Multivariate Nonlinear Regression Research Articles

Optimal sliding friction coefficients for isolated viaducts and bridges: A comparison study

The aim of this work is to evaluate the influence of the pier–abutment–deck interaction on the seismic response of bridges isolated by single concave sliding pendulum isolators (friction pendulum system [FPS]) through a comparison with the results of the seismic response of isolated bridges without considering the presence of the rigid abutment (i.e., isolated viaducts). Two different multidegree-of-freedom (mdof) models are properly defined to carry out this comparison. In the both mdof models, five vibrational modes are considered to describe the elastic behavior of the reinforced concrete pier, and an additional degree of freedom is adopted to analyze the response of the infinitely rigid deck isolated by the seismic devices. The FPS isolator behavior is described through a widespread velocity-dependent model. By means of a nondimensional formulation of the motion equations with respect to the seismic intensity, a parametric analysis for several structural properties is performed in order to investigate the differences between the two mdof models in relation to the relevant response parameters. The uncertainty in the seismic input is taken into account by means of a set of natural records with different characteristics. Finally, multivariate nonlinear regression relationships are provided to estimate the optimum values of the sliding friction coefficient able to minimize the pier displacements relative to the ground as a function of the structural properties considering or neglecting the presence of the abutment.

Nonisothermal Crystallization Kinetics: Studying the Validity of Different Johnson–Mehl–Avrami–Erofeev–Kolmogorov (JMAEK) Based Equations

The Johnson–Mehl–Avrami–Erofeev–Kolmogorov (JMAEK) equation modifications proposed by Jeziorny, by Atiqullah and by Ruitenberg and Woldt were applied on the nonisothermal crystallization data of Poly(ethylene 2,5-furandicarboxylate)/2.5 wt. % Carbon Nanotubes (PEF/2.5 CNTs) nanocomposite. The classic JMAEK equation and the JMAEK model using the multivariate nonlinear regression method were also used to evaluate the cold crystallization's kinetic parameters of PEF/2.5 CNTs. Differential Scanning Calorimetry (DSC) measurements were carried out at heating rates of 1, 2.5, 5, 10 and 15 K/min. The effective activation energy (Eα) of the process was calculated using the isoconversional method of Vyazovkin. The Eα values indicate a nonlinear Arrhenian behavior and a complex process. The multivariate nonlinear regression method suggested a consecutive two-mechanism reaction model described by the JMAEK equation for both mechanisms. This method provided better fitting results of the experimental data for all the heating rates simultaneously and kinetic parameters' values within the limits of physical meaning. In any case, Jeziorny's modification of the JMAEK model was the most inadequate since the extracted Avrami exponents exceed in value those with a physical meaning.

Space fractional kinetic model for different types of suspension profiles in turbulent flows with a neural network-based estimation of fractional orders

Estimation of particle concentration distribution profile is important to understand the inherent mixing process, changes of river morphology, and nature of transportation of particles. In this study, the effects of non-locality are investigated to analyze, distinguish and predict different types of profiles of the suspended concentration distribution in turbulent flows through open channels, pipes, and mixing boxes. The effects of non-locality are captured through fractional derivatives and integrals. Starting from the fractional Liouville equation and considering the power-law jumps of particles in the carrier fluid, the fractional kinetic equation is derived through mathematical analysis. The kinetic model is solved and an analytical model is proposed using the Laplace transformation. The proposed solution contains the two parameters (as α the order of fractional model and μ the shape parameter) Mittag-Leffler function and is a more general one that contains several previous models as special cases. The proposed model can predict four different types of suspension concentration distributions namely type-I, type-II, plume-like, and convex type profiles under different conditions. Further, the model is validated with existing experimental data of dilute and dense flow in open channels and pipes, flows in square pipes, slurry pipelines, and fluid mud gravity currents. Satisfactory results are obtained. A detailed non-linear multivariate regression analysis is carried out with a wide range of selected data and models are proposed to compute model parmeters. To increase the efficiency of the model, an artificial neural network (ANN) model is proposed to predict the model parameters. The ANN-based models show better results than the regression analysis. Validation results show that non-locality has significant effects on the structure of the concentration profiles. From the analysis of the model parameters, it is found that type-I profile occurs under the subdiffusion process where α⩽1 and μ=1 and other three types (type-II, plume-like and convex) of profiles are associated with the superdiffusion process when α>1 and μ>1.

