Comparative analysis of RSM, ANN and ANFIS and the mechanistic modeling in eriochrome black-T dye adsorption using modified clay

Prediction of g-C3N4-based photocatalysts in tetracycline degradation based on machine learning

Investigating the effects of g-C3N4-based photocatalysts on experimental parameters during tetracycline (TC) degradation can be helpful in discovering the optimal parameter combinations to improve the degradation efficiencies in general. Machine learning methods can avoid the problems of high cost, time-consuming and possible instrumental errors in experimental methods, which have been proven to be an effective alternative for evaluating the entire experimental process. Eight typical machine learning models were explored for their effectiveness in predicting the TC degradation efficiencies of g-C3N4 based photocatalysts. XGBoost (XGB) was the most reliable model with R2, RMSE and MAE values of 0.985, 4.167 and 2.900, respectively. In addition, XGB's feature importance and SHAP method were used to rank the importance of features to provide interpretability to the results. This study provided a new idea for developing g-C3N4-based photocatalysts for TC degradation and intelligent algorithms for predicting the photocatalytic activity of g-C3N4-based photocatalysts.

Application of Optimization Response Surface for the Adsorption of Methylene Blue Dye onto Zinc-coated Activated Carbon

The activated carbon was produced in the first phase of this investigation by chemically activating hazelnut shell waste with H3PO4. Composite materials were obtained by coating the activated carbon with zinc oxide, whose BET surface area was calculated as 1278 m2 g-1. ZnO-doped ZnO/AC composite was synthesized as an adsorbent for its possible application in the elimination of organic dyestuff MB, and its removal efficiency was investigated. Morphological properties of ZnO/AC were characterized using analytical methods such as XRD, SEM, and BET. The adsorption system and its parameters were investigated and modeled using the response surface method of batch adsorption experiments. The experimental design consisted of three levels of pH (3, 6.5, and 10), initial MB concentration (50, 100, and 150 mg L-1), dosage (0.1, 0.3, and 0.5 g 100 mL-1), and contact time (5, 50, and 95 min). The results from the RSM suggested that the MB removal efficiency was 98.7% under the optimum conditions of the experimental factors. The R2 value, which expresses the significance of the model, was determined as 99.05%. Adsorption studies showed that the equilibrium data fit well with the Langmuir isotherm model compared to Freundlich. The maximum adsorption capacity was calculated as 270.70 mg g-1.

Machine learning and statistical physics modeling of tetracycline adsorption using activated carbon derived from Cynometra ramiflora fruit biomass

The current investigation reports the usage of adaptive neuro-fuzzy inference system (ANFIS) and artificial neural network (ANN), the two recognized machine learning techniques in modelling tetracycline (TC) adsorption onto Cynometra ramiflora fruit biomass derived activated carbon (AC). Many characterization methods utilized, confirmed the porous structure of synthesized AC. ANN and ANFIS models utilized pH, dose, initial TC concentration, mixing speed, time duration, and temperature as input parameters, whereas TC removal percentage was designated as the output parameter. The optimized configuration for the ANN model was determined as 6-8-1, while the ANFIS model employed trimf input and linear output membership functions. The obtained results showed a strong correlation, indicated by high R2 values (ANNR2: 0.9939 & ANFISR2: 0.9906) and low RMSE values (ANNRMSE: 0.0393 & ANFISRMSE: 0.0503). Apart from traditional isotherms, the dataset was fitted to statistical physics models wherein, the double-layer with a single energy satisfactorily explained the physisorption mechanism of TC adsorption. The sorption energy was 21.06 kJ/mol, and the number of TC moieties bound per site (n) was found to be 0.42, conclusive of parallel binding of TC molecules to the adsorbent surface. The adsorption capacity at saturation (Qsat) was estimated to be 466.86 mg/g - appreciably more than previously reported values. These findings collectively demonstrate that the AC derived from C. ramiflora fruit holds great potential for efficient removal of TC from a given system, and machine learning approaches can effectively model the adsorption processes.

