摘要
在高原、山地等复杂地形条件下修建铁路时,弃渣场的选址成为不可避免的问题。科学合理的弃渣场选址对于实现铁路建设与环境保护的协调发展具有重要意义。研究以雅安至昌都段铁路沿线为研究区,基于遥感技术,选择河流、道路、断层、居民点、坡度、温度、降水、植被类型、土壤属性、土地利用、人口密度、自然保护区、冰川多个生态指标因子,采用层次分析法(Analytic Hierarchy Process, AHP)、综合权重法(Comprehensive Weight, CW)和随机森林法(Random Forest, RF)构建弃渣场选址适宜性评价体系,进而得出雅昌线沿线弃渣场生态选址适宜性格局图,并通过多元逐步回归与障碍度分析,确定弃渣选址主要影响因子。研究结果显示:(1)AHP、RF与CW的弃渣适宜性评价结果准确性均在85%以上,其中RF用于渣场选址最为可靠,精度为89.8%,其结果能够准确反映区域弃渣堆放适宜性程度;(2)通过多元逐步回归分析发现,弃渣堆放生态选址主要受土地利用、坡度、道路、居民点、河流、断层这6个主要因子的影响;(3)在研究区域内,不适宜区集中分布于中部自然保护区,占总面积的10%,基本不适宜区在东部零散分布于人口密集区、建筑及耕地区域,西部则受断层与河流影响呈集聚分布。研究表明,弃渣堆放生态选址体系能够有效指导渣场选址工作,可为大尺度范围交通工程建设的生态环境评价提供理论依据。
The construction of railways in complex terrains such as plateaus and mountainous regions requires careful slag dump site selection to minimize ecological impacts.This study focuses on the Ya an-Qamdo railway section as the research area and utilizes remote sensing technology to analyze multiple ecological indicator factors,including rivers,roads,faults,settlements,slope,temperature,precipitation,vegetation type,soil properties,land use,population density,nature reserves,and glaciers.By employing the Analytic Hierarchy Process(AHP),Comprehensive Weight(CW),and Random Forest(RF)methods,an evaluation system for slag dump site suitability was developed,generating an ecological suitability distribution map for slag dump sites along the Ya an-Qamdo railway.Further analysis using multiple stepwise regression and obstacle degree models identified the key factors influencing slag dump site selection.The results indicate that:(1)The accuracy of slag dump suitability assessments using AHP,RF,and CW all exceeded 85%,with the Random Forest method proving the most reliable,achieving an accuracy of 89.8%.The RF model s results closely matched actual conditions,effectively reflecting regional suitability for slag dumping.(2)Stepwise regression analysis revealed that ecological slag dump siting is primarily influenced by six critical factors:land use,slope,roads,settlements,rivers,and faults.(3)Within the study area,unsuitable zones are mainly concentrated in the central nature reserves,covering approximately 10%of the total area.The basically unsuitable zones are scattered across the densely populated eastern regions,urban built-up areas,and cultivated lands,while exhibiting clustered distribution in the western areas due to the influence of faults and rivers.This study demonstrates that the proposed ecological slag dump siting system can effectively guide railway construction waste disposal and provide a robust theoretical foundation for large-scale ecological environmental assessments in transportation infrastructure projects.The findings highlight the importance of integrating multi-factor analysis and machine learning techniques to optimize slag dump site selection while minimizing ecological disruption.
作者
姬翠翠
陈泊宇
韩风雷
陈立川
陈茂霖
杨志鑫
刘劲宏
JI Cuicui;CHEN Boyu;HAN Fenglei;CHEN Lichuan;CHEN Maolin;YANG Zhixin;LIU Jinhong(School of Smart City,Chongqing Jiaotong University,Chongqing 400074,China;Institute of Future Civil Engineering Sciences and Technology,Chongqing Jiaotong University,Chongqing 400074,China;Chongqing Institute of Geology and Mineral Resources,Chongqing 400042,China)
出处
《安全与环境学报》
北大核心
2025年第6期2398-2406,共9页
Journal of Safety and Environment
基金
国家自然科学基金青年基金项目(42301459)
重庆市自然科学基金面上项目(CSTB2023NSCQ-MSX0967,CSTB2023NSCQ-MSX0327,CSTB2022NSCQ-MSX1093)。
关键词
环境工程学
弃渣场选址
层次分析法
随机森林
综合权重
生态评价因子
environmental engineering
slag dump site selection
analytic hierarchy process
random forest
comprehensive weight
ecological evaluation factor
