检索结果-内蒙古大学图书馆

Subsurface Damage Detection and Quantitative Prediction of Thin-Walled Complex Curved-Surface Component

学校读者我要写书评

暂无评论

SSRN 2024年

作者： Qin, Biao Liu, Henan Cheng, Jian Tian, Jinchuan Sun, Jiangang Zhou, Zihan Ma, Chuanzhen Chen, Mingjun State Key Laboratory of Robotics and System Harbin Institute of Technology P.O. Box 413 Heilongjiang Harbin150001 China Center for Precision Engineering Harbin Institute of Technology P.O. Box 413 Heilongjiang Harbin150001 China

The hemispherical resonant gyroscope (HRG), distinguished by its simple structure, high reliability, and remarkable precision, stands out as one of the most promising high-performance inertial navigation devices in advanced equipment. However, the core sensitive component of the HRG, known as hemispherical resonator (HSR), with a complex and small curvature structure made of fused silica, experiences subsurface damage (SSD) during the grinding process. This not only leads to brittle fracture, which affects the operational performance of the gyroscope, but also significantly hampers the polishing efficiency of the component. This study was initiated by quantitatively characterizing the SSD of various structural features of the HSR (center rod, lip edge of the spherical shell, and spherical shell). Especially, for the curved-surface of the HSR, the average deviation of SSD detection result is within 6%. Subsequently, based on the characterized SSD, the grinding damage mechanisms of the HSR was explored. The grinding surface integrity is then elucidated from the perspective of maximum undeformed chip thickness. Finally, a predictive model for the SSD depth of the HSR after ultra-precision grinding was established according to the nonlinear relationship between the surface roughness and SSD depth. The maximum deviation between the predictive and experimental values of SSD depth is 9.4%, indicating a high level of predictive accuracy. This work fills the blank in SSD detection of the HSR, laying a theoretical foundation for the high-performance manufacturing of the HSR. © 2024, The Authors. All rights reserved.

关键词： Damage detection

A novel cross-scale nanopositioning stage based on dual piezoelectric stick-slip actuation with a shared flexible hinge

学校读者我要写书评

暂无评论

A novel cross-scale nanopositioning stage based on dual piez...

International Conference on Manipulation, Automation and robotics at Small Scales (MARSS)

作者： Siyuan Meng Junhui Zhu Yong Wang Luoning Shi Jia Yu Dongmei Wu Wei Dong Changhai Ru State Key Laboratory of Ro-botics and System Harbin Institute of Technology Harbin China School of Electronic and Infor-mation Engineering Suzhou University of Science and Technology Suzhou China Micro-Nano Automation Institute JITRI Suzhou China State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin China School of Electronic and Information Engineering Suzhou University of Science and Technology Suzhou China

ISBN: (数字)9798350376807

ISBN: (纸本)9798350376814

This paper presents a novel cross-scale nanopositioning stage that integrates the advantages of piezoelectric stick-slip positioners and piezoelectric scanners, utilizing the concept of macro-micro positioning. Initially, a unique structural design featuring a single flexible hinge shared by a small and a large PZT is proposed to reduce the stage's size effectively, enabling millimeter-scale strokes with nanometer-scale positioning accuracy. Subsequently, the cross-scale motion generation mechanism of the dual piezoelectric stick-slip drive is investigated, leading to the development of a system-level dynamics model for the proposed positioning stages. The mechanism design is further optimized to enhance performance. A prototype is manufactured, and a series of experiments are conducted to evaluate the stage's performance. Results indicate that the proposed positioning stage achieves a maximum motion range of 20 mm and a minimum step length of 70 nm in small piezoceramic macro-motion stepping mode, along with a maximum scanning range of 4.9 μm and a motion resolution of 16 nm in large pie-zoceramic micro-motion scanning mode. Moreover, the compact structure of the stage (30×17×8 mm 3 ) supports a maximum motion speed of 10 mm/s and a maximum load of 2 kg. Experi-mental validation confirms the feasibility of the proposed stage, demonstrating nanometer positioning resolution, high accuracy, high speed, and large travel range. These attributes highlight the significant potential of the proposed stage in various in-situ SEM nanorobotic instrument systems.

