节点文献
铁路弃渣场绿色施工等级综合评价研究
Study on Comprehensive Evaluation of Green Construction Grade of Railway Abandoned Slag Dump
【作者】 张薇;
【导师】 鲍学英;
【作者基本信息】 兰州交通大学 , 土木工程建造与管理, 2021, 硕士
【摘要】 我国铁路客运需求量大,铁路建设迅速发展,在建项目众多。近些年来,铁路建设项目愈发重视资源节约、生态环境保护等问题,因此铁路线路中桥隧比越来越高,弃渣产生量也越来越大。弃渣堆积体受自身重力或地震、降水等因素影响易引发水土流失、失稳滑坡和泥石流等自然灾害,这些灾害的发生对铁路的建设、运营和沿线生态危害极大。因此,弃渣场成为铁路建设项目中衍生的资源浪费较多对周边环境破坏较大的附属工程,与之相关的弃渣场绿色施工等级评价研究具有较大的科学价值和应用前景。本文在参考国内外建设项目弃渣场施工、水土保持及环境影响评价资料的基础上,针对铁路线路长、桥隧比高、弃渣量巨大的特点,详细分析铁路弃渣场施工过程中的各个施工行为,找出影响弃渣场绿色施工的影响因素及其产生的原因,提出弃渣场绿色施工的主要措施。遵循指标体系构建的基本原则,构建一套包含水土保持、弃渣综合处理、资源节约及高效利用和环境保护4个一级指标,18个二级指标在内的铁路弃渣场绿色施工评价指标体系。通过研究每个评价指标的度量方法,给出指标的量化公式,深入分析现有国内外评价标准,确定了铁路弃渣场绿色施工等级的评价标准和各个指标的评级标准。铁路弃渣场施工是一个动态随机的过程,施工期间受到诸多外部因素的影响。因此,考虑到铁路弃渣场绿色施工等级评价随机性、非线性的特点,采用一种自学习自调整模型的机器学习语言—支持向量回归(Support Vector Regression,SVR)作为评价模型。支持向量回归适用于小样本、非线性、高维数问题的回归拟合,能够更好的评价铁路弃渣场绿色施工等级,为铁路弃渣场绿色施工等级的评价提供一种新的选择。但支持向量回归的预测效果受参数影响较大,当评价模型的参数选取不够好时,将导致评价模型运行速度变慢,预测效果失去准确性。为此,本文首先对样本数据进行归一化处理,消除数据间量级影响,然后使用遗传算法选取最优参数正则化常数C和核函数高斯径向基函数(RBF)的gamma值,建立GA-SVR评价模型。最后,选取银西铁路甘宁段永乐隧道出口弃渣场、早胜二号隧道出口三号隧道进口弃渣场、上阁村隧道1#斜井弃渣场进行实例验证,在训练好的GA-SVR模型中输入无量纲化后的各指标实测值,运行得到输出值为3.483、3.292、4.203,即三座弃渣场的绿色施工等级分别为Ⅲ级、Ⅲ级、Ⅳ级。将评价结果与实际情况相对比,结论基本一致,证明了本文选取的评价指标及评价模型具有一定的合理性与可行性,能够用于解决实际问题,可以为后续铁路弃渣场的研究提供参考,具有一定的研究价值。
【Abstract】 In China,passenger rail service demand is large,the railway are developing rapidly,and there are a lot of projects under construction.Over the years,railway construction projects pay attention to resource conservation and ecological environmental protection.Accordingly,the ratio of bridge and tunnel in railway is getting higher,and the amount of abandoned slag is also increasing.The slag is affected by gravity,earthquake,precipitation and other factors,which can easily cause soil erosion,unstable landslides and debris flow.The occurrence of these disasters is unfavorable to the railway.Therefore,the abandoned slag dump is an auxiliary project derived from railway projects that wastes more resources and causes greater damage to the surrounding environment.The related green construction grade evaluation research of abandoned slag dump has scientific value and application prospect.This paper refers to the construction materials of the abandoned slag dump.Because there are many railway abandoned slags in railway construction,this paper resolves each construction activity of the abandoned slag dump,finds out the influencing factors and causes,and advances some main measures for the green construction.According to the principle of index construction,the evaluation index system was constructed,including 4 first-level indexes and18 second-level indexes.Study the index measurement method,give the formula of the index,analyze and determine the evaluation standard of the indicators and the grade of the railway abandoned dump.The construction of abandoned slag dump is dynamic and influenced by external factors.Support vector regression is suitable for green construction grade evaluation of abandoned slag dump.SVR can get a good regression fitting in small sample,nonlinear and high-dimensional problems.It can better evaluate the green construction grade and provide a new choice for the green construction grade evaluation of abandoned slag dump.But the prediction effect is related to the parameters.When the parameters are not well chosen,the running speed of the evaluation model will become slow and the prediction effect will lose accuracy.To this end,the paper first normalized the data to eliminate the magnitude of the influence between the data,and then use the GA to find the parameters C and gamma to establish the GA-SVR evaluation model.Finally,the abandoned slag dump at the exit of Yongle Tunnel,the abandoned slag dump at the exit of Zaosheng No.2 tunnel and at the entrance of No.3 tunnel and the abandoned slag dump at the 1 # inclined shaft of Shangge Village Tunnel are selected as examples,in GA-SVR model,the measured values of each index after dimensionless input are 3.483,3.292 and 4.203,that is to say,the green construction grade of three waste dump is grade III,III and IV.The evaluation result is similar to the actual situation,which proves that the evaluation index and model in this paper are reasonable and feasible.It can solve the practical problems and provide reference for the research of railway abandoned slag dump,and has certain research value.
【Key words】 Railway abandoned slag dump; green construction; genetic algorithm; support vector regression; comprehensive evaluation;