Identification of source information for sudden water pollution incidents in rivers and lakes based on variable-fidelity surrogate-DREAM optimization

For sudden water pollution incidents in rivers and lakes, the ability to quickly identify the pollution source is of great importance for providing early accident warning and implementing emergency control measures. Based on Bayesian reasoning, a variable-fidelity surrogate-differential evolution ad...

Full description

Saved in:
Bibliographic Details
Published in:Environmental modelling & software : with environment data news 2020-11, Vol.133, p.104811, Article 104811
Main Authors: Wu, Wei, Ren, Jucheng, Zhou, Xiaode, Wang, Jiawei, Guo, Mengjing
Format: Article
Language:English
Subjects:
Accuracy
Adaptive control
Algorithms
Bayesian analysis
Computer applications
Differential evolution adaptive metropolis optimization(DREAM)
Emergency warning programs
Evolutionary computation
Identification of the pollution source
Lakes
Optimization
Pollutants
Pollution effects
Pollution sources
Rivers
Sudden water pollution
Variable-fidelity surrogate
Water pollution
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:For sudden water pollution incidents in rivers and lakes, the ability to quickly identify the pollution source is of great importance for providing early accident warning and implementing emergency control measures. Based on Bayesian reasoning, a variable-fidelity surrogate-differential evolution adaptive metropolis optimization(DREAM) optimization model for coupled inversion process is established in the posterior space of the pollution source.In order to verify the effectiveness of the algorithm, this paper takes lake A as the research area, and gives a hypothetical water pollution emergency, the pollution source location, release time and released mass of water pollutants suddenly released into water bodies were determined according to the method proposed in this paper. The results show that in the case of ensuring the accuracy of calculation, the algorithm can accelerate more than 200 times and effectively improves the computational efficiency of the traditional method for obtaining the source information of sudden water pollution events. •A variable-fidelity surrogate model is used to determine the source of sudden water pollution incidents in rivers and lakes.•A new sample point is added during the update of the surrogate model.•Using the surrogate model instead of the physical model reduces the computation time by a factor 200.
ISSN:1364-8152
1873-6726
DOI:10.1016/j.envsoft.2020.104811