Differential DNA methylation by early versus late parenteral nutrition in the PICU: a biological basis for its impact on emotional and behavioral problems documented 4\xa0years later

BackgroundThe PEPaNIC multicenter randomized controlled trial (RCT) has shown that early administration of supplemental parenteral nutrition (early-PN) as compared with withholding PN for 1 week (late-PN) induced long-term internalizing, externalizing and total emotional/behavioral problems in critically ill children, as observed 4 years later. Early-PN was further shown to alter the methylation status of 37 CpG-sites in leukocyte DNA between admission and discharge from the pediatric intensive care unit (PICU). In a preplanned subanalysis of the PEPaNIC trial, we now investigated whether the altered methylation of these CpG-sites could statistically explain the negative impact of early-PN on emotion/behavior documented 4 years after PICU admission.ResultsThe combination of DNA methylation data and data on behavior 4 years after PICU admission was available for 403 of the 1440 patients (aged 0–17 years at PICU admission) who were included in the PEPaNIC RCT (192 early-PN and 211 late-PN patients). Mediation analyses with use of bootstrapped multivariable non-linear regression analyses adjusted for baseline risk factors revealed that the adverse alterations by early-PN in methylation of the 37 CpG-sites together statistically explained its harmful impact on internalizing, externalizing and total emotional/behavioral problems. When adding the methylation status of the 37 CpG-sites to the models, the explanatory power improved with a 1.710 to 1.851-fold increase, and the impact of the altered methylation status of the CpG-sites explained the impact of the randomization to early-PN versus late-PN.ConclusionsAbnormal DNA methylation induced by the early use of PN in the PICU provides a biological basis for its long-term harmful effect on emotion/behavior of critically ill children 4 years after PICU admission.Trial Registration ClinicalTrials.gov NCT01536275, registered February 17, 2012, https://clinicaltrials.gov/ct2/show/NCT01536275.

Assessment and modeling of the debonding failure of fabric-reinforced cementitious matrix (FRCM) systems

This paper aims at developing a model that is capable to accurately predict the debonding strains in reinforced concrete (RC) members strengthened with fabric-reinforced cementitious matrix (FRCM) systems. A large database consisting of 393 shear bond specimens strengthened with Polyparaphenylene Benzobisoxazole (PBO), Carbon (C), Glass (G), and Steel (S) FRCM systems was firstly compiled from the published literature. A sensitivity analysis was carried out to identify the key parameters that most affected the debonding mechanism in FRCM. The notable influence of the compressive and tensile strengths of the concrete substrate, the compressive strength of FRCM mortar, and the axial stiffness of FRCM system on the debonding strains in FRCM systems was evidenced. Contrarily, the tensile strength of FRCM mortars showed slight or no impact on the FRCM debonding strains. Based on the results of the sensitivity analysis, three simple models were developed using a multivariate nonlinear regression analysis. The models were then optimized and validated against the experimental results of 41 flexural members strengthened with different types of FRCM systems. Two of the three models proved an excellent prediction performance of the debonding strains with an average predicted–to–experimental strain ratios, εpred/εexp, of 0.99 ± 0.27 and 1.02 ± 0.27 with coefficients of variation (COV) of 0.28 and 0.26, respectively. Both models could safely predict the debonding strains in FRCM-strengthened members regardless of the type of FRCM system used. Neglecting the tensile strength of the concrete substrate in the third model resulted in an average εpred/εexp ratio of 0.85 ± 0.37 with a COV of 0.44.

Adjuvant surgical decision-making system for lumbar intervertebral disc herniation after percutaneous endoscopic lumber discectomy: a retrospective nonlinear multiple logistic regression prediction model based on a large sample

