Multivariate Optimization Method Research Articles

Technological Supportof Parts’ Fatigue Life by Modeling Their Turning Process

Background. The issue of technological support of the required part’s material fatigue life by creating a mathematical model of the finishing turning process is considered. This model includes, as a target function, the maximum process productivity and the set of feed constraints and cutting speeds, permissible force and cutting power, machining precision, tool stability, surface roughness, part fatigue life, and cutting condition optimization.Objective. The aim of the paper is to provide technological support for the required part fatigue life by determining the rational cutting conditions, taking into account the properties of the processed material, and to develop appropriate methodological recommendations.Methods. The objective of the research is achieved by creating a mathematical model of the finishing turning process, determining the coefficients of the generalized materials’ characteristics of the classification group taking into account the properties of the processed material, further model optimization by the method of sliding admission and the definition of rational cutting conditions.Results. The mathematical model of the finishing turning process, which takes into account the characteristics of the processed material and the properties of the technological processable system, was created. Relative coefficients of the materials’ generalized characteristics of the structural alloyed chromium steels group are calculated. A multivariate optimization method of the developed mathematical model is proposed.Conclusions. The proposed methodological recommendation of technological support of a part’s material fatigue life based on a mathematical model of the finishing turning process, which includes the maximum process productivity and set of restrictions as the target function, which allows determining the rational cutting condition by the chosen method of nonlinear optimization.

Multivariate optimization of mebeverine analysis using molecularly imprinted polymer electrochemical sensor based on silver nanoparticles

Thin film of a moleculary imprinted polymer (MIP) based on electropolymerization method with sensitive and selective binding sites for mebeverine (MEB) was developed. This film was cast on pencil graphite electrode (PGE) by electrochemical polymerization in solution of pyrrole (PY) and template MEB via cyclic voltammetry scans and further electrodeposition of silver nanoparticles (AgNPs). Several parameters controlling the performance of the silver nano particles MIP pencil graphite electrode (AgNPs-MIP-PGE) including concentration of PY(mM) concentration of mebeverine (mM), number of cycles in electropolymerization, scan rate of CV process (mV. s−1), deposition time of AgNPs on to the MIP surface (s), stirring rate of loading solution (rpm), electrode loading time (min), pH of Britton–Robinson Buffer (BRB) solution were examined and optimized using multivariate optimization methods such as Plackett–Burman design (PBD) and central composite design (CCD). Two dynamic linear ranges of concentration for the MIP sensor were obtained as. 1 × 10 −8 to 1 × 10 −6 and 1 × 10 −5 to1 × 10−3 M with the limit of detection (LOD) of 8.6 × 10 −9M (S/N = 3). The proposed method was successfully intended for the determination of MEB in real samples (serum, capsule). The sensor was showed highly reproducible response (RSD 1.1%) to MEB concentration.

Effects of fluid type and pressure order on performance of convergent–divergent nozzles: An efficiency model for supersonic separation

AbstractA deep analysis on the hydrodynamics of convergent–divergent nozzles is performed by changing the working conditions and fluid type. The nozzle geometry lowers the temperature of the flowing fluid, which contributes to condensation and phase change. The focus of this paper is to evaluate the nozzle performance and cooling capacity in terms of temperature, pressure, and gas type in a fixed geometry of Sajben Laval nozzle. The analysis has been conducted via a 2‐dimensional turbulent computational fluid dynamics simulation for illustrating the behavior of the fluid. A criterion for the nozzle performance is provided by prediction of exact shock wave position. An investigation on 6 different gas types demonstrates that heat capacity and thermal conductivity are the most rolling fluid properties of the nozzle performance. Furthermore, it is found that the shock wave position is unchanged during alteration of fluid type or pressure scale.A new model is provided for prediction of convergent–divergent nozzle performance in the supersonic conditions for dehydration of natural gas as a well‐known industrial application of the nozzles. The model is developed by the genetic algorithm as a multivariable optimization method.