Computational AI to predict and optimize the relationship between dye removal efficiency and Gibbs free energy in the adsorption process utilizing TiO2/chitosan-polyacrylamide composite

Building a model that can accurately anticipate and optimize the dynamics of dye removal and Gibbs free energy within the framework of an adsorption process is the main goal of this research. Furthermore, it has been determined that a correlation exists between the efficacy of dye removal and the behavior of Gibbs free energy throughout the process of adsorption. The study utilized a composite material consisting of chitosan-polyacrylamide/TiO2 as an adsorbent to remove anionic dye from a mainly aqueous solution. The parameters have been analyzed using response surface methodology (RSM), artificial neural networks (ANN), and machine learning (ML) techniques in this particular context. The obtained F-value of 814.62 for the RSM model, which assesses dye removal efficiency, suggests that the model under examination is statistically significant. Furthermore, based on the RSM data, the proposed model demonstrates a significant level of accuracy in predicting the performance of the TiO2/chitosan-polyacrylamide composite as an adsorbent during the dye removal adsorption process. The ANN model achieved a high level of accuracy, as evidenced by its R2 value of 0.999455. Through the utilization of neural networks and machine learning, the intended objective of forecasting dye removal efficiency and Gibbs free energy behavior in the adsorption process was effectively accomplished.

Intelligent algorithms-aided modeling and optimization of the deturbidization of abattoir wastewater by electrocoagulation using aluminium electrodes

The removal of turbidity from abattoir wastewater (AWW) by electrocoagulation (EC) was modeled and optimized using Artificial Intelligence (AI) algorithms. Artificial neural networks (ANN), adaptive neuro-fuzzy inference systems (ANFIS), particle swarm optimization (PSO), and genetic algorithms (GA) were the AI tools employed. Five input variables were considered: pH, current intensity, electrolysis time, settling time, and temperature. The ANN model was evaluated using the Levenberg-Marquardt (trainlm) algorithm, while the ANFIS modeling was accomplished using the Sugeno-type FIS. The ANN and ANFIS models demonstrated linear adequacy with the experimental data, with an R2 value of 0.9993 in both cases. The corresponding statistical error indices were RMSE (ANN = 5.65685E-05; ANFIS = 2.82843E-05), SSE (ANN = 1.60E-07; ANFIS = 3.4E-08), and MSE (ANN = 3.2E-09; ANFIS = 8E-10). The error indices revealed that the ANFIS model had the least performance error and is considered the most reliable of the two. The process optimization performed with GA and PSO considered turbidity removal efficiency, energy requirement, and electrode material loss. An optimal turbidity removal efficiency of 99.39 % was predicted at pH (3.1), current intensity (2 A), electrolysis time (20 min), settling time (50 min), and operating temperature (50 °C). This represents a potential for the delivery of cleaner water without the use of chemicals. The estimated power consumption and the theoretical mass of the aluminium electrode dissolved at the optimum condition were 293.33 kW h/m3 and 0.2237 g, respectively. The work successfully affirmed the effectiveness of the EC process in the removal of finely divided suspended particles from AWW and demonstrated the suitability of the AI algorithms in the modeling and optimization of the process.

Chitosan-minerals-based composites for adsorption of caesium, cobalt and europium

