LIULin-hao,CHENGQi-ming,CHENYao,et al.Identification of Influencing Factors for Removal of Typical Heavy Metals in Bioretention System[J].China Water & Wastewater,2023,39(23):124-132.
Click Copy

Identification of Influencing Factors for Removal of Typical Heavy Metals in Bioretention System

China Water & Wastewater[ISSN:1000-4062/CN:12-1073/TU] volume: 第39卷 Number: 第23期 Page: 124-132 Column: Date of publication: 2023-12-01
Author(s):
LIU　Lin-hao1;  CHENG　Qi-ming1;  CHEN　Yao1; 2;  YUAN　Shao-chun1; 2;  WU　Qiong1
（1. School of River and Ocean Engineering， Chongqing Jiaotong University， Chongqing 400074， China； 2. Engineering Laboratory of Environmental Hydraulic Engineering of Chongqing Municipal Development and Reform Commission， Chongqing Jiaotong University， Chongqing 400074， China）
Keywords:
bioretention;  CART algorithm;  machine learning;  binary tree model;  heavy metal;  design parameter;  environment variable
Abstract:
Bioretention system has dual functions of stormwater runoff reduction and pollution control. However, its removal performance of heavy metals is susceptible to be affected by design parameters and environmental variables. Based on the literature data, a binary tree machine learning model was constructed by CART algorithm for determining the design parameters and environmental variables of biological retention system, and the influencing factors of biological retention system for removal of Cu, Zn, Pb and other typical heavy metals were identified. The most sensitive factor affecting the removal of Cu and Pb was the inflow concentration, while the most sensitive factor affecting the removal of Zn was the depth of soil medium. The accuracy (p0) of the binary tree model for identification of the three heavy metal influencing factors was 0.86, 0.80 and 0.74, respectively, the classification consistency was above medium level, and the Cohen’s Kappa coefficient (Ka) was 0.72, 0.60 and 0.48, respectively. Univariate correlation analysis was difficult to identify the sensitive factors of bioretention system for the removal of typical heavy metals. In contrast, the machine learning method based on literature data could not only effectively mine the influence degree of sensitive factors in bioretention systems, but also identify the corresponding threshold, which could provide some reference for the subsequent optimization design and operation and maintenance management.
Last Update: 2023-12-01