Conventional Response Surface Methodology Research Articles

Optimizing the effect of using novel hydrogen enriched nano particles added emulsified waste mango seed biodiesel in diesel engine

Restricted fossil fuel resources, increased liquid fuel prices, and the pollution range of conventional fuel direct the research into novel technology. The application of biodiesel enhances engine characteristics. In this work, different concentrations of hybrid nanoparticles, namely Aluminum oxide and Titanium oxide (25 ppm, 50 ppm, and 75 ppm), were introduced as nanoadditives. In addition, 10 % share of hydrogen gas by volume was given to analyze the engine characteristics of hydrogen-fueled single cylinder diesel engine. The entire testing was conducted on a constant water content (10 % vol.) emulsified Mango Seed Methyl Ester (MSME) as base fuel blend. The experiment has been planned in Response Surface Methodology (RSM) based on Box Behnken Design (BBD) approach. The variable input parameters, engine load, speed and nanoparticles (NP) concentrations are optimized using RSM and Deep Recurrent Neural Network based Honey Badger Algorithm (DRNN-HBA). The addition of nanoparticles improved the thermal property of the fuel blend, thereby increasing Brake Thermal Efficiency (BTE) for the B20W10 fuel blend with nanoparticles. A maximum BTE of 32 % is achieved by adding 75 ppm nanoparticle to the fuel blend with some counter effect on NOx emission at maximum load. The optimized outcomes are 58 % engine load, 1800 rpm engine speed and 75 ppm NP concentration (B20W10NP75). The proposed DRNN-HBA prediction model achieved better regression correlation (above 0.99) for combustion and emission results with very lesser Root Mean Square Error (RMSE ≤ 6.6) compared to conventional DRNN and RSM.

Optimization of biodegradation yield of reactive blue 49: An integrated approach using response surface methodology based marine predators algorithm

Biodegradation involving the use of biological systems are proving to be more cost-effective, energy efficient and environmentally friendly method for treatment with dye. There are also some advantages of enzymatic degradation to remove pollutants from textile effluents. They are specific and selectable for the substrates. Moreover, the usage of enzymatic degradation with laccase has environmental and social dimensions of sustainable bio-economy when compared with commercial solutions. In this study, besides the conventional RSM (Response Surface Methodology) approaches, the performance of a new metaheuristic, MPA (Marine Predators Algorithm) integrated with RSM has also been tested. The effect of four independent variables time, enzyme amount, agitation speed, and initial dye concentration have been studied to obtain maximum decolorization. Two major techniques of RSM, namely Box Behnken Design (BBD) and Central Composite Design (CCD) have been used to determine optimal levels of parameters. The highest decolorization efficiency (%) of Reactive Blue 49 (RB49) has been achieved as 92.82% with BBD and 90.56% with CCD. Maximum efficiency for BBD based MPA as being 94.69% has been obtained at 70.96 min, 1.5 mL, 74.63 rpm, and 99.09 ppm. For CCD based MPA, the level of efficiency has been obtained as 93.1% at 39.69 min, 1 mL, 72.07 rpm, and 74.41 ppm. The results of MPA indicate that using metaheuristics combined with RSM is a suitable and sustainable way to optimize parameters of RB49% for decolorization.

Maximizing Adsorption Involving Three Solutes on Enhanced Adsorbents Using the Mixture-Process Variable Design.

Unmodified (UN), acid-treated (AT) and microwave-acid-treated (MAT) activated carbons were optimized for their solute removal efficacies by adjusting feed mixture compositions and process conditions. Acetaminophen, benzotriazole, and caffeine were used either individually or as binary/ternary mixtures in this study. The process conditions considered were the pH, adsorbent dosage, and type of adsorbent. Experimental responses such as total adsorbent loading (qtotal) and total percentage removal (PRtotal) were fitted with empirical models that had high adjusted R2 (>0.95), insignificant lack of fit (p-value > 0.22), and high model predictive R2 (>0.93). Mixture compositions of the feed were found to interact significantly not only among themselves but with process variables as well. Hence, adsorption optimization must simultaneously consider mixture as well as process variables. The conventional response surface methodology for mixtures, termed as ridge analysis, optimizes mixture compositions at specified values of process variables. An improved steepest ascent method which considers mixture and process variables simultaneously was developed in this work. This could track the path of steepest ascent toward globally optimal settings, from any arbitrary starting point within the design space. For the chosen adsorbent, optimal settings for feed mixture compositions and pH were found to change along this steepest ascent path. The feed compositions, pH, and adsorbent dosage identified for maximum adsorbent utilization were usually quite different from those identified for maximum total percentage removal. When both these objectives were optimized together, the most favorable compromise solutions for qtotal and PRtotal were, respectively, 264.1 mg/g and 43.4% for UN, 294.9 mg/g and 52.5% for AT, and 336.6 mg/g and 55.9% for MAT.

