检索结果-内蒙古大学图书馆

Topology inference for multi-agent cooperation under unmeasurable latent input

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Jiao, Qing Li, Yushan He, Jianping Shi, Ling Department of Automation Shanghai Jiao Tong University Key Laboratory of System Control and Information Processing Ministry of Education of China Shanghai China Department of Electronic and Computer Engineering Hong Kong University of Science and Technology Hong Kong

Topology inference is a crucial problem for cooperative control in multi-agent systems. Different from most prior works, this paper is dedicated to inferring the directed network topology from the observations that consist of a single, noisy and finite time-series system trajectory, where the cooperation dynamics is stimulated with the initial network state and the unmeasurable latent input. The unmeasurable latent input refers to intrinsic system signal and extrinsic environment interference. Considering the time-invariant/varying nature of the input, we propose two-layer optimization-based and iterative estimation-based topology inference algorithms (TO-TIA and IE-TIA), respectively. TO-TIA allows us to capture the separability of global agent state and eliminates the unknown influence of time-invariant input on system dynamics. IE-TIA further exploits the identifiability and estimability of more general time-varying input and provides an asymptotic solution with desired convergence properties, with higher computation cost compared with TO-TIA. Our novel algorithms relax the dependence of observation scale and leverage the empirical risk reformulation to improve the inference accuracy in terms of the topology structure and edge weight. Comprehensive theoretical analysis and simulations for various topologies are provided to illustrate the inference feasibility and the performance of the proposed algorithms. Copyright © 2020, The Authors. All rights reserved.

关键词： Multi agent systems

Energy efficient robust beamforming and cooperative jamming design for IRS-Assisted MISO networks

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Wang, Qun Zhou, Fuhui Hu, Rose Qingyang Qian, Yi Department of Electrical and Computer Engineering Utah State University LoganUT84322 United States College of Electronic and Information Engineering Nanjing University of Aeronautics and Astronautics Nanjing210000 China Key Laboratory of Dynamic Cognitive System of Electromagnetic Spectrum Space of Nanjing University of Aeronautics and Astronautics China Department of Electrical and Computer Engineering University of Nebraska–Lincoln OmahaNE68182 United States

Energy-efficient design and secure communications are of crucial importance in wireless communication networks. However, the energy efficiency achieved by using physical layer security can be limited by the channel conditions. In order to tackle this problem, an intelligent reflecting surface (IRS) assisted multiple input single output (MISO) network with independent cooperative jamming is studied. The energy efficiency is maximized by jointly designing the transmit and jamming beamforming and IRS phase-shift matrix under both the perfect channel state information (CSI) and the imperfect CSI. In order to tackle the challenging non-convex fractional problems, an algorithm based on semidefinite programming (SDP) relaxation is proposed for solving energy efficiency maximization problem under the perfect CSI case while an alternate optimization algorithm based on S-procedure is used for solving the problem under the imperfect CSI case. Simulation results demonstrate that the proposed design outperforms the benchmark schemes in term of energy efficiency. Moreover, the tradeoff between energy efficiency and the secrecy rate is found in the IRS-assisted MISO network. Furthermore, it is shown that IRS can help improve energy efficiency even with the uncertainty of the CSI. Copyright © 2020, The Authors. All rights reserved.

