Article
Computer Science, Interdisciplinary Applications
Haizhou Yang, Seong Hyeong Hong, Yi Wang
Summary: This paper presents a novel computation-aware multi-fidelity surrogate-based optimization methodology and a new sequential and adaptive sampling strategy based on expected improvement reduction. It improves the exploration and convergence rate of the optimization process under a fixed computational budget.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2022)
Article
Computer Science, Artificial Intelligence
Yong Pang, Xiaonan Lai, Yitang Wang, Xiwang He, Shuai Zhang, Xueguan Song
Summary: In this work, a new algorithm based on the Kriging surrogate model and a novel constraint handling method is proposed for expensive multiobjective optimization problems. The algorithm achieves superior results in highly heterogeneous optimization problems and bi-objective constrained scenarios, by reducing time complexity and increasing accuracy. The effectiveness of the proposed method is verified through benchmark comparison problems and practical applications.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Information Systems
Zan Yang, Haobo Qiu, Liang Gao, Liming Chen, Jiansheng Liu
Summary: This paper proposes an adaptive surrogate-assisted MOEA/D framework (ASA-MOEA/D) for efficiently solving expensive constrained multi-objective optimization problems. With three specific search strategies, ASA-MOEA/D achieved targeted searches for different subproblems based on their optimization states. The framework maintained feasibility, convergence, and diversity through the use of RBF surrogates and exploration of unexplored subregions. Empirical studies showed that ASA-MOEA/D with tchebycheff approach outperformed four state-of-the-art algorithms.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Junfeng Tang, Handing Wang, Lin Xiong
Summary: In preference-based multi-objective optimization, knee solutions are the implicit preferred promising solutions. However, finding knee solutions is difficult and computationally expensive. To address this issue, we propose a surrogate-assisted evolutionary multi-objective optimization algorithm that uses surrogate models to replace expensive evaluations. Experimental results show that our proposed algorithm outperforms state-of-the-art knee identification evolutionary algorithms on most test problems within a limited computational budget.
SWARM AND EVOLUTIONARY COMPUTATION
(2023)
Article
Computer Science, Interdisciplinary Applications
Leshi Shu, Ping Jiang, Yan Wang
Summary: This work proposes a multi-fidelity Bayesian optimization approach that utilizes hierarchical Kriging to reduce optimization costs, quantifies the impact of high and low-fidelity samples based on expected further improvement, and introduces a novel acquisition function to determine the location and fidelity level of the next sample simultaneously. The proposed approach is compared with state-of-the-art methods for multi-fidelity global optimization and shows that it can achieve global optimal solutions with reduced computational costs.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2021)
Article
Computer Science, Interdisciplinary Applications
Zhendong Guo, Qineng Wang, Liming Song, Jun Li
Summary: The study introduces a new infill criterion called Filter-GEI for addressing sample assignment issue in multi-fidelity optimization. By considering correlations between HF and LF models and adding an adaptive filter function on top of the GEI acquisition function, Filter-GEI efficiently allocates HF and LF samples to achieve a good balance between local and global search, with further improvement in efficiency through infilling multiple HF and LF samples in each iteration along with parallel computing. Tests on mathematical toy problems and an engineering problem demonstrate the effectiveness of the proposed algorithm.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2021)
Article
Engineering, Electrical & Electronic
Sen Yin, Ruitao Wang, Jian Zhang, Yan Wang
Summary: In this paper, we propose an efficient asynchronous parallel approach for multi-objective optimization with multiple constraints. Experimental results demonstrate that our method outperforms other approaches in terms of performance.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS
(2022)
Article
Computer Science, Artificial Intelligence
Qinghua Gu, Qian Wang, Xuexian Li, Xinhong Li
Summary: A new algorithm, RFMOPSO, is proposed in this paper to optimize constrained combinatorial optimization problems by combining multi-objective particle swarm optimization with a random forest model. Adaptive ranking strategy and novel rule are employed to improve search speed and adaptively update particle states for better objective balance and feasible solutions. Experimental results show promising performance on benchmark problems with discrete variables varying from 10 to 100.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Computer Science, Interdisciplinary Applications
Qingyu Wang, Takuji Nakashima, Chenguang Lai, Bo Hu, Xinru Du, Zhongzheng Fu, Taiga Kanehira, Yasufumi Konishi, Hiroyuki Okuizumi, Hidemi Mutsuda
Summary: The study proposes an enhanced multi-point infill criterion called PEHVInne, which improves the convergence efficiency of multi-objective optimization problems by adding a relatively large number of points in each cycle. The criterion considers both the potential for improvement of an evaluation point and the crowding distances between this point and its nearest-neighbor Pareto front points. Experimental results show that the proposed criterion enhances the efficiency and convergence of the optimization process for most benchmarks.
