Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
The Use of Artificial Neural Networks for Modeling Color and Chemical Oxygen Demand Removal from Olive Mill Wastewater Using Grape Molasses Soil
https://orcid.org/http://orcid.org/0000-0002-1286-7846
Olive mill wastewater (OMW) is classified as highly pollutant-containing wastewater that must be treated to an acceptable level before being discharged into receiving water environments. Recently, many chemical matters have been widely used in OMW treatment, but their harmful chemical structures have caused various health and environmental problems and have been economically challenging. Therefore, this research presents the use of natural grape molasses soil as a chemical coagulant for OMW treatment. An artificial neural network (ANN) modeling technique was used to predict effluent color and chemical oxygen demand (COD) results after the coagulation-precipitation process. The input variables of the ANN process were mixing time, mixing speed, amount of grape molasses soil, effluent pH, and effluent turbidity, while the responses of the ANN were effluent color (λ436, λ525, and λ620) and effluent COD. Twelve different training algorithms were used to develop the ANN model. Bayesian regularization (trainbr) indicated to be the best backpropagation training algorithm with the lowest mean-squared error (0.0089). The appropriate architecture of an ANN-based model for use in effluent color and COD prediction consists of 5 input variables, tangential sigmoid transfer function (tansig) in 10 hidden layer neurons, and a linear transfer function (purelin) in 4 output layer neurons. The ANN testing R2 for effluent color (λ436, λ525, and λ620) and effluent COD has been determined to be 0.9742, 0.9711, 0.9624, and 0.8773, respectively. It is necessary to model and optimize the chemical treatment process with grape molasses soil of OMW in the search for a more efficient and economical process. This study enabled a reduction in the electricity required and the amount of coagulant consumed by optimizing the mixing time, mixing speed, and the amount of grape molasses soil added. The results proved that the proposed ANN performed well in predicting the effluent color and effluent COD with the use of natural grape molasses soil while avoiding economic and environmental sustainability problems.
The Use of Artificial Neural Networks for Modeling Color and Chemical Oxygen Demand Removal from Olive Mill Wastewater Using Grape Molasses Soil
https://orcid.org/http://orcid.org/0000-0002-1286-7846
Olive mill wastewater (OMW) is classified as highly pollutant-containing wastewater that must be treated to an acceptable level before being discharged into receiving water environments. Recently, many chemical matters have been widely used in OMW treatment, but their harmful chemical structures have caused various health and environmental problems and have been economically challenging. Therefore, this research presents the use of natural grape molasses soil as a chemical coagulant for OMW treatment. An artificial neural network (ANN) modeling technique was used to predict effluent color and chemical oxygen demand (COD) results after the coagulation-precipitation process. The input variables of the ANN process were mixing time, mixing speed, amount of grape molasses soil, effluent pH, and effluent turbidity, while the responses of the ANN were effluent color (λ436, λ525, and λ620) and effluent COD. Twelve different training algorithms were used to develop the ANN model. Bayesian regularization (trainbr) indicated to be the best backpropagation training algorithm with the lowest mean-squared error (0.0089). The appropriate architecture of an ANN-based model for use in effluent color and COD prediction consists of 5 input variables, tangential sigmoid transfer function (tansig) in 10 hidden layer neurons, and a linear transfer function (purelin) in 4 output layer neurons. The ANN testing R2 for effluent color (λ436, λ525, and λ620) and effluent COD has been determined to be 0.9742, 0.9711, 0.9624, and 0.8773, respectively. It is necessary to model and optimize the chemical treatment process with grape molasses soil of OMW in the search for a more efficient and economical process. This study enabled a reduction in the electricity required and the amount of coagulant consumed by optimizing the mixing time, mixing speed, and the amount of grape molasses soil added. The results proved that the proposed ANN performed well in predicting the effluent color and effluent COD with the use of natural grape molasses soil while avoiding economic and environmental sustainability problems.
The Use of Artificial Neural Networks for Modeling Color and Chemical Oxygen Demand Removal from Olive Mill Wastewater Using Grape Molasses Soil
https://orcid.org/http://orcid.org/0000-0002-1286-7846
Olive mill wastewater (OMW) is classified as highly pollutant-containing wastewater that must be treated to an acceptable level before being discharged into receiving water environments. Recently, many chemical matters have been widely used in OMW treatment, but their harmful chemical structures have caused various health and environmental problems and have been economically challenging. Therefore, this research presents the use of natural grape molasses soil as a chemical coagulant for OMW treatment. An artificial neural network (ANN) modeling technique was used to predict effluent color and chemical oxygen demand (COD) results after the coagulation-precipitation process. The input variables of the ANN process were mixing time, mixing speed, amount of grape molasses soil, effluent pH, and effluent turbidity, while the responses of the ANN were effluent color (λ436, λ525, and λ620) and effluent COD. Twelve different training algorithms were used to develop the ANN model. Bayesian regularization (trainbr) indicated to be the best backpropagation training algorithm with the lowest mean-squared error (0.0089). The appropriate architecture of an ANN-based model for use in effluent color and COD prediction consists of 5 input variables, tangential sigmoid transfer function (tansig) in 10 hidden layer neurons, and a linear transfer function (purelin) in 4 output layer neurons. The ANN testing R2 for effluent color (λ436, λ525, and λ620) and effluent COD has been determined to be 0.9742, 0.9711, 0.9624, and 0.8773, respectively. It is necessary to model and optimize the chemical treatment process with grape molasses soil of OMW in the search for a more efficient and economical process. This study enabled a reduction in the electricity required and the amount of coagulant consumed by optimizing the mixing time, mixing speed, and the amount of grape molasses soil added. The results proved that the proposed ANN performed well in predicting the effluent color and effluent COD with the use of natural grape molasses soil while avoiding economic and environmental sustainability problems.
The Use of Artificial Neural Networks for Modeling Color and Chemical Oxygen Demand Removal from Olive Mill Wastewater Using Grape Molasses Soil
Environ Model Assess
Tufaner, Fatih (Autor:in)
Environmental Modeling & Assessment ; 27 ; 855-868
01.10.2022
14 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
| British Library Conference Proceedings | 2015
Olive oil mill wastewater for soil nitrogen and carbon conservation
| Online Contents | 2009
Olive oil mill wastewater for soil nitrogen and carbon conservation
| Online Contents | 2009