检索结果-内蒙古大学图书馆

38th Conference on Neural information Processing systems, NeurIPS 2024

作者： Guan, Jian Wu, Wei Wen, Zujie Xu, Peng Wang, Hongning Huang, Minlie The CoAI group DCST Institute for Artificial Intelligence China State Key Lab of Intelligent Technology and Systems China Beijing National Research Center for Information Science and Technology China Tsinghua University Beijing100084 China Ant Group China

The notable success of large language models (LLMs) has sparked an upsurge in building language agents to complete various complex tasks. We present AMOR, an agent framework based on open-source LLMs, which reasons with external knowledge bases and adapts to specific domains through human supervision to the reasoning process. AMOR builds reasoning logic over a finite state machine (FSM) that solves problems through autonomous executions and transitions over disentangled modules. This allows humans to provide direct feedback to the individual modules, and thus naturally forms process supervision. Based on this reasoning and feedback framework, we develop AMOR through two-stage fine-tuning: warm-up and adaptation. The former fine-tunes the LLM with examples automatically constructed from various public datasets, enabling AMOR to generalize across different knowledge environments, while the latter tailors AMOR to specific domains using process feedback. Extensive experiments across multiple domains demonstrate the advantage of AMOR to strong baselines, thanks to its FSM-based reasoning and process feedback mechanism. The code and data are publicly available at https://***/JianGuanTHU/AMOR. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Cooperative Multi-robot Target Searching and Tracking Using Velocity Inspired Robotic Fruit Fly Algorithm

引用

SN Computer Science 2021年第6期2卷 1-12页

作者： Garg, Vikram Robotics and Intelligent System Lab ABV- Indian Institute of Information Technology and Management Gwalior India

Target searching and tracking using a swarm of robots is a classical problem in the domain of robotics. The cooperation among the swarm robots has got increasing attention lately due to the different versions of the problem and complex environment. In this paper, a centralized control algorithm is proposed which utilizes cooperation among the swarm of robots for searching and tracking targets which is the velocity inspired robotic fruit fly algorithm (VRFA). The particle velocity concept of the particle swarm optimization is added to the fruit fly algorithm to improve the parameters such as local extremum and low convergence. The simulation results of the proposed technique in different scenarios demonstrate the effectiveness of the algorithm and its ability to keep tracking the targets until the exit condition matched. At last, the simulation result is shown with different environments and parameter settings. © 2021, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.

关键词： Cooperation Fruit fly optimization intelligent search Swarm intelligence Target searching Target tracking VRFA

来源：评论

学校读者我要写书评

暂无评论

Risk Assessment Based Time-adjusted Optimization for Safe and Fast Flight of Quadrotor UAVs

引用

IEEE Transactions on Vehicular technology 2025年

作者： Zhu, Hai Li, Baoquan Yin, Haolin Zhang, Xuebo Tiangong University School of Control Science and Engineering The Tianjin Key Laboratory of Intelligent Control of Electrical Equipment Tianjin300387 China Nankai University Institute of Robotics and Automatic Information Systems The Tianjin Key Laboratory of Intelligent Robotics Tianjin300071 China

Generating an executable reference trajectory in realtime is a core of autonomous navigation for a quadrotor unmanned aerial vehicle (UAV). In the paper, a timeadjusted optimization method is proposed for UAV motion planning to achieve safe and fast flight. Firstly, based on parameterization for uniform B-spline trajectories, relationship of knot spans and UAV speeds is analyzed, and an iterative velocity restrict algorithm is introduced to tackle nonexecutable flight. Subsequently, combining start and target points in local planning, an evaluation criterion function is designed with control points of the above trajectory and directions of collision avoidance. On this basis, risk assessment is conducted according to surrounding environments, where four phases of passing obstacles is constructed by considering risk level variation. Then, the weight function in each phase is determined by relationship of obstacle and planned local trajectory. Moreover, time intervals of B-spline trajectories are adjusted to obtain nonuniform trajectories, so as to affect UAV speeds. Simulation and experiment are conducted with comparison to verify performance of the proposed method. © 2025 IEEE.