Robust Process Parameter Design Methodology: A New Estimation Approach by Using Feed-Forward Neural Network Structures and Machine Learning Algorithms

In robust design (RD) modeling, the response surface methodology (RSM) based on the least-squares method (LSM) is a useful statistical tool for estimating functional relationships between input factors and their associated output responses. Neural network (NN)-based models provide an alternative means of executing input-output functions without the assumptions necessary with LSM-based RSM. However, current NN-based estimation methods do not always provide suitable response functions. Thus, there is room for improvement in the realm of RD modeling. In this study, a new NN-based RD modeling procedure is proposed to obtain the process mean and standard deviation response functions. Second, RD modeling methods based on the feed-forward back-propagation neural network (FFNN), cascade-forward back-propagation neural network (CFNN), and radial basis function network (RBFN) are proposed. Third, two simulation studies are conducted using a given true function to verify the proposed three methods. Fourth, a case study is examined to illustrate the potential of the proposed approach. In conclusion, a comparative analysis of the three feed-forward NN structure-based modeling methods and conventional LSM-based RSM proposed in this study showed that the proposed methods were significantly lower in the expected quality loss (EQL) and various variability indicators.

Artificial intelligence driven advanced optimization strategy vis-à-vis response surface optimization of production medium: Bacterial exopolysaccharide production as a case-study

In the present study, Artificial Intelligence (AI) driven multivariate-optimization strategy, hybrid-support vector machine (h-SVM) was employed and compared with conventional response surface methodology (RSM) in terms of its performance efficiency to optimize the medium composition for improved exopolysaccharide (EPS) production by a thermophilic isolate of Bacillus licheniformis. Firstly, the application of Plackett-Burman design and analysis indicated that sucrose, KH2PO4 and NaCl have significant positive effects, whereas NaNO3 showed a significant negative effect on EPS production. Then, experimental data of a central-composite design was leveraged by RSM and the h-SVM to optimize these significant variables. SVM showed better performance in data modeling (R2 = 0.9874, Root-mean-square error = 0.693) compared to conventional RSM. Subsequently, the SVM model was coupled with optimization algorithms, Teaching–learning-based optimization (TLBO) and Particle swarm optimization (PSO). As a result, a maximum EPS concentration of 64 ± 1.1 g L‐ was obtained from h-SVM predicted model (SVM-PSO, SVM-TLBO), which is about a 5-fold increase in EPS production to that obtained with the un-optimized medium. Within our knowledge, this is the first report on the implementation of SVM-TLBO to improve the production of bioproducts like bacterial EPS while comparing it with other techniques like SVM-PSO and RSM.

Predictive capability evaluation and optimization of Pb(II) removal by reduced graphene oxide-based inverse spinel nickel ferrite nanocomposite

Pb(II) is a heavy metal that is a prominent contaminant in water contamination. Among the different pollution removal strategies, adsorption was determined to be the most effective. The adsorbent and its type determine the adsorption process's efficiency. As part of this effort, a magnetic reduced graphene oxide-based inverse spinel nickel ferrite (rGNF) nanocomposite for Pb(II) removal is synthesized, and the optimal values of the independent process variables (like initial concentration, pH, residence time, temperature, and adsorbent dosage) to achieve maximum removal efficiency are investigated using conventional response surface methodology (RSM) and artificial neural networks (ANN). The results indicate that the initial concentration, adsorbent dose, residence time, pH, and process temperature are set to 15 mg/L, 0.55 g/L, 100 min, 5, and 30 °C, respectively, the maximum removal efficiency (99.8%) can be obtained. Using the interactive effects of process variables findings, the adsorption surface mechanism was examined in relation to process factors. A data-driven quadratic equation is derived based on the ANOVA, and its predictions are compared with ANN predictions to evaluate the predictive capabilities of both approaches. The R2 values of RSM and ANN predictions are 0.979 and 0.991 respectively and confirm the superiority of the ANN approach.

