基于BP人工神经网络的快速工程估算研究综述  

A Survey of Quick Estimation of Engineering Based on BP Artificial Neural Network

作　　者：韩思淼 黄剑[1] HAN Si-miao;HUANG Jian(School of Architectural Engineering,Huangshan University,Huangshan 245041,China)

机构地区：[1]黄山学院建筑工程学院,黄山245041

出　　处：《价值工程》2021年第25期191-193,共3页Value Engineering

基　　金：黄山学院2020年度大学生创新创业训练计划项目“基于人工神经网络的工程成本估算”(项目编号:S202010075)。

摘　　要：我国建筑行业发展长期受到“三超”现象困扰,因此近年来如何通过成本管理合理预测工程造价成为现阶段建筑行业的主流思想。有越来越多的研究希望通过人工神经网络模型应用于工程造价估算来达到合理建模,快速估价的目的。为了明确人工神经网络估算模型在工程估算中的应用和发展前景,需要对相关文献进行综述和梳理。首先,先从应用最广泛的BP神经网络估算模型入手,分析该BP神经网络模型应用过程中的优缺点,并对现有解决方法进行综述。其中讨论了通过强弱耦合弥补BP模型应用缺陷以及选取其他常见模型应用于工程估算,如:RBF神经网络模型、模糊神经网络模型。最后,我们对人工神经网络的研究前景进行展望,为实现工程造价智能化奠定了基础。The development of China′s construction industry has long been plagued by the phenomenon of exceed the budget.Therefore,how to reasonably predict the project cost through cost management has become the mainstream idea in the current construction in industry in recent years.To ensure reasonable modeling and quick appraisal,more and more researchers hope to accurate estimation through artificial neural network estimation model.In order to define the application and development prospect of artificial neural network estimation model in engineering estimation clearly,relevant researches should be thoroughly reviewed and sorted out.Firstly,Refer to BP artificial neural network estimation model,analyzing this model’s advantages and disadvantages in the application process and reviewing relevant researches in this article.Afterwords,remedying BP model’s defects in application through strong and weak coupling and applying other common models to engineering estimation are discussed,such as RBF neural network and fuzzy neural network.Finally,we look forward to the research prospects in the field of artificial neural network.These works will lay a foundation for realizing engineering cost intelligence.

关 键 词：人工神经网络 工程造价 快速估算 强弱耦合 综述 

分 类 号：TU723[建筑科学—建筑技术科学]