摘要
机器学习,作为人工智能的一个关键子领域,已在环境领域中发挥着越来越重要的作用,尤其在处理地表水水质管理中复杂问题时,机器学习显示出相对传统方法的显著优势.本综述重点探讨了多种机器学习算法在地表水水质管理方面的应用,分析了溶解氧、生物需氧量、化学需氧量、浊度、温度、pH等不同水质参数对地表水水质分类、监测及预测结果的影响,并列举了几种在实际工程应用中最常见的机器学习模型,如人工神经网络、支持向量机、随机森林及决策树等,最终归纳总结了几种用于提升输出精度的混合模型在地表水水质管理中的实际应用.综上所述,实现机器学习对地表水水质准确、高效管理,不仅取决于选用的水质参数是否能够作为特定算法的数据集,还依赖于合理使用多种机器学习模型进而增加输出结果的可信度.
Abstract
Machine learning,a key subfield of artificial intelligence,has been playing an increasingly important role in the environmental field.When dealing with complex problems in surface water quality management,it shows significant advantages over traditional methods.This review focused on the applications of various machine learning algorithms in surface water quality management.It analyzed the effects of different water quality parameters,such as dissolved oxygen,biological oxygen demand,chemical oxygen demand,turbidity,temperature,pH,etc.,for surface water quality classification,monitoring,and prediction.This review also provided an in-depth discussion of several machine learning models that were commonly used in real-world engineering applications,such as artificial neural networks,support vector machines,random forests,decision trees,and deep learning.In addition,this review explored the application of hybrid models for improving output accuracy in surface water quality management.In summary,the realization of machine learning for accurate and efficient management of surface water quality not only depends on suitability of selected parameters for specific algorithms but also relies on reasonable use of multiple machine learning models to increase the credibility of the output results.
国家科技期刊平台
NSTL
万方数据