Determination of the Optimal Neural Network Transfer Function for Response Surface Methodology and Robust Design

Response surface methodology (RSM) has been widely recognized as an essential estimation tool in many robust design studies investigating the second-order polynomial functional relationship between the responses of interest and their associated input variables. However, there is scope for improvement in the flexibility of estimation models and the accuracy of their results. Although many NN-based estimations and optimization approaches have been reported in the literature, a closed functional form is not readily available. To address this limitation, a maximum-likelihood estimation approach for an NN-based response function estimation (NRFE) is used to obtain the functional forms of the process mean and standard deviation. While the estimation results of most existing NN-based approaches depend primarily on their transfer functions, this approach often requires a screening procedure for various transfer functions. In this study, the proposed NRFE identifies a new screening procedure to obtain the best transfer function in an NN structure using a desirability function family while determining its associated weight parameters. A statistical simulation was performed to evaluate the efficiency of the proposed NRFE method. In this particular simulation, the proposed NRFE method provided significantly better results than conventional RSM. Finally, a numerical example is used for validating the proposed method.

State of the Art Review on Statistical Modelling and Optimization of Bioenergy Production from Oil Seeds

The article captures a brief state-of-the-art encompassing statistical modelling and optimization of bioenergy production from oil seeds. Exploring the research space of bioenergy production from biomass oil, various numerical approaches have been employed to model and optimize the process parameters for maximum response yields. Various performance studies have also been carried out to evaluate the predictive capabilities of emerging Artificial intelligence (AI) algorithms on modelling of bioenergy production from seed oil using the conventional Response Surface Methodology (RSM) as a reference base. For the records, the precision of measurement, management of uncertainties, more accurate data analysis and prediction are techniques which these methodologies have the capacity to do. The commonly used techniques for optimization and modelling studies of bioenergy from biomass oil and the analysis of their usage according to their performance metrics are detailed in the body of this work. Aside the relative limitation of RSM models in highly non-linear processes as compared to the robustness of AI models, RSM models still continue to hover on large scale applications when it comes to modelling and optimization studies on bioenergy production from oil-rich plant seeds. In the era of big data analysis in relation to the test and measure of analytical performances and prediction ability, AI models have begun to gain attention in the last two decades owing to their better estimation capabilities and ability to handle more data points in real-time prediction of bioenergy production. This study has been able to show successful applications of both RSM and AI models in bioenergy field, with a pointer to further adoption of the later in more related studies as a result of its relative advantages.

Multi-objective optimization of peel and shear strengths in ultrasonic metal welding using machine learning-based response surface methodology.

Ultrasonic metal welding (UMW) is a solid-state joining technique with varied industrial applications. Despite of its numerous advantages, UMW has a relative narrow operating window and is sensitive to variations in process conditions. As such, it is imperative to quantitatively characterize the influence of welding parameters on the resulting joint quality. The quantification model can be subsequently used to optimize the parameters. Conventional response surface methodology (RSM) usually employs linear or polynomial models, which may not be able to capture the intricate, nonlin-ear input-output relationships in UMW. Furthermore, some UMW applications call for simultaneous optimization of multiple quality indices such as peel strength, shear strength, electrical conductivity, and thermal conductivity. To address these challenges, this paper develops a machine learning (ML)- based RSM to model the input-output relationships in UMW and jointly optimize two quality indices, namely, peel and shear strengths. The performance of various ML methods including spline regression, Gaussian process regression (GPR), support vector regression (SVR), and conventional polynomial re-gression models with different orders is compared. A case study using experimental data shows that GPR with radial basis function (RBF) kernel and SVR with RBF kernel achieve the best prediction accuracy. The obtained response surface models are then used to optimize a compound joint strength indicator that is defined as the average of normalized shear and peel strengths. In addition, the case study reveals different patterns in the response surfaces of shear and peel strengths, which has not been systematically studied in the literature. While developed for the UMW application, the method can be extended to other manufacturing processes.

