Article
Engineering, Multidisciplinary
Wenke Qiu, Shaomeng Jin, Liang Xia, Tielin Shi
Summary: This work develops length scale control schemes for bi-directional evolutionary structural optimization method, enabling an enhanced and flexible control of structural member sizes. The schemes involve constraining local material volumes and post-processing modification of local features to control both maximum and minimum structural length scales, which are proven to be effective and efficient based on benchmark design results in both 2D and 3D cases.
INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING
(2022)
Article
Computer Science, Software Engineering
Aaditya Chandrasekhar, Krishnan Suresh
Summary: In this paper, an approximate length scale filter strategy for topology optimization (TO) is proposed by extending a density-based TO formulation using neural networks (TOuNN). The proposed method enhances TOuNN with a Fourier space projection to approximately control the minimum and/or maximum length scales. The method does not involve additional constraints and automates sensitivity computations using the neural net's library.
COMPUTER-AIDED DESIGN
(2022)
Article
Engineering, Multidisciplinary
Longlong Song, Jian Zhao, Tong Gao, Jiajia Li, Lei Tang, Yang Li, Weihong Zhang
Summary: In this paper, length scale control methods are proposed for density-based multi-material topology optimization, which are effective in controlling feature length scales and improving the manufacturability of optimized structures.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2022)
Article
Computer Science, Interdisciplinary Applications
Hui Liu, Peng Wei, Michael Yu Wang
Summary: This paper proposes an adaptive parameterized level set topology optimization method (APLSM) using a bilinear basis function and applies CPU parallel computing strategy. The method avoids solving an additional linear system and improves computational efficiency. It can also design structures with high geometric complexity and has been verified effective in 2D and 3D problems.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2022)
Article
Engineering, Multidisciplinary
Quhao Li, Guowei Liang, Yunfeng Luo, Fengtong Zhang, Shutian Liu
Summary: Topology optimization is widely used in engineering for innovative designs, but the optimized results often lack manufacturability. This study proposes an explicit and general method to control the minimum length scale in topology optimization, which is accurate and easily implemented. By computing the average density of elements in a small circular region, all local constraints are aggregated into a single constraint, and the sensitivity analysis of the constraint function is derived. Numerical examples demonstrate the effectiveness of the proposed algorithm.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2023)
Article
Engineering, Multidisciplinary
Fusheng Qiu, Hongliang Liu, Hongjuan Zhao
Summary: This paper proposes a feature modeling approach for 3D structural topology design optimization with feature constraints. The method is flexibly applied to structural design optimization with added holes by changing constraint factors, enabling more direct and easier design of structures with predetermined features.
CMES-COMPUTER MODELING IN ENGINEERING & SCIENCES
(2021)
Article
Computer Science, Interdisciplinary Applications
Yingchun Bai, Jiale Cai, Zixiang Wang, Siqi Li
Summary: This paper proposes a magneto-structural topology optimization method considering additive manufacturing constraints, which can generate and fabricate superior-performance yet lightweight magneto-structural components through the sequential use of topology optimization and additive manufacturing. The design manufacturability of optimized designs is improved by incorporating two important additive manufacturing constraints into the topology optimization model.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2022)
Article
Engineering, Multidisciplinary
Chen ShiTong, Wang YuSheng, Li DeChen, Chen FeiFei, Zhu XiangYang
Summary: Soft robots have advantages in adaptive and safe interactions, but also require proper stiffness to withstand external loads. This work proposes a computational design framework for soft grippers inspired by human hand dexterity, optimizing skeleton layouts for improved interaction performance in different scenarios.
SCIENCE CHINA-TECHNOLOGICAL SCIENCES
(2021)
Article
Engineering, Multidisciplinary
Ying Zhou, Jihong Zhu, Haifei Zhan, Weihong Zhang, Yuantong Gu
Summary: This paper introduces a feature-driven topology optimization method based on B-spline offset feature inspired by the geometric morphology of worms. The proposed method shows high deformability and ability to effectively handle irregular shapes, demonstrating advantages in stiffest structure design and compliant mechanism synthesis.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2021)
Article
Computer Science, Artificial Intelligence
Xu Weiyao, Wu Muqing, Zhu Jie, Zhao Min
Summary: This paper proposes a multi-scale skeleton adaptive weighted graph convolution network (MSAWGCN) for skeleton-based action recognition, which extracts more abundant spatial features of skeletons through multi-scale skeleton graph convolution network and adopts a simple graph vertex fusion strategy to learn the graph topology adaptively.
APPLIED SOFT COMPUTING
(2021)
Article
Computer Science, Information Systems
Kai Hu, Junlan Jin, Chaowen Shen, Min Xia, Liguo Weng
Summary: This paper proposes a novel attentional weighting strategy-based dynamic GCN (AWD-GCN) to extract discriminative action features by capturing the dynamic relationships among the three partitions of the human skeleton. The method uses a new dynamic adjacency matrix and attention weighting mechanism. Multi-scale position attention and multi-level attention are also proposed to differentiate human action in different spatial scales. Experimental results on challenging datasets demonstrate the effectiveness and superiority of the proposed method.
MULTIMEDIA SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Minsi Wang, Bingbing Ni, Xiaokang Yang
Summary: The paper proposes a multi-view interactional graph network (MV-IGNet) that can construct, learn, and infer multi-level spatial skeleton contexts. MV-IGNet utilizes different skeleton topologies as multi-views to generate complementary action features. Compared to mainstream methods, MV-IGNet has a smaller model size and faster inference.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2023)
Article
Robotics
Gaofeng Li, Fernando Caponetto, Edoardo Del Bianco, Vasiliki Katsageorgiou, Ioannis Sarakoglou, Nikos G. Tsagarakis
Summary: The use of Signed Distance Function (SDF) to describe the geometric information of the workspace and precompute the corresponding nearest points for unreachable points allows for a time-saving workspace limit approach. This approach effectively addresses the discontinuity problem inherited from the discretized SDF and has been shown to be effective in experiments.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Automation & Control Systems
Linhan Li, Guanci Yang, Yang Li, Dongying Zhu, Ling He
Summary: In this study, we propose an abnormal sitting posture recognition method based on the multi-scale spatiotemporal features of skeleton graph (ASPR) to fuse the spatial dimension, temporal dimension, and whole-body skeleton features. We first build a human abnormal sitting posture dataset (HASP) with multidimensional features. Then, we propose a multiple scale spatiotemporal feature extraction model based on graph convolutional network (M2SGCN) to capture the spatiotemporal features. Additionally, a feature extraction model of local skeletal angles based on recurrent neural network is used to capture the change rule of skeletal angles of human sitting posture. Experimental results demonstrate the superior performance of ASPR compared to state-of-the-art models.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Engineering, Multidisciplinary
Xiaojun Wang, Jiazheng Zhu, Bowen Ni
Summary: This article introduces a reliability-based structural optimization method and proposes a credible sequential optimization strategy. By quantifying credibility and establishing a reliability index, this method significantly improves the computational efficiency of the optimization process while maintaining a high level of credibility for structural parameters. Engineering examples are used to emphasize the necessity of considering credibility in structural optimization methods.
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING
(2024)