关键词： Risk assessment

来源：评论

学校读者我要写书评

暂无评论

Pedestrian Target Detection Model Based on Improved Yolov8 5

Pedestrian Target Detection Model Based on Improved Yolov8

引用

5th International Conference on Artificial Intelligence and Electromechanical Automation, AIEA 2024

作者： Huang, Jinglun Liang, Jianan Cao, Yongjun School of Electronics and Information Engineering Wuyi University Guangdong Jiangmen China Guangdong Key Laboratory of Modern Control Technology Institute of Intelligent Manufacturing Gdas Guangdong Guangzhou China South China Robotics Innovation Research Institute Guangdong Foshan China School of Mechanical & Automotive Engineering South China University of Technology Guangdong Guangzhou China

ISBN: (纸本)9798350366174

Dense pedestrian detection is a key technology in public intelligent surveillance. It identifies the location and number of pedestrians in videos through object detection methods, thereby achieving intelligent surveillance of crowds. However, in crowded scenes, due to occlusion and small pedestrian targets, it is easy to miss pedestrians. To solve this problem, this paper proposes an improved YOLOv8 detection algorithm, YOLOv-8FastTCD. Firstly, the FasterNet module is used to replace the backbone network. Secondly, a task collaborative dynamic detection head (TCDDH) is introduced after the neck network. The performance of this model on the CrowdHuman dense pedestrian detection dataset has been improved, with mAP50 and mAP50-95 reaching 78.4% and 51.5% respectively, which are 2.1% and 2.0% higher than the original YOLOv8n model. © 2024 IEEE.

关键词： FasterNet Pedestrian Detection TCDDH YOLOv8

来源：评论

学校读者我要写书评

暂无评论

Constructing Lithium-Free Anode/Separator Interface via 3D Carbon Fabric Scaffold for Ultrasafe Lithium Metal Batteries

引用

Research 2024年第3期2023卷 359-368页

作者： Dongdong Li Shengchen Yang Zijian Zheng Wen-Yong Lai State Key Laboratory of Organic Electronics and Information Displays(SKLOEID) Institute of Advanced Materials(IAM)School of Chemistry and Life SciencesNanjing University of Posts and Telecommunications9 Wenyuan RoadNanjing 210023China Department of Materials Chemistry Huzhou University1 Xueshi RoadHuzhou 313000China Laboratory for Advanced Interfacial Materials and Devices School of Fashion and TextilesDepartment of Applied Biology and Chemical TechnologyResearch Institute for Intelligent Wearable Systems(RI-IWEAR)Research Institute for Smart Energy(RISE)The Hong Kong Polytechnic UniversityHong Kong SAR 999077China

Metallic lithium represents a promising anode candidate to be utilized in future high-energy lithium ***,the undesirable dendrite growth and fragile solid-electrolyte interphase(SEI)pose critical challenge for pursuing further practical *** contrast to traditional approaches of using inert/lithiophilicity coating,here,we demonstrate a reverse strategy of introducing a highly conductive and lithophobic carbon fabric(CF)scaffold on lithium foil to guide a favorable nucleation site of lithium far away from the anode/separator *** CF scaffold with high conductivity can couple with inner electric field for achieving a uniform distribution of the lithium-ion flux,while the lithophobic feature offers the condition to guide the preferred deposition of lithium onto the underlying lithium foil,which greatly reduces the risk of dendrite-induced short ***,the SEI immersed in the CF scaffold is well supported by CF fibers and therefore exhibits extremely high stability during charge–discharge *** a result,the lithium/CF anodes show>2,000-h stable cycling at 0.5 mA cm−*** metal batteries equipped with our lithium/CF anode deliver a high capacity retention of~99.99%per cycle,i.e.,retain~97.3%capacity after 200 *** unique interface-regulation strategy is versatile for various conductive scaffolds(e.g.,ultrathin and ultralight conductive fabrics),exhibiting high superiority for highly safe lithium metal batteries.