JOURNAL OF COMPUTATIONAL SCIENCE
(2022)
Article
Engineering, Mechanical
Randall J. Kania, Shapour Azarm
Summary: Engineering design optimization problems often involve two competing objectives and uncertainty. This article proposes an approach using a Bayesian framework to solve multi-objective optimization problems under interval uncertainty. The method iteratively relaxes solutions to converge to a set of non-dominated, robust optimal solutions and uses a variation of the bi-objective expected improvement criterion to encourage variety and density of solutions. Several examples are tested and compared, showing that the proposed method performs well at finding robustly optimized feasible solutions with limited function evaluations.
JOURNAL OF MECHANICAL DESIGN
(2023)
Article
Thermodynamics
Lucas F. Santos, Caliane B. B. Costa, Jose A. Caballero, Mauro A. S. S. Ravagnani
Summary: A surrogate-based multi-objective optimization framework is used to design natural gas liquefaction processes and compare their energy consumption and heat exchanger area utilization. The surrogate-based framework provides better solutions than traditional multi-objective optimization methods. The results contribute to achieving lower energy consumption and higher heat exchanger utilization efficiency.
Article
Chemistry, Multidisciplinary
Lei Sheng, Weichao Zhao, Ying Zhou, Weimeng Lin, Chunyan Du, Hongwei Lou
Summary: An optimization model for optical imaging system is established, combining Latin hypercube sampling, Kriging surrogate model training, and multi-objective optimization algorithm NSGA-III. Compared with traditional optical system simulation methods, this model achieves high-accuracy results and significantly improves optimization efficiency. Case studies on Cooke triplet optical system show that the imaging quality is considerably improved, demonstrating the suitability of this model for optimal optical system design.
APPLIED SCIENCES-BASEL
(2022)
Article
Mathematics
Shande Li, Jian Wen, Jun Wang, Weiqi Liu, Shuai Yuan
Summary: This paper proposes a high-precision surrogate modeling method based on the parallel multipoint expected improvement point infill criteria for solving large-scale complex simulation problems. The method combines global search ability with local search ability, improving the overall accuracy of the fitting function.
Article
Computer Science, Artificial Intelligence
Qinghua Gu, Qian Wang, Neal N. Xiong, Song Jiang, Lu Chen
Summary: A surrogate-assisted evolutionary algorithm is proposed in this paper for solving expensive constrained multi-objective discrete optimization problems. By embedding random forest models and an improved stochastic ranking strategy, the algorithm makes significant progress in optimization efficiency and candidate solution quality.
COMPLEX & INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Interdisciplinary Applications
Maliki Moustapha, Alina Galimshina, Guillaume Habert, Bruno Sudret
Summary: Accounting for uncertainties is crucial for the safety of engineering structures. This study proposes a method for robust design optimization by considering quantiles of objective functions. By introducing the concept of common random numbers and using a surrogate-assisted approach, the computational cost of the optimization problem is reduced.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2022)
Article
Engineering, Mechanical
Jolan Wauters
Summary: This paper discusses robust design optimization (RDO) to account for variability in the design phase. The method is formulated as a multi-objective problem with the objective of minimizing both the mean and variance of the objective. The use of Bayesian optimization and surrogate modeling techniques are proposed to address computational cost and uncertainty. An analytical treatment of the problem is presented to obtain quantities of interest without sampling. The proposed method, which combines Bayesian optimization with analytical uncertainty quantification, is validated and proves to be efficient.