关键词： Energy efficiency

DeePMD-kit v2: A software package for Deep Potential models

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Zeng, Jinzhe Zhang, Duo Lu, Denghui Mo, Pinghui Li, Zeyu Chen, Yixiao Rynik, Marián Huang, Li'ang Li, Ziyao Shi, Shaochen Wang, Yingze Ye, Haotian Tuo, Ping Yang, Jiabin Ding, Ye Li, Yifan Tisi, Davide Zeng, Qiyu Bao, Han Xia, Yu Huang, Jiameng Muraoka, Koki Wang, Yibo Chang, Junhan Yuan, Fengbo Bore, Sigbjørn Løland Cai, Chun Lin, Yinnian Wang, Bo Xu, Jiayan Zhu, Jia-Xin Luo, Chenxing Zhang, Yuzhi Goodall, Rhys E.A. Liang, Wenshuo Singh, Anurag Kumar Yao, Sikai Zhang, Jingchao Wentzcovitch, Renata Han, Jiequn Liu, Jie Jia, Weile York, Darrin M. Weinan, E. Car, Roberto Zhang, Linfeng Wang, Han Laboratory for Biomolecular Simulation Research Institute for Quantitative Biomedicine Department of Chemistry and Chemical Biology Rutgers University PiscatawayNJ08854 United States AI for Science Institute Beijing100080 China DP Technology Beijing100080 China Academy for Advanced Interdisciplinary Studies Peking University Beijing100871 China HEDPS CAPT College of Engineering Peking University Beijing100871 China College of Electrical and Information Engineering Hunan University Changsha China Yuanpei College Peking University Beijing100871 China Program in Applied and Computational Mathematics Princeton University PrincetonNJ08540 United States Department of Experimental Physics Comenius University Mlynská Dolina F2 Bratislava842 48 Slovakia Center for Quantum Information Institute for Interdisciplinary Information Sciences Tsinghua University Beijing100084 China Center for Data Science Peking University Beijing100871 China ByteDance Research Zhonghang Plaza No. 43 North 3rd Ring West Road Haidian District Beijing China College of Chemistry and Molecular Engineering Peking University Beijing100871 China Baidu Inc. Beijing China Key Laboratory of Structural Biology of Zhejiang Province School of Life Sciences Westlake University Zhejiang Hangzhou China Westlake AI Therapeutics Lab Westlake Laboratory of Life Sciences and Biomedicine Zhejiang Hangzhou China Department of Chemistry Princeton University PrincetonNJ08544 United States SISSA Scuola Internazionale Superiore di Studi Avanzati Trieste34136 Italy Laboratory of Computational Science and Modeling Institute of Materials École Polytechnique Fédérale de Lausanne Lausanne1015 Switzerland Department of Physics National University of Defense Technology Hunan Changsha410073 China State Key Lab of Processors Institute of Computing Technology Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China School of Electronics Engineerin

DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics simulations using machine learning potentials (MLP) known as Deep Potential (DP) models. This package, which was released in 2017, has been widely used in the fields of physics, chemistry, biology, and material science for studying atomistic systems. The current version of DeePMD-kit offers numerous advanced features such as DeepPot-SE, attention-based and hybrid descriptors, the ability to fit tensile properties, type embedding, model deviation, Deep Potential - Range Correction (DPRc), Deep Potential Long Range (DPLR), GPU support for customized operators, model compression, non-von Neumann molecular dynamics (NVNMD), and improved usability, including documentation, compiled binary packages, graphical user interfaces (GUI), and application programming interfaces (API). This article presents an overview of the current major version of the DeePMD-kit package, highlighting its features and technical details. Additionally, the article benchmarks the accuracy and efficiency of different models and discusses ongoing developments. © 2023, CC BY-NC-ND.

关键词： Molecular dynamics

Ultrahigh-Q AlGaAs-on-insulator microresonators for integrated nonlinear photonics

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Xie, Weiqiang Chang, Lin Shu, Haowen Norman, Justin C. Peters, Jon D. Wang, Xingjun Bowers, John E. Department of Electrical and Computer Engineering University of California Santa BarbaraCA93106 United States State Key Laboratory of Advanced Optical Communications System and Networks Department of Electronics School of Electronics Engineering and Computer Science Peking University Beijing100871 China Institute for Energy Efficiency University of California Santa BarbaraCA93106 United States

Aluminum gallium arsenide (AlGaAs) and related III-V semiconductors have excellent optoelectronic properties. They also possess strong material nonlinearity as well as high refractive indices. In view of these properties, AlGaAs is a promising candidate for integrated photonics, including both linear and nonlinear devices, passive and active devices, and associated applications. For integrated photonics low propagation loss is essential, particularly in nonlinear applications. However, achieving low-loss and high-confinement AlGaAs photonic integrated circuits poses a challenge. Here we show an effective reduction of surface-roughness-induced scattering loss in fully etched high-confinement AlGaAs-on-insulator nanowaveguides, by using a heterogeneous wafer-bonding approach and optimizing fabrication techniques. We demonstrate ultrahigh-quality AlGaAs microring resonators and realize quality factors up to 3.52×106 and finesses as high as 1.4×104. We also show ultra-efficient frequency comb generations in those resonators and achieve record-low threshold powers on the order of ~20 µW and ~120 µW for the resonators with 1 THz and 90 GHz free-spectral ranges, respectively. Our result paves the way for the implementation of AlGaAs as a novel integrated material platform specifically for nonlinear photonics, and opens a new window for chip-based efficiency-demanding practical applications. Copyright © 2020, The Authors. All rights reserved.