关键词： lithium equipped versatile

来源：评论

学校读者我要写书评

暂无评论

Fault Diagnostic Opportunities for Robot Servo Motors Using the Physics of Failure Analysis 14

Fault Diagnostic Opportunities for Robot Servo Motors Using ...

引用

14th International Conference on Reliability, Maintainability and Safety, ICRMS 2023

作者： Zhang, Bo Liang, Shan Wang, Kai Key Laboratory of Networked Control Systems Shenyang Institute of Automation Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang China College of Computer Science and Technology University of Chemical Technology Chinese Academy of Sciences Shenyang China

ISBN: (纸本)9798350329988

Joint robots are widely used in various industries, such as aviation, aerospace, and automotive manufacturing. The performance degradation of servo motors can significantly affect the overall performance of the robots because of the core component of joint robots, and therefore, it is critical to detect and address servo motor failures promptly. In this study, we analyze some common causes and mechanisms of performance degradation in robot servo motors using the physics of failure analysis. We also discuss strategies for detecting and diagnosing faults using these mechanisms before the servo motors fail completely. © 2023 IEEE.

关键词： Servomotors

来源：评论

学校读者我要写书评

暂无评论

Unified Addressing Algorithm for Tasks in Cloud Control systems 41

Unified Addressing Algorithm for Tasks in Cloud Control Syst...

引用

第41届中国控制会议

作者： Guan Wang Yufeng Zhan Yuanqing Xia Liping Yan Key Laboratory of Intelligent Control and Decision of Complex Systems School of AutomationBeijing Institute of Technology School of Information Science and Engineering University of ZaoZhuang

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

With the rapid development of cloud computing and control theory,a new paradigm of networked control systems named cloud control systems has been proposed to satisfy the requirements of large-scale and complex *** is critical to manage large-scale tasks with personalized needs on the large number of heterogeneous ***,current TCP/IPbased addressing approach does not consider the runtime constraints of tasks,resulting highly concurrent service search,and communication between tasks will become the system *** this paper,we propose a task property encoding method which takes multi-dimensional properties into *** on the proposed encoding method,we design a uniform addressing method for tasks in the ***,we experimentally demonstrate the effectiveness of task property encoding,and the unified addressing method for tasks is more effective than other methods in the case of high task concurrency.

关键词： Cloud control systems address encoding cloud-edge-device cooperation

来源：评论

学校读者我要写书评

暂无评论

Comparison of ROS Local Planners for a Holonomic Robot in Gazebo Simulator 1

引用

Proceedings of the 8th International Conference on Interactive Collaborative robotics, ICR 2023

作者： Apurin, Artem Abbyasov, Bulat Martínez-García, Edgar A. Magid, Evgeni Intelligent Robotics Department Institute of Information Technology and Intelligent Systems Kazan Federal University 35 Kremlyovskaya street Kazan420111 Russia Institute of Engineering and Technology Department of Industrial Engineering and Manufacturing Autonomous University of Ciudad Juarez Manuel Díaz H. No. 518-B Zona Pronaf Condominio Chihuahua Cd Juárez32315 Mexico Tikhonov Moscow Institute of Electronics and Mathematics HSE University 34 Tallinn street Moscow123458 Russia

