Article
Construction & Building Technology
Zhicheng Wang, Zhenggang Cao, Feng Fan, Ying Sun
Summary: A novel shape optimization method is proposed in this paper to improve the mechanical performance of free-form grid structures while satisfying architectural requirements. By considering control point height as optimization variables and structural strain energy and a comprehensive quantitative index as optimization objectives, a sensitivity hybrid multi-objective evolutionary algorithm (SH-MOEA) is developed for shape optimization. The results demonstrate that the developed algorithm outperforms other algorithms in terms of accuracy, uniformity, and computational efficiency, effectively improving the similarity of surface, fluence, and regularity of free-form grids.
JOURNAL OF BUILDING ENGINEERING
(2021)
Article
Construction & Building Technology
Xingye Wang, Shaojun Zhu, Qiang Zeng, Xiaonong Guo
Summary: An improved multi-objective hybrid genetic algorithm is proposed for the shape and size optimization of large-scale free-form spatial latticed structures. The method uses B-spline theory and incorporates fully stressed design to obtain an optimal set of cross-sections of the members.
JOURNAL OF BUILDING ENGINEERING
(2021)
Article
Computer Science, Interdisciplinary Applications
Sejin Kim, Innyoung Kim, Donghyun You
Summary: A novel multi-condition multi-objective optimization method using deep reinforcement learning has been developed to find the Pareto front within a defined condition space. This method learns the correlations between conditions and optimal solutions, providing a unique capability in solving problems with nonlinear characteristics. The method has successfully determined the Pareto front with high resolution in both a modified Kursawe benchmark problem and an airfoil shape optimization problem. Compared to single-condition optimization methods, this multi-condition optimization method greatly accelerates the search for the Pareto front by reducing the required number of function evaluations. The analysis of the aerodynamic performance of optimized airfoils confirms the indispensability of multi-condition optimization in avoiding significant degradation of target performance under varying flow conditions.
JOURNAL OF COMPUTATIONAL PHYSICS
(2022)
Article
Engineering, Civil
Jinglan Cui, Yanfeng Zheng, Makoto Ohsaki, Yaozhi Luo
Summary: A new structural morphology method is proposed in this paper for assembling a free-form grid surface using plate elements by introducing a new directing condition and choosing control points of directing Be ' zier curves as design variables. The optimization problem aims to minimize the strain energy of the grid structure subjected to static loads under developability conditions, directing conditions, and continuity conditions as constraints. Numerical examples demonstrate the effectiveness of the proposed method in generating a new rational architectural shape integrating aesthetics, mechanics, and constructability for free-form grid surface structures.
ENGINEERING STRUCTURES
(2021)
Article
Mathematics, Applied
Feifei Yang, Tiantang Yu, Zhaowei Liu, Tinh Quoc Bui
Summary: The paper introduces a method of shape optimization for free-form surface structures using isogeometric analysis, and verifies the effectiveness and performance of the proposed method through numerical examples. The mechanical properties of the optimized structure are substantially improved.
FINITE ELEMENTS IN ANALYSIS AND DESIGN
(2023)
Article
Computer Science, Interdisciplinary Applications
Peter Gangl, Stefan Koethe, Christiane Mellak, Alessio Cesarano, Annette Muetze
Summary: This paper presents a gradient-based free-form shape optimization method for designing a synchronous reluctance machine used in an X-ray tube. The method utilizes the mathematical concept of shape derivatives to obtain optimal designs without introducing geometric parametrization. It also extends the optimization algorithm to handle multiple objective functions and demonstrates a way to obtain an approximate Pareto front. The results show that the presented method can achieve optimal designs with low computational cost compared to a stochastic optimization algorithm.
COMPEL-THE INTERNATIONAL JOURNAL FOR COMPUTATION AND MATHEMATICS IN ELECTRICAL AND ELECTRONIC ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Boyang Qu, Guosen Li, Li Yan, Jing Liang, Caitong Yue, Kunjie Yu, Oscar D. Crisalle
Summary: This paper proposes a grid-guided particle swarm optimizer for solving multimodal multi-objective optimization problems. By using a grid in the decision space, the algorithm is able to detect promising subregions and generate multiple subpopulations, maintaining diversity and improving search efficiency. Experimental results demonstrate that the proposed algorithm outperforms other evolutionary methods.