Lumbar disc herniation (LDH) is a common condition that can affects an individual' quality of life. In patients for whom conservative treatment is ineffective after 3 months, surgical treatment, such as percutaneous endoscopic lumbar discectomy (PELD), is recommended. Because PELD is minimally invasive and produces thorough nerve root decompression, both surgeons and patients often prefer it to other techniques. Surgeons find it challenging to prevent postoperative recurrent LDH (rLDH) when they use PELD. We created and verified a model for evaluating patients' recurrence risk factors before surgery so that surgeons can choose other surgical techniques when necessary. Retrospective study. One thousand eight hundred seven patients who underwent PELD at our hospital between 2012 and 2015 were enrolled. The main outcome measure was rLDH at any follow-up time point. Data were retrospectively analyzed for 1807 patients who underwent PELD at our hospital at some point between 2012 and 2015; all patients had been monitored for at least 5 years after surgery. They were divided into a recurrence group and a nonrecurrence group. Clinical and radiological risk factors were assessed over time to determine their correlations with recurrence and to exclude less important factors. A nonlinear multivariate logistic regression model was established to predict the recurrence rate before surgery. A total of 1706 patients were monitored after PELD; data were missing for 101 additional patients. The total recurrence rate was 10.38%, and the most common time from surgery to recurrence was 1 year. Ten risk factors were assessed and included in the analysis. Regarding clinical risk factors, patients with hypertension (p < .001; correlation coefficient R [R]=0.235; odds ratio [OR]=4.749), diabetes (p < .001; R=0.381; OR=16.797), a history of smoking (p < .001; R=0.347; OR=9.012), and a history of performing intense physical labor (p < .001; R=0.409; OR=19.592) had a higher recurrence rate. Regarding radiological risk factors, patients with disc degeneration (Pfirrmann grade III) (p < .001; R=0.228; OR=4.919), Modic changes (level 2) (p < .001; R=0.309; OR=7.934), herniation in the form of extrusion (p < .001; R=0.365; OR=12.228), a higher disc height index (DHI) (p < .001; R=0.336), and a larger segmental range of motion (p < .001; R=0.243) had a higher recurrence rate. When the lumbar motion angle was negative (p < .001; R=0.318; OR=13.680), the recurrence rate was high. The overall accuracy of the final model was 97.6% (1665 of 1706). The recognition rate for non-rLDH cases was 99.0% (1514 of 1529), and the rate for rLDH cases was 85.3% (151 of 177); the AUC was 0.9315. A simple model was used. For those patients with postoperative trauma (p < .001; R=0.382; OR=13.680), a comparison model was established, and the corresponding recurrence rate was 23.0% ± 25.0% (0-76%). A large cohort of patients underwent long-term monitoring, and 11 risk factors were verified for assessing each patient's risks before surgery to predict the postoperative recurrence of LDH following PELD. The risk of recurrence may be effectively reduced with the use of alternative surgical techniques in high risk cases.

Multiple nonlinear regression model of cutting force for C/SiC composites by laser‐assisted micromachining

AbstractBased on the response surface methodology (RSM), laser‐assisted micromachining (LAMM) experiments of carbon fiber reinforced silicon carbide matrix(C/SiC) composites have been carried out to obtain the data of three‐direction cutting forces Fx, Fy, Fz, and result‐force F in the cutting process. A stepwise regression method is used to screen out the factors that significantly affected the cutting force, and hence a multivariate nonlinear mathematical regression model has been established. Through the analysis of variance and regression analysis, the R2 of the regression model is the cutting force Fx: 0.975, the cutting force Fy: 0.970, the cutting force Fz: 0.972, the resultant force F: 0.987, and the probability p &gt; F‐values are all less than 0.001. The model analysis results show that the main effects and interactions of the three factors of laser power, cutting speed, and cutting depth significantly influence the cutting force. The main effect of the most considerable influence on the resultant F is cutting depth, followed by cutting speed and laser power, and the interactive effect is cutting speed and cutting depth. The experimental results show that the model prediction results are in good agreement with the experimental results.

Catalytic pyrolysis of mechanically non-recyclable waste plastics mixture: Kinetics and pyrolysis in laboratory-scale reactor