Multi-objective optimization of a novel crude lipase-catalyzed fatty acid methyl ester (FAME) production using low-order polynomial and Kriging models

ABSTRACTIn this paper, conventional response surface methodology (RSM) based on low-order polynomials and an alternative Kriging-based method are used for the model-based single and multi-objective optimization of fatty-acid methyl ester (FAME) production catalyzed by a novel crude lipase from the yeast Cryptococcus diffluens (D44). The coefficient of determination for the two modeling approaches was calculated as 0.97 for the Kriging method, and 0.86 for RSM; showing a more reliable representation of experimental data by Kriging. Both models were used to perform single (maximizing FAME titer and temporal productivity separately) and multi-objective (maximizing FAME titer and temporal productivity simultaneously) optimizations of four important operating conditions (reaction time and temperature; amount of crude enzyme; and volume of methanol used). In all cases, the highest temperature considered (60°C) gave the best results. A reduction of reaction time in half was seen to be necessary to achieve optimum productivity compared to titer, when the two objectives were considered separately. The observed trade-off between the two objectives was quantified via multi-objective optimization using Pareto-front analysis.

Synthesis and characterization of MFI zeolite nanosheets for optimizing the dominant operational conditions of methanol to propylene process

Nanosheets of MFI zeolite were successfully synthesized by hydrothermal method and were properly characterized by XRD, FE-SEM, EDX-dot mapping, TEM, FT-IR, N2 adsorption/desorption, NH3-TPD, ICP-AES and TGA techniques. The characterizations of MFI zeolite nanosheets suitably confirmed well synthesized MFI nanostructure, superior specific surface area (695.96m2.g-1) and great mesopores volume. Currently, the major concern in commercialized technologies of methanol to propylene (MTP) process is to discover the best operational conditions for maximizing propylene selectivity. Therefore, in present investigation the effectiveness of more significant operational conditions such as reaction temperature, methanol molar percent in feedstock and methanol WHSV on propylene selectivity was considered. For first time, conventional response surface methodology and an artificial neural network model coupled with genetic algorithm were appropriately employed to optimize MTP operational conditions. For providing the data bank of optimization, MTP reaction was carried out over the synthesized MFI zeolite nanosheets at various operational conditions. It was concluded that neuro-genetic method showed more reliable performance in optimizing MTP operational conditions with R2 value of 0.9998. Furthermore, time on stream examination was conducted over MFI zeolite nanosheets at optimal operational conditions which were suggested by neuro-genetic approach. At optimal operational conditions the catalytic life-time was sufficiently enhanced (101h).

Process optimization and adsorption modeling using activated carbon derived from palm oil kernel shell for Zn (II) disposal from the aqueous environment using differential evolution embedded neural network

Presence of toxic heavy metals like Zn (II) in an aqueous medium is growing environmental problem. Researchers are continuously thriving to develop efficient techniques to remove pollutants from wastewater and industrial effluents. Adsorption is found to be an effective treatment technique, but its applications in process industries limit itself due to the high cost of adsorbent. In this regard, a low-cost adsorbent produced from palm oil kernel shell based agricultural waste were examined for its efficiency to remove Zn (II) from the aqueous effluent. The performance of adsorption technique depends on the independent process variables. Therefore, the influence of parameters like initial solution concentration, pH, residence time, activated carbon dosage and process temperature on the removal of Zn (II) from the batch adsorption process were studied systematically. The optimal values of independent process variables to achieve maximum removal efficiency were studied using conventional response surface methodology (RSM) and data-driven modeling technique like artificial neural network (ANN). The meta-heuristic differential evolution optimization is embedded on to the ANN architecture to optimize the search space of neural network. The optimized trained neural network depicts the testing data and validation data with R2 equal to 0.995 for both cases. The multiple outcomes of this study indicate the superiority of ANN-DE based model predictions over the traditional quadratic model predictions provided by RSM.

