基于BP神经网络的地下管廊施工质量控制  

Construction quality control of underground pipe gallery based on BP neural network

作　　者：张军[1] 王帅 张玉丽 ZHANG Jun;WANG Shuai;ZHANG Yu-li(School of Civil Engineering,University of Science and Technology Liaoning,Anshan 114000,China)

机构地区：[1]辽宁科技大学土木工程学院,辽宁鞍山114000

出　　处：《沈阳工业大学学报》2020年第5期595-600,共6页Journal of Shenyang University of Technology

基　　金：辽宁省教育厅科学技术研究项目(201703135).

摘　　要：为了促进工程施工质量前期控制在地下管廊工程中的应用,结合BP神经网络基本理论,提出了一种基于人、材料、机械设备、施工方法和环境五大质量影响因素(即4M1E)的量化处理方法.通过MATLAB R2014a数值仿真模拟软件实现对4M1E的量化处理,建立了地下管廊施工质量前期预测模型,并对该模型结果进行了测试与验证.结果表明,模型中的预测值与期望值基本一致,证明了该模型是有效的、可靠的,进而对地下管廊施工质量前期控制有着重要的参考价值.In order to promote the application of preliminary control of construction quality for underground pipe gallery engineering,combined with the basic theories of BP neural network,a quantitative processing of five quality influencing factors(4M1E)concerning man,material,construction machine,construction method and environment was proposed.The quantitative processing of 4M1E was realized with MATLAB R2014a numerical simulation software.A preliminary prediction model for the construction quality of underground pipe gallery was established,and the results obtained by as-proposed model were tested and verified.The results show that the predicted values by the model are basically consistent with the expected values,indicating that the model is effective and reliable,and thus the model has important reference value for the preliminary construction quality control of underground pipe gallery.

关 键 词：地下综合管廊 BP神经网络 前期预测 模糊决策 4M1E量化处理 有限差分数值模拟 质量控制 施工优化 

分 类 号：TU71[建筑科学—建筑技术科学]