Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Scientific Reports volume 14, Article number: 13750 (2024 ) Cite this article Ao Sewage Treatment Plant
In this investigation, the modeling of the Aksaray industrial wastewater treatment plant was performed using artificial neural networks with various architectures in the MATLAB software. The dataset utilized in this study was collected from the Aksaray wastewater treatment plant over a 9-month period through daily records. The treatment efficiency of the plants was assessed based on the output values of chemical oxygen demand (COD) output. Principal component analysis (PCA) was applied to furnish input for the Feedforward Backpropagation Artificial Neural Networks (FFBANN). The model’s performance was evaluated using the Mean Squared Error (MSE), the Mean Absolute Error (MAE) and correlation coefficient (R2) parameters. The optimal architecture for the neural network model was determined through several trial and error iterations. According to the modeling results, the ANN exhibited a high predictive capability for plant performance, with an R2 reaching up to 0.9997 when comparing the observed and predicted output variables.
Water quality is a critical global concern, especially when considering the scarcity of clean water, necessitating the optimal utilization of limited resources. Pollution arising from the discharge of chemical pollutants in both industrial and domestic wastewaters poses a significant threat to aquatic biota1. Improper operation of wastewater treatment plants (WWTPs) can lead to environmental and public health problems. The discharge of treated effluent into water sources has the potential to spread disease and have adverse effects on aquatic ecosystems. Therefore, the implementation of effective monitoring and control techniques for wastewater systems is well-recognized by water and environmental engineers. However, modeling a WWTP presents challenges due to the complexity of the processes involved2. The relevant physical, biological, and chemical processes exhibit nonlinear behaviours, making them challenging to describe using conventional linear mathematical models3. Therefore, reliable and easy monitoring of treatment systems can be achieved by developing robust, nonlinear methods capable of predicting WWTP performance using measured values of wastewater parameters.
In recent years, the use of machine learning algorithms has become increasingly appealing as an alternative to the induction strategy for deriving models from experimental data in wastewater processes4. AI-based models are used to benchmark the performance of WWTPs, because their process control systems can be seamlessly integrated with AI models, enabling fully automated operation with minimal human intervention5. AI-based models for wastewater treatment systems can effectively handle the nonlinearity and complexity of these processes. In particular, the ANN model overcomes the limitations of earlier numerical and mathematical models developed for wastewater treatment systems6. There are many statistical and AI-based predictive models used for this purpose. Among the widely and successfully used methods are various Artificial neural networks (ANNs)5, response surface methodology (RSM)7,8, radial basis function (RBF) 8, and fuzzy logic9.
ANNs, as computational techniques, are nonlinear models designed to mimic the functionality and decision-making processes of the human brain10. ANNs have been increasingly applied in various environmental modeling studies11,12 and investigations into water quality issues13,14. In the domain of wastewater treatment plant (WWTP) modeling, ANNs have been successfully employed for the prediction of WWTP parameters15,16,17, process control18,19,20,21, and the estimation of output parameters and characteristics22,23. However, many of these studies necessitate diverse input data, contributing to the costliness and time-consuming nature of the modeling process.
This study demonstrated the use of an ANN for predicting and forecasting chemical oxygen demand (COD) within Aksaray Industrial wastewater Treatment Plant (AIWWTP), where intricate and dynamic processes are concealed within the monitored datasets. The training process focused on estimating and predicting parameters, specifically COD, by minimizing the error function while adhering to the permissible limits set by the AIWWTP license. This technique serves as a valuable complement to conventional mathematical modeling methods commonly employed for the prediction and forecasting of wastewater treatment plants.
PCA stands out as a widely adopted statistical technique in the domain of dimensionality reduction and multivariate data analysis and has gained increasing popularity over the past two decades across various fields24. Successful applications of PCA analysis have been observed in modeling diverse industrial processes, including the modeling of the IWWTP25 within this period. The fundamental principle behind PCA lies in leveraging the collinear nature of the data to effectively reduce the dimensionality of the measurement space. This is achieved by introducing a few essential pseudovariables known as principal components (PCs). These components serve to elucidate the primary mechanisms steering the underlying process and are typically fewer in number than the measured variables. PCA represents one of the multivariate statistical method that can simplify the complexity of input variables, particularly when dealing with extensive information volumes, aiming for an enhanced interpretation of variables26.