JOURNAL OF MECHANICAL DESIGN
(2022)
Article
Computer Science, Interdisciplinary Applications
Nicolas Delaisse, Toon Demeester, Dieter Fauconnier, Joris Degroote
Summary: This article presents a new framework for incorporating surrogate models in quasi-Newton methods and compares several existing methods with it. This approach can expedite the convergence by providing an initial solution and a Jacobian matrix.
COMPUTERS & STRUCTURES
(2022)
Article
Engineering, Mechanical
Henri Dolfen, Jeroen De Ridder, Landon Brockmeyer, Elia Merzari, Graham Kennedy, Jeroen De Ridder, Katrien Van Tichelen
Summary: MYRRHA, a prototype of a Generation IV reactor, is being constructed in Mol, Belgium. It is a liquid metal fast reactor that uses lead-bismuth eutectic as a coolant. This research investigates turbulence-induced vibrations and proposes a methodology consisting of multiple steps for accurate prediction of cylinder vibration. The study confirms the moderating effect of friction on fretting and provides important parameters for fretting wear prediction.
JOURNAL OF FLUIDS AND STRUCTURES
(2022)
Article
Engineering, Aerospace
Andy J. Keane, Ivan I. Voutchkov
Summary: The use of conditional generative adversarial networks has become popular in recent years, with companies like Google, IBM, and Facebook experimenting in areas such as facial recognition and image classification. However, similar advancements have been limited in the engineering sector due to the high cost and effort required to generate suitable engineering images, as well as the challenges of integrating synthetic image data into traditional engineering workflows.
Article
Engineering, Aerospace
Jolan Wauters, Joris Degroote, Ivo Couckuyt, Guillaume Crevecoeur
Summary: In this work, a method is proposed to address optimization-under-uncertainty problems by simultaneously minimizing both the mean and variance of the objective in a multi-objective setting. This method allows for choosing the desired level of robustness without the need for repeating the optimization process. The method also ensures that the system will not fail with a prescribed probability of constraint failure.
Article
Materials Science, Textiles
Axel Bral, Lode Daelemans, Joris Degroote
Summary: Textiles and their production machines are increasingly using simulations of the production process, which require reliable structural yarn models. However, these models are often based on simplifying assumptions, and more realistic fiber models have not been used. This paper proposes a new method to obtain a structural yarn model through numerical simulations, using a high-fidelity geometrical yarn model. The method is validated by comparing the results with experimental data.
TEXTILE RESEARCH JOURNAL
(2023)
Article
Biotechnology & Applied Microbiology
Sarah Vandenbulcke, Tim De Pauw, Frank Dewaele, Joris Degroote, Patrick Segers
Summary: Cerebrospinal fluid (CSF) dynamics are essential for maintaining a stable central nervous system environment, but the impact of different physiological processes on CSF flow is not well understood. This study presents a computational model that incorporates physiological processes as boundary conditions to simulate CSF pressures and velocities. The model's accuracy is validated by comparing the simulation results with in vivo measurements.
FRONTIERS IN BIOENGINEERING AND BIOTECHNOLOGY
(2022)
Review
Computer Science, Interdisciplinary Applications
Nicolas Delaisse, Toon Demeester, Rob Haelterman, Joris Degroote
Summary: Fluid-structure interaction simulations can be performed in a partitioned way to couple a flow solver with a structural solver. However, without additional stabilization efforts, the Gauss-Seidel iterations between the solvers can converge slowly or not at all under common conditions. Quasi-Newton methods can stabilize and accelerate the coupling iterations, while still treating the solvers as black boxes and accessing data only at the fluid-structure interface. This review focuses on reformulating various coupling methods in the generalized Broyden framework to highlight their similarities and differences, and also compares their performance.
