机器学习方法在污水处理系统中的应用

APPLICATION OF MACHINE LEARNING METHODS IN WASTEWATER TREATMENT SYSTEMS

由于污水处理系统一般较为复杂且受外界因素影响较多,对其进行精准调控一直是环境领域的难题之一。传统方法无法满足日益复杂的工程项目需求。近年来发展起来的机器学习方法为此类问题提供了一系列有效的解决方案。介绍了人工神经网络、支持向量机、随机森林等机器学习方法的特点,并从水质预测预警、污水处理系统故障诊断和智能控制3个方面阐述了机器学习方法在污水处理领域的应用,分析了机器学习方法相较于传统方法的优势及其应用于污水处理系统中存在的问题,展望了机器学习方法未来在污水处理领域应用的前景和趋势。

Wastewater treatment systems are usually very complicated and may be affected by many external factors. Therefore, control and management of these systems are always one of the great challenges in environmental engineering. Traditional controlling and managing approaches could not meet the needs of the increasingly complex wastewater treatment facilities. While the recently-developed machine learning methods provide a series of effective solutions for such problems. This article introduces the characteristics of machine learning methods, including artificial neural networks, support vector machines, random forests, etc., and explains the application of machine learning methods in the field of wastewater treatment systems from three aspects, i.e. water quality prediction and early warning, wastewater treatment system fault diagnosis and intelligent control. The advantages of machine learning methods and the challenges of their applications in wastewater treatment systems are also presented. In addition, the future development trends of machine learning methods in the field of wastewater treatment are outlined.