ANNs, which are extensively employed for predicting valuable data from nonlinear variables, are shaped by three fundamental components: architecture, activation functions, and the training algorithms27. This network comprises input, hidden, and output layers. The input layer accepts variables and facilitates transmission, while the hidden layer conveys variations to the output layer. The output layer generates the final output of the structure. Each layer is interconnected by nodes (neurons), each performing a distinct nonlinear activation function. A hidden node produces an intermediate output by performing a weighted sum of inputs and then transforming it with a transfer function. Hidden nodes transfer data to connected nodes in the next layer until the output layer completes the process by producing the final output28. Factors such as the number of layers, neurons, and the type of activation function employed significantly influence the performance of the ANN model structure. However, careful consideration and selection of these parameters are crucial when constructing a model for specific applications29,30,31.
ANNs represent a complex computational framework with a distributed nature, characterized by multiple processing elements operating concurrently. Within this intricately structured system, interconnected components possess the inherent capability to autonomously adapt their connection strengths during the learning process. The primary aim of this research was to predict COD concentrations in industrial wastewater by creating ANN structure using MATLAB (MATRIX LABORATORY) Online (basic version) mathematical software32. The ANN employed in this analysis consists of an input layer, a hidden layer, and an output layer, each comprising multiple neurons, as depicted in Fig. 1. To prevent numerical overflows from arising from excessively large or small weights, normalization of the input and output data was conducted, constraining them to a range between 0 and 1, as exemplified in Eq. (1)33:
Schematic view of a feedforward neural network with an input layer, a hidden layer and an output layer.
In this context, the normalized value (xnorm) is calculated based on the original data (xi) using the maximum (xmax) and minimum (xmin) values. This process ensures that the scaled data fall within the range of 0–1.
For this study, the FFBPANN algorithm, initially proposed by Rumelhart, Hinton, and Williams27 was employed. The FFBPANN algorithm is highly effective at learning in ANNs; it operates by propagating the error from the output layer back through the hidden layer and to the input layer of the network to achieve the desired final outputs. The algorithm utilizes the gradient descent technique to calculate the network's weight and adjust the interconnection weights to minimize the output error, as shown in Eq. (2)34:
In Eq. (2), the connective weight (Wix) represents the weight associated with a particular connection, while η denotes the learning rate that influences the weight adjustment process. The error signal (\({\delta }_{x}^{n}\) ) and the output value of the sublayer (\({A}_{i}^{n-1}\) ) also play crucial roles in determining the new weight values. The summation function is employed to compute the weighted sum of all the input signals, serving as the initial step in the network's computation process, as described in Eq. (3)35:
In this paper, a hyperbolic tangent sigmoid transfer function was used in the hidden layer and a linear transfer function was used for the output layer36. To determine the ideal architecture, neural networks were trained using varying iteration numbers (epochs). The dataset was subjected to random partitioning, resulting in three separate subsets: 70% for training, 15% for validation, and the remaining 15% for testing purposes.
In this study, measurement data for wastewater parameters monitored at the AIWWTP were collected over a 9-month period. Wastewater samples for these parameters were collected daily by experts from the facility and analyzed in an accredited laboratory. A total of 19 parameters were utilized for the ANN. These parameters are listed in Table 1.
The purpose of the performance evaluation of the trained ANN model was to assess the quality of the developed model. To achieve this, several statistical measurements were considered when evaluating the performance of the ANN model. These include the mean squared error (MSE) Eq. (4), the mean absolute error (MAE) Eq. (5) and the coefficient of multiple determination (R2) Eq. (6)37. These metrics offer insights into the model's predictive efficiency across different datasets; lower MSE and MAE values indicate better performance, while higher R2 scores suggest improved explanatory power.
Prior to predicting COD concentrations, the wastewater characteristics in the AIWWTP were consistently monitored daily from March to November 2023. The values of the monitored parameters are described in Table 1.
A wastewater treatment facility monitors various physical, chemical, and microbiological parameters38. Table 1 presents the parameters that are tracked daily, influencing CODeffluent and indicating treatment efficiency. Additionally, the facility conducts less frequent microbiological analyses. However, these analyses are performed primarily in instances of unusual declines in treatment efficiency and are therefore not considered within the model limits.
