检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Xu, Chongyang Huang, Buzhen Zhang, Chengfang Feng, Ziliang Wang, Yangang College of Computer Science Sichuan University Chengdu610065 China Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education The School of Automation Southeast University Nanjing210096 China Intelligent Policing Key Laboratory of Sichuan Province Sichuan Police College Luzhou646000 China

Human mesh recovery can be approached using either regression-based or optimization-based methods. Regression models achieve high pose accuracy but struggle with model-to-image alignment due to the lack of explicit 2D-3D correspondences. In contrast, optimization-based methods align 3D models to 2D observations but are prone to local minima and depth ambiguity. In this work, we leverage large vision-language models (VLMs) to generate interactive body part descriptions, which serve as implicit constraints to enhance 3D perception and limit the optimization space. Specifically, we formulate monocular human mesh recovery as a distribution adaptation task by integrating both 2D observations and language descriptions. To bridge the gap between text and 3D pose signals, we first train a text encoder and a pose VQ-VAE, aligning texts to body poses in a shared latent space using contrastive learning. Subsequently, we employ a diffusion-based framework to refine the initial parameters guided by gradients derived from both 2D observations and text descriptions. Finally, the model can produce poses with accurate 3D perception and image consistency. Experimental results on multiple benchmarks validate its effectiveness. The code will be made publicly available. © 2025, CC BY.

关键词： Mesh generation

来源：评论

学校读者我要写书评

暂无评论

Join multiple Riemannian manifold representation and multi-kernel non-redundancy for image clustering

引用

CAAI Transactions on Intelligence Technology 2024年第5期9卷 1305-1319页

作者： Mengyuan Zhang Jinglei Liu School of Computer and Control Engineering Yantai UniversityYantaiShandongChina

Image clustering has received significant attention due to the growing importance of image *** have explored Riemannian manifold clustering,which is capable of capturing the non-linear shapes found in real-world ***,the complexity of image data poses substantial challenges for modelling and feature *** methods such as covariance matrices and linear subspace have shown promise in image modelling,and they are still in their early stages and suffer from certain ***,these include the uncertainty of representing data using only one Riemannian manifold,limited feature extraction capacity of single kernel functions,and resulting incomplete data representation and *** overcome these limitations,the authors propose a novel approach called join multiple Riemannian manifold representation and multi-kernel non-redundancy for image clustering(MRMNR-MKC).It combines covariance matrices with linear subspace to represent data and applies multiple kernel functions to map the non-linear structural data into a reproducing kernel Hilbert space,enabling linear model analysis for image ***,the authors use matrix-induced regularisation to improve the clustering kernel selection process by reducing redundancy and assigning lower weights to identical ***,the authors also conducted numerous experiments to evaluate the performance of our approach,confirming its superiority to state-of-the-art methods on three benchmark datasets.

关键词： clustering data mining image representation machine learning manifolds

来源：评论

学校读者我要写书评

暂无评论

Curiosity and Affect-Driven Cognitive Architecture for HRI

引用

IEEE Transactions on Affective Computing 2025年

作者： Berto, Leticia Tanevska, Ana Cirne, Azamor Costa, Paula Simoes, Alexandre Gudwin, Ricardo Rea, Francesco Colombini, Esther Sciutti, Alessandra State University of Campinas Institute of Computing Campinas Brazil Italian Institute of Technology CONTACT Unit Genoa Italy Uppsala University Department of Information Technology Uppsala Sweden Federal University of Paraiba Department of Finance and Accounting João Pessoa Brazil State University of Campinas School of Electrical and Computer Engineering Campinas Brazil São Paulo State University Department of Control and Automation Engineering Sorocaba Brazil

This study explores how humans and cognitive robots with different value systems and motivations understand each other's needs in free-form interactions. We developed a cognitive architecture that links sensing and perception to internal motivation and an intrinsic value system for determining actions. Inspired by young children's needs, this architecture includes three drives: learning, interaction, and recharging, each with varying dependence on the human partner. We aimed to assess how experimentally changing the importance of these drives within a fixed architecture affects interaction dynamics with human partners (acting as caregivers) and their understanding of the robot's needs. By adjusting the learning and interaction drives, we created two robot profiles: Playful, which prioritizes environmental exploration and playfulness to reduce boredom, and Social, which focuses on social interaction through touch and visual contact to increase comfort. Our findings show that changing the importance of these drives produces distinct behaviors and human perceptions. Robot behaviors matched their profiles, and participants adapted their responses accordingly. Participants identified and attributed distinct traits to each robot without knowing the specific profiles. Despite variability among human partners, the robots, especially the playful one, were generally well understood by most participants. © 2025 IEEE. All rights reserved.

