Article
Public, Environmental & Occupational Health
Wenbo Sun, Jiacheng Liu, Jingwen Hu, Judy Jin, Kevin Siasoco, Rongrong Zhou, Robert Mccoy
Summary: This study develops an adaptive restraint system using population-based simulations and machine-learning algorithms to consider occupant anthropometry variations and enhance safety balance. By conducting 2000 crash simulations and using a Gaussian-process-based surrogate model, an optimal adaptive restraint design policy is obtained. The results show that the adaptive design can further reduce joint injury risk among the whole population, with higher reductions observed for vulnerable subgroups such as tall obese males and short obese females.
FRONTIERS IN PUBLIC HEALTH
(2023)
Article
Engineering, Multidisciplinary
Wei Cong, Yong Zhao, Bingxiao Du, Senlin Huo, Xianqi Chen
Summary: This paper studies the spacecraft equipment layout optimization design problems with complicated performance constraints and diversity. By introducing similarity measures and optimization algorithms, the geometric diversity of layout schemes is considered to generate diversified layout schemes. The validity and effectiveness of the proposed methodology are demonstrated by two practical applications.
CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Jan-Hendrik Bastek, Dennis M. Kochmann
Summary: In this study, video diffusion generative models are used to predict and tune the nonlinear deformation and stress response of periodic stochastic cellular structures, including buckling and contact, which greatly simplifies and accelerates the identification of complex material properties.
NATURE MACHINE INTELLIGENCE
(2023)
Article
Mechanics
Liwei Wang, Anton van Beek, Daicong Da, Yu-Chin Chan, Ping Zhu, Wei Chen
Summary: This paper proposes a data-driven topology optimization approach for multiscale cellular design, utilizing a newly proposed latent-variable Gaussian process enhanced with separable kernels to achieve continuous transition of stiffness matrices between different microstructure classes. The method enables better performance in maximizing natural frequencies compared to single-scale designs, and can be easily extended to other multiscale optimization problems.
COMPOSITE STRUCTURES
(2022)
Article
Computer Science, Interdisciplinary Applications
Sourabh Shende, Andrew Gillman, Philip Buskohl, Kumar Vemaganti
Summary: Predicting global optima for non-convex and expensive objective functions is a challenge in various engineering applications. Bayesian optimization is a powerful method for solving such problems, but selecting the right surrogate model and hyperparameters for optimal convergence speed is a difficult task. This study systematically analyzes the computational costs of Bayesian optimization and evaluates two different modifications for improved performance. The results provide insights into the trade-offs and cost distribution between the modifications, as well as guidelines for implementation in new problems.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2022)
Article
Engineering, Multidisciplinary
Bobin Guan, Min Wan, Xiangdong Wu, Xuexi Cui, Yisheng Zhang
Summary: This study proposes a novel lightweight design process that incorporates the influence of assembly connection and non-probabilistic uncertainty in material properties. The proposed method is applied to the structural design of a forming machine, and the results demonstrate the importance of considering these factors in ensuring safe and lightweight structures.
ENGINEERING OPTIMIZATION
(2023)
Article
Computer Science, Interdisciplinary Applications
Sourabh Shende, Andrew Gillman, David Yoo, Philip Buskohl, Kumar Vemaganti
Summary: Bayesian optimization (BO) is a popular method for solving optimization problems with expensive objective functions, although its application in structural optimization is still in early stages. By using a Gaussian process (GP) as a surrogate model, BO demonstrates the ability to efficiently solve origami-inspired design problems, showing better computational efficiency compared to traditional methods.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2021)
Article
Energy & Fuels
Alexander Epp, Johannes Christofer Hahn, Dirk Uwe Sauer
Summary: This study proposes an optimization strategy for the holistic design of battery systems by utilizing coupled simulations and combining Gaussian Process Regression and Classification methods. The results show promising outcomes in terms of meeting different technical requirements.
JOURNAL OF ENERGY STORAGE
(2022)
Article
Materials Science, Multidisciplinary
Fanping Sui, Ruiqi Guo, Zhizhou Zhang, Grace X. Gu, Liwei Lin
Summary: This paper introduces the concept of digital materials and their application in composite material design. Through a deep reinforcement learning scheme, an automated process for digital material design is achieved, resulting in improved design quality and significant computational advantages.
ACS MATERIALS LETTERS
(2021)
Article
Multidisciplinary Sciences
Sanmun Kim, Chanhyung Park, Shinho Kim, Haejun Chung, Min Seok Jang
Summary: This work reports on the influence of design parameters on the optical efficiency of metasurface-based color splitters, as well as the possibility of fabricating them in legacy fabrication facilities with low structure resolutions.
Article
Economics
Hadis Anahideh, Lulu Kang, Nazanin Nezami
Summary: This article aims to design a fair resource allocation approach that maximizes geographical diversity and avoids unfairness based on demographic differences. Given the COVID-19 pandemic, it is important to prioritize the allocation of medical resources to vulnerable populations, such as poor communities and minority groups, to effectively prevent further spread of the virus.
SOCIO-ECONOMIC PLANNING SCIENCES
(2022)
Article
Computer Science, Software Engineering
Seowoo Jang, Soyoung Yoo, Namwoo Kang
Summary: Generative design is a computational design method that automatically explores designs under predefined constraints. This study proposes a reinforcement learning-based generative design process to maximize the diversity of topology designs. By approximating the optimization process with neural networks, the study demonstrates the effectiveness of the proposed method in an automotive wheel design case study.
COMPUTER-AIDED DESIGN
(2022)
Article
Business
Cristina O. Vlas, Orlando C. Richard, Goce Andrevski, Alison M. Konrad, Yang Yang
Summary: This study finds that implementing diversity management routines, such as mentoring programs, network groups, internship-based recruiting practices, and succession planning for racial minorities, enhances a firm's ability to compete with a wider range of competitive actions, leading to increased financial performance. The presence of diversity cognition routines also plays a moderating role in the indirect effect of diversity management routines on company performance.
JOURNAL OF BUSINESS RESEARCH
(2022)
Article
Mechanics
Kalyana B. Nakshatrala
Summary: This study provides a theoretical analysis of material design problems for fluid flow through porous media using the adjoint state method. The results offer rigorous answers to how to pose such design problems and have significant implications for computational material design.
Article
Engineering, Electrical & Electronic
Yanwen Xu, Anand Vikas Lalwani, Kanika Arora, Zhuoyuan Zheng, Anabel Renteria, Debbie G. G. Senesky, Pingfeng Wang
Summary: This article proposes a physics-informed machine learning technique to optimize the geometry design of Hall magnetic sensors. By establishing multiphysics-based finite element models and Gaussian process-based surrogate models, the performance of Hall sensors can be effectively studied and optimized. The consistency between the research results and experimental results demonstrates the feasibility and effectiveness of this method.
IEEE SENSORS JOURNAL
(2022)