The way we choose the input parameters truly matters for how well the ANN model works. In this study, we used PCA to identify a smaller set of important new variables (components and dimensions) by combining different physical and chemical factors that might affect the amount of CODeffluent. The PCA analysis is commonly employed with the aim of reducing the number of input variables provided to the ANN model and minimizing deviations in the dataset39.
PCA was used to calculate component scores for each analyzed parameter in order to predict COD concentrations. The PCA analysis, as illustrated in Fig. 2, revealed the components. Beyond the fifth point, the contribution of components to the variance decreases, and the contributions of additional variances become comparable to each other. Therefore, five variables were utilized as inputs for the ANN to predict COD concentrations. The first five components illuminate the variation in COD concentration examined in our study. In the subsequent step, component scores were determined for each water sample, and the score coefficient matrix is detailed in Table 2.
Total variance explained by PCA analyses.
The components were identified by considering component loadings greater than 0.4540. In the predicted first principal component (PC1), the copper, zinc, iron, fluoride, cadmium, total, chromium, chromium+6, lead, CODinfluent and TSS parameters were grouped. Therefore, the first principal component primarily explained the variation in the CODeffluent, the parameter under consideration in our study, accounting for 36.98% of the total parameter variance. The second principal component (PC2) encompasses color, total phosphorus, TKN and fat-grease parameters and contributed to 14.67% of the total variance. Moreover, the third principal component (PC3) included sulfate and fish bioassay parameters, explaining 10.42% of the overall parameter variance. PC4 involves pH and total cyanide content, whereas PC5 included the mercury content. In conclusion, the first five components collectively elucidate 77.95% of the total variance. The positive or negative component loadings of the input parameters provide insights into the direction of the impact on the CODeffluent parameter.
In prediction models, the structure of the ANN model is crucial, and hence, the number of neurons in the hidden layer plays a significant role. Table 3 presents the results of the study conducted to determine the optimal number of neurons for the network structure, with R2, MSE and MAE outcomes obtained. In the study conducted as described in Fig. 3, utilizing a single hidden layer and employing the Levenberg–Marquardt (trainlm) training function, the R2 values corresponding to varying numbers of neurons are meticulously presented. Figure 4 illustrates how the predictive capacity of a FFBPANN with 10 neurons in each hidden layer is influenced by the number of hidden layers. The transfer function applied to neurons was the tangent sigmoid, while the training function utilized was trainlm. As demonstrated, the optimal prediction capability for the dataset was achieved with a single hidden layer.
The R2 values across networks with varying numbers of neurons for the complete dataset.
The R2 values across networks with different hidden layers.
In ANN studies, the process of achieving optimal learning and results close to the ground truth involves running the constructed network structure iteratively. During each iteration, the network generates an average mean squared error and correlation coefficient. The network's operation concludes when the model with the lowest mean squared error and the highest correlation coefficient is attained, and the network is then saved. In this research, Fig. 5 presents the regression results from the network that produced the lowest mean squared error and the highest correlation coefficient. The best determination coefficient was achieved when 70% of the data were used for training the neural network, 15% for validation, and 15% for testing. The R2 values obtained for training, validation, and testing were 0.99, 0.99, and 0.99, respectively.
The correlation coefficients (R) for training, validation, testing, and overall results obtained from the analysis using the ANN structure.
As observed in Fig. 6, the comparison between the ANN predictions and the measured COD values demonstrates that the ANN prediction model performs exceptionally well. When comparing the values, the R2 value is calculated as 0.9997. This correlation coefficient indicates that the model is suitable for this study and can be reliably utilized in similar research. In the literature, Rashidi and Moghaddam41 used a multi-layer feedforward neural network for COD removal and developed a predictive model with a high correlation coefficient. Moghaddam et al.42 used a backpropagation neural network and found an R2 of 0.9843 for predicting the output parameter. In their study conducted at a wastewater treatment facility, Khatri et al. 5 compared various ANN models. The best predictive model was found to be the deep feedforward backpropagation (DFFBP) algorithm, which was implemented with 3 hidden layers and 11 neurons. This model achieved an initial R correlation coefficient of 0.997 on the training data set.
Correlation between the measured and predicted COD data.
