| dc.description.abstract | Coagulation is vital in water treatment, but predicting coagulant doses using mechanistic models is challenging due to complex processes. Traditionally, jar testing is used, but it struggles with varying water quality. This study explores machine learning (ML) techniques to predict effluent quality and coagulant requirements. It found that random data splits yield higher similarity scores and better performance than time-based splits. Random forest regressor (RFR) and linear support vector regressor (Linear SVR) performed well in high and low similarity scenarios, respectively, while multilayer perceptron (MLP) showed consistent performance across all conditions. A hybrid ML model combining RFR, Linear SVR, and MLP outperformed individual models, reducing errors by 1%–45%. Converted to ONNX format, the model was integrated into the digital twin of a water purification plant in Texas, improving real-time predictions. This study demonstrates the potential of ML in optimizing water treatment and reducing the need for traditional testing methods. This study demonstrates how machine learning (ML) can improve the coagulation process in water treatment, offering a more efficient alternative to traditional jar testing. By using ML models, water treatment plants can predict the necessary coagulant doses more accurately and quickly, without relying on time-consuming manual tests. The hybrid ML model developed in this study combines multiple techniques to handle varying water quality conditions, improving the precision of coagulant predictions. The model has practical applications in real-time operations. For example, it has been successfully integrated into the digital twin of a water purification plant in Texas, where it provides real-time predictions to optimize treatment processes. This results in more efficient water treatment, reduced chemical usage, and lower operational costs. By reducing the need for traditional testing methods, this approach allows water treatment plants to respond faster to changes in water quality, ensuring more consistent and reliable water supply. This research shows that machine learning can be a valuable tool for improving water treatment processes, offering both operational and economic benefits for plants worldwide. | |
