通过积极学习策略进行领域适应,用于废水处理厂的异常分类
Francesca Bellamoli1, Marco Vian2, Mattia Di Iorio3
1Department of Information Engineering and Computer Science, University of Trento, via Sommarive 9, Trento 38123, Italy; ETC Sustainable Solutions Srl, via dei Palustei 16, Trento 38121, Italy
概括
本研究引入了用于废水处理厂 (WWTP) 的主动学习机器学习算法,以检测异常. 该方法可以有效地适应具有有限数据的新工厂,降低成本并优化管理.
科学领域:
- 环境工程 环境工程
- 废水处理技术 废水处理技术
- 机器学习应用 机器学习应用
背景情况:
- 废水处理厂 (WWTP) 的间歇通风控制器使用传感器数据来降低成本,但在异常检测方面遇到了困难.
- 在WWTP中用于异常检测的机器学习是数据密集型的,对新工厂的实施构成挑战.
- 传感器和过程异常可能导致WWTP的运营效率低下和成本增加.
研究的目的:
- 开发一种可适应的机器学习算法,用于识别间歇性通风WWTP中的异常.
- 为了使新的WWTP能够有效地适应新的WWTP,而可用的数据有限.
- 为优化WWTP管理和标签工作提供决策支持系统.
主要方法:
- 开发一种机器学习算法,利用主动学习来代地选择样本进行模型微调.
- 基于17个现有水电站数据的梯度增强模型的初始训练.
- 评估三个采样策略 (低概率,高) 以实现高效的模型适应.
主要成果:
- 积极学习显著提高了异常检测模型适应新WWTP的适应性,使用最小的数据.
- 低概率和高取样策略在早期模型适应方面被证明是有效的.
- 实现了接近最佳的F2得分,样本要求大大减少.
结论:
- 提出的积极学习方法为在新的WWTP中部署异常检测模型提供了一个有效的策略.
- 这种方法优化了标签工作,并通过有效的决策支持系统支持WWTP管理.
- 该算法有助于降低成本,提高废水处理过程的运行稳定性.
更多相关视频
11:31Continuously-stirred Anaerobic Digester to Convert Organic Wastes into Biogas: System Setup and Basic Operation
Published on: July 13, 2012
33.7K
08:49Vegetated Treatment Systems for Removing Contaminants Associated with Surface Water Toxicity in Agriculture and Urban Runoff
Published on: May 15, 2017
10.5K
相关概念视频
Environmental Applications of Microorganisms
1
Microorganisms play a pivotal role in maintaining ecosystem balance by recycling essential elements such as carbon, nitrogen, and phosphorus, as well as supporting processes like bioremediation, wastewater treatment, and biofuel production.Microbes in Elemental CyclesIn the carbon cycle, microorganisms decompose organic matter, releasing carbon dioxide via aerobic respiration. This carbon dioxide is subsequently used by photosynthetic organisms to synthesize organic compounds, closing the...
1
Bioremediation
18.2K
Bioremediation is the use of prokaryotes, fungi, or plants to remove pollutants from the environment. This process has been used to remove harmful toxins in groundwater as a byproduct of agricultural run-off and also to clean up oil spills.
18.2K
