Search results

The prolonged training time of the neural network (NN) has sparked considerable debate regarding their application in the field of optimization. The purpose of this paper is to explore the beneficial assistance of NN-based alternative models in inductance design, with a particular focus on multi-objective optimization and uncertainty analysis processes.

Under Gaussian-distributed manufacturing errors, this study predicts error intervals for Pareto points and select robust solutions with minimal error margins. Furthermore, this study establishes correlations between manufacturing errors and inductance value discrepancies, offering a practical means of determining permissible manufacturing errors tailored to varying accuracy requirements.

The NN-assisted methods are demonstrated to offer a substantial time advantage in multi-objective optimization compared to conventional approaches, particularly in scenarios where the trained NN is repeatedly used. Also, NN models allow for extensive data-driven uncertainty quantification, which is challenging for traditional methods.

Three objectives including saturation current are considered in the multi-optimization, and the time advantages of the NN are thoroughly discussed by comparing scenarios involving single optimization, multiple optimizations, bi-objective optimization and tri-objective optimization. This study proposes direct error interval prediction on the Pareto front, using extensive data to predict the response of the Pareto front to random errors following a Gaussian distribution. This approach circumvents the compromises inherent in constrained robust optimization for inductance design and allows for a direct assessment of robustness that can be applied to account for manufacturing errors with complex distributions.

Parameter estimation in complex engineering structures typically necessitates repeated calculations using simulation models, leading to significant computational costs. This paper aims to introduce an adaptive multi-output Gaussian process (MOGP) surrogate model for parameter estimation in time-consuming models.

The MOGP surrogate model is established to replace the computationally expensive finite element method (FEM) analysis during the estimation process. We propose a novel adaptive sampling method for MOGP inspired by the traditional expected improvement (EI) method, aiming to reduce the number of required sample points for building the surrogate model. Two mathematical examples and an application in the back analysis of a concrete arch dam are tested to demonstrate the effectiveness of the proposed method.

The numerical results show that the proposed method requires a relatively small number of sample points to achieve accurate estimates. The proposed adaptive sampling method combined with the MOGP surrogate model shows an obvious advantage in parameter estimation problems involving expensive-to-evaluate models, particularly those with high-dimensional output.

A novel adaptive sampling method for establishing the MOGP surrogate model is proposed to accelerate the procedure of solving large-scale parameter estimation problems. This modified adaptive sampling method, based on the traditional EI method, is better suited for multi-output problems, making it highly valuable for numerous practical engineering applications.

The purpose of this study is to propose a systematic model selection procedure for long-term load forecasting (LTLF) for ex-ante and ex-post cases considering uncertainty in exogenous predictors.

The different variants of regression models, namely, Polynomial Regression (PR), Generalised Additive Model (GAM), Quantile Polynomial Regression (QPR) and Quantile Spline Regression (QSR), incorporating uncertainty in exogenous predictors like population, Real Gross State Product (RGSP) and Real Per Capita Income (RPCI), temperature and indicators of breakpoints and calendar effects, are considered for LTLF. Initially, the Backward Feature Elimination procedure is used to identify the optimal set of predictors for LTLF. Then, the consistency in model accuracies is evaluated using point and probabilistic forecast error metrics for ex-ante and ex-post cases.

From this study, it is found PR model outperformed in ex-ante condition, while QPR model outperformed in ex-post condition. Further, QPR model performed consistently across validation and testing periods. Overall, QPR model excelled in capturing uncertainty in exogenous predictors, thereby reducing over-forecast error and risk of overinvestment.

These findings can help utilities to align model selection strategies with their risk tolerance.

To propose the systematic model selection procedure in this study, the consistent performance of PR, GAM, QPR and QSR models are evaluated using point forecast accuracy metrics Mean Absolute Percentage Error, Root Mean Squared Error and probabilistic forecast accuracy metric Pinball Score for ex-ante and ex-post cases considering uncertainty in the considered exogenous predictors such as RGSP, RPCI, population and temperature.

This paper aims to present a predictive model approach to estimate the tensile behavior of polylactic acid (PLA) under uncertainty using the fused deposition modeling (FDM) and American Society for Testing and Materials (ASTM) D638’s Types I and II test standards.