关键词： Accuracy Automation Nanopositioning Instruments Dynamics Prototypes Mechanism design

Biomimetic Composite Tendon and Pneumatic Actuation for Soft Manipulator: Balancing High Load-Bearing Capacity, Repetitive Precision, and Active Softening Contact Safety

学校读者我要写书评

暂无评论

SSRN 2024年

作者： Li, Xianglong Zhang, Shouyi Xiong, Quan Sui, Dongbao Zheng, Tianjiao Zhang, Qinghua Wang, Ziqi Luan, Lingkai Zhao, Jie Zhu, Yanhe State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin150001 China Department of Biomedical Engineering National University of Singapore Singapore117583 Singapore Ji Hua Laboratory Foshan528200 China Harbin University of Science and Technology Harbin150001 China

Soft manipulator is a prominent research area in the field of soft robotics, but designing a soft manipulator capable of serving human life remains challenging, facing issues such as low motion robustness and poor accuracy. The trunk of an elephant, a unique biological tissue in nature, is characterized by its flexibility, strong load-bearing capacity, and dexterous yet soft tip, distinct from skeletal-dependent movements. Inspired by the thrust-deformation mechanism of the elephant trunk under multi-muscle action, this paper reports a novel composite tendon and pneumatic hybrid-driven tapered soft manipulator that mimics the multi-muscle action of the elephant trunk. In the design, a common Bowden tube is introduced as an artificial tendon to simulate the thrust of elephant trunk muscles. This paper first establishes an equivalent finite element model for Bowden tube tendons, along with an exploration of the impact of constraint ring distribution intervals on tendon buckling during thrust generation. This paper also establishes a finite element model for the entire experimental prototype, validating the effectiveness of Bowden tendon tension and thrust and demonstrating the large deformation curvature and flexible motion space of the soft manipulator. Finally, the paper experimentally tests the stability of repetitive motion, load-bearing capacity, and safety of active softening contact with humans of the soft manipulator, demonstrating that this biomimetic design exhibits not only flexibility and agility but also robustness in motion and precision in accuracy, closely resembling the capabilities of an elephant trunk, and holds promise as flexible supernumerary robotic limbs for mobile or humanoid robots. © 2024, The Authors. All rights reserved.

关键词： Tendons

Tsun: A Traversability-Based Safety-Ensured Uneven Terrain Navigation system

学校读者我要写书评

暂无评论

SSRN 2024年

作者： Wang, Han Wang, Gongcheng Zhu, Hongbiao Du, Chengjin Bai, Hua Ding, Pengchao Xu, Wenda Du, Zhijiang Wang, Weidong The State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin150080 China Istituto Italiano di Tecnologia The BioRobotics Institute Scuola Superiore Sant’Anna Genova16163 Italy

The application of autonomous navigation in mobile robots has seen notable success in indoor, structured 2D environments. However, applying this technology to outdoor, 3D unstructured environments — especially on uneven terrains, presents significant challenges. To address this, we propose a navigation framework for uneven terrain based on continuous traversability analysis. Our methodology unfolds in three stages: First, Bayesian generalized kernel inference is employed to estimate attributes of the unobserved point cloud, enabling the computation of continuous terrain traversability values by using dispersed point cloud data. Second, an improved A* (TRA-A*) algorithm is proposed for global path determination. In the final stage, nonlinear model predictive control (NMPC) is utilized for path following. This involves integrating the traversable exponents and uncertainty into the cost function, resulting in a safety-focused motion planning algorithm that encourages the navigation stability of the robot. The effectiveness of this method was verified through simulation and actual scene experiments. This algorithm improves the robot's navigation success rate by 8% while sacrificing a small amount of 9.83% of the path length in the actual environment, preventing the robot from tipping over or getting stuck in rough terrain, and improving the stationarity of the robot during driving. It reduces the cost of the terrain the robot travels on by 42.6%, reduces the robot's longitudinal acceleration fluctuation and lateral angular velocity fluctuation by 28.52% and 28.9% respectively, and improves the robot's driving stability. © 2024, The Authors. All rights reserved.