ARCHIVES OF COMPUTATIONAL METHODS IN ENGINEERING
(2023)
Article
Engineering, Mechanical
Peyman Havaej, Joris Degroote, Dieter Fauconnier
Summary: This paper investigates the effects of double-sided surface waviness on Thermo-Elastohydrodynamic Lubrication (TEHL) in line contacts through numerical simulations. The study examines the surface amplitude, wavelength, relative position, and slide-to-roll ratio. The results show that the classic equivalent deformable body method produces 17%, 21%, and 26% deviations in maximum pressure, temperature, and friction compared to the full two-body simulations. The tangential elastic deformation of surface asperities is identified as the fundamental cause of these deviations, which cannot be described by the equivalent geometry.
TRIBOLOGY INTERNATIONAL
(2023)
Article
Engineering, Mechanical
Peyman Havaej, Joris Degroote, Dieter Fauconnier
Summary: This study compares the film thickness, lubricant temperature, and traction curves of two groups of commonly used constitutive models for lubricants in thermo-elastohydrodynamic lubrication (TEHL) modelling. The results show significant deviations in central film thickness, coefficient of friction (CoF), and maximum lubricant temperature between the different constitutive models. The study highlights the sensitivity of TEHL simulation results to the choice of lubricant constitutive models and the importance of carefully selecting appropriate models for specific applications.
Article
Nuclear Science & Technology
Henri Dolfen, Stefan Vandewalle, Joris Degroote
Summary: The design evaluation of nuclear components using numerical methods typically focuses on ideal conditions, but in reality, the geometry and operating conditions may differ. Understanding and ensuring the safety of nuclear energy systems requires investigating more realistic conditions, such as the deformation of fuel assemblies due to thermal and irradiation effects. A paradigm shift is needed to move from deterministic simulations to simulations involving stochastic processes.
NUCLEAR ENGINEERING AND DESIGN
(2023)
Article
Engineering, Biomedical
Federico Cane, Lucas Delcour, Alberto Cesare Luigi Redaelli, Patrick Segers, Joris Degroote
Summary: Cardiovascular diseases can lead to chronic pathological conditions, and early diagnosis is crucial for treatment. Image-based patient-specific Computational Fluid Dynamics (CFD) models provide detailed information about cardiac blood flow, supporting clinicians in diagnosis and treatment planning. This study developed a Chimera-based method to build a patient-specific 3D CFD model of the heart, considering the challenges of left ventricle and mitral valve motion. The results suggest that the mitral valve plays a crucial role in intraventricular flow development.
FRONTIERS IN MEDICAL TECHNOLOGY
(2022)
Proceedings Paper
Automation & Control Systems
Jolan Wauters, Tom Lefebvre, Guillaume Crevecoeur
Summary: Until recently, model-based design of dynamic mechatronic systems has followed a sequential approach that hinders finding systems with concurrent optimal design and trajectory. To address this, co-design methods have emerged that simultaneously consider design and trajectory optimization. This paper examines two co-design strategies, the simultaneous approach and the nested approach, for the optimal design of a quadcopter. A comparison with a sequential approach shows the added value of co-design in the design phase of dynamical systems and the impact of objective and constraint function formulation.
2022 IEEE/ASME INTERNATIONAL CONFERENCE ON ADVANCED INTELLIGENT MECHATRONICS (AIM)
(2022)
Article
Engineering, Aerospace
Jolan Wauters
Summary: This paper examines the design optimization-under-uncertainty of a forward swept wing blended wing body unmanned aerial vehicle. The traditional blended wing bodies often use a tailless structure, requiring a backward swept wing for longitudinal static stability, which can lead to flow separation and a loss of lift. To solve this problem, a conceptual redesign with a forward swept wing is proposed. The paper introduces a novel framework called SAMURAI, which considers the computational cost by using surrogate modeling and treats the problem as a multi-objective problem to obtain a series of robust and reliable UAV designs.
INTERNATIONAL JOURNAL OF MICRO AIR VEHICLES
(2022)
Article
Engineering, Aerospace
Juraj Mihalik, Andy J. Keane
Summary: This study experimentally investigates the feasibility of using Custer channel wings for slow flight and STOL of small fixed-wing unmanned aerial vehicles, and demonstrates the advantages of channel wings, including significantly increased lift and reduced takeoff and landing distance.
JOURNAL OF AIRCRAFT
(2022)