Application of a multi-variable optimization method to determine lift-up design for optimum wind comfort

The lift-up building design has been demonstrated to provide favorable wind comfort, but there is a lack of investigation on optimum wind comfort condition. This study coupled computational fluid dynamics (CFD) technique and response surface methodology (RSM) to determine the most desirable wind comfort around an isolated building with lift-up design. A multi-variable optimization method is proposed to determine optimum wind comfort and the corresponding lift-up design variables, namely, lift-up height (HL), core aspect ratio (AR) and core number (N). To better illustrate wind comfort around the building, the wind comfort in the lift-up area and the podium area are investigated separately. The Detached Eddy Simulation (DES) approach is employed throughout the whole CFD simulation process. The quality and goodness of the established RSM models are examined by analysis of variance and genetic algorithm is applied to generate optimal design solution. The generated results illustrate good performance of the established RSM model. Results show that the optimum wind comfort is obtained when HL is 8 m, AR is 10%, and N is 6. The lift-up core aspect ratio is subsequently found to have greatest effect on wind comfort among the three design variables in both the lift-up area and the podium area. In addition, the proposed method is applicable to other similar environmental design conditions and the outcomes of study can also be of great value in the improvement of wind comfort in compact urban cities.

Determination of bisphenols with estrogenic activity in plastic packaged baby food samples using solid-liquid extraction and clean-up with dispersive sorbents followed by gas chromatography tandem mass spectrometry analysis

Bisphenols (BPs) are a family of chemicals with known endocrine disrupting activity. Bisphenol A (BPA) is the most representative prototype of this group of chemicals. Recently, the use of BPA, a prototype of endocrine disruptors, has been reduced and replaced with structural analogs due to its negative effects on both the environment and consumers. In this work, a new method is presented for the determination of seven BPs, with estrogenic activity in ready-to-eat plastic packaged baby foods. The procedure involves the isolation of the analytes using solid-liquid phase extraction with acetonitrile followed by a clean-up step with a mixture of dispersive-SPE sorbents (C18 and PSA) and magnesium sulphate, to reduce matrix effect from proteins, sugars and lipids. Extraction parameters were optimized using multivariate optimization methods. The compounds were detected and quantified by gas chromatography tandem mass spectrometry (GC–MS/MS). The limits of quantification were between 0.1 and 1.2ngg−1 for the studied analytes. The method was validated using matrix–matched calibration and recovery assays with spiked samples. Recovery rates were between 91% and 110% and % RSD was lower than 13% in all cases. The method has been successfully applied for the determination of these endocrine disrupting chemicals (EDCs) in samples of a novel type of food consumed by pre-schoolers. This is the first study to analyze EDCs in plastic packaged foods consumed by this target group.

Gas chromatographic-mass spectrometric analysis of urinary volatile organic metabolites: Optimization of the HS-SPME procedure and sample storage conditions

Non-targeted metabolomics research of human volatile urinary metabolome can be used to identify potential biomarkers associated with the changes in metabolism related to various health disorders. To ensure reliable analysis of urinary volatile organic metabolites (VOMs) by gas chromatography-mass spectrometry (GC-MS), parameters affecting the headspace-solid phase microextraction (HS-SPME) procedure have been evaluated and optimized. The influence of incubation and extraction temperatures and times, coating fibre material and salt addition on SPME efficiency was investigated by multivariate optimization methods using reduced factorial and Doehlert matrix designs. The results showed optimum values for temperature to be 60°C, extraction time 50min, and incubation time 35min. The proposed conditions were applied to investigate urine samples’ stability regarding different storage conditions and freeze-thaw processes. The sum of peak areas of urine samples stored at 4°C, −20°C, and −80°C up to six months showed a time dependent decrease over time although storage at −80°C resulted in a slight non-significant reduction comparing to the fresh sample. However, due to the volatile nature of the analysed compounds, more than two cycles of freezing/thawing of the sample stored for six months at −80°C should be avoided whenever possible.