关键词： Cost functions

A Novel Multi-Fingered Hand for Robotic Grasp 7

学校读者我要写书评

暂无评论

A Novel Multi-Fingered Hand for Robotic Grasp

7th Annual International Conference on Network and Information systems for Computers, ICNISC 2021

作者： Zang, Xizhe Wang, Chao Zhang, Pu Heng, Shuai Zhao, Jie State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin150001 China Zhejiang Lab Hangzhou311121 China

ISBN: (纸本)9781665402323

As robots are gradually used in more and more scenarios, the requirements for robots to complete a variety of complex tasks are becoming higher and higher. Robotic hands, as the important tool for robot to complete the task, has attracted more and more attention. However, the robotic grasping tool usually has a single function and few degrees of freedom, which makes it difficult to adapt to more complex application scenarios. This paper presented a novel robotic hand with four fingers, which is designed based on the characteristics of human hands. We introduced the design process of the finger structure, analyzed the kinematics and statics of the designed hand, and further manufactured the prototype. To verify the function of the designed multi-fingered hand, we designed a serial of grasping experiments for different objects. The experimental results show that the designed multi-fingered hand can successfully grasp different experimental objects and meet the grasping requirements of the robot. © 2021 IEEE.

关键词： Degrees of freedom (mechanics)

DeepMatcher: A Deep Transformer-based Network for Robust and Accurate Local Feature Matching

学校读者我要写书评

暂无评论

arXiv

arXiv 2023年

作者： Xie, Tao Dai, Kun Wang, Ke Li, Ruifeng Zhao, Lijun State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin150006 China SenseTime Group Inc. Beijing100080 China

Local feature matching between images remains a challenging task, especially in the presence of significant appearance variations, e.g., extreme viewpoint changes. In this work, we propose DeepMatcher, a deep Transformer-based network built upon our investigation of local feature matching in detector-free methods. The key insight is that local feature matcher with deep layers can capture more human-intuitive and simpler-to-match features. Based on this, we propose a Slimming Transformer (SlimFormer) dedicated for DeepMatcher, which leverages vector-based attention to model relevance among all keypoints and achieves long-range context aggregation in an efficient and effective manner. A relative position encoding is applied to each SlimFormer so as to explicitly disclose relative distance information, further improving the representation of keypoints. A layer-scale strategy is also employed in each SlimFormer to enable the network to assimilate message exchange from the residual block adaptively, thus allowing it to simulate the human behaviour that humans can acquire different matching cues each time they scan an image pair. To facilitate a better adaption of the SlimFormer, we introduce a Feature Transition Module (FTM) to ensure a smooth transition in feature scopes with different receptive fields. By interleaving the self- and cross-SlimFormer multiple times, DeepMatcher can easily establish pixel-wise dense matches at coarse level. Finally, we perceive the match refinement as a combination of classification and regression problems and design Fine Matches Module to predict confidence and offset concurrently, thereby generating robust and accurate matches. Experimentally, we show that DeepMatcher significantly outperforms the state-of-the-art methods on several benchmarks, demonstrating the superior matching capability of DeepMatcher. The code is available at https://***/XT1997/DeepMatcher. © 2023, CC BY.

