Wastewater Treatment Plants (WWTPs) are intricate systems involving numerous physical, chemical, and biological processes. Effective management and supervision of these systems are crucial to enhancing productivity and producing high-quality water. This research employs data-driven analysis, utilising various machine learning algorithms and optimisation techniques, to predict critical wastewater parameters in both the influent and effluent of WWTPs. By accurately predicting these parameters, the study aims to improve the performance, efficiency, and reliability of WWTPs.
In the subsequent phase, real-time data will be integrated into the model to develop control strategies. These strategies will be simulated using Benchmark Simulation Models to evaluate their effectiveness in maximising pollutant removal while minimising energy consumption and greenhouse gas (GHG) emissions. The research will explore different machine learning approaches to refine the predictive accuracy and optimise control mechanisms.
The research outputs will include advanced predictive models for influent wastewater characteristics, as well as models for predicting BOD, COD, and NH4 levels in the primary and secondary treatment effluent. Additionally, the study will estimate GHG emissions and energy consumption, ultimately leading to the development of optimised control strategies. These strategies will focus on reducing operational costs and minimising environmental impact, thereby offering a significant potential for enhancing the sustainability of WWTP operations.
School of Engineering
Email: h.khoshvaght@ecu.edu.au