Post-consumer waste plastics that cannot be mechanically recycled represent a concerning environmental issue. According to the latest available data for Europe, as much as 25% of collected post-consumer waste plastics are landfilled, 43% is energy recovered, and 32% is recycled. One possible way of recovering non-recyclable plastics is pyrolysis, which is considered environmentally friendly technology for obtaining fuel or chemicals from plastic waste. To tackle the challenge of recovering non-recyclable plastics via pyrolysis, it is necessary to determine their actual composition. Visual separation of collected non-recyclable plastics was performed, and Fourier-transform infrared spectroscopy was used to confirm the accuracy of visual separation. A significant amount of plastics labelled as “other” was found. Since the composition of “other” waste plastics has not been sufficiently investigated, relatively few studies on their pyrolysis have been conducted. Therefore, they were characterised and added to the mixture with other found polymer types of non-recyclable plastics. Thermogravimetric analysis was conducted to determine thermochemical behaviour and kinetic parameters required for laboratory pyrolysis investigation. Kinetic analysis was conducted using the Friedman isoconversional model-free method and non-linear multivariate regression method. The goal of this paper was to analyse the kinetics, determine the product yield and characteristics of the pyrolysis process of non-recyclable plastics over zeolite catalysts. It was found how the decomposition of non-recyclable plastics occurs in two decomposition steps. The activation energy of non-recyclable plastics was 144 kJ/mol in the first stage of decomposition and 262 kJ/mol in the second stage of decomposition. It decreased by 34% and 6.5% after fresh fluid catalytic cracking catalyst was added and 41% and 18.3% with iron-modified Zeolite Socony Mobil-5 catalyst. The yield of condensate was 55% (wax) for the original sample, and it decreased to 50% (wax and oil) and 27% (mostly oil) with fresh fluid catalytic cracking and iron modified Zeolite Socony Mobil-5 catalysts. Processes with catalysts promoted the formation of olefins and aromatic compounds in pyrolytic oil. All pyrolysis products had a high value of higher heating value ranging from 39 MJ/kg to 43 MJ/kg showing good potential for further energy use.

Estimation of higher heating values (HHVs) of biomass fuels based on ultimate analysis using machine learning techniques and improved equation

To have a sustainable economy and environment, several countries have widely inclined to the utilization of non-fossil fuels like biomass fuels to produce heat and electricity. The advantage of employing biomass for combustion is emerging as a potential renewable energy, which is regarded as a cheap fuel. Chemical constituents or elements are essential properties in biomass applications, which would be costly and labor-intensive to experimentally estimate them. One of the criteria to evaluate the energy of biomass from an economic perspective is the higher heating value (HHV). In the present work, we have applied multilayer perceptron artificial neural network (MLP-ANN), least-squares support vector machine (LSSVM), ant colony-adaptive neuro-fuzzy inference system (ACO-ANFIS), particle swarm optimization- ANFIS (PSO-ANFIS), genetic algorithm-radial basis function (GA-RBF) and new multivariate nonlinear regression (MNR) as accurate correlation methods to estimate HHVs of biomass fuels based on the ultimate analysis. 535 experimental data were gathered from literature and categorized into eight classes of by-products of fruits, agri-wastes, wood chips/tree species, grasses/leaves/fibrous materials, other waste materials, briquettes/charcoals/pellets, cereal and Industrial wastes. In the term of statistical analysis, average absolute relative deviation (AARD) authenticates that MNR and GA-RBF algorithm with %AARD of 3.5 and 3.4 could be used to estimate HHV. In addition, developed models results were compared to the results of 69 recently previously published empirical correlations and it confirms the reliability of our results. Relevency factor shows the impact of biomass elements on HHV and outlier analysis indicates the unreliable experimental data. The results of this study can be used by researchers to design and optimize biomass combustion systems.

Improving Snow Estimates Through Assimilation of MODIS Fractional Snow Cover Data Using Machine Learning Algorithms and the Common Land Model

AbstractIn this study, an innovative MODIS fractional snow cover (SCF) data assimilation (DA) prototype framework that invokes machine learning (ML) techniques and Common land model (CoLM) is proposed to improve the estimation of the snow depth (SD) and the SCF. To validate our new framework, we analyzed two snow seasons from 2013 to 2015 at 46 stations in Northern Xinjiang in China. We developed 12 SCF DA schemes that represent different DA methods (direct insertion (DI) and Ensemble Kalman Filter (EnKF)), observational data (original data and gap‐filled MODIS SCF data), and observation operators (five new snow depletion curves (SDCs) defined using traditional multivariate nonlinear regression and four ML methods). While improving the frequency of the SCF observations in the DI‐based DA scheme only resulted in a marginal improvement in the snow estimates, by adding new SDCs fitted by ML techniques (e.g., deep belief network), and the gap‐filled MODIS SCF data to the EnKF‐based DA scheme, we were able to reduce model structural uncertainties of CoLM and achieve marked improvement in the accuracy of the snow estimates (RMSE = 5.92 cm, mean bias error = −1.94 cm, and average degree of improvement = 32.18% for SD estimates and RMSE = 15. 79%, mean bias error = −1.21%, and average degree of improvement = 47.95% for SCF estimates). Our results demonstrate the feasibility of improving snow estimates by combining the ML techniques with physically based snowpack model in a SCF DA framework.

