A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Optimising Adsorption-Based Distillery Wastewater Treatment by Predicting Effluent Characteristics Using Machine Learning
Distillery wastewater is highly polluted and hazardous material resulting from alcohol production. Adsorption has proven to be an effective method used for treating distillery wastewater. In recent years, Machine Learning (ML) has emerged as a powerful technique in the field of wastewater treatment offering optimization capabilities. ML models are useful in predictions, monitoring, image recognition, and many more. This study work focuses on predicting the removal rate of pollutants namely, chemical oxygen demand, biochemical oxygen demand, total suspended solids, and colour. Batch adsorption was used for this study. This paper uses three sets of input parameters viz. (a) influent characteristics of distillery wastewater, (b) adsorbent particle size (c) adsorption conditions for the development of ML models. This research makes use of widely recognized ML models Random Forest (RF) and Extreme Gradient Boosting (XGBoost) to improve adsorption-based wastewater treatment processes. The models were constructed using Jupyter Notebook from Anaconda Navigator using Python programming language. The root mean square error (RMSE), mean absolute error, and coefficient of determination (R2) were used in the model's validation phase. The most accurately predicted parameter is the colour removal rate by XGBoost with an R2 value of 0.9991 and RMSE value of 0.5342. XGBoost model demonstrates prediction accuracy ranging from 0.9910 to 0.9991, while RF shows accuracy from 0.9756 to 0.9887. XGBoost model outperformed the RF model in all parameters and is best fitted for predicting removal rates of pollutants over the RF model. This paper explores the use of machine learning for predicting effluent characteristics, paving the way for future research and advancements in sustainable distillery wastewater treatment.
Optimising Adsorption-Based Distillery Wastewater Treatment by Predicting Effluent Characteristics Using Machine Learning
Distillery wastewater is highly polluted and hazardous material resulting from alcohol production. Adsorption has proven to be an effective method used for treating distillery wastewater. In recent years, Machine Learning (ML) has emerged as a powerful technique in the field of wastewater treatment offering optimization capabilities. ML models are useful in predictions, monitoring, image recognition, and many more. This study work focuses on predicting the removal rate of pollutants namely, chemical oxygen demand, biochemical oxygen demand, total suspended solids, and colour. Batch adsorption was used for this study. This paper uses three sets of input parameters viz. (a) influent characteristics of distillery wastewater, (b) adsorbent particle size (c) adsorption conditions for the development of ML models. This research makes use of widely recognized ML models Random Forest (RF) and Extreme Gradient Boosting (XGBoost) to improve adsorption-based wastewater treatment processes. The models were constructed using Jupyter Notebook from Anaconda Navigator using Python programming language. The root mean square error (RMSE), mean absolute error, and coefficient of determination (R2) were used in the model's validation phase. The most accurately predicted parameter is the colour removal rate by XGBoost with an R2 value of 0.9991 and RMSE value of 0.5342. XGBoost model demonstrates prediction accuracy ranging from 0.9910 to 0.9991, while RF shows accuracy from 0.9756 to 0.9887. XGBoost model outperformed the RF model in all parameters and is best fitted for predicting removal rates of pollutants over the RF model. This paper explores the use of machine learning for predicting effluent characteristics, paving the way for future research and advancements in sustainable distillery wastewater treatment.
Optimising Adsorption-Based Distillery Wastewater Treatment by Predicting Effluent Characteristics Using Machine Learning
Lecture Notes in Civil Engineering
Nehdi, Moncef (editor) / Rahman, Rahimi A. (editor) / Davis, Robin P. (editor) / Antony, Jiji (editor) / Kavitha, P. E. (editor) / Jawahar Saud, S. (editor) / Bhoye, Dipak (author) / Vyas, Gayatri S. (author) / Nikhar, Chaitali K. (author) / Dalvi, Rupa S. (author)
International Conference on Structural Engineering and Construction Management ; 2024 ; Angamaly, India
2024-12-29
15 pages
Article/Chapter (Book)
Electronic Resource
English
Integrated treatment of molasses distillery wastewater using microfiltration (MF)
| Online Contents | 2015
Wet air oxidation induced enhanced biodegradability of distillery effluent
| Online Contents | 2014
Prediction of effluent concentration in a wastewater treatment plant using machine learning models
| Online Contents | 2015