中国安全科学学报 ›› 2019, Vol. 29 ›› Issue (5): 91-96.doi: 10.16265/j.cnki.issn1003-3033.2019.05.016

• 安全工程技术科学 • 上一篇    下一篇

DEA-BP神经网络下地铁车站深基坑施工安全评价

宋博 讲师   

  1. 郑州信息科技职业学院 建筑工程学院,河南 郑州 450008
  • 收稿日期:2019-02-22 修回日期:2019-04-15 发布日期:2020-11-02
  • 作者简介:宋博(1982—),女,河南商丘人,硕士,讲师,主要从事建筑工程技术与工程管理方面的研究。E-mail:songbohnzz@163.com。
  • 基金资助:
    河南省科技厅科技公关项目(182102210575)。

Safety evaluation for deep foundation pit construction in metro station based on DEA-BP neural network

SONG Bo   

  1. School of Architecture and Engineering, Zhengzhou Vocational University of Information and Technology, Zhengzhou Henan 450008, China
  • Received:2019-02-22 Revised:2019-04-15 Published:2020-11-02

摘要: 为明确地铁车站深基坑施工安全等级,考虑评价指标的非线性关系和复杂动态性,基于数据包络法(DEA)-反向传播(BP)神经网络,提出一种地铁车站深基坑施工安全评价方法。首先,从人员、设备、环境、管理、技术5个方面系统构建安全评价指标体系;然后,利用DEA计算指标权重,运用BP神经网络网络评价地铁车站深基坑施工安全等级;最后,以重庆地铁1号线小什车站为例,运用该方法评价车站深基坑施工安全。结果表明:该车站深基坑施工安全等级为高,与实际情况吻合;安全意识、机械伤人、周边环境、技术交底、渗流破坏是影响深基坑施工安全的主要指标。

关键词: 地铁车站, 深基坑施工, 安全评价, 数据包络法(DEA), 反向传播(BP)神经网络

Abstract: In order to determine the safety level of deep foundation pit construction in metro station, a safety evaluation method for deep foundation pit construction in metro station based on DEA-BP neural network is proposed with the nonlinear relationship and complex dynamics of evaluation index into consideration. Firstly, the safety evaluation index system was established from five aspects including people, devices, environment, management and technology. Then, the index weight was calculated by use of DEA, and the safety level of construction was evaluated by using BP neural network. Lastly, with Xiaoshi station of Chongqing metro line 1 as an example, the construction safety of deep foundation pit in metro station was evaluated by adopting the proposed method. The results show that the safety level of deep foundation pit construction in this station is high and consistent with the actual situation, and the main indicators that affect the safety of construction are safety awareness, mechanical injury, surrounding environment, technical disclosure and seepage damage.

Key words: metro station, deep foundation pit construction, safety evaluation, data envelopment analysis (DEA), back propagation (BP) neural network

中图分类号: