Feedforward neural network model estimating pollutant removal process within mesophilic upflow anaerobic sludge blanket bioreactor treating industrial starch processing wastewater

An integrated review on the role of different biocatalysts, process parameters, bioreactor technologies and data-driven predictive models for upgrading biogas

As energy consumption and waste generation from human activities continue to rise, the technology of anaerobic digestion (AD), which converts waste into bioenergy, has gained popularity. Biogas produced from AD commonly contains 60 % CH4, 40 % CO2 and a minor fraction of impurities. Currently, several anaerobic reactors have been designed to upgrade the biogas with biomethane content above 90 %. This review summarizes the current trends in the biological upgradation of biogas from a bio-circular economy perspective to achieve sustainable energy goals. Examples of applications reporting the latest advancements in treating industrial effluents using high-rate anaerobic reactors have been mentioned. The integrated anaerobic-aerobic hybrid reactor offers a solution to the limitations of traditional methods in treating diverse effluents. A special focus on biological upgradation techniques such as in-situ, ex-situ, and hybrid mechanisms have been briefed. The key advantage of hybrid upgradation is its ability to address the pH rise during in-situ process. Additionally, the applications of artificial neural networks and optimization to upgrade biogas production have been discussed. The review concludes with future research directives with emphasis on the economic viability of the approaches.

Intelligent FA/FNA alternating strategy for nitrite-oxidizing bacteria inhibition: Data-driven prediction and process control

Alternating treatment with free ammonia (FA) and free nitrous acid (FNA) is an effective strategy to inhibit nitrite-oxidizing bacteria (NOB) in partial nitrification (PN) process. However, the current alternating treatment relies on manual assessment of nitrite accumulation rate (NAR), which poses challenges for automation. This study proposed an intelligent control system for automatic real-time switching between FA/FNA inhibition strategy and real-time control of inhibition concentration to realize stable NOB inhibition. By comparing the prediction performance of three models with different complexities for both classification and regression methods, support vector machine (SVM) was used to determine whether to alternate the strategies based on whether NAR was below 96 % and multilayer perceptron (MLP) was used for real-time control of FNA concentration by predicting the nitrite concentration. The high prediction accuracy of these two sub-models provides a solid foundation for the automatic control model of FA/FNA. Both models were trained and tested using an experimental dataset with manual alternating FA/FNA strategy over 180 days of a continuous flow PN reactor. After optimizing algorithms, the SVM had a 91.67 % classification accuracy, while the MLP showed an R2 of 0.96 and an RMSE of 53.16. During the real-time control of the intelligent control system, the SVM showed a classification accuracy of 97.5 % compared to actual measurements, and the R2 between the controlled FNA and the actual FNA is 0.83, with an RMSE of 0.04. The real-time operation demonstrates that the intelligent control system can promptly realize FA/FNA alternating and accurately control FNA concentration, maintaining NAR above 95.55 % while ammonium removal efficiency above 52.99 %. Compared to the current alternating treatment, the intelligent control strategy simplifies the manual operations and enables the automation of the FA/FNA alternating inhibition strategy, contributing to the stable and efficient operation of the PN process and promoting the application of PN/A process.

Carbon source dosage intelligent determination using a multi-feature sensitive back propagation neural network model

The carbon reduction concept drives the development of low-carbon and sustainable wastewater treatment plant (WWTP) operation technologies. In the denitrification stage of WWTPs in China, there are widespread problems of uneconomical dosage consumption and unstable total nitrogen (TN) concentration in effluent through manual experience to add external carbon sources. Deep learning methods can deal with these problems. However, the methods often require a large amount of data. This paper establishes a multi-feature sensitive back propagation neural network (BPNN) based on Shapley additive explanations (SHAP) and sensitivity analysis (MFS-BPNN-SSA) model to predict carbon source dosage in WWTPs and address short-term and limited data. The model also incorporates theoretical formulas to enhance prediction accuracy and feedback regulation to handle anomalous data. The prediction performance of the MFS-BPNN-SSA model surpasses traditional machine learning and deep learning models. R and R2 reach 0.9999, 1.75% and 3.48% higher, respectively, compared to the best-performing traditional model. The model has been operating safely in the WWTP for over two years, achieving a 9% improvement in effluent TN concentration and a 14% reduction in carbon source dosage. This study provides a novel strategy for pollution reduction and carbon mitigation in WWTPs.

Application of artificial intelligence tools in wastewater and waste gas treatment systems: Recent advances and prospects

The non-linear complex relationships among the process variables in wastewater and waste gas treatment systems possess a significant challenge for real-time systems modelling. Data driven artificial intelligence (AI) tools are increasingly being adopted to predict the process performance, cost-effective process monitoring, and the control of different waste treatment systems, including those involving resource recovery. This review presents an in-depth analysis of the applications of emerging AI tools in physico-chemical and biological processes for the treatment of air pollutants, water and wastewater, and resource recovery processes. Additionally, the successful implementation of AI-controlled wastewater and waste gas treatment systems, along with real-time monitoring at the industrial scale are discussed.

Environmental resilience through artificial intelligence: innovations in monitoring and management

The rapid rise of artificial intelligence (AI) technology has revolutionized numerous fields, with its applications spanning finance, engineering, healthcare, and more. In recent years, AI's potential in addressing environmental concerns has garnered significant attention. This review paper provides a comprehensive exploration of the impact that AI has on addressing and mitigating critical environmental concerns. In the backdrop of AI's remarkable advancement across diverse disciplines, this study is dedicated to uncovering its transformative potential in the realm of environmental monitoring. The paper initiates by tracing the evolutionary trajectory of AI technologies and delving into the underlying design principles that have catalysed its rapid progression. Subsequently, it delves deeply into the nuanced realm of AI applications in the analysis of remote sensing imagery. This includes an intricate breakdown of challenges and solutions in per-pixel analysis, object detection, shape interpretation, texture evaluation, and semantic understanding. The crux of the review revolves around AI's pivotal role in environmental control, examining its specific implementations in wastewater treatment and solid waste management. Moreover, the study accentuates the significance of AI-driven early-warning systems, empowering proactive responses to environmental threats. Through a meticulous analysis, the review underscores AI's unparalleled capacity to enhance accuracy, adaptability, and real-time decision-making, effectively positioning it as a cornerstone in shaping a sustainable and resilient future for environmental monitoring and preservation.

Artificial intelligence-enabled optimization of Fe/Zn@biochar photocatalyst for 2,6-dichlorophenol removal from petrochemical wastewater: A techno-economic perspective

While numerous studies have addressed the photocatalytic degradation of 2,6-dichlorophenol (2,6-DCP) in wastewater, an existing research gap pertains to operational factors' optimization by non-linear prediction models to ensure a cost-effective and sustainable process. Herein, we focus on optimizing the photocatalytic degradation of 2,6-DCP using artificial intelligence modeling, aiming at minimizing initial capital outlay and ongoing operational expenses. Hence, Fe/Zn@biochar, a novel material, was synthesized, characterized, and applied to harness the dual capabilities of 2,6-DCP adsorption and degradation. Fe/Zn@biochar exhibited an adsorption energy of -21.858 kJ/mol, effectively capturing the 2,6-DCP molecules. This catalyst accumulated photo-excited electrons, which, upon interaction with adsorbed oxygen and/or dissolved oxygen generated •O2-. The •OH radicals could also be produced from h+ in the Fe/Zn@biochar valence band, cleaving the C-Cl bonds to Cl- ions, dechlorinated byproducts, and phenols. An artificial neural network (ANN) model, with a 4-10-1 topology, "trainlm" training function, and feed-forward back-propagation algorithm, was developed to predict the 2,6-DCP removal efficiency. The ANN prediction accuracy was expressed as R2 = 0.967 and mean squared error = 5.56e-22. The ANN-based optimized condition depicted that over 90% of 2,6-DCP could be eliminated under C0 = 130 mg/L, pH = 2.74, and catalyst dosage = 168 mg/L within ∼4 h. This optimum condition corresponded to a total cost of $7.70/m3, which was cheaper than the price estimated from the unoptimized photocatalytic system by 16%. Hence, the proposed ANN could be employed to enhance the 2,6-DCP photocatalytic degradation process with reduced operational expenses, providing practical and cost-effective solutions for petrochemical wastewater treatment.

Prediction of biological nutrients removal in full-scale wastewater treatment plants using H2O automated machine learning and back propagation artificial neural network model: Optimization and comparison

The effective control of total nitrogen (ETN) and total phosphorus (ETP) in effluent is challenging for wastewater treatment plants (WWTPs). In this work, automated machine learning (AutoML) (mean square error = 0.4200 ∼ 3.8245, R2 = 0.5699 ∼ 0.6219) and back propagation artificial neural network (BPANN) model (mean square error = 0.0012 ∼ 6.9067, R2 = 0.4326 ∼ 0.8908) were used to predict and analyze biological nutrients removal in full-scale WWTPs. Interestingly, BPANN model presented high prediction performance and general applicability for WWTPs with different biological treatment units. However, the AutoML candidate models were more interpretable, and the results showed that electricity carbon emission dominated the prediction. Meanwhile, increasing data volume and types of WWTP hardly affected the interpretable results, demonstrating its wide applicability. This study demonstrated the validity and the specific advantages of predicting ETN and ETP using H2O AutoML and BPANN model, which provided guidance on the prediction and improvement of biological nutrients removal in WWTPs.

Machine learning model optimization for removal of steroid hormones from wastewater

In the past few decades, there has been a significant focus on detecting steroid hormones in aquatic environments due to their influence on the endocrine system. Most compounds of these pollutants are the natural steroidal estrogens, i.e., estrone (E1), 17β-Estradiol (E2), and the synthetic estrogen 17α-Ethinylestradiol (EE2). The Moving-Bed Biofilm Reactor (MBBR) technique is appropriate for eliminating steroid hormones. This study centers on creating a model to estimate the effectiveness of the MBBR system regarding its ability to eliminate E1, E2, and EE2. The results were modeled with artificial neural networks (ANNs). The Particle Warm Optimization (PSO) and Levenberg Marquardt (LM) algorithms were selected for network training. The models incorporated five input parameters, encompassing the COD loading rate, initial levels of E1, E2, and EE2 steroid hormones, and Hydraulic Retention Time (HRT). The optimum removal conditions (three steroid hormones and COD) were determined using the optimized ANN based on both PSO and LM algorithms. The optimal transfer functions for the hidden and output layers were identified as tan-sigmoid and linear, respectively. The best ANN structures (Neurons in input, hidden, and output layers) and correlation coefficients (R) were 5:9:4, with R = 0.9978, and 5:10:4, with R = 0.9982 for the trained networks with LM and PSO algorithms, respectively. Eventually, the input parameters' importance was ranked using sensitivity analysis (SA) through Pearson correlation and developed ANNs.

Alternative states in microbial communities during artificial aeration: Proof of incubation experiment and development of recurrent neural network models

Artificial aeration, a widely used method of restoring the aquatic ecological environment by enhancing the re-oxygenation capacity, typically relies upon empirical models to predict ecological dynamics and determine the operating scheme of the aeration equipment. Restoration through artificial aeration is involved in oxic-anoxic transitions, whether these transitions occurred in the form of a regime shift, making the development of predictive models challenging. Here, we confirmed the existence of alternative states in microbial communities during artificial aeration through aeration incubation experiment for the first time and considered its existence in neural network modeling in order to improve model performance. By aeration incubation experiment, it was confirmed that the alternative states exist in microbial communities during artificial aeration by two independent approaches, potential analysis and "enterotyping" approach. Comparing neural network models with and without considering the existence of alternative states, it was found that considering the existence of alternative states in modeling could improve the performance of neural network model. Our study provides a reference for the prediction of systems containing time series data where the current state will have an impact on later states. The developed model could be used for optimizing the operating scheme of the artificial aeration.

Machine learning-based model construction and identification of dominant factor for simultaneous sulfide and nitrate removal process

Accurate water quality prediction models are essential for the successful implementation of the simultaneous sulfide and nitrate removal process (SSNR). Traditional models, such as regression and analysis of variance, do not provide accurate predictions due to the complexity of microbial metabolic pathways. In contrast, Back Propagation Neural Networks (BPNN) has emerged as superior tool for simulating wastewater treatment processes. In this study, a generalized BPNN model was developed to simulate and predict sulfide removal, nitrate removal, element sulfur production, and nitrogen gas production in SSNR. Remarkable results were obtained, indicating the strong predictive performance of the model and its superiority over traditional mathematical models for accurately predicting the effluent quality. Furthermore, this study also identified the crucial influencing factors for the process optimization and control. By incorporating artificial intelligence into wastewater treatment modeling, the study highlights the potential to significantly enhance the efficiency and effectiveness of meeting water quality standards.

Predictive modelling and optimization of an airlift bioreactor for selenite removal from wastewater using artificial neural networks and particle swarm optimization

Selenite (Se⁴⁺) is the most toxic of all the oxyanion forms of selenium. In this study, a feed forward back propagation (BP) based artificial neural network (ANN) model was developed for a fungal pelleted airlift bioreactor (ALR) system treating selenite-laden wastewater. The performance of the bioreactor, i.e., selenite removal efficiency (RE_selenite) (%) was predicted through two input parameters, namely, the influent selenite concentration (IC_selenite) (10 mg/L - 60 mg/L) and hydraulic retention time (HRT) (24 h - 72 h). After training and testing with 96 sets of data points using the Levenberg-Marquardt algorithm, a multi-layer perceptron model (2-10-1) was established. High values of the correlation coefficient (0.96 ≤ R ≤ 0.98), along with low root mean square error (1.72 ≤ RMSE ≤ 2.81) and mean absolute percentage error (1.67 ≤ MAPE ≤ 2.67), clearly demonstrate the accuracy of the ANN model (> 96%) when compared to the experimental data. To ensure an efficient and economically feasible operation of the ALR, the process parameters were optimized using the particle swarm optimization (PSO) algorithm coupled with the neural model. The RE_selenite was maximized while minimizing the HRT for a preferably higher range of IC_selenite. Thus, the most favourable optimum conditions were suggested as: IC_selenite - 50.45 mg/L and HRT - 24 h, resulting in RE_selenite of 69.4%. Overall, it can be inferred that ANN models can successfully substitute knowledge-based models to predict the RE_selenite in an ALR, and the process parameters can be effectively optimized using PSO.

Machine learning-assisted exploration for carbon neutrality potential of municipal sludge recycling via hydrothermal carbonization

In the context of advocating carbon neutrality, there are new requirements for sustainable management of municipal sludge (MS). Hydrothermal carbonization (HTC) is a promising technology to deal with high-moisture MS considering its low energy consumption (without drying pretreatment) and value-added products (i.e., hydrochar). This study applied machine learning (ML) methods to conduct a holistic assessment with higher heating value (HHV) of hydrochar, carbon recovery (CR), and energy recovery (ER) as model targets, yielding accurate prediction models with R² of 0.983, 0.844 and 0.858, respectively. Furthermore, MS properties showed positive (e.g., carbon content, HHV) and negative (e.g., ash content, O/C, and N/C) influences on the hydrochar HHV. By comparison, HTC parameters play a critical role for CR (51.7%) and ER (52.5%) prediction. The primary sludge was an optimal HTC feedstock while anaerobic digestion sludge had the lowest potential. This study provided a comprehensive reference for sustainable MS treatment and industrial application.

Deep learning-based prediction of effluent quality of a constructed wetland

Data-driven approaches that make timely predictions about pollutant concentrations in the effluent of constructed wetlands are essential for improving the treatment performance of constructed wetlands. However, the effect of the meteorological condition and flow changes in a real scenario are generally neglected in water quality prediction. To address this problem, in this study, we propose an approach based on multi-source data fusion that considers the following indicators: water quality indicators, water quantity indicators, and meteorological indicators. In this study, we establish four representative methods to simultaneously predict the concentrations of three representative pollutants in the effluent of a practical large-scale constructed wetland: (1) multiple linear regression; (2) backpropagation neural network (BPNN); (3) genetic algorithm combined with the BPNN to solve the local minima problem; and (4) long short-term memory (LSTM) neural network to consider the influence of past results on the present. The results suggest that the LSTM-predicting model performed considerably better than the other deep neural network-based model or linear method, with a satisfactory R². Additionally, given the huge fluctuation of different pollutant concentrations in the effluent, we used a moving average method to smooth the original data, which successfully improved the accuracy of traditional neural networks and hybrid neural networks. The results of this study indicate that the hybrid modeling concept that combines intelligent and scientific data preprocessing methods with deep learning algorithms is a feasible approach for forecasting water quality in the effluent of actual engineering.

A back propagation neural network model for accurately predicting the removal efficiency of ammonia nitrogen in wastewater treatment plants using different biological processes

Accurately predicting the water quality of treated water from a water treatment plant (WWTP) based on the obtained operating database is of great significance. However, it is difficult for common mechanistic models to work well. In this study, a back propagation artificial neural network (BPANN) model with high accuracy was developed to predict the denitrification efficiency based on a 1-year operating database. Standardized principal component analysis (PCA) methods were used to address the data, and the PCA processed data exhibited the best accuracy. In three WWTPs adopting the anaerobic/anoxic/oxic (A²O) process, the ammonia nitrogen removal efficiency of WWTPs was successfully predicted by using five variables: inlet flow rate, pH value, original ammonia nitrogen concentration, Chemical oxygen demand (COD) concentration, and total phosphorus concentration. Importantly, the obtained BPANN model can be effectively used for other widely used treatment processes, such as oxidation ditch (OD), sequencing batch reactor activated sludge process (SBR), membrane bioreactor (MBR), and cyclic activated sludge technology (CAST), by simply optimizing the training data ratios between 50/50 and 90/10. This is the first trial to set up a universal model for predicting the denitrification efficiency of WWTPs adopting common biological processes. The model could be used to choose the optimum treatment process in the new WWTP design or take action in advance to avoid the risk of excessive emissions when the already built WWTPs are subjected to sudden shocks.

An Artificial Neural Network for Simulation of an Upflow Anaerobic Filter Wastewater Treatment Process

The purpose of this work was to develop a problem-solving approach and a simulation tool that is useful for the specification of wastewater treatment process equipment design parameters. The proposition of using an artificial neural network (ANN) numerical model for supervised learning of a dataset and then for process simulation on a new dataset was investigated. The effectiveness of the approach was assessed by evaluating the capacity of the model to distinguish differences in the equipment design parameters. To demonstrate the approach, a mock dataset was derived from experimentally acquired data and physical effects reported in the literature. The mock dataset comprised the influent flow rate, the bed packing material dimension, the type of packing material and the packed bed height-to-diameter ratio as predictors of the calorific value reduction. The multilayer perceptron (MLP) ANN was compared to a polynomial model. The validation test results show that the MLP model has four hidden layers, each having 256 units (nodes), accurately predicts calorific value reduction. When the model was fed previously unseen test data, the root-mean-square error (RMSE) of the predicted responses was 0.101 and the coefficient of determination (R2) was 0.66. The results of simulation of all 125 possible combinations of the 3 mechanical parameters and identical influent wastewater flow profiles were ranked according to total calorific value reduction. A t-test of the difference between the mean calorific value reduction of the two highest ranked experiments showed that the means are significantly different (p-value = 0.011). Thus, the model has the capacity to distinguish differences in the equipment design parameters. Consequently, the values of the three mechanical feature parameters from the highest ranked simulated experiment are recommended for use in the design of the industrial scale upflow anaerobic filter (UAF) for wastewater treatment.

Fundamentals and Applications of Artificial Neural Network Modelling of Continuous Bifidobacteria Monoculture at a Low Flow Rate

The application of artificial neural networks (ANNs) to mathematical modelling in microbiology and biotechnology has been a promising and convenient tool for over 30 years because ANNs make it possible to predict complex multiparametric dependencies. This article is devoted to the investigation and justification of ANN choice for modelling the growth of a probiotic strain of Bifidobacterium adolescentis in a continuous monoculture, at low flow rates, under different oligofructose (OF) concentrations, as a preliminary study for a predictive model of the behaviour of intestinal microbiota. We considered the possibility and effectiveness of various classes of ANN. Taking into account the specifics of the experimental data, we proposed two-layer perceptrons as a mathematical modelling tool trained on the basis of the error backpropagation algorithm. We proposed and tested the mechanisms for training, testing and tuning the perceptron on the basis of both the standard ratio between the training and test sample volumes and under the condition of limited training data, due to the high cost, duration and the complexity of the experiments. We developed and tested the specific ANN models (class, structure, training settings, weight coefficients) with new data. The validity of the model was confirmed using RMSE, which was from 4.24 to 980% for different concentrations. The results showed the high efficiency of ANNs in general and bilayer perceptrons in particular in solving modelling tasks in microbiology and biotechnology, making it possible to recommend this tool for further wider applications.

Machine learning exhibited excellent advantages in the performance simulation and prediction of free water surface constructed wetlands

Optimizing the design and operation parameters of free water surface constructed wetlands (FWS CWs) in runoff regulation and wastewater treatment is necessary to improve the comprehensive performance. In this study, nine machine learning (ML) algorithms were successfully developed to optimize the parameter combinations for FWS CWs. The scale effect of surface area on wetland performance was determined based on consistently smaller predictions (-6.2% to -28.9%) of the nine well-established ML algorithms. The models most suitable for FWS CW performance simulation and prediction were random forest and extra trees algorithms because of their high R² values (0.818 in both) with the training set and low mean absolute relative errors (4.7% and 3.8%, respectively) with the test set. Results from feature analysis of the six tree-based algorithms emphasized the importance of water depth and layout of inlet and outlet, and revealed the negligible effect of the aspect ratio. Feature importance and partial dependence analysis enhanced the interpretability of the tree-based algorithms. The proposed ML algorithms enabled the implementation of an extended scenario at a low cost in real time. Therefore, ML algorithms are suitable for expressing the complex and uncertain effects of the design and operation parameters on the performance of FWS CWs. Acquiring datasets consisting of more extensive, uniform, and unbiased parameter combinations is crucial for developing more robust and practical ML algorithms for the optimal design of FWS CWs.

Artificial Neural Network (ANN) Modelling for Biogas Production in Pre-Commercialized Integrated Anaerobic-Aerobic Bioreactors (IAAB)

The use of integrated anaerobic-aerobic bioreactor (IAAB) to treat the Palm Oil Mill Effluent (POME) showed promising results, which successfully overcome the limitation of a large space that is needed in the conventional method. The understanding of synergism between anaerobic digestion and aerobic process is required to achieve maximum biogas production and COD removal. Hence, this work presents the use of artificial neural network (ANN) to predict the COD removal (%), purity of methane (%), and methane yield (LCH4/gCODremoved) of anaerobic digestion and COD removal (%), biochemical oxygen demand (BOD) removal (%), and total suspended solid (TSS) removal (%) of aerobic process in a pre-commercialized IAAB located at Negeri Sembilan, Malaysia. MATLAB R2019b was used to develop the two ANN models. Bayesian regularization backpropagation (BR) showed the best performance among the 12 training algorithms. The trained ANN models showed high accuracy (R2 > 0.997) and demonstrated good alignment with the industrial data obtained from the pre-commercialized IAAB over a 6-month period. The developed ANN model is subsequently used to create the optimal operating conditions which maximize the output parameters. The COD removal (%) was improved by 33.9% (from 68.7% to 92%), while the methane yield was improved by 13.4% (from 0.23 LCH4/gCODremoved to 0.26 LCH4/gCODremoved). Sensitivity analysis shows that COD inlet is the most influential input parameters that affect the methane yield, anaerobic COD, BOD and TSS removals, while for aerobic process, COD removal is most affected by mixed liquor suspended solids (MLSS). The trained ANN model can be utilized as a decision support system (DSS) for operators to predict the behavior of the IAAB system and solve the problems of instability and inconsistent biogas production in the anaerobic digestion process. This is of utmost importance for the successful commercialization of this IAAB technology. Additional input parameters such as the mixing time, reaction time, nutrients (ammonium nitrogen and total phosphorus) and concentration of microorganisms could be considered for the improvement of the ANN model.

Predicting membrane fouling in a high solid AnMBR treating OFMSW leachate through a genetic algorithm and the optimization of a BP neural network model

Anaerobic membrane bioreactors are a promising technology in the treatment of high-strength wastewater; however, unpredictable membrane fouling largely limits their scale-up application. This study, therefore, adopted a backpropagation neural network model to predict the membrane filtration performance in a submerged system, which treats leachate from the organic fraction of municipal solid waste. Duration time, water yield flow, influent COD, pH, bulk sludge concentration, and the ratio of ΔTMP to filtration time were selected as input variables to simulate membrane permeability. The membrane pressure slightly increased by 1.1 kPa within 62 days of operation. The results showed that the AnMBR membrane filtration performance was acceptable when treating OFMSW leachate under a flux of 6 L/(m²·h). The model results indicated that the sludge concentration largely determined the membrane fouling with a contribution of 33.8%. Given the local minimization problem in the BP neural network process, a genetic algorithm was introduced to optimize the simulation process, and the relative error of the results was reduced from 5.57% to 3.57%. Conclusively, the artificial neural network could be a useful tool for the prediction of an AnMBR that is so far under development.

A comprehensive review on application of bioreactor for industrial wastewater treatment

In the recent past, wastewater treatment processes performed a pivotal role in accordance with maintaining the sustainable environment and health of mankind at a proper hygiene level. It has been proved indispensable by government regulations throughout the world on account of the importance of preserving freshwater bodies. Human activities, predominantly from industrial sectors, generate an immeasurable amount of industrial wastewater loaded with toxic chemicals, which not only cause dreadful environmental problems, but also leave harmful impacts on public health. Hence, industrial wastewater effluent must be treated before being released into the environment to restrain the problems related to industrial wastewater discharged to the environment. Nowadays, biological wastewater treatment methods have been considered an excellent approach for industrial wastewater treatment process because of their cost-effectiveness in the treatment, high efficiency and their potential to counteract the drawbacks of conventional wastewater treatment methods. Recently, the treatment of industrial effluent through bioreactor has been proved as one of the best methods from the presently available methods. Reactors are the principal part of any biotechnology-based method for microbial or enzymatic biodegradation, biotransformation and bioremediation. This review aims to explore and compile the assessment of the most appropriate reactors such as packed bed reactor, membrane bioreactor, rotating biological contactor, up-flow anaerobic sludge blanket reactor, photobioreactor, biological fluidized bed reactor and continuous stirred tank bioreactor that are extensively used for distinct industrial wastewater treatment.

Application of Various Machine Learning Models for Process Stability of Bio-Electrochemical Anaerobic Digestion

The application of a machine learning (ML) model to bio-electrochemical anaerobic digestion (BEAD) is a future-oriented approach for improving process stability by predicting performances that have nonlinear relationships with various operational parameters. Five ML models, which included tree-, regression-, and neural network-based algorithms, were applied to predict the methane yield in BEAD reactor. The results showed that various 1-step ahead ML models, which utilized prior data of BEAD performances, could enhance prediction accuracy. In addition, 1-step ahead with retraining algorithm could improve prediction accuracy by 37.3% compared with the conventional multi-step ahead algorithm. The improvement was particularly noteworthy in tree- and regression-based ML models. Moreover, 1-step ahead with retraining algorithm showed high potential of achieving efficient prediction using pH as a single input data, which is plausibly an easier monitoring parameter compared with the other parameters required in bioprocess models.

Valorisation of Starch Wastewater by Anaerobic Fermentation

Starch production is mainly focused on feedstocks such as corn, wheat and potato in the EU, whereas cassava, rice, and other feedstocks are utilised worldwide. In starch production, a high amount of wastewater is generated, which accumulates from different process steps such as washing, steeping, starch refining, saccharification and derivatisation. Valorisation of these wastewaters can help to improve the environmental impact as well as the economics of starch production. Anaerobic fermentation is a promising approach, and this review gives an overview of the different utilisation concepts outlined in the literature and the state of the technology. Among bioenergy recovery processes, biogas technology is widely applied at the industrial scale, whereas biohydrogen production is used at the research stage. Starch wastewater can also be used for the production of bulk chemicals such as acetone, ethanol, butanol or lactic acids by anaerobic microbes.

Estimation of in-situ biogas upgrading in microbial electrolysis cells via direct electron transfer: Two-stage machine learning modeling based on a NARX-BP hybrid neural network

With the increasing of data in wastewater treatment, data-driven machine learning models are useful for modeling biological processes and complex reactions. However, few data-driven models have been developed for simulating the microbial electrolysis cells (MECs) and traditional models are too ambiguous to comprehend the mechanisms. In this study, a new general data-driven two-stage model was firstly developed to predict CH₄ production from in-situ biogas upgrading in the biocathode MECs via direct electron transfer (DET), named NARX-BP hybrid neural networks. Compared with traditional one-stage model, the model could well predict methane production via DET with excellent performance (all R² and MES of 0.918 and 6.52 × 10^-2, respectively) and reveal the mechanisms of biogas upgrading, for the new systematical modeling approach could improve the versatility and applicability by inputting significant intermediate variables. In addition, the model is generally available to support long-term prediction and optimal operation for anaerobic digestion or complex MEC systems.

High efficiency in-situ biogas upgrading in a bioelectrochemical system with low energy input

Biogas produced from anaerobic digestion usually contains 30%-50% CO_2, much of which must be removed, before utilization. Bioelectrochemical biogas upgrading approaches show promise, however, they have not yet been optimized for practical applications. In this study, a bioelectrochemical system with low energy input (applied cathode potential of -0.5 V vs. standard hydrogen electrode, SHE) was used for in-situ biogas upgrading. High efficiency CO₂ conversion (318.5 mol/d/m²) was achieved when the system was operated with an organic load of 1.7 kgCOD/(m³ d). Methane content in the upgraded biogas was 97.0% and CO₂ concentrations stayed below 3%, which is comparable to biogas upgraded with more expensive and less sustainable physiochemical approaches. The high efficiency of this approach could likely be attributed to a significant enrichment of Methanothrix (92.7%) species on the cathode surface that were expressing genes involved in both acetogenic methanogenesis and direct electron transfer (DET). Electromethanogenesis by these organisms also increased proton consumption and created a higher pH that increased the solubility of CO₂ in the bioreactor. In addition, CO₂ removal from the biogas was likely further enhanced by an enrichment of Actinobacillus species known to be capable of CO₂ fixation. Artificial neural network (ANN) models were also used to estimate CH₄ production under different loading conditions. The ANN architecture with 10 neurons at hidden layers fit best with a mean square error of 6.06 × 10^-3 and R² of 0.99.

Estimating the chemical oxygen demand of petrochemical wastewater treatment plants using linear and nonlinear statistical models - A case study

In this research, twelve linear and nonlinear regression models were performed and evaluated to formulate the best one for the estimation of chemical oxygen demand level in the effluent of the clarifier unit of a petrochemical wastewater treatment plant. The input variables measured twice a day in the influent of the biological unit over a period of 13 months using standard methods. The piece-wise linear regression with breakpoint method, with a mean squared error value equal to 0.041, mean absolute error of 0.144, and correlation coefficient equal to 0.835 was found to estimate the output chemical oxygen demand parameter more sustainable rather than other linear and nonlinear methods. However, some of the other applied models such as radial basis function neural network and gene expressing programming models achieved good performance considering their correlation coefficient, robustness in presence of outliers, mean squared error and mean absolute error test. Mathematical and intelligent modeling proved useful as an accurate alternative to estimate the amount of chemical oxygen demand rather than spending time and cost for its laboratory tests.

Moving average convergence and divergence indexes based online intelligent expert diagnosis system for anaerobic wastewater treatment process

Anaerobic wastewater treatment process is efficient but unstable due to various disturbances, such as refractory organics and influent organic overloading. Therefore, sensitive and accurate status diagnosis is important for reasonable control to improve the stability of anaerobic process. In this study, an online intelligent expert diagnosis system for anaerobic process was established based on moving average convergence and divergence (MACD) indexes of gas- and liquid-phase parameters, combined with online monitoring system and expert diagnosis database. The effect of this diagnosis system was verified through refractory organics and organic overloading shock experiments. Results showed that this diagnosis system could make rapid, accurate and comprehensive diagnosis, predictions and early-warning. MACD algorithm could enhance pattern recognition capacity of status parameters, overcome the lagging of anaerobic process and filter irregular noisy fluctuations of status parameters. MACD index of H₂ partial pressure is suitable as sensitive early-warning indicator in the initial shock stage.

Simulating a combined lysis-cryptic and biological nitrogen removal system treating domestic wastewater at low C/N ratios using artificial neural network

In this study, a combined alkaline (ALK) and ultrasonication (ULS) sludge lysis-cryptic pretreatment and anoxic/oxic (AO) system (AO + ALK/ULS) was developed to enhance biological nitrogen removal (BNR) in domestic wastewater with a low carbon/nitrogen (C/N) ratio. A real-time control strategy for the AO + ALK/ULS system was designed to optimize the sludge lysate return ratio (R_SLR) under variable sludge concentrations and variations in the influent C/N (⩽ 5). A multi-layered backpropagation artificial neural network (BPANN) model with network topology of 1 input layer, 3 hidden layers, and 1 output layer, using the Levenberg-Marquardt algorithm, was developed and validated. Experimental and predicted data showed significant concurrence, verified with a high regression coefficient (R² = 0.9513) and accuracy of the BPANN. The BPANN model effectively captured the complex nonlinear relationships between the related input variables and effluent output in the combined lysis-cryptic + BNR system. The model could be used to support the real-time dynamic response and process optimization control to treat low C/N domestic wastewater.

Estimate of Head Posture Based on Coordinate Transformation with MP-MTM-LSTM Network

Fatigue monitoring can effectively reduce or even avoid traffic accidents. Head posture estimation is one of the focuses in the field of fatigue monitoring. In this paper, according to the coordinate rotation transformation and neural network theory, a method for predicting the change of head posture with sight-line coordinates is proposed. First, the coordinate rotation transformation theory is used to replace the head posture change amount with the coordinate change amount, and the first-order difference value of the sight-line point coordinate is obtained by the difference method. Then, under the unified Cartesian coordinate system, the MP-MTM-LSTM neural network is established with the input information of first-order difference value and the output information of coordinate change amount. The innovation of this method is that the Cartesian coordinate change is employed instead of the Euler angle transformation. In the model verification phase, the true value of the head pose is collected by the posture meter. The experimental results show that the absolute error between the predicted value and the true value estimated by the new method is less than 15%. In the field of fatigue monitoring, the proposed method can estimate the amount of head posture change effectively, which is suitable for the case where the head center point is not fixed.

Tackling environmental challenges in pollution controls using artificial intelligence: A review

This review presents the developments in artificial intelligence technologies for environmental pollution controls. A number of AI approaches, which start with the reliable mapping of nonlinear behavior between inputs and outputs in chemical and biological processes in terms of prediction models to the emerging optimization and control algorithms that study the pollutants removal processes and intelligent control systems, have been developed for environmental clean-ups. The characteristics, advantages and limitations of AI methods, including single and hybrid AI methods, were overviewed. Hybrid AI methods exhibited synergistic effects, but with computational heaviness. The up-to-date review summarizes i) Various artificial neural networks employed in wastewater degradation process for the prediction of removal efficiency of pollutants and the search of optimizing experimental conditions; ii) Evaluation of fuzzy logic used for intelligent control of aerobic stage of wastewater treatment process; iii) AI-aided soft-sensors for precisely on-line/off-line estimation of hard-to-measure parameters in wastewater treatment plants; iv) Single and hybrid AI methods applied to estimate pollutants concentrations and design monitoring and early-warning systems for both aquatic and atmospheric environments; v) AI modelings of short-term, mid-term and long-term solid waste generations, and various ANNs for solid waste recycling and reduction. Finally, the future challenges of AI-based models employed in the environmental fields are discussed and proposed.

Modeling the performance of Single-stage Nitrogen removal using Anammox and Partial nitritation (SNAP) process with backpropagation neural network and response surface methodology

Two novel feedforward backpropagation Artificial Neural Networks (ANN)-based-models (8:N_H:1 and 7:N_H:1) combined with Box-Behnken design of experiments methodology was proposed and developed to model NH₄⁺ and Total Nitrogen (TN) removal within an upflow-sludge-bed (USB) reactor treating nitrogen-rich wastewater via Single-stage Nitrogen removal using Anammox and Partial nitritation (SNAP) process. ANN were developed by optimizing network architecture parameters via response surface methodology. Based on the goodness-of-fit standards, the proposed three-layered NH₄⁺ and TN removal ANN-based-models trained with Levenberg-Marquardt-algorithm demonstrated high-performance as computations exhibited smaller deviations-(±2.1%) as well as satisfactory coefficient of determination (R²), fractional variance-(FV), and index of agreement-(IA) ranging 0.989-0.997, 0.003-0.031 and 0.993-0.998, respectively. The computational results affirmed that the ANN architecture which was optimized with response surface methodology enhanced the efficiency of the ANN-based-models. Furthermore, the overall performance of the developed ANN-based models revealed that modeling intricate biological systems (such as SNAP) using ANN-based models with the view to improve removal efficiencies, establish process control strategies and optimize performance is highly feasible. Microbial community analysis conducted with 16S rRNA high-throughput approach revealed that Candidatus Kuenenia was the most pronounced genera which accounted for 13.11% followed by Nitrosomonas-(6.23%) and Proteocatella-(3.1%), an indication that nitrogen removal pathway within the USB was mainly via partial-nitritation/anammox process.

Estimation of soil pH with geochemical indices in forest soils

Soil pH is a critical soil quality index and controls soil microbial activities, soil nutrient availability, and plant roots growth and development. The current study aims to evaluate various pedotransfer functions for predicting soil pH using different geochemical indices (CaO, ratios of Al2O3, Fe2O3, TiO2, SiO2, MgO, and K2O to CaO) in forest soils. Various models including empirical functions (quadratic, cubic, sigmoid, logarithmic) and artificial neural network with these geochemical indices were assessed by independent testing set. Mean bias error (MBE), root mean square error (RMSE), mean absolute percentage error (MAPE), mean absolute error (MAE), coefficient of determination (R2), t-statistics (t-stat), and Akaike's Information Criterion (AIC) were applied to evaluate the model performances. Additionally, a new indicator (global performance indictor, GPI) was originally introduced in this study and was used to rank these models. According to GPI, the sigmoid functions and ANNs performed better than others. On average, they could explain above 70% of the variability in soil pH. Both model structure and dataset shape impact on model performance. The best input was CaO for ANNs, sigmoid, and logarithmic functions. The ratios of K2O to CaO and Al2O3 to CaO were the best inputs for quadratic and cubic equations, respectively.

Recovery of High Value-Added Nutrients from Fruit and Vegetable Industrial Wastewater

The industrial processing water of fruit and vegetables has raised serious environmental concerns due to the presence of many important bioactive compounds being disposed in the wastewater. Bioactive compounds have great potential for the food industry to optimize their process and to recover these compounds in order to develop value-added products and to reduce environmental impacts. However, to achieve this goal, some challenges need to be addressed such as safety assurance, technology request, product regulations, cost effectiveness, and customer factors. Therefore, this review aims to summarize the recent advances of bioactive compounds recovery and the current challenges in wastewater from fruit and vegetable processing industry, including fruit and beverage, soybean by-products, starch and edible oil industry. Moreover, future direction for novel and green technology of bioactive compounds recovery are discussed, and a prospect of bioactive compounds reuse and sustainable development is proposed.

Performance, microbial community evolution and neural network modeling of single-stage nitrogen removal by partial-nitritation/anammox process

Single-stage nitrogen removal by anammox/partial-nitritation (SNAP) process was proposed and explored in a packed-bed-EGSB reactor to treat nitrogen-rich wastewater. With dissolved oxygen (DO) maintained within 0.2-0.5 mg/L, reactor performance and microbial community dynamics were evaluated and reported. To ascertain whether control/prediction of the SNAP process was feasible with mathematical modeling, a novel 3-layered backpropagation-artificial-neural-network-(BANN) was also developed to model nitrogen removal efficiencies. When NLR of 300 gN/m³·d and DO of <0.3 mg/L was employed, the SNAP-process demonstrated autotrophic nitrogen removal pathways with NH₄⁺-N and TN removal of 91.1% and 81.9%, respectively. Microbial community succession revealed by 16S rRNA high-throughput gene-sequencing indicated that Candidatus-Kuenenia-(33.83%), Nitrosomonas-(3.4%) Armatimonadetes_gp5-(1.39%), Ignavibacterium-(1.80%), Thiobacillus-(1.33%), and Nitrospira-(1.17%) were the most pronounced genera at steady-state. The proposed BANN-model demonstrated high-performance as computational results revealed smaller deviations (±3%) and satisfactory coefficient of determination-(R² = 0.989), fractional variance-(FV = 0.0107), and index of agreement-(IA = 0.997). Thus, forecasting the efficiency of a SNAP-process with neural-network modeling was highly feasible.

Evaluation and optimization of anammox baffled reactor (AnBR) by artificial neural network modeling and economic analysis

Anammox baffled reactor (AnBR) had a moderate start-up period of 53 days. Interestingly, tangled relationships between key parameters affecting anammox performance were observed, i.e., polynomial function for nitrogen loading rate (NLR) with extracellular polymeric substances (EPS), linear relationships between EPS with granules diameter, granules diameter with settling velocity, and settling velocity with biomass concentration. The correlation coefficients (R2) were 0.97, 0.84, 0.86, and 0.88, respectively. Furthermore, a multi-layered feed forward artificial neural network (ANN) was utilized for simulating and predicting the performance of AnBR. An ANN structure of two hidden layers with four neurons at 1st layer and eight neurons at 2nd layer achieved the best goodness of fit with the minimum mean squared error (MSE) and maximum R² of 0.002 and 0.99, respectively. Additionally, economic assessment stated that using AnBR at NLR of 4.04 ± 0.10 kg-N/m³/day achieved the maximum net present value of $48100.9.