In both the world and our country, wastewater from various sectors is pre-treated before entering organized industrial wastewater treatment facilities, resulting in a homogenous pollution load based on the monitored parameters. Consequently, the flow schemes of industrial wastewater treatment plants are generally similar. In our study, the ANN model, which exhibits high predictive capability, can be applied to numerous domestic-industrial wastewater treatment facilities. For specialized industrial wastewater treatment plants, ANN prediction can be conducted using parameters specific to that industry.
In the present investigation, the FFBPANN model was applied to determine the daily CODeffluent concentrations (R2) for the year 2023 at the inlet of the AIWWTP. Notably, the obtained R2 value for CODeffluent was 0.9997, with an associated MSE value of 0.0624, revealing the efficacy of the network structures. The optimal performance was achieved with the 5-10-1 ANN model, demonstrating its superiority in both the test and training datasets. These outcomes underscore the remarkable success of the ANN model employed in this research. Furthermore, the application of ANNs in the COD estimation was identified as a crucial aspect of PCA analysis aimed at determining the most effective model. The effectiveness of the FFBPANN model relies significantly on the input parameters, leading to superior results compared to traditional models.
Well-trained ANN parameters are crucial for providing reliable predictions in the wastewater treatment processes employed in WWTPs. In this study, it was concluded that the ANN model is successful at predicting the COD levels of WWTPs in terms of reliable and realistic results.
The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.
Jami, M. C., Husain, I. A. & Abdullah, N. A. Multiple inputs artificial neural network model for the prediction of wastewater treatment plant performance. J. AJBAS. 6(1), 62–69 (2012).
Ye, Z. et al. Tackling environmental challenges in pollution controls using artificial intelligence: A review. Sci. Total Environ. 699, 134279. https://doi.org/10.1016/j.scitotenv.2019.134279 (2020).
Article ADS CAS PubMed Google Scholar
Baleanu, D., Karaca, Y., Vázquez, L. & Macías-Díaz, J. E. Advanced fractional calculus, differential equations and neural networks: Analysis, modeling and numerical computations. Phys. Scr. 98(11), 110201. https://doi.org/10.1088/1402-4896/acfe73 (2023).
Khatri, N., Khatri, K. K. & Sharma, A. Artificial neural network modelling of faecal coliform removal in an intermittent cycle extended aeration system-sequential batch reactor based wastewater treatment plant. J. Water Process Eng. 37, 101477. https://doi.org/10.1016/j.jwpe.2020.101477 (2020).
Khatri, N., Vyas, A. K., Abdul-Qawy, A. S. H. & Rene, E. R. Artificial neural network based models for predicting the effluent quality of a combined upflow anaerobic sludge blanket and facultative pond: Performance evaluation and comparison of different algorithms. Environ. Res. 217, 114843. https://doi.org/10.1016/j.envres.2022.114843 (2023).
Article CAS PubMed Google Scholar
Khatri, N., Khatri, K. K. & Sharma, A. Prediction of effluent quality in ICEAS-sequential batch reactor using feedforward artificial neural network. Water Sci. Technol. 80(2), 213–222. https://doi.org/10.2166/wst.2019.257 (2019).
Article CAS PubMed Google Scholar
Moghaddam, A. H., Hazrati, H., Sargolzaei, J. & Shayegan, J. Assessing and simulation of membrane technology for modifying starchy wastewater treatment. Appl. Water Sci. 7, 2753–2765. https://doi.org/10.1007/s13201-016-0503-3 (2017).
Article ADS CAS Google Scholar
Vaziri, H., Moghaddam, A. H. & Mirmohammadi, S. A. Optimization of distillation column in phenol production process for increasing the isopropyl benzene concentration using response surface methodology and radial basis function (RBF) coupled with leave-one-out validation method. Chem. Pap. 74, 3311–3324. https://doi.org/10.1007/s11696-020-01162-w (2020).
Yelagandula, S. & Ginuga, P. Control of a waste water treatment plant using fuzzy logic controller. J. Inst. Eng. India Ser. E 103, 167–177. https://doi.org/10.1007/s40034-022-00241-9 (2022).
Obianyo, J. I., Udeala, R. C. & Alaneme, G. U. Application of neural networks and neuro-fuzzy models in construction scheduling. Sci. Rep. 13, 8199. https://doi.org/10.1038/s41598-023-35445-5 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Gulati, S. et al. Estimating PM2.5 utilizing multiple linear regression and ANN techniques. Sci. Rep. 13, 22578. https://doi.org/10.1038/s41598-023-49717-7 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Seo, Y., Kim, S. & Singh, V. P. Comparison of different heuristic and decomposition techniques for river stage modelling. Environ. Monit. Assess. 190(7), 392. https://doi.org/10.1007/s10661-018-6768-2 (2018).