Currently, there is a growing interest in the use of natural materials in various fields of science, technology and environmental protection due to their availability, low-cost, non-toxicity and biodegradability. Chitosan, natural clay of local origin, montmorillonite, zeolite, cross-linking agents (epichlorohydrin, sodium tripolyphosphate, glutaraldehyde) and plasticisers (glycerol) were used to synthesise composites. The composites were characterised by attenuated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR), X-ray diffraction analysis (XRD) and scanning electron microscope (SEM), tested for their antibacterial activity and used in batch experiments to study the adsorption of caesium, cobalt and europium ions. The maximum capacities for adsorption of caesium, cobalt and europium on the composites were 1400 mg/g, 900 mg/g and 18 mg/g, respectively. The experimental data fit better the Langmuir isotherm model and indicate favourable monolayer adsorption of Cs+, Co2+ and Eu3+ at homogeneous sites of the composites. The experimental data were in better agreement with the pseudo-second-order non-linear kinetic model for most elements and adsorbents. Adaptive neuro-fuzzy inference system proved to be a practical tool with good performance and generalisation capability for predicting the adsorption capacity of composites for caesium, cobalt, and europium ions. It was found that the predicted data were very close to the experimental data.

Removal of an agricultural herbicide (2,4-Dichlorophenoxyacetic acid) using magnetic nanocomposite: A combined experimental and modeling studies

This study focused on modeling the removal of one of the widely used agricultural herbicides known as 2,4-Dichlorophenoxyacetic acid (2,4-D) using polypyrrole-coated Fe2O3 nanoparticles (Fe2O3@PPy). The Fe2O3@PPy nanocomposite was synthesized by surface-coating the Tabebuia aurea leaf extract synthesized Fe2O3 nanoparticles with polypyrrole. After characterization, the adsorptive potential of the nanocomposite for removing 2,4-D from aqueous solution was examined. Central composite design (CCD) was employed for optimizing the adsorption, revealing an adsorption efficiency of 90.65% at a 2,4-D concentration of 12 ppm, a dosage of 3.8 g/L, an agitation speed of 150 rpm, and 196 min. Adsorption dataset fitted satisfactorily to Langmuir isotherm (R2: 0.984 & χ2: 0.054) and pseudo-second-order kinetics (R2: 0.929 & χ2: 0.013) whereas the exothermic and spontaneous nature were confirmed via the thermodynamic study. The predictive models, including adaptive neuro-fuzzy inference system (ANFIS), artificial neural network (ANN), and response surface methodology (RSM), demonstrated good precision for the prediction of 2,4-D adsorption, with respective R2 of 0.9719, 0.9604, and 0.9528. Nevertheless, statistical analysis supported ANFIS as the better forecasting tool, while RSM was the least effective. The maximum adsorption capacity of 2,4-D onto the Fe2O3@PPy nanocomposite was 7.29 mg/g, significantly higher than a few reported values. Therefore, the Fe2O3@PPy nanocomposite could serve as a competent adsorbent to remove 2,4-D herbicide from aqueous streams.

Optimal pretreatment of plantain peel waste valorization for biogas production: Insights into neural network modeling and kinetic analysis

This work proposed a model for the substrate treatment stage of biogas production process in an anaerobic digestion system. Adaptive neuro-fuzzy inference system (ANFIS), response surface method (RSM), and artificial neural network (ANN) were comparatively used in the simulation and modeling of the treatment process for improved biogas yield. Waste plantain peels were pretreated and used as substrate. FTIR and SEM results revealed that the pretreatment improved the substrate's desirable qualities. The amount of biogas yield was controlled by time, NaOH concentration, and temperature of the substrate pretreatment. Optimum pretreatment conditions obtained were a temperature of 102.7 °C, time of 31.7 min and NaOH concentration of 0.125 N. RSM, ANN, and ANFIS modeling techniques were proficient in simulating the biogas production, as evidenced by high R2values of 0.9281, 0.9850, and 0.9852, respectively. Furthermore, the values of the calculated error terms such as RMSE (RSM = 0.04799, ANN = 0.00969, and ANFIS = 0.00587) and HYBRID (RSM = 18.556, ANN = 0.803, and ANFIS = 0.0447) were low, indicating a satisfactory correlation between experimental and predicted values. Scrubbing of the biogas with caustic soda and activated charcoal increased the methane content to 94 %. The kinetics of the cumulative biogas yield were best fit with the Logistics and Modified Logistics models. The low C/N ratio in addition to the presence of potassium, nitrogen, and phosphorus suggested that the spent plantain peel slurry can be utilized as an agricultural fertilizer in crop production. The observations of this study therefore recommends the pre-treatment of biodigestion substrates as a key means to enhance beneficiation of methane production.