Application of multivariate optimization method in nanomolar simultaneous determination of morphine and codeine in the presence of uric acid using a glassy carbon electrode modified with a hydroxyapatite-Fe3O4 nanoparticle/multiwalled carbon nanotubes composite

The electrochemical oxidation of morphine (MO) and codeine (COD) has been investigated by the application of a novel glassy carbon electrode modified with a hydroxyapatite-Fe3O4 nanoparticles/multiwalled carbon nanotubes composite (HA-FeNPs-MWCNTs/GCE). The modified electrode worked as an efficient sensor for simultaneous determination of MO and COD in the presence of uric acid. Response surface methodology was utilized to optimize the voltammetric response of the modified electrode for the determination of MO and COD. The amount of HA-FeNPs in the modifier matrix (%HA-FeNPs), the solution pH and the accumulation time were chosen as the three important operating factors through the experimental design methodology. The central composite design as a response surface approach was applied for obtaining the optimum conditions leading to maximum oxidation peak currents for MO and COD. The differential pulse voltammetry results showed that the obtained anodic peak currents were linearly proportional to concentration in the range of 0.08–32 µM with a detection limit (S/N = 3.0) of 14 nM for MO and in the range of 0.1–28 µM and with a detection limit of 22 nM for COD. The proposed method was successfully applied to determine these compounds in human urine and blood serum samples.

A statistical multivariable optimization method using improved orthogonal algorithm based on large data

ABSTRACTMultivariable optimization under large data environment concerns with how to reliably obtain a set of optimization results from a mass of data that influence the object function complexly. This is an important issue in statistical calculation because the complexity between variable parameters leads to repeated statistical calculation analysis and a significant amount of data waste. A statistical multivariable optimization method using improved orthogonal algorithm based on large data is proposed. Considering the optimization problem with multi-parameters under large data environment, a multi-parameter optimization model used for improved orthogonal algorithm is established based on large data. Furthermore, an extensive simulation study on temperature field distribution of anti-/de-icing component was conducted to verify the validity of the statistical calculation analysis optimization method. The optimized temperature field distribution meets the anti-/de-icing requirements through numerical simulation. Simulation results show that the optimization effect is more evident and accurate than the non-optimized temperature distribution with the optimized results of the proposed method. Results verify the effectiveness of the proposed method.

Effects of structural defects and polarization charges in InGaN-based double-junction solar cell

The performance of a double heterojunction solar cell based on Indium Gallium Nitride (InGaN) including a tunnel junction was simulated. The most challenging aspects of InGaN solar cells development being the crystal polarization and structural defects detrimental effects, their impact on the solar cell performances has been investigated in detail. The solar cell simulation was performed using physical models and InGaN parameters extracted from experimental measurements. The optimum efficiency of the heterojunction solar cell was obtained using a multivariate optimization method which allows to simultaneously optimize eleven parameters. The optimum defect free efficiency obtained is 24.4% with a short circuit current JSC=12.92mA/cm2, an open circuit voltage VOC=2.29V and a fill factor FF=82.55%. The performances evolution as functions of the polarization and the defects types and parameters was studied from their maximum down to as low as a 2% efficiency.

Removal of Pb (II) and Cu (II) from aqueous solutions by NaA zeolite coated magnetic nanoparticles and optimization of method using experimental design

Magnetic nanozeolites were synthesized through a new method and their abilities were evaluated toward adsorption of Pb (II) and Cu (II) from aqueous solutions. A multivariate optimization method using central composite design (CCD) was designed to obtain the best conditions of the adsorption. NaA zeolite because of its high affinity for the investigated ions was selected and was coated onto magnetic nanoparticles. The morphology and characterization of the adsorbent were studied by different techniques such as SEM imaging, FT-IR spectrometry and AGFM measurements. The most favorable conditions for Pb (II) removal was achieved in pH = 5.6 and 8 mg sorbent in a 750 mg L−1 Pb (II) aqueous solution when removal process was done in 8 min. Maximum removal of Cu (II) was attained under the optimal conditions with pH = 2.9, adsorbent amounts 16 mg, initial ion concentration 22.1 mg L−1 and sonication time 24 min. The predicted model of Design Expert software was indicated the significant main factors and the possible interactions of the factors. Based on the results, some interactions between the effecting factors were found for Pb (II) and Cu (II) ions and R2 was attained 0.99 and 0.97 for these ions respectively. Our findings clearly showed that magnetic nanozeolites can be made successfully by entering silica-coated magnetic nanoparticles into the initial solution of zeolite synthesis. Also, the coating of the magnetic nanoparticles with zeolite NaA improved effectively the adsorption of the selected ions while we can conveniently separate the nanozeolites with an external magnet.