关键词： Robot learning

来源：评论

学校读者我要写书评

暂无评论

Flexibility Improvement of CHP Unit for Wind Power Accommodation

引用

Journal of Modern Power Systems and Clean Energy 2022年第3期10卷 731-742页

作者： Dongming Zhang Yong Hu Yaokui Gao School of Control and Computer Engineering North China Electric Power UniversityBeijingChina

Improving the flexibility of combined heat and power(CHP)units is an important way to solve the problem of wind power accommodation in northern ***,this paper analyzes the principle of an extraction-type CHP unit,calculates its safe operation range,and analyzes its contradiction between heating and ***,the safe operation ranges of the CHP unit with several flexibility modifications are further calculated,which involve two-stage bypass,low-pressure cylinder(LPC)removal,heat storage tank,and electric ***,based on the safe operation ranges,their effects on improving the capabilities of deep peak shaving and wind power accommodation are compared,and their adaptabilities to different wind scenarios are *** results show that:①all flexibility modifications can improve the deep peak shaving capability of the CHP unit,especially for the two-stage bypass and the electric boiler;②LPC removal modification can accommodate wind power to some extent,but most of wind power is still abandoned;③heat storage tank modification is unstable in different wind scenarios,which is determined by the surplus heating capability during the daytime.

关键词： Combined heat and power(CHP) flexibility bypass low-pressure cylinder(LPC)removal heat storage tank electric boiler

来源：评论

学校读者我要写书评

暂无评论

Multi-Objective Task Assignment for Maximizing Social Welfare in Spatio-Temporal Crowdsourcing

引用

China Communications 2021年第11期18卷 11-25页

作者： Shengnan Wu Yingjie Wang Xiangrong Tong School of Computer and Control Engineering Yantai UniversityYantai 264005China

With the development of the Internet of Things(IoT),spatio-temporal crowdsourcing(mobile crowdsourcing)has become an emerging paradigm for addressing location-based sensing ***,the delay caused by network transmission has led to low data processing ***,edge computing can solve this problem,effectively reduce the delay of data transmission,and improve data processing capacity,so that the crowdsourcing platform can make better decisions ***,this paper combines spatio-temporal crowdsourcing and edge computing to study the Multi-Objective Optimization Task Assignment(MOO-TA)problem in the edge computing *** proposed online incentive mechanism considers the task difficulty attribute to motivate crowd workers to perform sensing tasks in the unpopular *** this paper,the Weighted and Multi-Objective Particle Swarm Combination(WAMOPSC)algorithm is proposed to maximize both platform’s and crowd workers’utility,so as to maximize social *** algorithm combines the traditional Linear Weighted Summation(LWS)algorithm and Multi-Objective Particle Swarm Optimization(MOPSO)algorithm to find pareto optimal solutions of multi-objective optimization task assignment problem as much as possible for crowdsourcing platform to *** comparison experiments on real data sets,the effectiveness and feasibility of the proposed method are evaluated.

关键词： spatio-temporal crowdsourcing edge computing task assignment multi-objective optimization particle swarm optimization Pareto optimal solution

来源：评论

学校读者我要写书评

暂无评论

Two-level deep reinforcement learning approach for energy management of integrated energy system considering deep peak shaving for combined heat and power and coordinated scheduling of electric vehicles

引用

Applied Thermal engineering 2025年 275卷

作者： Zhang, Pengchao Yang, Tingting Yan, Huiru Zhu, Yansong Hu, Yong Zeng, Deliang Liu, Jizhen School of Control and Computer Engineering North China Electric Power University Beijing102206 China

