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

Learning a neural network-based soft sensor with double-errors parallel optimization towards effluent variable prediction in wastewater treatment plants

With the development of machine learning and artificial intelligence (ML/AI) models, data-driven soft sensors, especially the neural network-based, have widespread utilization for the prediction of key water quality indicators in wastewater treatment plants (WWTPs). However, recent research indicates that the prediction performance and computational efficiency are greatly compromised due to the time-varying, nonlinear and high-dimensional nature of the wastewater treatment process. This paper proposes a neural network-based soft sensor with double-errors parallel optimization to achieve more accurate prediction for effluent variables timely. Firstly, relying on the Activity Based Classification (ABC) principle, an ensemble variable selection method that combines Pearson correlation coefficient (PCC) and mutual information (MI) is introduced to select the optimal process variables as auxiliary variables, thereby reducing the data dimensionality and simplifying the model complexity. Subsequently, a double-errors parallel optimization methodology with minimizing both point prediction error and distribution error simultaneously is proposed, aiming to enhancing the training efficiency and the fitting quality of neural networks. Finally, the effectiveness is quantitatively assessed in two datasets collected from the Benchmark Simulation Model no. 1 (BMS1) and an actual oxidation ditch WWTP. The experimental results illustrate that the proposed soft sensor achieves precise effluent variable prediction, with RMSE, MAE and R2 values being 0.0606, 0.0486, 0.99930, and 0.06939, 0.05381, 0.98040, respectively. Consequently, this soft sensor can expedite the convergence speed in the neural network training process and enhance the prediction performance, thereby contributing to the effective optimization management of WWTPs.

Artificial neural network modeling for the prediction, estimation, and treatment of diverse wastewaters: A comprehensive review and future perspective

The application of artificial neural networks (ANNs) in the treatment of wastewater has achieved increasing attention, as it enhances the efficiency and sustainability of wastewater treatment plants (WWTPs). This paper explores the application of ANN-based models in WWTPs, focusing on the latest published research work, by presenting the effectiveness of ANNs in predicting, estimating, and treatment of diverse types of wastewater. Furthermore, this review comprehensively examines the applicability of the ANNs in various processes and methods used for wastewater treatment, including membrane and membrane bioreactors, coagulation/flocculation, UV-disinfection processes, and biological treatment systems. Additionally, it provides a detailed analysis of pollutants viz organic and inorganic substances, nutrients, pharmaceuticals, drugs, pesticides, dyes, etc., from wastewater, utilizing both ANN and ANN-based models. Moreover, it assesses the techno-economic value of ANNs, provides cost estimation and energy analysis, and outlines promising future research directions of ANNs in wastewater treatment. AI-based techniques are used to predict parameters such as chemical oxygen demand (COD) and biological oxygen demand (BOD) in WWTP influent. ANNs have been formed for the estimation of the removal efficiency of pollutants such as total nitrogen (TN), total phosphorus (TP), BOD, and total suspended solids (TSS) in the effluent of WWTPs. The literature also discloses the use of AI techniques in WWT is an economical and energy-effective method. AI enhances the efficiency of the pumping system, leading to energy conservation with an impressive average savings of approximately 10%. The system can achieve a maximum energy savings state of 25%, accompanied by a notable reduction in costs of up to 30%.

Spatio-temporal evolution mechanism and dynamic simulation of nitrogen and phosphorus pollution of the Yangtze River economic Belt in China