Production planning of new and remanufacturing products in hybrid production systems

As recycling and remanufacturing become more common, companies may simultaneously operate two production processes for producing new and remanufactured products, and these are called hybrid production systems. However, such systems bring greater managerial complexity, and thus the determination of an appropriate operation strategy to set the recycling ratio of the old product and allocate the capacities of production to the new and remanufactured products is an important issue. This study mainly focuses on a hybrid production system in which the determination of the optimal operation strategy is based on the consideration of the related costs, the uncertainty about recycling, the demand substitution, the capacity limitation, and the component durability. Moreover, the competition between new and remanufactured products is also considered to construct the proposed model, with the aim of maximizing the profit of the manufacturer. The mathematical model are designed from several possible scenarios to estimate the expected revenue and related costs of the hybrid production system. A Hessian matrix and multivariate optimization methods are used to examine and obtain the optimal solution of the mathematical model. The impacts of related parameters are investigated by utilizing sensitivity analysis, and the results of this can give insights to the related production managers.

Process Parameters Optimization for Producing AA6061/TiB2 Composites by Friction Stir Processing

Abstract Friction stir processing (FSP) is solid state novel technique developed to refine microstructure and to improve the mechanical properties and be used to fabricate the aluminium alloy matrix composites. An attempt is made to fabricate AA6061/TiB2 aluminium alloy composite (AMCs) and the influence of process parameters like rotational speed, transverse feed, axial load and percentage reinforcement on microstructure and mechanical properties were studied. The microstructural observations are carried out and revealed that the reinforcement particles (TiB2) were uniformly dispersed in the nugget zone. The Tensile strength and Hardness of composites were evaluated. It was observed that tensile strength, and hardness were increased with increased the rotational speed and percentage reinforcement of particles. The process parameters were optimized using Taguchi analysis (Single Variable) and Grey analysis (Multi Variable). The most influential parameter was rotational speed in single variable method and multi variable optimization method. The ANOVA also done to know the percentage contribution of each parameter.

Removal of Nickel (II) and Cobalt (II) from Wastewater Using Vinegar-Treated Eggshell Waste Biomass

The use of waste materials as low-cost adsorbents is attractive due to their contribution in the reduction of costs for waste disposal, therefore contributing to environmental protection and most importantly, offers an attractive potential alternative to their conventional methods of removal of toxic ions from wastewater. Eggshells are naturally occurring and an abundant biomass that has proven to offer an economic solution for toxic ions removal. The eggshell biomass was treated with acetic acid (vinegar). Nickel (II) and Cobalt (II) ions were selected as model ions to demonstrate the potential of eggshell waste in removing excess toxic heavy metal ions from wastewater. All the experiments were carried out in batch process with laboratory prepared samples. Multivariate optimization method was used to identify factors affecting adsorption. These factors included metal ion concentration, pH, contact time and biomass dosage on removal of nickel and cobalt from wastewater effluent was investigated. Two-level fraction factorial and central composite design were used for optimization methods. Fourier Transform Infrared Spectroscopy, Raman Spectroscopy, and Scanning Electron Microscopy coupled with Energy-dispersive X-ray spectroscopy were used to study physical properties of the waste material. The percentage removal of Nickel (II) and Cobalt (II) was 78.70 ± 1.02 and 76.53 ± 1.21 respectively. Vinegar-treated eggshells were proposed as eco-friendly, cheap, easily available and an efficient method for removal of heavy metals from the environment.

Assessing pretreatment reactor scaling through empirical analysis.