The development and utilization of renewable energy have effectively reduced dependence on limited fossil fuel resources and greenhouse gas emissions. However, the accompanying uncertainty of renewable energy generation presents substantial challenges to the stability of the power system. Existing energy management algorithms of Integrated Energy Systems (IES) often rely on precise system models and load forecasting, which frequently fail to adapt to the high level of uncertainty of realistic scenarios, thereby limiting scheduling flexibility and efficiency. To address these issues, a two-stage scheduling strategy for energy management based on two-level Deep Reinforcement Learning (DRL), referred to as TLPPO, is proposed. The results indicate that the TLPPO effectively enhances the peak-shaving capabilities of the IES, with a 10.24% and 16.16% increase in the self-balancing rate compared to other multi-agent algorithms. Additionally, the two-stage dispatch strategy enhances the peak-shaving capabilities of the IES, allowing it to locally absorb a high proportion of renewable energy, reaching up to 82.63% of the total load. This strategy also significantly reduces the duration of deep peak-shaving for coal-fired power plant units. Validation under typical scenarios demonstrates that it completely eliminates operations below 40%, limits the 40%–50% operating range to just 3 h, and ensures the economic and safe operation of the power plant. These research findings serve as a reference for the management of IES with high renewable energy penetration. © 2025 Elsevier Ltd

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Effective Two-Stage Double Auction for Dynamic Resource Trading in Edge Networks via Overbooking

arXiv

引用

arXiv 2025年

作者： Wu, Sicheng Liwang, Minghui Wang, Deqing Wang, Xianbin Wu, Chao Tang, Junyi Li, Li Jiao, Zhenzhen The School of Informatics Xiamen University Fujian China The Department of Control Science and Engineering The National Key Laboratory of Autonomous Intelligent Unmanned Systems Tongji University Shanghai China Frontiers Science Center for Intelligent Autonomous Systems Ministry of Education Tongji University Shanghai China The Department of Electrical and Computer Engineering Western University ON Canada The School of Public Affairs Zhejiang University Hangzhou310027 China School of Information Science and Engineering Lanzhou University Lanzhou China The iF-Labs Beijing Teleinfo Technology Co. Ltd. CAICT China

To facilitate responsive and cost-effective computing resource scheduling and service delivery over edge-assisted mobile networks, this paper investigates a novel two-stage double auction methodology via utilizing an interesting idea of resource overbooking to overcome dynamic and uncertain nature from edge servers (sellers) and demand from mobile devices (as buyers). The proposed auction integrates multiple essential factors such as social welfare maximization and decision-making latency (e.g., the time for determining winning seller-buyer pairs) reduction, by introducing a stagewise strategy: an overbooking-driven pre-double auction (OPDAuction) for determining long-term cooperations between sellers and buyers before practical resource transactions as Stage I, and a real-time backup double auction (RBDAuction) for handling residual resource demands during actual transactions. In particular, by applying a proper overbooking rate, OPDAuction helps with facilitating trading contracts between appropriate sellers and buyers as guidance for future transactions, by allowing the booked resources to exceed supply. Then, since pre-auctions may cause risks, our RBDAuction adjusts to real-time market changes, further enhancing the overall social welfare. More importantly, we offer an interesting view to show that our proposed two-stage auction can support significant design properties such as truthfulness, individual rationality, and budget balance. Through extensive experiments, we demonstrate good performance in social welfare, time efficiency, and computational scalability, outstripping conventional methods in dynamic edge computing settings. Copyright © 2025, The Authors. All rights reserved.

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

Active disturbance rejection control for load frequency control systems under uncertainty and resonance attacks

引用

International Journal of Electrical Power & Energy Systems 2025年 169卷

作者： Xingqi Hu Wen Tan Guolian Hou School of Control & Computer Engineering North China Electric Power University Beijing 102206 China School of Electrical and Control Engineering North China University of Technology Beijing 100144 China