关键词： III-V semiconductors

Adaptive Dynamic Surface Control of Constrained Nonlinear systems

学校读者我要写书评

暂无评论

Adaptive Dynamic Surface Control of Constrained Nonlinear Sy...

第32届中国控制与决策会议

作者： Ailiang Xing Biao Song Yixin Yin School of Automation and Electrical Engineering University of Science and Technology Beijing Key Laboratory of Knowledge Automation for Industrial Processes Ministry of Education Demonstration and Research Department of Network Information System Beijing Institute of Computer Application

This paper presents an adaptive dynamic surface control for uncertain nonlinear strict feedback systems with constraints on both input and *** technology is employed in the controller *** performance method with asymmetric barrier Lyapunov function(BLF) is used to deal with the output *** input saturation is solved by using the Nussbaum *** dynamic surface control(DSC) technique is employed to reduce the complexity of the virtual *** is illustrated that all closed-loop signals are semi-global uniformly ultimately bounded(SGUUB).Finally,a simulation example proves the proposed results.

关键词： Input saturation Output constraints Nussbaum function Dynamic surface control

Online reinforcement learning control by direct heuristic dynamic programming: from time-driven to event-driven

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Zhao, Qingtao Si, Jennie Sun, Jian Key Laboratory of Intelligent Control and Decision of Complex System Beijing Institute of Technology Beijing100081 China Department of Electrical Computer and Energy Engineering Arizona State University TempeAZ85287 United States

In this paper time-driven learning refers to the machine learning method that updates parameters in a prediction model continuously as new data arrives. Among existing approximate dynamic programming (ADP) and reinforcement learning (RL) algorithms, the direct heuristic dynamic programming (dHDP) has been shown an effective tool as demonstrated in solving several complex learning control problems. It continuously updates the control policy and the critic as system states continuously evolve. It is therefore desirable to prevent the time-driven dHDP from updating due to insignificant system event such as noise. Toward this goal, we propose a new event-driven dHDP. By constructing a Lyapunov function candidate, we prove the uniformly ultimately boundedness (UUB) of the system states and the weights in the critic and the control policy networks. Consequently we show the approximate control and cost-to-go function approaching Bellman optimality within a finite bound. We also illustrate how the event-driven dHDP algorithm works in comparison to the original time-driven dHDP. Copyright © 2020, The Authors. All rights reserved.

关键词： Learning algorithms

Toward building next-generation Geocoding systems: a systematic review

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Yin, Zhengcong Goldberg, Daniel W. Lin, Binbin Zhou, Bing Li, Diya Ma, Andong Ming, Ziqian Cai, Heng Zhang, Zhe Wang, Shaohua Gao, Shanzhen Lee, Joey Ying Li, Xiao Huo, Da Department of Geography Texas A&M University 797 Lamar St. College StationTX77840 United States Department of Geography Pennsylvania State University 201 Old Main University ParkPA16802 United States Department of Earth and Atmospheric Sciences Metropolitan State University of Denver Science Building 2014 DenverCO80217 United States Esri Inc 380 New York St. RedlandsCA92373 United States State Key Laboratory of Remote Sensing and Digital Earth Aerospace Information Research Institute Chinese Academy of Sciences No.9 Dengzhuang South Rd. Beijing China Department of Computer Information Systems Virginia State University 1 Hayden St. PetersburgVA23806 United States LABI Education 143 Keelung Rd. Taipei Taiwan Transport Studies Unit University of Oxford South Parks Rd. OxfordOX1 3QY United Kingdom Department of Civil & Mineral Engineering University of Toronto 35 St George St. TorontoON Canada

Geocoding systems are widely used in both scientific research for spatial analysis and everyday life through location-based services. The quality of geocoded data significantly impacts subsequent processes and applications, underscoring the need for next-generation systems. In response to this demand, this review first examines the evolving requirements for geocoding inputs and outputs across various scenarios these systems must address. It then provides a detailed analysis of how to construct such systems by breaking them down into key functional components and reviewing a broad spectrum of existing approaches, from traditional rule-based methods to advanced techniques in information retrieval, natural language processing, and large language models. Finally, we identify opportunities to improve next-generation geocoding systems in light of recent technological advances. © 2025, CC BY.

