025-83701931 设为首页 加入收藏
文章摘要
基于SOM网络的山地城市径流污染影响因素研究
Study on Influencing Factors of Runoff Pollution in Mountain City Based on SOM
  
DOI:
中文关键词: 降雨径流  场次降雨污染物平均浓度  影响因素  自组织映射神经网络  山地城市
英文关键词: Rainfall runoff  Event mean concentration (EMC)  Influencing factor  Self organizing map(SOM)  Mountainous city
基金项目:重庆市自然科学基金资助项目(cstc2020jcyj-msxmX1000);重庆市留创计划基金资助项目(cx2017065);重庆市青少年创新人才培养雏鹰计划基金资助项目(CY200701);万州区城乡建设委员会海绵城市专题研究基金资助项目(17A2739)
作者单位
冯力柯 重庆交通大学河海学院 
陈垚 重庆交通大学河海学院重庆交通大学环境水利工程重庆市工程实验室 
袁绍春 重庆交通大学河海学院重庆交通大学环境水利工程重庆市工程实验室 
朱子奇 重庆交通大学河海学院 
摘要点击次数: 448
全文下载次数: 33
中文摘要:
      以万州区海绵城市建设试点为研究对象,采用自组织映射神经网络(SOM网络),通过径流污染指标变量与影响因素变量的U〖CD*2〗矩阵识别降雨径流污染的主要影响因素及其相关性,并利用构建的场次降雨污染物平均浓度(EMC)预测模型分析确定各因素的影响强度。结果表明,有限的监测数据难以识别径流污染的显著影响因素,而采用SOM组分面(矩阵)可直观地半定量解译各污染物间的相关关系,与Pearson定量相关性分析结果吻合。SOM网络可有效识别径流污染物的主要影响因素,其中SS、COD和TP污染物主要受降雨量和径流系数影响,NO-3-N和NH3-N受降雨量和降雨历时影响,而TN则主要受径流系数驱动影响。
英文摘要:
      In this study, the main factors influencing runoff pollutants and their correlations were identified by self organizing map (SOM) and the U matrix of runoff pollution index variables and influencing factor variables, taking the pilot project of Wanzhou sponge city construction as research objects. Meanwhile, the intensity of each factors was analyzed and determined by a prediction model of event mean concentration (EMC) of rainfall runoff pollutants built in this study. Results showed that the significant influencing factors could hardly be identified by limited monitoring data, while the correlation between the pollutants could be interpreted intuitively and semi quantitatively by the SOM component planes (U matrix), and the results were consistent with that by Pearson quantitative correlation analysis. It was confirmed that the key factors influencing runoff pollution could be identified through SOM. SS, COD and TP were affected by rainfall and runoff coefficient, NO-3-N and NH3-N were affected by rainfall and rainfall duration, while TN was affected by runoff coefficient.
查看全文   查看/发表评论  下载PDF阅读器
关闭