ISBN: (数字)9783031431111

ISBN: (纸本)9783031431104

A safe robot navigation in a dynamic environment is an essential part of an autonomous exploration path planning. A path planning part of a navigation involves global and local planners. While a global planner finds an optimal path with a prior knowledge of an environment and static obstacles, a local planner recalculates the path to avoid dynamic obstacles. The main goal of a local planning is adjusting an initial plan produced by a global planner in an online fashion. It is a crucial step to ensure a robot operation in dynamic environments because in real world scenarios an environment usually contains people and thus, a dynamic obstacles avoidance must respond quickly and recalculate an actual route. Holonomic robotic platforms are robotic vehicles that use omni-wheels to move in any direction, at any angle, without an additional rotation. These robotic platforms are ideal for working zones with a limited space access. This paper provides a comparison of ROS local planners that support omni-wheel mobile robots: Trajectory Rollout, DWA, EBand, and TEB. The algorithms were compared using a path length, a travelling time and a number of obstacle collisions. Gazebo simulator was used for modeling virtual scenes with dynamic obstacles. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Mobile robots

来源：评论

学校读者我要写书评

暂无评论

Robot Grasp with Multi-object Detection based on RGB-D Image

Robot Grasp with Multi-object Detection based on RGB-D Image

引用

2021 China Automation Congress, CAC 2021

作者： Ouyang, Weixing Huang, Wenhui Min, Huasong Institute of Robotics and Intelligent Systems Wuhan University of Science and Technology Wuhan China

ISBN: (纸本)9781665426473

An autonomous grasping system based on the method of depth learning is built in this paper, which is composed of a depth camera, a two-finger parallel gripper and a six-degree-of-freedom mechanical arm. To solve the poor performances of recognition and location of the grasping system in complex scenes with multiple irregular objects of different size, a multi-object grasping detection method based on RGB-D images is proposed. This method uses the combination of FPN and RPN methods to design the grasp proposal network, which can simultaneously use the shallow and deep information in the network to improve the detection effect of objects with different sizes, especially small objects. Secondly, the ROI Align method is introduced as a regional feature aggregation method, and the bilinear interpolation method can significantly improve the detection accuracy of the grasping model. Finally, the multiobject dataset containing small objects is used to test the proposed model, and the result illustrates the generalization ability of the proposed model. Experimental results show that the accuracy of the method proposed in this paper achieves 97% accuracy on single-object recognition and 92% accuracy on multi-object recognition. The robot grasping system based on Kinect camera and UR5 robotic arm achieved 91.0% grasping success rate on single-object, and 89% grasping success rate on multi-object. Comparing with some popular grasping methods, the results show the effectiveness and superiority of the proposed method in this paper. © 2021 IEEE

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

An Underwater Acoustic Signal Denoising Method for Broadband Interference from Underwater Platforms Based on the Deep Neural Network Model 6

An Underwater Acoustic Signal Denoising Method for Broadband...

引用

6th International Conference on information Communication and Signal Processing, ICICSP 2023

作者： Cao, Linxuan Lei, Zhixiong Zhou, Zihao Sun, Jie Shi, Yang School of Marine Science and Technology Northwestern Polytechnical University Shaanxi Key Laboratory of Underwater Information Technology Xi'an China Southwest China Institute of Electronic Technology Chengdu610036 China Shenyang Institute of Automation Chinese Academy of Sciences State Key Laboratory of Robotics Shenyang110016 China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110169 China

ISBN: (纸本)9798350339994

As underwater vehicles widely used in ocean observation, the self-noise generated from the underwater platforms will affect the underwater acoustic signal observation, impeding signal detection. In this paper, we proposed an ideal binary mask (IBM) based Deep Neural Network (DNN) denoising method to reduce the broadband transient interference that underwater platforms caused. First, we analysis the interference produced by the underwater glider, illustrating its hazard by a common noisy situation. Then the IBM based DNN model was train to reduce this interference. Compared with spectral subtraction and Wiener filtering method, the method we proposed has short processing time of 0.15 $s$ in average, 7 dB higher signal-to-interference ratio (SIR) gain, and more complete retention of the original signal components in low SIR situations. © 2023 IEEE.

关键词： broadband interference deep neural network ideal binary mask transient signal underwater acoustic signal denoising

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：