The pervasive presence of hackers has seriously constrained the development of national cyber security. The necessity of addressing the issue of counteracting cyber-attacks is of paramount importance to the power grid system. In this study, we propose a novel attack compensation structure, namely the linear active disturbance rejection (LADRC) controller with a compensation loop. This enhances the existing second-order LADRC structure through incorporating a compensation loop that multiplies the difference between the output of the linear extended state observer (LESO) and the frequency deviation factor multiplied with RoCoF with a compensation factor controlled by a specific control law. The compensating attack loop drawing inspiration from LESO, possesses robust real-time online observation and estimation capabilities; effectively estimating and compensating for disturbances to stabilize system frequency within acceptable limits. The compensation loop employs three model-based resonance strategies to counteract disturbances, with detailed theoretical derivations and simulations conducted to determine the compensation factors. Simulation testing has been expanded to include two and three-area systems to validate the proposed approach. In addition, a comparison is made between the proposed compensation structure and the existing control scheme. The simulation results demonstrate that the proposed attack compensation scheme exhibits superior anti-attack capabilities while maintaining flexibility against resonance attacks.

关键词： Resonance attacks Power system Linear active disturbance rejection Estimate Anti-interference

来源：评论

学校读者我要写书评

暂无评论

Rescheduling costs and adaptive asymmetric errors guided closed-loop prediction of power loads in mine integrated energy systems

引用

Energy and AI 2025年 21卷

作者： Miao, Qi Sun, Xiaoyan Ma, Chen Zhang, Yong Gong, Dunwei School of Information and Control Engineering China University of Mining and Technology No. 1 Daxue Road Jiangsu Xuzhou221116 China School of Artificial Intelligence and Computer Science Jiangnan University No. 1800 Lihu Avenue Jiangsu Wuxi214122 China Department of Computer Science City University of Hong Kong Tat Chee Avenue 518057 Hong Kong School of Information Science and Technology Qingdao University of Science and technology NO. 99 Songling Road Shandong Qingdao266061 China

The development of an integrated energy system for mining that efficiently recycles multiple resources is a crucial strategy for achieving dual carbon reduction targets in the mining sector. Precise load forecasting is fundamental to ensuring the safe and efficient scheduling of this system. However, existing studies often overlook the coupling between load forecasting and scheduling results, treating them independently, which frequently leads to high rescheduling costs due to forecasting errors. To address this issue, we propose a closed-loop load forecasting algorithm that incorporates rescheduling costs and asymmetric errors. We first proposed a data generation and model construction strategy by using real load, predicted load, and rescheduling costs to capture the relationship between load forecasting and rescheduling costs. Considering the different impacts of under-forecasting and over-forecasting on scheduling results, the rescheduling cost model is further integrated with asymmetric prediction errors to define the loss function of the Bi-LSTM based forecasting model. Additionally, an optimization strategy for self-tuning asymmetric prediction error fusion coefficients is designed to ensure the accuracy of load forecasting. The proposed algorithm is applied to the power load forecasting of an integrated energy system in a coal mine in Shanxi. The results demonstrate the effectiveness of the algorithm in reducing system rescheduling costs while ensuring forecasting accuracy, highlighting its potential application in power load forecasting for mine integrated energy systems. © 2025

关键词： Electric load forecasting

来源：评论

学校读者我要写书评

暂无评论

Viewpoint Optimization Using Genetic Algorithm for Flying Robot Inspection of Electricity Transmission Tower Equipment

引用

Chinese Journal of Electronics 2014年第2期23卷 426-431页

作者： YANG Guotian DONG Ruifang WU Hua LIU Chang'an School of Control and Computer Engineering North China Electric Power University

To fully exploit the limited flight-time of the flying robot, and ensure the successful visibility of target, viewpoint optimization is proposed in this paper for the inspection of electricity transmission tower equipment with an optimization function to determine the best viewpoints in a local viewpoint region. The local viewpoint regions are generated from the local objective regions which are determined by the geometrical structure of a priori3 D model for the electricity transmission tower *** optimization function is structured based on three factors including visibility, viewing quality and observation distance. In addition, the fitness function of genetic algorithm is used to find the optimal viewpoint. The experimental results demonstrate the effectiveness and efficiency of the proposed viewpoint selection algorithm.

关键词： Flying robot inspection Electricity transmission tower equipment Viewpoint optimization Genetic algorithm 3D model

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：