Agro waste as a source of bioactive glass for targeted drug delivery and bone implantation

The present work describes the synthesis of biocompatibe alkali-free glasses utilizing the agricultural waste rice husk and eggshells. A series of glasses with basic composition SiO2–P2O5–MgO–CaO has been synthesized using melt-quench technique, where P2O5 is partially replaced by MnO. These glasses are characterized to assess their in-vitro reactivity with the help of X-Ray diffraction, scanning electron microscopy, energy dispersive X-ray spectroscopy, fourier transform infrared spectroscopy, and microwave plasma atomic emission spectroscopy techniques before and after soaking in the simulated body fluid solution. Further, drug loading and release studies are carried out with the help of UV–Visible spectrophotometer. Wherefore, a multivariable non-linear regression model has been developed to predict the release rate at different manganese oxide concentrations. The proposed model shows a high Pearson correlation coefficient between the observed and predicted absorbance. These results designate that low-cost agro-waste derived alkali-free glasses have potential in medical applications in drug delivery as well as it helps in environmental pollution management.

Long-term evaluation and calibration of three types of low-cost PM2.5 sensors at different air quality monitoring stations

To evaluate the performance of low-cost PM2.5 sensors and develop calibration models for correcting for the PM2.5 sensor data (PM2.5,S), field comparison tests were conducted based on Met One BAM-1020 data at various locations using long-term (≥one year) data of Plantower PMS5003, Sensirion SPS30, and Honeywell HPMA115S0 PM2.5 sensors. Both multivariate linear regression (MLR) and non-linear regression (NLR) models using hourly RHs and original sensor PM2.5 data as parameters were able to obtain accurate calibrated hourly PM2.5 values with MNBs (mean normalized biases) less than about ±10% and MNEs (mean normalized errors) less than about 30% for all three types of PM2.5 sensors at all monitoring locations. On the other hand, the MNB and MNE of the calibrated 24-hr average PM2.5 data for the two models were less than ±13% and 20%, respectively. Moreover, the slope, intercept, and R2 of the linear regression line of the calibrated 24-hr average PM2.5 and BAM-1020 data were as good as 1.0 ± 0.1, 0.0 ± 2.0 μg/m3, and ≥0.88, respectively. Therefore, these well-calibrated sensors can well be served for education and information (MNE<50%), hotspot identification and characterization (MNE<30%), and personal exposure study (MNE<30%) purposes, and even supplement the existing daily PM2.5 data of the air quality monitoring stations (MNE<20%).

Prediction of solitary wave attenuation by emergent vegetation using genetic programming and artificial neural networks

Analyzing the attenuation of extreme waves by coastal emergent vegetation provides crucial information on revetment planning. In this study, three kinds of laboratory experiments of wave attenuation by rigid vegetation are performed. Transmission coefficient (Kt) was used to characterize the effect of wave attenuation. The influence of dimensionless factors including relative wave height (H/h), relative width (B/L), relative height (hv/h) and solid volume fraction (ϕ) on the Kt under the action of solitary wave was explored by Genetic Programming (GP), Artificial Neural Networks (ANNs) and multivariate non-linear regression (MNLR). Prediction formulae (R2 is up to 0.95) of the Kt in different models were established by GP method, and the sensitivity of each dimensionless factor was analyzed by statistical analysis. ANNs were used to compare the weight of each factor. The power function relationships between Kt and factors was obtained by MNLR. The results show that GP can qualitatively acquire the sensitivity of parameters and is suitable for the sensitivity analysis of the vegetation wave disspation model, providing a more efficient and accurate prediction method. The results can provide guidelines for vegetation planting as well as the scientific basis for vegetation revetment engineering.

Multivariate Quadratic Nonlinear Regression Model of the Ultimate Pull-Out Load of Electrohydraulic Expansion Joints Based on Response Surface Methodology

Electrohydraulic expansion joining has great potential for joining the light weight and high strength thin-walled pipes due to its high strain rate. Based on the central composite design (CCD) of response surface methodology, multiple experiments of electrohydraulic expansion joining process were performed. The multivariate quadratic nonlinear regression model between process parameters (discharge voltage, wire length, and wire diameter) and the ultimate pull-out load of the joints was established. The results revealed that discharge voltage, wire length and wire diameter all had a significant effect on the ultimate pull-out load. The discharge voltage had the most significant effect. The interaction between the discharge voltage and the wire diameter had a significant effect on the ultimate pull-out load. The optimal parameter combination (discharge voltage = 6 kV, wire length = 10 mm, wire diameter = 0.833 mm) was obtained and verified through the experiments. This study would provide guidance for the choice of the process parameters in real applications.

