摘要
本文针对城市水环境综合整治决策中的多目标问题,以中山市南朗流域13条河涌为研究对象,通过最小二乘支持向量回归算法(LSSVM)分别建立起流域河涌水质高锰酸钾指数预测模型和溶解氧预测模型。同时,为了提高模型的鲁棒性和运算速度,使用PCA算法降低输入变量维数。结果表明:基于PCA-LSSVM的预测模型具有较好的预测能力,模型测试数据与实际数据的相关系数分别为0.8290和0.8126;结合快速非支配排序遗传算法(NSGA-Ⅱ)建立河涌水质优化模型,优化结果表明,在区域流域的水环境综合治理中,应重点关注居民生活污染源的治理和工业污染源的治理。该研究可为此类水环境综合整治措施的设计和操作提供参考和指导。
In this paper,aiming at the multi-objective problem in the decision-making of comprehensive improvement of urban water environment,taking 13 rivers in Nanlang watershed of Zhongshan as the research object,the potassium permanganate index prediction model and dissolved oxygen prediction model of river water quality in the watershed were established by least squares support vector regression algorithm(LSSVM).At the same time,in order to improve the robustness and operation speed of the model,PCA algorithm is used to reduce the dimension of input variables.The results show that the prediction model based on PCA-LSSVM has good prediction ability,and the correlation coefficients between the model test data and the actual data are 0.8290 and 0.8126 respectively.Combined with the fast non-dominated sorting genetic algorithm(NSGA-Ⅱ),a river water quality optimization model is established.The optimization results show that in the comprehensive treatment of water environment in regional river basins,attention should be paid to the treatment of domestic pollution sources and industrial pollution sources.This study can provide reference and guidance for the design and operation of comprehensive improvement measures of this kind of water environment.
作者
谢彬
黄柳祯
罗旌生
XIE Bin;HUANG Liuzhen;LUO Jingsheng(Zhongshan Environmental Protection Technology Center,Zhongshan 528400,China;Jinan University Research Center of Aquatic Biology,Guangzhou 510632,China)
出处
《水资源开发与管理》
2021年第10期32-41,共10页
Water Resources Development and Management
基金
国家“十二五”科技支撑计划《珠三角复合污染型村镇环境整治和修复技术集成及工程示范》(2012BAJ21B07)
国家自然科学基金资助项目(41201506)
广东省应用型科技研发专项“村镇综合废水生态处理集成工艺与再生水农业利用技术研发及示范”(2015B020235008)。
关键词
城市水环境综合整治
最小二乘法支持向量机
快速非支配遗传算法
多目标优化
混合智能算法
comprehensive improvement of urban water environment
least squares support vector machine:fast non-dominated genetic algorithm
multi-objective optimization
hybrid intelligent algorithm