Excess nitrogen and phosphorus inputs are the main causes of aquatic environmental deterioration. Accurately quantifying and dynamically assessing the regional nitrogen and phosphorus pollution emission (NPPE) loads and influencing factors is crucial for local authorities to implement and formulate refined pollution reduction management strategies. In this study, we constructed a methodological framework for evaluating the spatio-temporal evolution mechanism and dynamic simulation of NPPE. We investigated the spatio-temporal evolution mechanism and influencing factors of NPPE in the Yangtze River Economic Belt (YREB) of China through the pollution load accounting model, spatial correlation analysis model, geographical detector model, back propagation neural network model, and trend analysis model. The results show that the NPPE inputs in the YREB exhibit a general trend of first rising and then falling, with uneven development among various cities in each province. Nonpoint sources are the largest source of land-based NPPE. Overall, positive spatial clustering of NPPE is observed in the cities of the YREB, and there is a certain enhancement in clustering. The GDP of the primary industry and cultivated area are important human activity factors affecting the spatial distribution of NPPE, with economic factors exerting the greatest influence on the NPPE. In the future, the change in NPPE in the YREB at the provincial level is slight, while the nitrogen pollution emissions at the municipal level will develop towards a polarization trend. Most cities in the middle and lower reaches of the YREB in 2035 will exhibit medium to high emissions. This study provides a scientific basis for the control of regional NPPE, and it is necessary to strengthen cooperation and coordination among cities in the future, jointly improve the nitrogen and phosphorus pollution tracing and control management system, and achieve regional sustainable development.

A multi-subsystem collaborative Bi-LSTM-based adaptive soft sensor for global prediction of ammonia-nitrogen concentration in wastewater treatment processes

Ammonia-nitrogen concentration is a key water quality indicator, which reflects changes in pollutant components during wastewater treatment processes. The timely and accurate detection results contribute to optimizing control and operational management of wastewater treatment plants (WWTPs), but current detection methods only focus on the effluent location. This paper proposes a multi-subsystem collaborative Bi-LSTM-based adaptive soft sensor to achieve the global prediction of ammonia-nitrogen concentration. Firstly, the wastewater treatment process is divided into several independent subsystems depending on the reaction mechanism, and the variable selection is performed using mutual information. Subsequently, the bidirectional long short-term memory network (Bi-LSTM) is employed to construct a model for predicting ammonia-nitrogen concentration within each subsystem, and the outputs between neighboring subsystems are incorporated as a set of new variables added into the training dataset to strengthen their connection. Finally, to address performance degradation caused by environmental factors, a probability density function (PDF)-based dynamic moving window method is proposed to enhance the robustness. The effectiveness and superiority of the proposed soft sensor are validated in the Benchmark Simulation Model no. 1 (BSM1). The experimental results demonstrate that the proposed soft sensor can accurately predict the global ammonia-nitrogen concentration in the face of different weather conditions including sunny, rainy, and stormy days. This study contributes to the stable operation of WWTPs with higher treatment efficiency and lower economic costs.

Direct and indirect monitoring methods for nitrous oxide emissions in full-scale wastewater treatment plants: A critical review

Mitigation of nitrous oxide (N2O) emissions in full-scale wastewater treatment plant (WWTP) has become an irreversible trend to adapt the climate change. Monitoring of N2O emissions plays a fundamental role in understanding and mitigating N2O emissions. This paper provides a comprehensive review of direct and indirect N2O monitoring methods. The techniques, strengths, limitations, and applicable scenarios of various methods are discussed. We conclude that the floating chamber technique is suitable for capturing and interpreting the spatiotemporal variability of real-time N2O emissions, due to its long-term in-situ monitoring capability and high data acquisition frequency. The monitoring duration, location, and frequency should be emphasized to guarantee the accuracy and comparability of acquired data. Calculation by default emission factors (EFs) is efficient when there is a need for ambiguous historical N2O emission accounts of national-scale or regional-scale WWTPs. Using process-specific EFs is beneficial in promoting mitigation pathways that are primarily focused on low-emission process upgrades. Machine learning models exhibit exemplary performance in the prediction of N2O emissions. Integrating mechanistic models with machine learning models can improve their explanatory power and sharpen their predictive precision. The implementation of the synergy of nutrient removal and N2O mitigation strategies necessitates the calibration and validation of multi-path mechanistic models, supported by long-term continuous direct monitoring campaigns.