Optimisation of process parameters of a thermal digester for the rapid conversion of food waste into value-added soil conditioner

A novel thermal digester for converting food waste (FW) into nutrient-rich soil conditioner was designed and explored. The process variables, that is, temperature, the volume of the digestion chamber and the rotational speed of the digester were optimised using response surface methodology (RSM). The study revealed that the digester temperature of 150°C and rotational speed of 40 RPM required minimum time (180 minutes) for attaining the equilibrium moisture with a minimum energy consumption of 0.218 kWh kg-1. The process resulted in 80 ± 2.5% reduction in total volume of the FW. Detailed characterisation revealed that the end product was comparable to the organic fertiliser as per the Fertiliser Association of India norms. The digestion helps in breakdown of cellulose content of FW into hemicellulose which supports formation of primary and secondary walls, seed storage carbohydrates, and facilitates plant growth. 1H-Nuclear magnetic resonance (1H-NMR) spectra of the end product revealed mineralisation of organics during digestion. Decrease in ultraviolet (UV) absorbance value at 280 nm also revealed the humification of the end product. X-ray diffraction (XRD) analysis disclosed extremely low crystallinity and non-recalcitrant nature of the end product. A low humification index value (HI-3.43), high fertilising index (FI-4.8), and clean index (CI-5.0) revealed that the end product could safely be utilised as an organic fertiliser. The cost-benefit analysis revealed that thermal digestion technique is profitable and economically viable with benefit-cost ratio (BCR) of 1.35. The study offers a unique approach for the rapid and hassle-free production of value-added soil conditioner from FW.

Artificial neural networks for performance prediction of full-scale wastewater treatment plants: a systematic review

Wastewater treatment plants (WWTPs) are complex systems that must maintain high levels of performance to achieve adequate effluent quality to protect the environment and public health. Artificial intelligence and machine learning methods have gained attention in recent years for modeling complex problems, such as wastewater treatment. Although artificial neural networks (ANNs) have been identified as the most common of these methods, no study has investigated the development and configuration of these models. We conducted a systematic literature review on the use of ANNs to predict the effluent quality and removal efficiencies of full-scale WWTPs. Three databases were searched, and 44 records of the 667 identified were selected based on the eligibility criteria. The data extracted from the papers showed that the majority of studies used the feedforward neural network model with a backpropagation training algorithm to predict the effluent quality of plants, particularly in terms of organic matter indicators. The findings of this research may help in the search for an optimum design modeling process for future studies of similar prediction problems.

Comparative evaluation and sensitivity analysis of multi-modelling and optimization of milling Ti-6Al-4V alloy with high-pressure coolant jets

An effective cooling method with the proper selection of process parameters can intensify the machining performance by reducing the loss of resources with better quality products. In this regard, modelling is an appropriate way of predicting responses in changing environment and optimization is an efficient tool of selecting the best process parameters based on the specific desire. With a view to enhance the machinability of Ti-6Al-4V alloy, the first attempt of the current study was to predict the performance characteristics of milling such as cutting force (N), specific cutting energy (J/mm³) and surface roughness (μm) with the variation of speed (m/min), feed (mm/min), depth of cut (mm) and cooling approach by developing mathematical models. For the present work, three different predictive models such as response surface methodology (RSM), artificial neural network (ANN), and adaptive neuro fuzzy inference system (ANFIS) was followed. Additionally, a comparative assessment of the used predictive models was carried out and ANFIS was noticed as the most accurate predictive model. After that, optimization of the selected responses was conducted by multiple-objective optimization on the basis of ratio analysis (MOORA) method where the relative weights of each response were defined by principal component analysis (PCA). For milling Ti-6Al-4V alloy within the specific boundary conditions, PCA-MOORA suggested an optimal parameter setting at 32 m/min speed, 22 mm/min feed rate, and 0.75 mm depth of cut with rotary high-pressure cooling. Finally, the sensitivity of the used MOORA method with the variation of unitary ratio was checked out to take a robust decision.