The prediction approach combines artificial neural network (ANN) and finite element analysis (FEA), Monte Carlo simulation (MCS) and experimental testing for estimating tensile behavior for FDM considering uncertainties of input parameters. FEA with variance-based sensitivity analysis is used to quantify the impacts of uncertain variables, resulting in determining the significant variables for use in the ANN model. ANN surrogates FEA models of ASTM D638’s Types I and II standards to assess their prediction capabilities using MCS. The developed model is applied for testing the tensile behavior of PLA given probabilistic variables of geometry and material properties.

The results demonstrate that Type I is more appropriate than Type II for predicting tensile behavior under uncertainty. With a training accuracy of 98% and proven presence of overfitting, the tensile behavior can be successfully modeled using predictive methods that consider the probabilistic nature of input parameters. The proposed approach is generic and can be used for other testing standards, input parameters, materials and response variables.

Using the proposed predictive approach, to the best of the authors’ knowledge, the tensile behavior of PLA is predicted for the first time considering uncertainties of input parameters. Also, incorporating global sensitivity analysis for determining the most contributing parameters influencing the tensile behavior has not yet been studied for FDM. The use of only significant variables for FEA, ANN and MCS minimizes the computational effort, allowing to simulate more runs with reduced number of variables within acceptable time.

Prior investigations concerning misalignment resulting from journal deformation typically relied on predefined misaligned angles. Nevertheless, scant attention has been devoted to the determination of these misaligned angles. Furthermore, existing studies commonly treat the journal as rigid under such circumstances. Therefore, the present study aims to introduce a framework for determining misaligned angles and to compare outcomes between rigid and flexible journal configurations.

The bearing forces are considered as an external load leading to journal deformation. This deformation is calculated using the finite element method. The pressure distribution producing the bearing force is solved using the finite difference method. The mesh grids in the finite element and finite difference methods are matched for coupling calculation. By iteration, the pressure distribution of the lubricant film at the equilibrium position is determined.

Results show that the deformation-induced misalignment has a significant influence on the performance of the bearing when the journal flexibility is taken into account. The parametric study reveals that the misalignment relies on system parameters such as bearing length-diameter ratio and static load.

The investigation of this work provides a quantification method of misalignment of hydrodynamic bearings considering the elastic deformation of the journal, which assists in the design of bearing in a rotor-bearing system.

The peer review history for this article is available at: https://publons.com/publon/10.1108/ILT-10-2023-0337/

Industrial simulations of turbulent flows often rely on Reynolds-averaged Navier-Stokes (RANS) turbulence models, which contain numerous closure coefficients that need to be calibrated. This paper aims to address this issue by proposing a semi-automated calibration of these coefficients using a new framework (referred to as turbo-RANS) based on Bayesian optimization.

The authors introduce the generalized error and default coefficient preference (GEDCP) objective function, which can be used with integral, sparse or dense reference data for the purpose of calibrating RANS turbulence closure model coefficients. Then, the authors describe a Bayesian optimization-based algorithm for conducting the calibration of these model coefficients. An in-depth hyperparameter tuning study is conducted to recommend efficient settings for the turbo-RANS optimization procedure.

The authors demonstrate that the performance of the k-ω shear stress transport (SST) and generalized k-ω (GEKO) turbulence models can be efficiently improved via turbo-RANS, for three example cases: predicting the lift coefficient of an airfoil; predicting the velocity and turbulent kinetic energy fields for a separated flow; and, predicting the wall pressure coefficient distribution for flow through a converging-diverging channel.

To the best of the authors’ knowledge, this work is the first to propose and provide an open-source black-box calibration procedure for turbulence model coefficients based on Bayesian optimization. The authors propose a data-flexible objective function for the calibration target. The open-source implementation of the turbo-RANS framework includes OpenFOAM, Ansys Fluent, STAR-CCM+ and solver-agnostic templates for user application.

This study aims to explore the evolving field of physics-informed neural networks (PINNs) through an analysis of 996 records retrieved from the Web of Science (WoS) database from 2019 to 2022.