BackgroundPretreatment is a critical step in the biochemical conversion of lignocellulosic biomass to fuels and chemicals. Due to the complexity of the physicochemical transformations involved, predictively scaling up technology from bench- to pilot-scale is difficult. This study examines how pretreatment effectiveness under nominally similar reaction conditions is influenced by pretreatment reactor design and scale using four different pretreatment reaction systems ranging from a 3 g batch reactor to a 10 dry-ton/days continuous reactor. The reactor systems examined were an automated solvent extractor (ASE), steam explosion reactor (SER), ZipperClave®Reactor (ZCR), and large continuous horizontal screw reactor (LHR). To our knowledge, this is the first such study performed on pretreatment reactors across a range of reaction conditions and at different reactor scales.ResultsThe comparative pretreatment performance results obtained for each reactor system were used to develop response surface models for total xylose yield after pretreatment and total sugar yield after pretreatment followed by enzymatic hydrolysis. Near- and very-near-optimal regions were defined as the set of conditions that the model identified as producing yields within one and two standard deviations of the optimum yield. Optimal conditions identified in the smallest scale system (the ASE) were within the near-optimal region of the largest scale reactor system evaluated. The maximum total sugar yields for the ASE and LHR were 95,%, while 89,% was the optimum observed in the ZipperClave.ConclusionsThe optimum condition identified using the automated and less costly to operate ASE system was within the very-near-optimal space for the total xylose yield of both the ZCR and the LHR, and was within the near-optimal space for total sugar yield for the LHR. This indicates that the ASE is a good tool for cost effectively finding near-optimal conditions for operating pilot-scale systems. Additionally, using a severity factor approach to optimization was found to be inadequate compared to a multivariate optimization method. Finally, the ASE and the LHR were able to enable significantly higher total sugar yields after enzymatic hydrolysis relative to the ZCR, despite having similar optimal conditions and total xylose yields. This underscores the importance of mechanical disruption during pretreatment to improvement of enzymatic digestibility.Electronic supplementary materialThe online version of this article (doi:10.1186/s13068-016-0620-0) contains supplementary material, which is available to authorized users.

Computational design and multivariate optimization of an electrochemical metoprolol sensor based on molecular imprinting in combination with carbon nanotubes

This work describes the development of an electrochemical sensor based on a new molecularly imprinted polymer for detection of metoprolol (MTP) at ultra-trace level. The polypyrrole (PPy) was electrochemically synthesized on the tip of a pencil graphite electrode (PGE) which modified whit functionalized multi-walled carbon nanotubes (MWCNTs). The fabrication process of the sensor was characterized by cyclic voltammetry (CV) and the measurement process was carried out by differential pulse voltammetry (DPV). A computational approach was used to screening functional monomers and polymerization solvent for rational design of molecularly imprinted polymer (MIP). Based on computational results, pyrrole and water were selected as functional monomer and polymerization solvent, respectively. Several significant parameters controlling the performance of the MIP sensor were examined and optimized using multivariate optimization methods such as Plackett–Burman design (PBD) and central composite design (CCD). Under the selected optimal conditions, MIP sensor was showed a linear range from 0.06 to 490 μmol L−1 MTP, a limit of detection of 2.88 nmol L−1, a highly reproducible response (RSD 3.9%) and a good selectivity in the presence of structurally related molecules. Furthermore, the applicability of the method was successfully tested with determination of MTP in real samples (tablet, and serum).

Multivariate optimization of a capillary zone electrophoresis assay method for simultaneous quantification of metformin and vildagliptin from a formulation

ABSTRACTA rapid, accurate, precise, and optimized capillary zone electrophoresis assay was established and validated for the simultaneous quantification of metformin and vildagliptin in tablets. The electrophoretic separation was achieved on an untreated bonded silica capillary with a background electrolyte comprising 25 mM of borate buffer at pH 7.5 at 207 nm. The concentration of the buffer and the pH of BGE were optimized using the multivariate optimization method for determining the retention time and peak area. Furthermore, the sample injection time, capillary oven temperature, and applied voltage were optimized. The capillary zone electrophoresis technique was validated for all required parameters as per the International Conference on Harmonization recommendations. The linearity ranged in the concentrations of 5–500 µg/mL and 5–100 µg/mL with the limit of detections of 0.22 µg/mL and 0.40 µg/mL for metformin and vildagliptin, respectively. In addition, the percent relative standard error for repeatability and inter-day precision was within the acceptable range. The mean recoveries determined by the capillary zone electrophoresis method were 99.2% and 100.4% for metformin and vildagliptin, respectively. Finally, the capillary zone electrophoresis process was effectively used for the assays of metformin and vildagliptin in their solid dosage form, and statistical outcomes were in agreement with the outcomes of the previously validated RP-HPLC method.