Real-time regulation of filling speed based on multiparameter synergistic coupling effect in fused deposition modeling

This paper aimed to perform a preliminary study on real-time regulation of filling speed by considering multiparameter synergistic coupling effect in fused deposition modeling. Bivariate experiments and sampling printing tests on designed samples, under different parameter combinations which are made up by filling speed, layer thickness, nozzle temperature and corner angle, were carried out to build and validate a real-time regulation model of filling speed by employing error analysis and multivariate nonlinear regression method. Results showed that the equivalent dimensional deviation values and area deviation values of Sample A, Sample B, Sample C by using real-time regulation algorithm of filling speed are lower than or close to the concentrated averages by using the optimal constant filling speed, with the maximum of 0.32 mm and 0.46 mm2. Actual and theoretical build time of Sample D at real-time regulated filling speed are less than those at 20 mm/s (“low speed”), and close to those at 45 mm/s (“medium speed”) or 70 mm/s (“high speed”), which indicates that the method of real-time dynamic adjusting filling speed can not only improve printing accuracy, but also give a full consideration to printing efficiency at the same time.

A Novel Load Forecasting Model for Automatic Fault Clearance in Secondary Distribution Electric Power Grid using an Extended-Multivariate Nonlinear Regression

Smart grid is an emerging platform adopted by many electric power utility companies to enhance proper service delivery as well as cost-effective operations. Automatic Fault Detection and Clearance (AFDC) is a part of intelligent technology initiatives established on Tanzania’s grid aiming at detecting, managing, and handling fault with little or without human intervention. Being one of the components of AFDC, the Load Forecasting agency plays important role in feeding restoration and distributed energy resource agencies. Forecasting load demand profile in the AFDC requires a robust mechanism for accommodating both fault-based and random-walk situations. Unfortunately, existing LF mechanisms do not address these two issues robustly. Therefore, this research work aims to propose an AFDC-based LF (LF-AFDC) model whose time series data is characterized by random-walk distribution. Firstly, we establish design requirements using focus group discussion and literature review. Secondly, iterative validation is conducted to assess the suitability of the proposed LF mechanism in the intended context. Thirdly, the core forecasting part of the mechanism is achieved using the extended Multivariate Nonlinear Regression (e-MNLR) model. Findings indicate the capability of the proposed model to forecast the next load profile from the given fault-date, fault-time, restoration duration, Gross-Domestic Product (GDP), number of customers, and population information. Furthermore, the designed e-MNLR seems to outperform the ANN, SVM, LSTM, and MNLR models.

Calibration of \u03ba-\u03b5 turbulence model for thermal\u2013hydraulic analyses in rib-roughened narrow rectangular channels using genetic algorithm

Nowadays, applications of turbulent fluid flow in removing high heat flux in rib-roughened narrow channels are drawing much interest. In this work, an improved version of the κ-ε turbulence model is proposed for better prediction of thermal–hydraulic characteristics of flow inside rib-roughened (pitch-to-rib height (p/k) ratio = 10 and 20) narrow channels (channel height, H = 1.2 mm and 3.2 mm). For this, the four turbulence model parameters, Cμ, Cε1, Cε2, and σk, are calibrated. These parameters are adjustable empirical constants provided for controlling the accuracy of the turbulence model results when needed. The simulated data are used to develop correlations between the relative errors in predicting the friction factor (f), Nusselt number (Nu), and the model parameters using a multivariate nonlinear regression method. These correlations are used to optimize the errors using genetic algorithm. Results reveal that the calibrated parameters are not the same for all the narrow channel configurations. After calibration, the overall predictive improvements are up to 35.83% and 27.30% for p/k = 10 and p/k = 20 respectively when H = 1.2 mm. Also, up to 15.48% and 18.05% improvements are obtained for p/k = 10 and p/k = 20 respectively when H = 3.2 mm. The role of the two parameters Cε1 and Cε2 are found to be of primary importance. Furthermore, three types of nanofluids i.e. Al2O3-water, CuO-water, and TiO2-water are studied using the calibrated model to check the potentiality of heat transfer enhancement. Among them, CuO-water nanofluid is predicted to have around 1.32 times higher value of Nu than pure water for the same narrow channel configuration.Article Highlightsκ-ε turbulence model is calibrated for rib-roughened narrow rectangular channels using genetic algorithm.Cε1 and Cε2 are the most influential parameters on the performance of the model inside rib-roughened narrow channel.Suggested calibration process is more effective for channel height of 1.2 mm than 3.2 mm.