Process optimization and adsorption modeling of Pb(II) on nickel ferrite-reduced graphene oxide nano-composite

The presence of toxic heavy metals such as Pb(II) in aqueous environments is causing adverse health risks to human beings, motivating researchers to develop efficient techniques to remove pollutants from potable and effluent water. Adsorption was found to be an efficient technique among the various pollutant removal techniques. Efficiency of the adsorption process depends on the adsorbent and its nature. In this regard, a synthesized nickel ferrite-reduced graphene oxide (NFRGO) nano-composite is developed and utilized as an adsorbent to remove heavy metal ions. Since the performance of the adsorption technique depends on independent process variables, the influence of parameters such as the initial solution concentration, adsorbent dosage, and contact residence time on the removal of Pb(II) by NFRGO using a batch adsorption process are systematically studied in this work. Based on the design of experiments approach and central composite design (CCD), twenty experimental runs are performed with each process variable segregated in the experimental range. The optimal values of the independent process variables to achieve maximum removal efficiency are examined using conventional response surface methodology (RSM). A quadratic model, which consists of a first-order and second-order degree regressive model is developed using the analysis of variance (ANOVA) and RSM - CCD framework. Based on the desirability index, the optimum values were found to be an initial concentration of 18.38mg/L, an adsorbent dose of 0.55g/L, and a contact residence time of 83min in order to achieve 99% removal of Pb(II) ions with 0.953 desirability.

A nested-ANN prediction model for surface roughness considering the effects of cutting forces and tool vibrations

Abstract This paper demonstrates a nested-ANN (Artificial Neural Network) model predicting surface roughness ( R a ). The special ANN includes enclosed-ANNs and an output-ANN. The enclosed-ANN models use cutting parameters as inputs to predict the values of cutting forces and tool vibrations respectively, and then forward all outputs to the output-ANN model. Subsequently, the output-ANN adopts the forward values and cutting parameters as inputs to predict R a . To verify the effectiveness of the nested-ANN model, it is compared with mathematical and statistical models based on conventional ANN and RSM (Response Surface Methodology) using the same experimental data. The results show that the nested-ANN uses less input variables to obtain superior prediction accuracy than other models. Additionally, the statistical analyses show that R a is mostly affected by the feed rate and has a signification correlation with the feed rate, the cutting force in both radial and tangential directions as well as the tool vibrations.

Optimization of conditions for hydrogen production from complex dairy wastewater by anaerobic sludge using desirability function approach

Present study deals with the multiple-response optimization for biohydrogen production using anaerobic sludge and outstanding approach to overcome the drawbacks of conventional response surface methodology (RSM). Dairy wastewater was used as source in batch fermentation was followed for this study. Response surface methodology (RSM), based on a three level, four variable Box–Behnken design, was employed to obtain the best possible combination of substrate concentration, pH, COD/N ratio and COD/P ratio for maximum H2 yield (HY) and specific hydrogen production rate (SHPR). Experimental data were evaluated by applying RSM integrating a desirability function approach. The optimum H2 yield and SHPR conditions were: substrate concentration 15.3 g COD/L, pH 5.5, COD/N ratio 100.5 and COD/P ratio 120 with maximum overall desirability D of 0.94. The confirmation experiment under these optimal condition showed a HY and SHPR of 13.54 mmol H2/g COD and 29.91 mmol H2/g-VSS.d, respectively. This was only 0.22% and 0.20%, respectively, different from the predicted values, suggesting that the desirability function approach with RSM was a useful technique to get the maximum H2 yield and SHPR simultaneously.

Effectiveness and limits of response surface methodology in application to discrete and stochastic simulation of manufacturing plants

