基于机器学习的生物炭吸附重金属建模研究进展  

Research progress on modeling of heavy metal adsorption by biochar based on machine learning

作　　者：冯丁 刘晶静[1,2] 马文丹 刘玉学 杨尘[1] 张萌萌 FENG Ding;LIU Jingjing;MA Wendan;LIU Yuxue;YANG Chen;ZHANG Mengmeng(School of Energy and Machinery Engineering,Jiangxi University of Science and Technology,Nanchang 330000,China;Jiangxi Key Laboratory of Mining&Metallurgy Environmental Pollution Control,Ganzhou 341000,China;Institute of Environmental Resources,Soil and Fertilizer,Zhejiang Academy of Agricultural Sciences,Hangzhou 310021,China)

机构地区：[1]江西理工大学能源与机械工程学院,江西南昌330000 [2]江西省矿冶环境污染控制重点实验室,江西赣州341000 [3]浙江省农业科学院环境资源与土壤肥料研究所,浙江杭州310021

出　　处：《工业水处理》2024年第12期1-11,共11页Industrial Water Treatment

基　　金：国家自然科学基金项目(52304187);江西省自然科学基金面上项目(20224BAB203041)。

摘　　要：生物炭吸附重金属试验面临参数众多、污染状况复杂、生物炭特性丰富以及研究成本高且周期长等问题,传统吸附模型已无法满足当前研究需求。近年来机器学习在高维数据处理和复杂问题分析方面展现出巨大潜力,其建模流程清晰且成熟,在重金属污染预测方面精准而稳定,对发掘隐藏的吸附机制具有独特价值,是生物炭吸附重金属建模研究的优质选择。阐述了机器学习建模的工作流程及优势;从吸附效率预测、促进优化实验、洞察吸附机理三方面综述了机器学习在生物炭吸附重金属中的应用;分析了机器学习在生物炭吸附重金属研究领域面临的挑战,并对跨学科合作的前景与发展趋势进行了展望,如构建更加全面可靠的吸附数据库、引入表面官能团等深层影响因素、关注模型准确性和计算成本平衡等以深化对生物炭吸附重金属建模的研究。Traditional adsorption models are faced with challenges due to various parameters,complex pollution conditions,diverse biochar properties,and the high cost and lengthy duration of research.Machine learning has demonstrated significant potential in handling high-dimensional data and analyzing complex problems.Its clear and mature modeling process provides precise and stable predictions for heavy metal pollution and holds unique value in uncovering hidden adsorption mechanisms,making it an excellent choice for studying heavy metal adsorption by biochar.This paper described the workflow and advantages of machine learning modeling.The application of machine learning in heavy metal adsorption across three aspects of predicting adsorption efficiency,aiding in optimizing experiments,and gaining insights into adsorption mechanisms was summarized.Furthermore,the challenges that machine learning faced in the field of biochar-mediated heavy metal adsorption and interdisciplinary collaboration anticipation were analyzed.Strategies such as establishing a more comprehensive and reliable adsorption database,incorporating surface functional groups as influential factors,emphasizing model accuracy,and balancing computational costs were proposed to deepen the research on modeling heavy metal adsorption by biochar.

关 键 词：生物炭 机器学习 重金属 预测模型 环境修复 

分 类 号：X703[环境科学与工程—环境工程] TP181[自动化与计算机技术—控制理论与控制工程]