Estimating the methane flux of the Dajiuhu subalpine peatland using machine learning algorithms and the maximal information coefficient technique

The eddy covariance (EC) technique has emerged as the most widely used method for long-term continuous methane flux (FCH4) observations. However, the completeness of the FCH4 time series is limited by instrumental failures and data quality issues, resulting in missing data gaps ranging from 20 % to 90 %. In this situation, the excellent performance of machine learning (ML) algorithms in filling missing FCH4 data has provided a foundation for developing regional-scale FCH4 models. In this study, we established estimation models for FCH4 utilizing random forest (RF), support vector machine (SVM), back propagation (BP) and nonlinear multiple regression (MLR) algorithms. The maximal information coefficient (MIC) technique was employed to identify and rank the environmental factors that were correlated with FCH4. Our findings revealed that soil temperature (Ts), soil water content (SWC) and air temperature (Ta) were the primary environmental factors influencing FCH4. Among the four algorithms, from perspectives of model accuracy and relatively small number of driving factors, the RF models exhibited the best performance, followed by BP and SVM, whereas MLR demonstrated the lowest performance. Among the 144 RF models established using nine datasets, RF model with 8 driving factors in all-year (RFall-year8) could capture seasonal variations. Ultimately, we recommend (RFall-year8 as the optimal model for estimating FCH4 in the Dajiuhu subalpine peatland.

Modeling nitrogen removal performance based on novel microbial activity indicators in WWTP by machine learning and biological interpretation

Modeling the pollutant removal performance of wastewater treatment plants (WWTPs) plays a crucial role in regulating their operation, mitigating effluent anomalies and reducing operating costs. Pollutants removal in WWTPs is closely related to microbial activity. However, there is extremely limited knowledge on the models accurately characterizing pollutants removal performance by microbial activity indicators. This study proposed a novel specific oxygen uptake rate (SOURATP) with adenosine triphosphate (ATP) as biomass. Firstly, it was found that SOURATP and total nitrogen (TN) removal rate showed similar fluctuated trends, and their correlation was stronger than that of TN removal rate and common SOURMLSS with mixed liquor suspended solids (MLSS) as biomass. Then, support vector regressor (SVR), K-nearest neighbor regressor (KNR), linear regressor (LR), and random forest (RF) models were developed to predict TN removal rate only with microbial activity as features. Models utilizing the novel SOURATP resulted in better performance than those based on SOURMLSS. A model fusion (MF) algorithm based on the above four models was proposed to enhance the accuracy with lower root mean square error (RMSE) of 2.25 mg/L/h and explained 75% of the variation in the test data with SOURATP as features as opposed to other base learners. Furthermore, the interpretation of predictive results was explored through microbial community structure and metabolic pathway. Strong correlations were found between SOURATP and the proportion of nitrifiers in aerobic pool, as well as between heterotrophic bacteria respiratory activity (SOURATP_HB) and the proportion of denitrifies in anoxic pool. SOURATP also displayed consistent positive responses with most key enzymes in Embden-Meyerhof-Parnas pathway (EMP), tricarboxylic acid cycle (TCA) and oxidative phosphorylation cycle. In this study, SOURATP provides a reliable indication of the composition and metabolic activity of nitrogen removal bacteria, revealing the potential reasons underlying the accurate predictive result of nitrogen removal rates based on novel microbial activity indicators. This study offers new insights for the prediction and further optimization operation of WWTPs from the perspective of microbial activity regulation.

Prediction of effluent total nitrogen and energy consumption in wastewater treatment plants: Bayesian optimization machine learning methods

The control of effluent total nitrogen (TN) and total energy consumption (TEC) is a key issue in managing wastewater treatment plants. In this study, effluent TN and TEC predictive models were established by selecting influent water quality and process control indicators as input features. The prediction performance of machine learning methods under different random seeds was explored, the moving average method was used for data amplification, and the Bayesian algorithm was used for hyperparameter optimization. The results showed that compared with the traditional hyperparameter optimization method for effluent TN prediction, the coefficient of determination (R2) increased by 0.092 and 0.067, reaching 0.725, and the root mean square error (RMSE) decreased by 0.262 and 0.215 mg/L, reaching 1.673 mg/L, respectively, after Bayesian optimization and data amplification. During TEC prediction, R2 increased by 0.068 and 0.042, reaching 0.884, and the RMSE decreased by 232.444 and 197.065 kWh, reaching 1305.829 kWh, respectively.