The aim of the Response Surface Methodology (RSM), originally designed for chemical, physical and biological applications and experiments, but then extended also to simulations in the industrial field, is the construction of reliable response surfaces characterized by high adherence to the experimental data describing the reality being studied [1]-[2]-[11]. In order to achieve this result the most important scholars in this fields focus their attention on experimental projects capable of providing regression meta-models, that fit well the initial experimental data and have a sort of stability of the width of the confidence intervals on the average response and the prediction intervals [11]. At the same time it is also necessary to choose experimental projects allowing a good estimate of the Experimental Error divided into the two components called pure error and lack of fit by the literature on this subject. One of the key assumptions of the conventional RSM approach is that the Experimental Error strictly linked to the system under examination is fixed and it cannot be controlled by experimenters. Nevertheless, for applications in complex industrial plants, it is not possible to carry out the experimental phase directly on the real system and the object of study needs to be transferred into a simulation model [3]-[8]. Therefore, this assumption becomes meaningless. The transfer of the real physical model into a simulation model implies a substantial change in the nature of the Experimental Error, which, compared to the one in the conventional experiments, changes from fixed to dependent on the length of the simulation run and hence time-variant [4]-[6]-[7]-[9]-[10]. Therefore, it is no longer enough to try to improve the quality of the response surface using the traditional RSM concepts of Optimal Variance, Orthogonality and Rotatability, as these are strictly dependent on the value of the experimental error. In the applications on the discrete and stochastic simulation models experimenters must hence try to reduce, each time it is possible, the magnitude of the simulation model’s outgoing Experimental Error extending the simulation run and consequently the computation time. Only once this has been done will it be possible to fine-tune the study examining the response surface to find the aforementioned properties whose influence on the quality of the outgoing surface is substantially lower compared to the reduction of the Experimental Error. This article illustrates an application to an industrial case that not only shows the validity of what is affirmed, but also highlights other limits linked to the use of the conventional RSM approach to simulation experiments on complex industrial plants. These limits challenge the efficacy of linear regression meta-models as descriptors of the relation between the independent and dependent variables considered.

An Engineering Approach to Improve the Stamping Robustness of High Strength Steels

High strength steels (HSSs) are one of the light-weight sheet metals well suited for reducing vehicle weight due to their higher strength-to-weight ratio. However, HSS tend to have bigger variations in their mechanical properties due to more complex rolling techniques involved in the steel-making process. Such uncertainties, when combined with variations in the process parameters such as friction and blank holder force, pose a significant challenge in maintaining the robustness of HSS sheet metal stamping. The paper presents a systematic and robust approach, combining the power of the finite element method and stochastic statistics to decrease the sensitivity of HSS stamping in the presence of above-mentioned uncertainties. First, the statistical distribution of sheet metal properties of selected HSS is characterized from a material sampling database. Then a separate interval adaptive response surface methodology (RSM) is applied in modeling sheet metal stamping. The new method significantly improves the model accuracy when compared with the conventional RSM within a single interval. Finally, the Monte Carlo method is employed to simulate the stochastic response of material/process variations to stamping quality and to provide optimal process parameter designs to reduce the sensitivity of these effects. The experiment with the obtained optimal process design demonstrates the improvements of stamping robustness using small-batch experiments.

Global Optimization With Multivariate Adaptive Regression Splines

This paper presents a novel procedure for approximating the global optimum in structural design by combining multivariate adaptive regression splines (MARS) with a response surface methodology (RSM). MARS is a flexible regression technique that uses a modified recursive partitioning strategy to simplify high-dimensional problems into smaller yet highly accurate models. Combining MARS and RSM improves the conventional RSM by addressing highly nonlinear high-dimensional problems that can be simplified into lower dimensions, yet maintains a low computational cost and better interpretability when compared to neural networks and generalized additive models. MARS/RSM is also compared to simulated annealing and genetic algorithms in terms of computational efficiency and accuracy. The MARS/RSM procedure is applied to a set of low-dimensional test functions to demonstrate its convergence and limiting properties.

Using surrogate models and response surfaces in structural optimization ? with application to crashworthiness design and sheet metal forming

The aim of this paper is to determine if the Space Mapping technique using surrogate models together with response surfaces is useful in the optimization of crashworthiness and sheet metal forming. In addition, the efficiency of optimization using Space Mapping will be compared to traditional structural optimization using the Response Surface Methodology (RSM). Five examples are used to study the algorithm: one optimization of an analytic function and four structural optimization problems. All examples are constrained optimization problems. In all examples, the algorithm converged to an improved design with all constraints fulfilled, even when a conventional RSM optimization failed to converge. For the crashworthiness design problems, the total computing time for convergence was reduced by 53% using Space Mapping compared to conventional RSM. For the sheet metal forming problems the total computing time was reduced by 63%. The conclusions are that optimization using Space Mapping and surrogate models can be used for optimization in crashworthiness design and sheet metal forming applications with a significant reduction in computing time.