Use of artificial intelligence for optimizing biosorption of textile wastewater using agricultural waste

Most of the dyes are toxic and non-biodegradable in textile industry wastewaters. Therefore, removal of textile dye using agriculture waste becomes crucial for the environment. This can be accomplished by the biosorption process which is the passive uptake of pollutants by agricultural waste. In this study, Response Surface Methodology (RSM), Artificial Neural Network (ANN) and Adaptive Neuro-Fuzzy Inference System (ANFIS) were used to obtain optimum conditions for Methylene Blue (MB) removal using sugarcane bagasse and peanut hulls as low-cost agricultural waste. The experimental design was carried out to study the effect of temperature, pH, biosorbent amount and dye concentration. The maximum MB dye removal considering the effect of total dissolved solids from aqueous solutions of 74.49% and 67.99% by sugarcane bagasse and peanut hulls, respectively. The models specify that they could predict biosorption with high accuracy having R²-value above 0.9. Statistical studies for RSM, ANFIS and ANN models were compared. Further, the models were optimized for maximum dye removal was at 1.21 g of biosorbent, pH 5.24, 31.24 mg/L MB concentration, 22.29°C of dye solution using sugarcane bagasse and at 1.37 g of biosorbent, pH 5.77, 36.7 mg/L MB concentration, 26.8°C of dye solution using peanut hulls. Additionally, Fourier Transform Infra-Red (FTIR) spectral analysis was also carried out to confirm the biosorption.

Preparation of Graphene Oxide-Maghemite-Chitosan Composites for the Adsorption of Europium Ions from Aqueous Solutions

The adsorption of Eu(III) on composites synthesised from graphene oxide (GO), maghemite (MGH), and chitosan (CS) has been studied using different approaches. The physicochemical and morphological characteristics of the composites GO-MGH, GO-CS, GO-MGH-CS I, II, and III were determined by XRD, Mössbauer spectroscopy, FTIR, Raman spectroscopy, and TEM. According to the results of batch experiments, the maximum experimental adsorption capacity was 52, 54, 25, 103, and 102 mg/g for GO-MGH, GO-CS, GO-MGH-CS I, II, and III, respectively. The data obtained are in better agreement with the Langmuir, pseudo-second-order, and pseudo-first-order models only for GO-MGH. Thus, the adsorption of Eu(III) on the composites was a favourable, monolayer, and occurred at homogeneous sites. The nature of adsorption is chemical and, in the case of GO-MGH, physical. Tests of the composites in natural waters showed a high removal efficiency for Eu(III), Pu(IV), and Am(III), ranging from 74 to 100%. The ANFIS model has quite good predictive ability, as shown by the values for R², MSE, SSE, and ARE. The GO-MGH-CS composites with the high adsorption capacity could be promising candidates for the removal of Eu(III) and the pre-concentration of Pu(IV) and Am(III) from natural waters.

Optimization of the Pentachlorophenol Adsorption by Organo-Clays Based on Response Surface Methodology

The aim of this study is to optimize the adsorption of pentachlorophenol (PCP) using an organo-clay under the response surface methodology. The adsorbent was selected from a montmorillonite exchanged by various cations, such as Fe³⁺, Al³⁺, Zn²⁺, Mg²⁺, Na⁺, and modified by bromide cetyltrimethylammonium (CTAB) as surfactant. The obtained organo-montmorillonite was characterized using several techniques, such as Fourier-transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), thermogravimetric analysis (TGA), scanning electron microscopy (SEM), and nitrogen adsorption, performed at -196 °C. The results showed an increase in basal space from 1.65 to 1.88 nm and a decrease in the specific surface and pore volume, with an increase in pore diameter, including the presence of characteristic bands of -CH₂- and -CH₃- groups at 2926 and 2854 cm^-1 in the FTIR spectrum after the modification. The optimization of PCP removal by clay adsorbents is achieved using the response surface methodology (RSM) with a four-factor central composite model, including pH of solution, mass of adsorbent, contact time, and initial concentration. The results proved the validity of the regression model, wherein the adsorption capacity reaches its maximum value of 38 mg/g at a lower adsorbent mass of 20 mg, pH of 6, contact time (t_c) of 5 h, and initial concentration of 8 mg/L.

