一个多模型组合,用于在全尺度零液体排放系统中先进预测反透性能
Haojie Ding1, Ning Hao2, Qilin Cao2
1State Key Laboratory of Regional Environment and Sustainability, School of Environment, Tsinghua University, Beijing 100084, P. R. China.
本研究介绍了一种机器学习框架,用于预测反透 (RO) 性能,预测膜污染,以改善工业用水管理. ConvLSTM模型准确预测流体和盐排放,在零液体排放应用中提高系统稳定性.
科学领域:
- 环境工程 环境工程
- 水处理技术水处理技术
- 机器学习应用 机器学习应用
背景情况:
- 反透 (RO) 对于零液体排放 (ZLD) 和海水淡化至关重要,但面临着膜污染的挑战.
- 需要预测工具来积极管理RO系统的性能.
- 工业废水再利用和可持续的ZLD应用取决于稳定的RO运行.
研究的目的:
- 开发一个多维机器学习 (ML) 框架,用于预测工业ZLD系统中的RO性能.
- 准确预测污染相关的趋势,包括流体和盐排斥.
- 为了实现主动的运营调整,并提高RO系统的稳定性.
主要方法:
- 一个新的多维ML框架,包括数据采集,特征工程,ML建模,评估和决策.
- 评估六个ML模型,包括卷积长短期内存 (ConvLSTM) 和长短期内存 (LSTM) 网络.
- 在各种工业场景中进行外部验证,以证明框架的适应性.
主要成果:
- ConvLSTM网络在短期 (1天,R2=0.960) 和中期 (7天,R2=0.942) RO绩效预测方面表现出色.
- 对于长期 (30天) 预测,LSTM和ConvLSTM模型表现出相似的适用性.
- 该框架成功预测了污染趋势,并为各种运行条件选择了最佳模型.
结论:
- 拟议的ML框架有效地预测了工业ZLD系统的RO性能和污染趋势.
- 数据驱动的策略,特别是使用ConvLSTM,增强RO系统的稳定性并支持操作决策.
- 这种方法对于优化工业废水再利用和推进可持续的ZLD应用非常有价值.
更多相关视频
10:37Procedure to Evaluate the Efficiency of Flocculants for the Removal of Dispersed Particles from Plant Extracts
Published on: April 9, 2016
10:19Three-Dimensionally Printed Microfluidic Cross-flow System for Ultrafiltration/Nanofiltration Membrane Performance Testing
Published on: February 13, 2016
相关概念视频
Typical Model Studies
Osmosis and Osmotic Pressure of Solutions
Parameters Affecting Nonlinear Elimination: Zero-Order Input, First-Order Absorption and Two-Compartment Model
When a drug is administered through a constant intravenous infusion and eliminated via nonlinear pharmacokinetics, it follows zero-order input. For example, oral drugs undergo first-order absorption upon administration and are eliminated through nonlinear pharmacokinetics.
In the case of subcutaneously administered drugs,...
Mechanistic Models: Compartment Models in Individual and Population Analysis
One-Compartment Open Model for Extravascular Administration: Zero-Order Absorption Model
Zero-order absorption maintains a steady rate irrespective of the amount of drug left to be absorbed, making it a constant process. In the...
One-Compartment Open Model: Urinary Excretion Data and Determination of k