WoS database was analyzed for PINNs using an inhouse python code. The author’s collaborations, most contributing institutes, countries and journals were identified. The trends and application categories were also analyzed.

The papers were classified into seven key domains: Fluid Dynamics and computational fluid dynamics (CFD); Mechanics and Material Science; Electromagnetism and Wave Propagation; Biomedical Engineering and Biophysics; Quantum Mechanics and Physics; Renewable Energy and Power Systems; and Astrophysics and Cosmology. Fluid Dynamics and CFD emerged as the primary focus, accounting for 69.3% of total publications and witnessing exponential growth from 22 papers in 2019 to 366 in 2022. Mechanics and Material Science followed, with an impressive growth trajectory from 3 to 65 papers within the same period. The study also underscored the rising interest in PINNs across diverse fields such as Biomedical Engineering and Biophysics, and Renewable Energy and Power Systems. Furthermore, the focus of the most active countries within each application category was examined, revealing, for instance, the USA’s significant contribution to Fluid Dynamics and CFD with 319 papers and to Mechanics and Material Science with 66 papers.

This analysis illuminates the rapidly expanding role of PINNs in tackling complex scientific problems and highlights its potential for future research across diverse domains.

Predictive digital twin technology, which amalgamates digital twins (DT), the internet of Things (IoT) and artificial intelligence (AI) for data collection, simulation and predictive purposes, has demonstrated its effectiveness across a wide array of industries. Nonetheless, there is a conspicuous lack of comprehensive research in the built environment domain. This study endeavours to fill this void by exploring and analysing the capabilities of individual technologies to better understand and develop successful integration use cases.

This study uses a mixed literature review approach, which involves using bibliometric techniques as well as thematic and critical assessments of 137 relevant academic papers. Three separate lists were created using the Scopus database, covering AI and IoT, as well as DT, since AI and IoT are crucial in creating predictive DT. Clear criteria were applied to create the three lists, including limiting the results to only Q1 journals and English publications from 2019 to 2023, in order to include the most recent and highest quality publications. The collected data for the three technologies was analysed using the bibliometric package in R Studio.

Findings reveal asymmetric attention to various components of the predictive digital twin’s system. There is a relatively greater body of research on IoT and DT, representing 43 and 47%, respectively. In contrast, direct research on the use of AI for net-zero solutions constitutes only 10%. Similarly, the findings underscore the necessity of integrating these three technologies to develop predictive digital twin solutions for carbon emission prediction.

The results indicate that there is a clear need for more case studies investigating the use of large-scale IoT networks to collect carbon data from buildings and construction sites. Furthermore, the development of advanced and precise AI models is imperative for predicting the production of renewable energy sources and the demand for housing.

This paper makes a significant contribution to the field by providing a strong theoretical foundation. It also serves as a catalyst for future research within this domain. For practitioners and policymakers, this paper offers a reliable point of reference.

How to embed artificial intelligence (AI) in human resource management (HRM) is one of the core challenges of digital HRM. Despite regulations demanding humans in the loop to ensure human oversight of AI-based decisions, it is still unknown how much decision-makers rely on information provided by AI and how this affects (personnel) selection quality.

This paper presents an experimental study using vignettes of dashboard prototypes to investigate the effect of AI on decision-makers’ overreliance in personnel selection, particularly the impact of decision-makers’ information search behavior on selection quality.

Our study revealed decision-makers’ tendency towards status quo bias when using an AI-based ranking system, meaning that they paid more attention to applicants that were ranked higher than those ranked lower. We identified three information search strategies that have different effects on selection quality: (1) homogeneous search coverage, (2) heterogeneous search coverage, and (3) no information search. The more applicants were searched equally often (i.e. homogeneous) as when certain applicants received more search views than others (i.e. heterogeneous) the higher the search intensity was, resulting in higher selection quality. No information search is characterized by low search intensity and low selection quality. Priming decision-makers towards carrying responsibility for their decisions or explaining potential AI shortcomings had no moderating effect on the relationship between search coverage and selection quality.

Our study highlights the presence of status quo bias in personnel selection given AI-based applicant rankings, emphasizing the danger that decision-makers over-rely on AI-based recommendations.