Adsorption of caesium and cobalt ions on the muscovite mica clay-graphene oxide-γ-Fe2O3-Fe3O4 composite

The muscovite mica clay-graphene oxide-maghemite-magnetite (γ-Fe₂O₃-Fe₃O₄) composite was first used for the adsorption of caesium(I) and cobalt(II). The presence of clay minerals, graphene oxide, maghemite, and magnetite was detected in the prepared composite by XRD, WD-XRF, Mössbauer spectroscopy, and ATR-FTIR. The SEM and TEM results show that the composite has a layered structure with irregularly shaped pores on the surface. It was found that the adsorption of ions depends on the initial concentration, pH (except for caesium), mass of adsorbent, temperature, and contact time. The maximum adsorption capacity for Cs(I) and Co(II) was 2286 mg/g and 652 mg/g, respectively, and was obtained at concentrations (Cs(I) = 12,630 mg/L; Co(II) = 3200 mg/L), adsorbent mass of 0.01 g, pH (Cs(I) = 7; Co(II) = 5), temperature of 20 ± 1 °C, and contact time of 24 h. The high adsorption capacity of the composite could be due to a diversity of functional groups, a large number of active sites or the multilayer adsorption of caesium and cobalt ions on the surface of the composite. The Freundlich, Langmuir isotherms, and the pseudo-second-order kinetic model better describe the adsorption of these ions on the composite. The adsorption was non-spontaneous endothermic for Cs(I) and spontaneous endothermic for Co(II). The proposed mechanism of adsorption of Cs and Co ions on the composite is complex and involves electrostatic interactions and ion exchange. The ANFIS model proved to be quite effective in predicting the adsorption of Cs(I) and Co(II), as shown by the obtained values of R², MSE, SSE, and ARE.

Methylene Blue Adsorption BY UV-Treated Graphene Oxide Nanoparticles (UV/n-GO): Modeling and Optimization Using Response Surface Methodology and Artificial Neural Networks

To mitigate the negative effects of pollution produced by the growing levels of pollutants in the environment, research and development of novel and more effective materials for the treatment of pollutants originating from a variety of industrial sources should be prioritized. In this research, a UV-irradiated nano-graphene oxide (UV/n-GO) was developed and studied for methylene blue (MB) adsorption. Furthermore, the batch adsorption studies were modelled using response surface modelling (RSM) and artificial neural networks (ANNs). Investigations employing FTIR, XRD, and SEM were carried out to characterize the adsorbent. The best MB removal of 95.81% was obtained at a pH of 6, a dose of 0.4 g/L, an MB concentration of 25 mg/L, and a period of 40 min. This was accomplished with a desirability score of 0.853. A three-layer backpropagation network with an ideal structure of 4-4-1 was used to create an ANN model. The R2 and MSE values determined by comparing the modelled data with the experimental data were 0.9572 and 0.00012, respectively. The % MB removal predicted by ANN was 94.76%. The kinetics of adsorption corresponded well with the pseudo-second-order model (R2 > 0.97). According to correlation coefficients, the order of adsorption isotherm models is Redlich–Peterson > Temkin > Langmuir > Freundlich. Thermodynamic investigations show that MB adsorption was both spontaneous and endothermic.