A Model of Multi-Finger Coordination in Keystroke Movement

In multi-finger coordinated keystroke actions by professional pianists, movements are precisely regulated by multiple motor neural centers, exhibiting a certain degree of coordination in finger motions. This coordination enhances the flexibility and efficiency of professional pianists' keystrokes. Research on the coordination of keystrokes in professional pianists is of great significance for guiding the movements of piano beginners and the motion planning of exoskeleton robots, among other fields. Currently, research on the coordination of multi-finger piano keystroke actions is still in its infancy. Scholars primarily focus on phenomenological analysis and theoretical description, which lack accurate and practical modeling methods. Considering that the tendon of the ring finger is closely connected to adjacent fingers, resulting in limited flexibility in its movement, this study concentrates on coordinated keystrokes involving the middle and ring fingers. A motion measurement platform is constructed, and Leap Motion is used to collect data from 12 professional pianists. A universal model applicable to multiple individuals for multi-finger coordination in keystroke actions based on the backpropagation (BP) neural network is proposed, which is optimized using a genetic algorithm (GA) and a sparrow search algorithm (SSA). The angular rotation of the ring finger's MCP joint is selected as the model output, while the individual difference information and the angular data of the middle finger's MCP joint serve as inputs. The individual difference information used in this study includes ring finger length, middle finger length, and years of piano training. The results indicate that the proposed SSA-BP neural network-based model demonstrates superior predictive accuracy, with a root mean square error of 4.8328°. Based on this model, the keystroke motion of the ring finger's MCP joint can be accurately predicted from the middle finger's keystroke motion information, offering an evaluative method and scientific guidance for the training of multi-finger coordinated keystrokes in piano learners.

Identification of Burkholderia gladioli pv. cocovenenans in Black Fungus and Efficient Recognition of Bongkrekic Acid and Toxoflavin Producing Phenotype by Back Propagation Neural Network

Burkholderia gladioli pv. cocovenenans is a serious safety issue in black fungus due to the deadly toxin, bongkrekic acid. This has triggered the demand for an efficient toxigenic phenotype recognition method. The objective of this study is to develop an efficient method for the recognition of toxin-producing B. gladioli strains. The potential of multilocus sequence typing and a back propagation neural network for the recognition of toxigenic B. cocovenenans was explored for the first time. The virulent strains were isolated from a black fungus cultivation environment in Qinba Mountain area, Shaanxi, China. A comprehensive evaluation of toxigenic capability of 26 isolates were conducted using Ultra Performance Liquid Chromatography for determination of bongkrekic acid and toxoflavin production in different culturing conditions and foods. The isolates produced bongkrekic acid in the range of 0.05-6.24 mg/L in black fungus and a highly toxin-producing strain generated 201.86 mg/L bongkrekic acid and 45.26 mg/L toxoflavin in co-cultivation with Rhizopus oryzae on PDA medium. Multilocus sequence typing phylogeny (MLST) analysis showed that housekeeping gene sequences have a certain relationship with a strain toxigenic phenotype. We developed a well-trained, back-propagation neutral network for prediction of toxigenic phenotype in B. gladioli based on MLST sequences with an accuracy of 100% in the training set and an accuracy of 86.7% in external test set strains. The BP neutral network offers a highly efficient approach to predict toxigenic phenotype of strains and contributes to hazard detection and safety surveillance.

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-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.

Optimization of catalytic wet air oxidation process in microchannel reactor for TBBS wastewater treatment

Catalytic wet air oxidation (CWAO) process is employed for the treatment of N-tert-butyl-2-benzothiazolesulfenamide (TBBS) wastewater in a microchannel reactor that enables continuous operation of the reaction and allows for thorough mixing of oxygen and pollutants. To achieve the optimal process performance, four key parameters of pressure, temperature, time, and the mass ratio of input oxygen to wastewater COD are optimized using both response surface methodology (RSM) and backpropagation artificial neural network (BP-ANN). According to the correlation coefficients of model results and experimental data, BP-ANN performs better than RSM in simulation and prediction. The analysis of variance in RSM shows that all parameters are significant for the obtained quadratic model, but their interactions with each other are not significant. Connection weights algorithm is used to determine the relative importance of these parameters for the process efficiency, and it is demonstrated that temperature is the most influential parameter with a relative importance of 35.61%, followed by pressure (29.74%), time (19.53%) and ROC (15.12%).