Hayder, G., Solihin, M. I. & Kushiar, K. F. B. A Performance comparison of various artificial intelligence approaches for estimation of sediment of river systems. J. Ecol. Eng. 22(7), 20–27. https://doi.org/10.12911/22998993/137847 (2021).
Shamshirband, S. et al. Ensemble models with uncertainty analysis for multi-day ahead forecasting of chlorophyll a concentration in coastal waters. Eng. Appl. Comp. Fluid. 13(1), 91–101 (2019).
Dağtekin, M. & Yelmen, M. Modeling wastewater treatment plant (WWTP) performance using artificial neural networks: Case of Adana (Seyhan). Int. J. Agric. Environ. Food Sci. 6(4), 579–584. https://doi.org/10.31015/jaefs.2022.4.10 (2022).
Lou, I. & Zhao, Y. Sludge bulking prediction using principal component regression and artificial neural network. Math. Probl. Eng. https://doi.org/10.1155/2012/237693 (2012).
Duarte, M. S. et al. A review of computational modeling in wastewater treatment processes. ACS EST Water 4(3), 784–804 (2023).
Tumer, A. E. & Edebali, S. An artificial neural network model for wastewater treatment plant of Konya. Int. J. Intel. Syst. Appl. Eng. 3(4), 131–213. https://doi.org/10.18201/ijisae.65358 (2015).
Vyas, M., Modhera, B., Vyas, V. & Sharma, A. K. Performance forecasting of common effluent treatment plant parameters by artificial neural network. J. Environ. Appl. Sci. 6(1), 38–42 (2011).
Akratos, C. S., Papaspyros, J. N. & Tsihrintzis, V. A. An artificial neural network model and design equations for BOD and COD removal prediction in horizontal subsurface flow constructed wetlands. Chem. Eng. J. 143(1–3), 96–110. https://doi.org/10.1016/j.cej.2007.12.029 (2008).
Hamed, M. M., Khalafallah, M. G. & Hassanien, E. A. Prediction of wastewater treatment plant performance using artificial neural networks. Environ. Model Softw. 19(10), 919–928. https://doi.org/10.1016/j.envsoft.2003.10.005 (2004).
Abba, S. I. & Elkiran, G. Effluent prediction of chemical oxygen demand from the WWTP using artificial neural network application. Procedia Comput. Sci. 120, 156–163. https://doi.org/10.1016/j.procs.2017.11.223 (2017).
Han, H. G., Qiao, J. F. & Chen, Q. L. Model predictive control of dissolved oxygen concentration based on a self-organizing RBF neural network. Control Eng. Pract. 20(4), 465–476. https://doi.org/10.1016/j.conengprac.2012.01.001 (2012).
Bruni, V., Cardinali, M. L. & Vitulano, D. a short review on minimum description length: An application to dimension reduction in PCA. Entropy 24(2), 269. https://doi.org/10.3390/e24020269 (2022).
Article ADS MathSciNet PubMed PubMed Central Google Scholar
Arismendy, L. et al. Intelligent system for the predictive analysis of an industrial wastewater treatment process. Sustainability 12(16), 6348. https://doi.org/10.3390/su12166348 (2020).
Greenacre, M. et al. Principal component analysis. Nat. Rev. Methods Primers 2, 100. https://doi.org/10.1038/s43586-022-00184-w (2022).
García-Alba, J., Barcena, J. F., Ugarteburu, C. & García, A. Artificial neural networks as emulators of process-based models to analyze bathing water quality in estuaries. Water Res. 150, 283–295. https://doi.org/10.1016/j.watres.2018.11.063 (2019).
Article CAS PubMed Google Scholar
Kang, J. H., Song, J., Yoo, S. S., Lee, B. J. & Ji, H. W. Prediction of odor concentration emitted from wastewater treatment plant using an artificial neural network (ANN). Atmosphere 11, 784. https://doi.org/10.3390/atmos11080784 (2020).