关键词：

Reviewing continual learning from the perspective of human-level intelligence

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Chang, Yifan Li, Wenbo Peng, Jian Tang, Bo Kang, Yu Lei, Yinjie Gui, Yuanmiao Zhu, Qing Liu, Yu Li, Haifeng The State Key Laboratory of Fire Science Department of Automation and Institute of Advanced Technology University of Science and Technology of China Hefei China Institute of Intelligent Machines Chinese Academy of Sciences Hefei230031 China The Department of Electrical and Computer Engineering Mississippi State University United States College of Electronics and Information Engineering Sichuan University Chengdu China Faculty of Geosciences and Environmental Engineering The Southwest Jiaotong University Chengdu China Institute of Remote Sensing and Geographic Information System Peking Universit Beijing China School of Geosciences and Info-Physics Central South University South Lushan Road Changsha410083 China

Humans continual learning (CL) ability is closely related to Stability Versus Plasticity Dilemma that describes how humans achieve ongoing learning capacity and preservation for learned information. The notion of CL has always been present in artificial intelligence (AI) since its births. This paper proposes a comprehensive review of CL. Different from previous reviews that mainly focus on the catastrophic forgetting phenomenon in CL, this paper surveys CL from a more macroscopic perspective based on the Stability Versus Plasticity mechanism. Analogous to biological counterpart, "smart" AI agents are supposed to i) remember previously learned information (information retrospection);ii) infer on new information continuously (information prospection:);iii) transfer useful information (information transfer), to achieve high-level CL. According to the taxonomy, evaluation metrics, algorithms, applications as well as some open issues are then introduced. Our main contributions concern i) rechecking CL from the level of artificial general intelligence;ii) providing a detailed and extensive overview on CL topics;iii) presenting some novel ideas on the potential development of CL. © 2021, CC BY.

关键词： Petroleum reservoir evaluation

Learning of long-horizon sparse-reward robotic manipulator tasks with base controllers

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Wang, Guangming Xin, Minjian Wu, Wenhua Liu, Zhe Wang, Hesheng Department of Automation Key Laboratory of System Control and Information Processing of Ministry of Education Key Laboratory of Marine Intelligent Equipment and System of Ministry of Education Shanghai Engineering Research Center of Intelligent Control and Management Shanghai Jiao Tong University Shanghai200240 China Department of Computer Science and Engineering University of California San Diego La Jolla CA92093 United States Department of Computer Science and Technology University of Cambridge CambridgeCB2 1TN United Kingdom

Deep Reinforcement Learning (DRL) enables robots to perform some intelligent tasks end-to-end. However, there are still many challenges for long-horizon sparse-reward robotic manipulator tasks. On the one hand, a sparse-reward setting causes exploration inefficient. On the other hand, exploration using physical robots is of high cost and unsafe. In this paper, we propose a method of learning long-horizon sparse-reward tasks utilizing one or more existing traditional controllers named base controllers in this paper. Built upon Deep Deterministic Policy Gradients (DDPG), our algorithm incorporates the existing base controllers into stages of exploration, value learning, and policy update. Furthermore, we present a straightforward way of synthesizing different base controllers to integrate their strengths. Through experiments ranging from stacking blocks to cups, it is demonstrated that the learned state-based or image-based policies steadily outperform base controllers. Compared to previous works of learning from demonstrations, our method improves sample efficiency by orders of magnitude and improves the performance. Overall, our method bears the potential of leveraging existing industrial robot manipulation systems to build more flexible and intelligent controllers. Copyright © 2020, The Authors. All rights reserved.

关键词： Deep learning