Prediction of cadmium contents in rice grains from Quaternary sediment-distributed farmland using field investigations and machine learning

The Quaternary sediment-distributed regions of South China are suitable for rice cultivation, which is crucial for ensuring food security. Spatial correlations between soil cadmium (Cd) and rice Cd contents are generally poor, making the evaluation of rice quality and associated health risks challenging. In this study, we developed machine learning algorithms for predicting rice Cd contents using 654 data pairs of soil-rice samples collected in Guangxi province, China. After a comprehensive comparison, our results showed that the random forest (RF) had the better performance than artificial neural network (ANN) based on all the data (RMSEtesting 0.066 vs. 0.099 and R2testing 0.860 vs. 0.688). The feature importance analysis showed that soil CaO, Cd, elevation, and rainfall were the four most important features affecting the rice Cd contents in the study area. Using the established RF-predicated model, the rice Cd contents were predicted at the provincial level with an additional dataset of 1176 paddy soil samples. The prediction result revealed about 23 % of farmland cultivated rice with Cd content over 0.2 mg kg-1 in the study area. Therefore, it is recommended to implement strict measures by local agricultural departments to reduce rice Cd contents and ensure food safety in these areas. Our study provides valuable insights into the prediction of rice Cd contents, thus contributing to ensuring food safety and preventing Cd exposure-associated health risks.

The effect of biocarriers on the nitrification and microbial community in moving bed biofilm reactor for anaerobic digestion effluent treatment

The performance of a moving bed biofilm reactor (MBBR) depends largely on the type of biofilm carrier used. However, how different carriers affect the nitrification process, particularly when treating anaerobic digestion effluents, is not completely understood. This study aimed to evaluate the nitrification performance of two distinct biocarriers in MBBRs over a 140-d operation period, with a gradually decreasing hydraulic retention time (HRT) from 20 to 10 d. Reactor 1 (R1) was filled with fiber balls, whereas a Mutag Biochip was used for reactor 2 (R2). At an HRT of 20 d, the ammonia removal efficiency of both reactors was >95%. However, as the HRT was reduced, the ammonia removal efficiency of R1 gradually declined, ultimately dropping to 65% at a 10-d HRT. In contrast, the ammonia removal efficiency of R2 consistently exceeding 99% throughout the long-term operation. R1 exhibited partial nitrification, whereas R2 exhibited complete nitrification. Analysis of microbial communities showed that the abundance and diversity of bacterial communities, particularly nitrifying bacteria such as Hyphomicrobium sp. And Nitrosomonas sp., in R2 was higher than that in R1. In conclusion, the choice of biocarrier significantly impact the abundance and diversity of microbial communities in MBBR systems. Therefore, these factors should be closely monitored to ensure the efficient treatment of high-strength ammonia wastewater.

Operational parameter prediction of electrocoagulation system in a rural decentralized water treatment plant by interpretable machine learning model

Electrocoagulation (EC) is a promising alternative for decentralized drinking water treatment in rural areas as a chemical-free technology. However, seasonal fluctuations of water quality in influent remain a significant challenge for rural decentralized water supply, which was a potential threat to water safety. The frequent operation was required to ensure the effluent water quality by the experienced technicians, who were in shortage in rural areas. If the operational parameters prediction model based on water quality could be established, it might reduce the dependence on technicians. Therefore, an artificial neural network (ANN) model combined with genetic algorithm (GA) was used to establish a prediction model for unattended intelligent operation. Data on water quality and operational parameters were collected from a practical EC system in a decentralized water treatment plant. Seven water quality parameters (e.g., turbidity, temperature, pH and conductivity) were selected as input variables and the operational current was employed as the output. A non-linear relationship between water quality parameters and the operational current was verified by correlation analysis and principal component analysis (PCA). The mean squared error (MSE) and coefficient of determination (R²) were used as evaluation indexes to optimize the structure of the GA-ANN model. Influent turbidity was identified to be crucial in the GA-ANN model by model interpretation using sensitivity analysis and scenario analysis. The Garson weight of turbidity in the seven input variables achieved 45.4%. The predictive accuracy of the GA-ANN model sharply declined from 90% to 67.1% when influent turbidity data were absent. In addition, it was estimated that energy consumption savings of the GA-ANN method declined by 14.2% in comparison with the gradient control method. This study verifies the feasibility and stability of machine learning strategy for unattended operation in the rural decentralized water treatment plant.