关键词： Behavioral research

Laser Implicit Interaction technology Based on Object Affordance for Wmra

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Liu, Yaxin Liu, Yan Yao, Yufeng Zhong, Ming State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin150001 China

It is particularly important for the disabled to operate the wheelchair mounted robotic arm (WMRA) with as few limb movements as possible. This human-robot implicit interaction problem for the elderly and handicapped has attracted extensive attention. Several studies have documented that laser pointer interaction is an attractive method enabling humans to unambiguously point out objects and pick them up. However, how to further convey user intention and instruct the robot to complete complex tasks based on this method has yet to be studied. Therefore, based on the idea of object affordance, this paper proposes an implicit intention interaction technology using a laser pointer. The user could point to the object in the scene through the residual limb movement ability and the WMRA will reason the action intention and assemble the task content according to the intention. Firstly, a laser semantic identification algorithm based on the YOLOv4 & the SVM is proposed to identify laser points in the scene, which can effectively avoid the error operation caused by hand twitch. Then, an automatic execution technology based on the concept of object affordance is proposed to perform household tasks, which would identify the implicit action intention of the selected objects based on the conditional random field and assemble the mapping action of intention based on the finite state machine. Finally, the feasibility of the proposed implicit interaction technology is verified on the WMRA platform. The evaluation results show that the user only needs a simple operation and the WMRA can recognize the user's desired intentions and perform household tasks automatically. Compared with previous methods, the proposed method can output the desired intention faster and significantly reduce the user's limb involvement time (about 85%) in operating the WMRA under the same task. © 2022, The Authors. All rights reserved.

关键词： Random processes

Simulation and Experimental Study of Piezoelectric Stack Energy Harvester with Arch Beam Applied to the Railway system

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Min, Zhaowei Hou, Chengwei Sui, Guangdong Shan, Xiaobiao Xie, Tao State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin150001 China

As one of the most important modes of transportation, the safety of running vehicles and railway track is significant. It is essential to power sensors that detect and track health in remote areas. The vibration energy of the track structure is enormous and stable, which is not limited by weather factors such as solar and wind power. A new type of arch beam piezoelectric stack energy harvester for railway systems is studied in this paper. Through simulation analysis and experimental verification of the energy harvester, the influences of external resistance, load, pre-pressure, and load frequency on the energy harvesting performance of the piezoelectric energy harvester are discussed. When the frequency is less than 6Hz, the energy capture efficiency is greatly affected by the frequency. When the frequency exceeds 6Hz, the frequency has little effect, and the load dramatically affects the energy capture efficiency. The pre-pressure is not the more significant, the better, but there is an optimal value at 4.5kN. The energy capture circuit can be combined with the supercapacitor to power the sensor, forming a self-power device for real-time monitoring of track health and driving safety. © 2022, The Authors. All rights reserved.

关键词： Arches

YOWO-Plus: An Incremental Improvement

学校读者我要写书评

暂无评论

arXiv

arXiv 2022年

作者： Yang, Jianhua State Key Laboratory of Robotics and System Harbin Institute of Technology Harbin150001 China

In this technical report, we would like to introduce our updates to YOWO, a real-time method for spatio-temporal action detection. We make a bunch of little design changes to make it better. For network structure, we use the same ones of official implemented YOWO, including 3D-ResNext-101 and YOLOv2, but we use a better pretrained weight of our reimplemented YOLOv2, which is better than the official YOLOv2. We also optimize the label assignment used in YOWO. To accurately detection action instances, we deploy GIoU loss for box regression. After our incremental improvement, YOWO achieves 84.9% frame mAP and 50.5% video mAP on the UCF101-24, significantly higher than the official YOWO. On the AVA, our optimized YOWO achieves 20.6% frame mAP with 16 frames, also exceeding the official YOWO. With 32 frames, our YOWO achieves 21.6 frame mAP with 25 FPS on an RTX 3090 GPU. We name the optimized YOWO as YOWO-Plus. Moreover, we replace the 3D-ResNext-101 with the efficient 3D-ShuffleNet-v2 to design a lightweight action detector, YOWO-Nano. YOWO-Nano achieves 81.0 % frame mAP and 49.7% video frame mAP with over 90 FPS on the UCF101-24. It also achieves 18.4 % frame mAP with about 90 FPS on the AVA. As far as we know, YOWO-Nano is the fastest state-of-the-art action detector. Our code is available on https://***/yjh0410/PyTorch_YOWO. © 2022, CC BY.

关键词：