样本，把训练好的网络用于该电厂应急指挥中心电气工程的造价预测，经测算其精度满足要求，能够用于工程造价快速估算。Through composition of the project cost and its influencing factors, the use of neural net- works to deal effectively with the advantages of MIMO nonlinear mapping of the main features of the project as the input of the network, as the output of the project cost, the project to establish a cost estimation based on artificial neural networks model. Collected six nuclear power plant BOP electrical engineering as samples, the trained network is used to predict the cost of the emergency command center electrical engineering, the estimated accuracy to meet the requirements, can be used to quickly estimate the project cost.

关 键 词：神经网络:工程造价 估算模型 

分 类 号：TU723.3[建筑科学—建筑技术科学]