Development of a Real-Time Intelligent Platform for Wastewater Influent Forecasting

城市污水处理厂(WWTP)需要可靠的进水预测来设计、运行和控制污水处理工艺。对进水流量和负荷的不准确估计很容易导致处理效果差和运行成本高。城市污水的复杂收集和运输过程，以及基础设施老化带来的不确定性，使污水处理厂管理人员面临着编制可靠的废水流入预测的艰巨挑战。在本项目中，将开发一个污水进水预测的实时智能平台。这需要四项任务：1)开发模式挖掘层，以表征废水流入和污染负荷的时间波动；2)开发智能预测层，以生成可靠的进水预测；3)开发实时可视化层，以提供废水数据和预测的可视化；以及4)集成所开发的预测平台并将其商业化。这项拟议的研究是对发展用于支持污水处理厂管理的进水预测方法的独特贡献。它将为高素质的人员提供专业培训。学生将获得数据处理、时间序列分析、模式挖掘和机器学习技术方面的专业知识，这些技术在咨询和学术界非常重要。这项工作将由海德曼蒂斯环境软件解决方案公司和申请人在麦克马斯特大学的研究小组合作完成。海德曼蒂斯是加拿大领先的污水处理模拟公司之一。该项目将使他们能够将开发的预测平台整合到他们目前的软件产品中，这将使他们获得显著的市场优势。这项工作将产生重大的经济影响，因为它将为最大限度地利用该国现有和未来的废水处理设施提供宝贵的技术支持。

Municipal wastewater treatment plants (WWTPs) require reliable influent forecast for the design, operation, and control of wastewater treatment processes. Inaccurate estimates of influent flow and load could easily lead to poor treatment performance and high operating cost. The complex collection and transport processes of urban effluent, as well as uncertainties arising from aging infrastructure, make it a daunting challenge for WWTP managers to generate reliable wastewater influent forecasts. In this project, a real-time intelligent platform for wastewater influent forecasting will be developed. This entails four tasks: 1) develop a pattern mining layer to characterize the temporal fluctuations in wastewater inflow and pollution loading; 2) develop an intelligent forecasting layer to generate reliable influent forecasts; 3) develop a real-time visualization layer to provide visualization of wastewater data and forecasts; and 4) integrate and commercialize the developed forecasting platform. The proposed research represents a unique contribution to the development of influent forecasting methodologies for supporting WWTP management. It will provided specialized training for highly qualified personnel. The students will gain expertise in data processing, time series analysis, pattern mining, and machine learning techniques, which are highly valued in consulting and academia. This work will be completed collaboratively between Hydromantis Environmental Software Solutions, Inc. and the applicant's research group at McMaster University. Hydromantis is one of the Canada's leading companies in wastewater treatment simulation. This project will allow them to integrate the developed forecasting platform into their current software products, which will give them a significant market advantage. The impacts of this work will be economically significant, as it will provide valuable technical support for making the maximum use of the country's existing and future wastewater treatment facilities.