Article ADS CAS Google Scholar
Yaseen, Z. M., El-Shafie, A., Jaafar, O., Afan, H. A. & Sayl, K. N. Artificial intelligence based models for stream-flow forecasting: 2000–2015. J. Hydrol. 530, 829–844. https://doi.org/10.1016/j.jhydrol.2015.10.038 (2015).
Jamro, I. A. et al. Investigation of enhanced H2 production from municipal solid waste gasification via artificial neural network with data on tar compounds. J. Environ. Manag. 328, 117014. https://doi.org/10.1016/j.jenvman.2022.117014 (2023).
Mathur, R. et al. Modeling of two-stage anaerobic onsite wastewater sanitation system to predict effluent soluble chemical oxygen demand through machine learning. Sci. Rep. 14, 1835. https://doi.org/10.1038/s41598-023-50805-x (2014).
Article ADS CAS Google Scholar
URL-1, MATLAB Online (basic-free), Retrieved September 2023 from https://matlab.mathworks.com
Yang, S., Wang, X., Yang, Q., Dong, E. & Du, S. Instance segmentation based on improved self-adaptive normalization. Sensors 22, 4396. https://doi.org/10.3390/s22124396 (2022).
Article ADS PubMed PubMed Central Google Scholar
Rumelhart, D. E., Hinton, G. E. & Williams, R. J. Learning representations by back-propagating errors. Nature 323, 533–536. https://doi.org/10.1038/323533a0 (1986).
Kim, T. W. & Valdés, J. B. Nonlinear model for drought forecasting based on a conjunction of wavelet transforms and neural networks. J. Hydrol. Eng. 8(6), 319–328. https://doi.org/10.1061/(ASCE)1084-0699(2003)8:6(319) (2003).
Li, P., Zha, Y., Zuo, B. & Zhang, Y. A family of soil water retention models based on sigmoid functions. Water Resour. Res. 59(3), e2022WR033160. https://doi.org/10.1029/2022WR033160 (2023).
Ozel, H. U. et al. Application of artificial neural networks to predict the heavy metal contamination in the Bartin River. Environ. Sci. Pollut. Res. 27(34), 42495–42512. https://doi.org/10.1007/s11356-020-10156-w (2020).
Bourgeois, W., Burgess, J. E. & Stuetz, R. M. On-line monitoring of wastewater quality: A review. J. Chem. Technol. Biotechnol. 76, 337–348. https://doi.org/10.1002/jctb.393 (2001).
Granato, D., Santos, J. S., Escher, G. B., Ferreira, B. L. & Maggio, R. M. Use of principal component analysis (PCA) and hierarchical cluster analysis (HCA) for multivariate association between bioactive compounds and functional properties in foods: A critical perspective. Trends Food Sci. Technol. 72, 83–90. https://doi.org/10.1016/j.tifs.2017.12.006 (2018).
Tezbaşaran, E. & Gelbal, S. An investigation on usability of principal component analysis and artificial neural network models in the process of scale development. Mersin Univ. J. Fac. Educ. 14(1), 225–252. https://doi.org/10.17860/mersinefd.338879 (2008).
Rashidi, S. & Moghaddam, A. H. Investigation and optimization of anaerobic system for treatment of seafood processing wastewater. Chem. Pap. 75, 4649–4660. https://doi.org/10.1007/s11696-021-01675-y (2021).
Moghaddam, A. H., Shayegan, J. & Sargolzaei, J. Investigating and modeling the cleaning-in-place process for retrieving the membrane permeate flux: Case study of hydrophilic polyethersulfone (PES). J. Taiwan Ins. Chem. Eng. 62, 150–157. https://doi.org/10.1016/j.jtice.2016.01.024 (2016).
Environmental Engineering Department, Aksaray University, Aksaray, Turkey
Özgül Çimen Mesutoğlu & Oğuzhan Gök
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
Ö.Ç.M.: software, conceptualization, methodology, investigation, resources, writing—original draft preparation. O. G.: supervision, resources, investigation, visualization.
Correspondence to Özgül Çimen Mesutoğlu.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Çimen Mesutoğlu, Ö., Gök, O. Prediction of COD in industrial wastewater treatment plant using an artificial neural network. Sci Rep 14, 13750 (2024). https://doi.org/10.1038/s41598-024-64634-z
DOI: https://doi.org/10.1038/s41598-024-64634-z
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
Scientific Reports (Sci Rep) ISSN 2045-2322 (online)
Mbbr Sewage Treatment Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.