基于BP神经网络法地铁工程绿色施工评价指标体系研究  

Study on Green Construction Evaluation Index System of Metro Engineering based on BP Neural Network Method

作　　者：郭飞 闫宏锦 傅睿智 Guo Fei;Yan Hongjin;Fu Ruizhi(Beijing Municipal Construction Group Co.,Ltd.,Beijing 100089,China)

机构地区：[1]北京市政建设集团有限责任公司,北京100089

出　　处：《市政技术》2022年第10期113-119,共7页Journal of Municipal Technology

基　　金：北京市政建设集团有限责任公司科技创新项目(编号:2020-16)。

摘　　要：基于PSR框架模型逻辑思维,从压力、状态、响应3个方面构建了27项地铁工程绿色施工评价指标体系;借助MATLAB软件建立了3层BP神经网络的基本模型,通过训练样本模拟了非线性规律,并用校验样本进行了验证,证明了BP神经网络法的可行性;以北京地铁27号线与地铁19号线换乘站上清桥站为例,再次验证了BP神经网络模型的可行性且符合项目绿色施工的实际情况,为地铁工程提供了一种新的绿色施工评价方法。Based on pressure-state-response(PSR) framework model,27 green construction evaluation indexes system of metro engineering are constructed from 3 aspects of pressure, state and response. The basic model of threelayer BP neural network is established by MATLAB software. The nonlinear law is simulated by training samples and verified by calibration samples. The feasibility of BP neural network method is proved. Taking Shangqingqiao Station, the transfer station of Beijing Metro Line 27 and Metro Line 19 as an example, the feasibility of BP neural network model is verified again and meet the actual situation of green construction. It provides a new green construction evaluation method for metro engineering.

关 键 词：地铁工程 绿色施工评价 BP神经网络法 指标体系 

分 类 号：TU71[建筑科学—建筑技术科学]