Optimizing Multivariate Performance Measures from Multi-View Data

To date, many machine learning applications have multiple views of features, and different applications require specific multivariate performance measures, such as the F-score for retrieval. However, existing multivariate performance measure optimization methods are limited to single-view data, while traditional multi-view learning methods cannot optimize multivariate performance measures directly. To fill this gap, in this paper, we propose the problem of optimizing multivariate performance measures from multi-view data, and an effective method to solve it. We propose to learn linear discriminant functions for different views, and combine them to construct an overall multivariate mapping function for multi-view data. To learn the parameters of the linear discriminant functions of different views to optimize a given multivariate performance measure, we formulate an optimization problem. In this problem, we propose to minimize the complexity of the linear discriminant function of each view, promote the consistency of the responses of different views over the same data points, and minimize the upper boundary of the corresponding loss of a given multivariate performance measure. To optimize this problem, we develop an iterative cutting-plane algorithm. Experiments on four benchmark data sets show that it not only outperforms traditional single-view based multivariate performance optimization methods, but also achieves better results than ordinary multi-view learning methods.

1,4-dihydroxyanthraquinone electrochemical sensor based on molecularly imprinted polymer using multi-walled carbon nanotubes and multivariate optimization method

The present work describes development of a simple and cost-effective electrochemical sensor for determination of 1,4-dihydroxyanthraquinone (1,4-DAQ) based on molecularly imprinted polypyrrole (PPY). The molecularly imprinted polymer (MIP) was electrochemically synthesized onto surface of multi-walled carbon-nanotubes-modified pencil graphite electrode (MWCNs–PGE). A computational approach was employed to select the best functional monomer for rational design of MIP. Based on the computational results, pyrrole (PY) was selected as functional monomer. Plackett–Burman design (PBD) was used for selecting the variables which affected the analytical response (current). After screening, the main factors that affect on the MIP–MWCNTs–PGE response efficiency were also optimized using central composite design (CCD). Under the optimized conditions, calibration curve of the imprinted sensor showed a linear concentration range from 10nmolL−1 to 100µmolL−1, with the limit of detection (LOD) of 4.15nmolL−1. The imprinted sensor showed the advantages of high porous surface structure, ease of preparation, good reproducibility and repeatability, high selectivity and sensitivity. Furthermore, the proposed method was successfully extended for the determination of 1,4-DAQ in serum and plasma real samples.

A novel multivariate performance optimization method based on sparse coding and hyper-predictor learning

In this paper, we investigate the problem of optimization of multivariate performance measures, and propose a novel algorithm for it. Different from traditional machine learning methods which optimize simple loss functions to learn prediction function, the problem studied in this paper is how to learn effective hyper-predictor for a tuple of data points, so that a complex loss function corresponding to a multivariate performance measure can be minimized. We propose to present the tuple of data points to a tuple of sparse codes via a dictionary, and then apply a linear function to compare a sparse code against a given candidate class label. To learn the dictionary, sparse codes, and parameter of the linear function, we propose a joint optimization problem. In this problem, the both the reconstruction error and sparsity of sparse code, and the upper bound of the complex loss function are minimized. Moreover, the upper bound of the loss function is approximated by the sparse codes and the linear function parameter. To optimize this problem, we develop an iterative algorithm based on descent gradient methods to learn the sparse codes and hyper-predictor parameter alternately. Experiment results on some benchmark data sets show the advantage of the proposed methods over other state-of-the-art algorithms.

Optimization using the gradient and simplex methods

Traditionally optimization of analytical methods has been conducted using a univariate method, varying each parameter one-by-one holding fixed the remaining. This means in many cases to reach only local minima and not get the real optimum. Among the various options for multivariate optimization, this paper highlights the gradient method, which involves the ability to perform the partial derivatives of a mathematical model, as well as the simplex method that does not require that condition.The advantages and disadvantages of those two multivariate optimization methods are discussed, indicating when they can be applied and the different forms that have been introduced. Different cases are described on the applications of these methods in analytical chemistry.