Salt sealing induced in situ N-doped porous carbon derived from wheat bran for the removal of doxycycline from aqueous solution

In situ N-doped porous carbon (NPC) derived from wheat bran via a convenient salt sealing and air-assisted strategy was prepared for the removal of doxycycline (DOX) from aqueous solution. The NPC was precisely characterized by SEM, FTIR, XPS and BET analysis. Additionally, the experimental variables including contact time, adsorbent dosage of NPC and pH were optimized by using Box-Behnken design (BBD) under response surface methodology (RSM). The predicted adsorption capacity of DOX was found to be 291.14 mg g-1 under optimalizing experimental conditions of 196 min contact time, 0.2 g L-1 adsorbent dosage and pH 5.78. The adsorption experimental data fitted Langmuir, Koble-Corrigan and Redlich-Peterson models well, and the pseudo-second-order model perfectly described the DOX adsorption process onto NPC. Thermodynamic parameters of DOX adsorbed onto NPC indicated that the adsorption process was spontaneous and endothermic. Moreover, the adsorption of DOX on NPC was mostly controlled by electrostatic interaction, π-π electron-donator-acceptor (EDA) interaction, hydrogen-bonding and Lewis acid-base effect. Besides, the N element of NPC also played a role in capturing DOX. The maximum monolayer adsorption capacity of DOX was turn out to be 333.23 mg g-1 at 298 K, which suggested that the NPC could be a prospectively adsorbent for the removal of DOX from wastewater.

Artificial neural network (ANN) and response surface methodology (RSM) algorithm-based improvement, kinetics and isotherm studies of electrocoagulation of oily wastewater

The work reported here focuses on the oil and grease removal from wastewater by the electrocoagulation process and using modeling and optimization for obtaining the results considering four major operating parameters, viz. current density, pH, electrode distance and reaction time. 31 experiments were designed by design of experiments (DOE) of response surface methodology (RSM) and the analysis of variance (ANOVA) studies confirmed the agreement of the experimental results. Artificial neural network (ANN) was also utilized to determine predicted response using neural networks for 4-10-1 arrangement. Both the responses predicted by RSM and ANN were in alignment with the experimental results. Maximum removal of 78% was attained under the working parameters of 80 A m^-2, 3.6 pH, electrode distance of 0.005 m and reaction time of 20 min.

Experimental Investigation and Prediction of Mechanical Properties in a Fused Deposition Modeling Process

Additive manufacturing, also known as three-dimensional printing, is a computer-controlled advanced manufacturing process that produces three-dimensional items by depositing materials directly from a computer-aided design model, usually in layers. Due to its capacity to manufacture complicated objects utilizing a wide range of materials with outstanding mechanical qualities, fused deposition modeling is one of the most commonly used additive manufacturing technologies. For printing high-quality components with appropriate mechanical qualities, such as tensile strength and flexural strength, the selection of adequate processing parameters is critical. Experimentally, the influence of process parameters such as the raster angle, printing orientation, air gap, raster width, and layer height on the tensile strength of fused deposition modeling printed items was examined in this work. Through analysis of variance, the impact of each parameter was measured and rated. The system’s response was predicted using an adaptive neuro-fuzzy technique and an artificial neural network. In Minitab software, the Box-Behnken response surface experimental design was used to generate 46 experimental trials, which were then printed using acrylonitrile butadiene styrene polymer materials on a three-dimensional forge dreamer II fused deposition modelling printing machine. The results revealed that the raster angle, air gap, and raster width had significant impacts on the tensile strength. The adaptive neuro-fuzzy approach and artificial neural network predicted tensile strength accurately with an average percentage error of 0.0163 percent and 1.6437 percent, respectively. According to the findings, the model and experimental data are in good agreement.