APPLIED SOFT COMPUTING
(2022)
Article
Engineering, Civil
Baoshi Jiang, Jingyao Zhang, Makoto Ohsaki
Summary: This study presents four shape optimization problems to obtain rational shapes for free-form shell structures with high static and dynamic performances. The measure of static performance is the strain energy under static loads, while the measure of dynamic performance is the lower bound on the fundamental natural frequency or strain energy under seismic loads. Multiple numerical examples are provided to demonstrate the effectiveness of the proposed method.
Article
Computer Science, Artificial Intelligence
Junzhong Ji, Yannan Weng, Cuicui Yang, Tongxuan Wu
Summary: This paper presents a multi-resolution grid-based bacterial foraging optimization algorithm (MRBFO) to solve multiobjective optimization problems (MOPs). MRBFO redesigns four tailored optimization mechanisms and introduces a multi-resolution grid strategy to search for optimal nondominated solutions. The empirical results demonstrate the advantages of MRBFO.
SWARM AND EVOLUTIONARY COMPUTATION
(2022)
Article
Computer Science, Information Systems
Yule Wang, Wanliang Wang, Ijaz Ahmad, Elsayed Tag-Eldin
Summary: This paper proposes a multi-objective quantum-inspired seagull optimization algorithm (MOQSOA) to optimize the convergence and distribution of solutions in multi-objective optimization problems. The algorithm utilizes opposite-based learning, seagull behavior simulation, and principles of quantum computing to enhance the performance of MOPs in terms of distribution and convergence.
Article
Computer Science, Interdisciplinary Applications
Yan Su, Yue Wu, Wei Ji, Xiaoying Sun
Summary: The development of free-form grid structures has been facilitated by computer-aided design and advanced construction technology, however, grid generation remains a challenge. The use of Voronoi diagrams in computational morphogenesis helps form irregular grids for these structures, with the approach proving to be effective in generating visually appealing and structurally sound grids. The consideration of symmetry during the grid generation process can help accelerate it, while different grid distributions can be achieved by varying support conditions.
COMPUTER-AIDED CIVIL AND INFRASTRUCTURE ENGINEERING
(2021)
Article
Engineering, Mechanical
Ying Zhou, Liang Gao, Hao Li
Summary: This study proposes an optimization method for the design of infill for free-form surfaces. The method simultaneously optimizes the microstructural geometry and macroscopic distribution of the infills. It utilizes a local level sets approach to generate graded and connectable infill microstructures, and employs cubic polynomials interpolation to predict their effective properties. A computational conformal mapping technology is used to map the geometry between the 3D surface and a 2D parameter space, allowing the infill units to fit the curved surface without losing any geometric feature. The infill optimization problem is then recast as a 2D optimization problem defined in the parameter domain.
INTERNATIONAL JOURNAL OF MECHANICAL SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Guoqing Li, Wanliang Wang, Weiwei Zhang, Zheng Wang, Hangyao Tu, Wenbo You
Summary: This paper proposes a grid search based multi-population particle swarm optimization algorithm (GSMPSO-MM) to handle multimodal multi-objective optimization problems (MMOPs), aiming to balance diversity and convergence by adopting multiple populations and grid search methods. The environmental selection operator updates the non-dominated solution archive to improve solution quality.
SWARM AND EVOLUTIONARY COMPUTATION
(2021)
Article
Computer Science, Artificial Intelligence
C. Haider, F. O. de Franca, B. Burlacu, G. Kronberger
Summary: We describe and analyze algorithms for shape-constrained symbolic regression, which incorporate prior knowledge about the shape of the regression function. These algorithms are tested on physics models and compared to previous results achieved with single objective algorithms. The results show that the multi-objective algorithms can find mostly valid models even when using a soft-penalty approach. NSGA-II achieves the best overall rankings on instances with high noise.
APPLIED SOFT COMPUTING
(2023)
Article
Computer Science, Artificial Intelligence
Wei Chen, Faez Ahmed
Summary: This paper introduces a new generative adversarial network called MO-PaDGAN, which improves performance and coverage by introducing a loss function based on Determinantal Point Processes to model diversity and performance simultaneously in solving design problems.
APPLIED SOFT COMPUTING
(2021)