New methods based on a genetic algorithm back propagation (GABP) neural network and general regression neural network (GRNN) for predicting the occurrence of trihalomethanes in tap water

Monitoring trihalomethanes (THMs) levels in water supply systems is of great significance in ensuring drinking water safety. However, THMs detection is a time-consuming task. Developing predictive THMs models using parameters that are easier to obtain is an alternative. To date, there is still no application of optimization algorithms and general regression neural networks in predicting disinfection by-products levels. This study was to explore the feasibility of back propagation neural network (BPNN), genetic algorithm back propagation (GABP) neural network and general regression neural network (GRNN) for predicting THMs occurrence in real water supply systems. The results showed that the BPNN models' prediction ability was limited (test r_p = 0.571-0.857, N₂₅ = 61.5 %-91.5 %). Optimized by the genetic algorithm (GA), GABP models were generated and exhibited better prediction performance (test r_p = 0.573 and 0.696-0.863, N₂₅ = 68.2 %-93.6 %). However, GABP models took a lot of time and their prediction performance was unstable. A GRNN was then used to generate simpler neural network models, and the resulting prediction performance was excellent (total trihalomethanes and bromodichloromethane: test r_p = 0.657-0.824, N₂₅ = 81.8 %-100 %). In general, GRNN was the best at predicting THMs concentrations among the three models. However, it is worth noting that the prediction accuracy of bromodichloromethane (BDCM) was not high, which may be due to the absence of key parameters affecting BDCM formation. Accurate predictions of THMs by GRNN with these nine water parameters made THMs monitoring in real water supply systems possible and practical.

Elucidating the performance of integrated anoxic/oxic moving bed biofilm reactor: Assessment of organics and nutrients removal and optimization using feed forward back propagation neural network

A lab-scale integrated anoxic and oxic (A/O) moving bed biofilm reactor (MBBR) was investigated for the removal of organics and nutrients by varying chemical oxygen demand (COD) to NH₄-N ratio (C/N ratio: 3.5, 6.75, and 10), hydraulic retention time (HRT: 6 h, 15 h, and 24 h), and recirculation ratio (R: 1, 2, and 3). The use of activated carbon coated carriers prepared from waste polyethylene material and polyurethane sponges attached to a cylindrical frame in the integrated A/O MBBR increased the attached growth biomass significantly. >95 % of COD removal was observed under the C/N ratio of 10 at an HRT of 24 h. While the low C/N ratio favored the removal of NH₄-N (∼98 %) and PO₄^3--P (∼90 %) with an optimal R of 1.75. Using the experimental dataset, to predict and forecast the performance of integrated A/O MBBR, a feed-forward-backpropagation-neural-network model was developed.

Quantification of interfacial interaction related with adhesive membrane fouling by genetic algorithm back propagation (GABP) neural network

Since adhesive membrane fouling is critically determined by the interfacial interaction between a foulant and a rough membrane surface, efficient quantification of the interfacial interaction is critically important for adhesive membrane fouling mitigation. As a current available method, the advanced extended Derjaguin-Landau-Verwey-Overbeek (XDLVO) theory involves complicated rigorous thermodynamic equations and massive amounts of computation, restricting its application. To solve this problem, artificial intelligence (AI) visualization technology was used to analyze the existing literature, and the genetic algorithm back propagation (GABP) artificial neural network (ANN) was employed to simplify thermodynamic calculation. The results showed that GABP ANN with 5 neurons could obtain reliable prediction performance in seconds, versus several hours or even days time-consuming by the advanced XDLVO theory. Moreover, the regression coefficient (R) of GABP reached 0.9999, and the error between the prediction results and the simulation results was less than 0.01%, indicating feasibility of the GABP ANN technique for quantification of interfacial interaction related with adhesive membrane fouling. This work provided a novel strategy to efficiently optimize the thermodynamic prediction of adhesive membrane fouling, beneficial for better understanding and control of adhesive membrane fouling.