Data-Driven Machine Learning Intelligent Tools for Predicting Chromium Removal in an Adsorption System

This study investigates chromium removal onto modified maghemite nanoparticles in batch experiments based on a central composite design. The effect of modified maghemite nanoparticles on the adsorptive removal of chromium was quantitatively elucidated by fitting the experimental data using artificial neural network (ANN) and adaptive neuro-fuzzy interference system (ANFIS) modeling approaches. The ANN and ANFIS models, relating the inputs, i.e., pH, adsorbent dose, and initial chromium concentration to the output, i.e., chromium removal efficiency (RE), were developed by comparing the predicted value with that of the experimental values. The RE of chromium ranged from 49.58% to 92.72% under the influence of varying pH (i.e., 2.6–9.4) and adsorbent dose, i.e., 0.8 g/L to 9.2 g/L. The developed ANN model fits the experimental data exceptionally well with correlation coefficients of 1.000 and 0.997 for training and testing, respectively. In addition, the Pearson’s Chi-square measure (χ2) of 0.0004 and 0.0673 for the ANN and ANFIS models, respectively, indicated the superiority of ANN over ANFIS. However, a small discrepancy in the predictability of the ANFIS model was observed owing to the fuzzy rule-based complexity and overtraining of data. Thus, the developed models can be used for the online prediction of RE onto synthesized maghemite nanoparticles with different sets of input parameters and it can also predict the operational errors in the system.

Modelling of adsorption of anionic azo dye using Strychnos potatorum Linn seeds (SPS) from aqueous solution with artificial neural network (ANN)

Synthetic dyes used in the textile and paper industries pose a major threat to the environment. In the present research work, the adsorption efficiency of the natural adsorbent Strychnos potatorum Linn (Fam: Loganiaceae) seeds were examined against the reactive orange-M2R dye from aqueous solution by varying the process conditions such as contact time, pH, adsorbent dosage, and initial dye concentration on adsorption of anionic azo dye. This study compares different types of artificial neural networks which are feedforward artificial neural network (FANN) and nonlinear autoregressive exogenous (NARX) model to predict the efficiency of a cost-effective natural adsorbent Strychnos potatorum Linn seeds on removing reactive orange-M2R dye from aqueous solution. Twelve training algorithms of neural network were compared, and the prediction on the adsorption performance of anionic azo dye from aqueous solution using Strychnos potatonum Linn seeds was evaluated by using the root mean squared error (RMSE), mean absolute error (MAE), coefficient of determination (R²), and accuracy. For FANN model, Levenberg-Marquardt (LM) backpropagation with 19 hidden neurons was selected as the optimum FANN model, with R² of 0.994 and accuracy of 87.20%, 98.21%, and 66.60% for training, testing, and validation datasets, respectively. For NARX model, LM with 8 hidden neurons was selected as the most suitable training algorithm, with R² value of more than 0.99 and accuracy of 88.00%, 90.91%, and 75.00% for training, testing, and validation datasets, respectively. NARX model accurately predicted the adsorption of anionic azo dye from aqueous solution using Strychnos potatonum Linn seeds with better performance than FANN model.

A hybrid RSM-ANN-GA approach on optimisation of extraction conditions for bioactive component-rich laver (Porphyra dentata) extract

This research established the optimal conditions for infusion extraction (IE) and ultrasound-assisted extraction (UAE) of bioactive components from laver (Porphyra dentata) using response surface methodology (RSM) and artificial neural network coupled with genetic algorithm (RSM-ANN-GA). The variables, temperatures (60, 80, and 100 ℃) and times (10, 15, and 20 min) were designed to optimise total phenolic, total flavonoid, total amino acid, a* value, and R-phycoerythrin content of laver extract. The optimised condition for IE and UAE was achieved at 60 ℃ for 18.08 min and 80.66℃ for 14.76 min in RSM while showing 60 ℃ for 19 min and 80℃ for 15 min in the RSM-ANN-GA mode, respectively. Results revealed that RSM-ANN-GA provided better predictability and greater accuracy than the RSM model and laver extract from UAE gave the higher values of responses compared to those from IE. These findings highlight the high-efficient extraction method along with better statistical approach.