Correlation Between Intraventricular Pressure Difference and Indexed Flow of a Left Ventricular Assist Device

There is no definitive parameter for left ventricular (LV) preload in patients with a continuous-flow left ventricular assist device (LVAD). The intraventricular pressure difference (IVPD) is the maximum pressure difference between the mitral valve and LV apex during diastole; and, in past studies, the IVPD was influenced by volume loading. The authors hypothesized that IVPD in LVAD patients correlates with indexed LVAD flow and that IVPD can serve as a novel parameter of LV preload in this population. A single-center, retrospective, observational study. A tertiary-care hospital from August 2019 to July 2020. Sixteen ramp tests for adjustment of LVAD pump speed in 14 adult patients undergoing continuous-flow LVAD implantation. Measurement of IVPD during ramp tests. LVAD flow and IVPD were measured at each LVAD pump speed during the ramp test for the adjustment of LVAD pump speed after patients came off cardiopulmonary bypass during LVAD implantation. A straight, longitudinal view of the left atrium and left ventricle was obtained, and the pressure difference between the mitral valve and LV apex during diastole was measured by transesophageal echocardiography. The maximum pressure difference during diastole was recorded as IVPD. The relationship between indexed LVAD flow (LVAD flow/body surface area) and IVPD was assessed by a multivariate nonlinear regression analysis with the Huber-White sandwich estimator. IVPD correlated with indexed LVAD flow (p < 0.001). IVPD is a useful indicator of LV preload during LVAD implantation.

Fuzzy neural network and coupled gene expression programming/multivariate non-linear regression approach on mechanical features of hydroxyapatite/graphene oxide/epoxy: Empirical and optimization study

One way to enhance the mechanical properties of nanocomposites has been to use different fillers. In this study, ternary hybrid composites of graphene oxide/hydroxyapatite/epoxy resin were investigated. An experimental design was performed based on the central composite design (CCD). Epoxy resin was modified by incorporating different graphene oxide and hydroxyapatite weight from 0 to 0.5 wt.% and 0 to 7 wt.%, respectively. Experimental results showed that Young’s modulus, yield strength and impact strength improved up to 25.64%, 5.95% and 100.05% compared to the neat epoxy resin, respectively. In addition, gene expression programming (GEP), multivariate non-linear regression (MNLR) and fuzzy neural network (FNN) methods were employed to determine the effects of nanoparticles on the mechanical properties. Based on the modelling results, optimization process was investigated by using particle swarm optimization (PSO). Finally, the fracture surface morphologies of the nanocomposites were analyzed by scanning electron microscopy.

Modified artificial neural network model with an explicit expression to describe flow behavior and processing maps of Ti2AlNb-based superalloy

The elevated-temperature deformation behavior of Ti2AlNb superalloy was observed by isothermal compression experiments in a wide range of temperatures (950–1200 °C) and strain rates (0.001–10 $${\mathrm{s}}^{-1}$$ ). The flow behavior is nonlinear, strongly coupled, and multivariable. The constitutive models, namely the double multivariate nonlinear regression model, artificial neural network model, and modified artificial neural network model with an explicit expression, were applied to describe the Ti2AlNb superalloy plastic deformation behavior. The comparative predictability of those constitutive models was further evaluated by considering the correlation coefficient and average absolute relative error. The comparative results show that the modified artificial network model can describe the flow stress of Ti2AlNb superalloy more accurately than the other developed constitutive models. The explicit expression obtained from the modified artificial neural network model can be directly used for finite element simulation. The modified artificial neural network model solves the problems that the double multivariate nonlinear regression model cannot describe the nonlinear, strongly coupled, and multivariable flow behavior of Ti2AlNb superalloy accurately, and the artificial neural network model cannot be embedded into the finite element software directly. However, the modified artificial neural network model is mainly dependent on the quantity of high-quality experimental data and characteristic variables, and the modified artificial neural network model has not physical meanings. Besides, the processing maps were applied to obtain the optimum processing parameters.