基于DBO+DMAMH算法的河流突发水污染溯源研究  

Research on traceability of sudden water pollution in rivers based on DBO+DMAMH algorithm

作　　者：刘晓晴 贾冬艳 宋金玲[1,2] 樊刘炎 张思萱 LIU Xiaoqing;JIA Dongyan;SONG Jinling;FAN Liuyan;ZHANG Sixuan(School of Mathematics and Information Technology,Hebei Normal University of Science,Qinhuangdao 066004,China;Hebei Agricultural Data Intelligent Perception and Application Technology Innovation Center,Qinhuangdao 066004,China)

机构地区：[1]河北科技师范学院数学与信息科技学院,河北秦皇岛066004 [2]河北省农业数据智能感知与应用技术创新中心,河北秦皇岛066004

出　　处：《工业用水与废水》2025年第1期64-70,共7页Industrial Water & Wastewater

基　　金：河北省省级科技计划资助项目(21370103D);2023年度河北省高等学校科学研究项目(ZC2023123);河北省软件工程重点实验室项目(22567637H);河北省软件工程重点实验室开放课题(KF2307);河北省农业数据智能感知与应用技术创新中心开放课题(ADIC2023Y006,ADIC2023Y004,ADIC2023Y005)。

摘　　要：为了准确获得河流突发性水污染事件的污染物排放时刻、排放位置以及排放量等参数,首先,根据监测点位的污染物实测数据和水污染迁移扩散模型的预测数据建立目标优化函数,使用蜣螂优化(DBO)算法对水污染迁移扩散模型的参数进行率定;然后,将水污染溯源问题转化为贝叶斯估计问题,推导得到关于污染源信息的后验概率密度函数,使用扩散模型(Diffusion Model)辅助MH算法(DMAMH)对污染源信息进行抽样,从而反演出污染源的各个参数;最后,利用仿真试验对所提的水污染溯源方法进行验证。试验结果表明,利用DBO算法对水污染迁移扩散模型参数进行率定准确性较高,使用DMAMH算法进行溯源的结果也更加精确、高效,因此该方法对于河流突发水污染溯源事件具有较好的适用性。In order to accurately obtain the information of pollutant discharge moment, discharge location and the discharge volume of sudden water pollution events in the river, firstly, the objective optimizer function was established based on the measured pollutant data of the monitoring sites and the predicted data of the water pollution transport diffusion model, and the parameters of the water pollution transport diffusion model were calibrated using the dung beetle optimizer(DBO) algorithm. Then, the water pollution traceability problem was transformed into a Bayesian estimation problem, the posteriori probability density function about the pollution source information was derived, and the diffusion model-assisted Metropolis Hastings(DMAMH) algorithm was used to sample the information of the pollution source, so as to inversely perform the parameters of the pollution source, and finally, the proposed water pollution traceability method was verified using simulation experiments.The experimental results showed that the rate determination of the parameters of the water pollution transport diffusion model using DBO algorithm was more accurate, and the results of the traceability using DMAMH algorithm were more precise and efficient, so the method had better applicability for sudden water pollution traceability events in rivers.

关 键 词：水污染溯源 水污染迁移扩散模型 参数率定 扩散模型 MH算法 

分 类 号：X522[环境科学与工程—环境工程] X703.1