检索结果-内蒙古大学图书馆

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

作者： Gorodetskiy, Andrey Mironov, Konstantin Panov, Aleksandr Center of Cognitive Modeling Moscow Institute of Physics and Technology 9 Institutskij per. Dolgoprudny141701 Russia Artificial Intelligence Research Institute Moscow105064 Russia Federal Research Center "Computer Science and Control" 9 60-letiya Oktyabrya pr. Moscow117312 Russia

ISBN: (纸本)9783031431104

Applying learning-based control methods to real robots presents hard challenges, including the low sample efficiency of model-free reinforcement learning algorithms. The widely adopted approach to tackling this problem uses an environment dynamics model. We propose to use the Neural Ordinary Differential Equations to approximate transition dynamics as this allows for finer control of a trajectory generation process. NODE offers a continuous-time formulation that captures the temporal dependencies. We evaluate our approach on various tasks from simulation environment including learning 6-DoF robotic arm to open the door, which represents particular challenges for policy search. The NODE model is trained to predict movement of the arm and the door, and is used to generate trajectories for the model-based policy optimization. Our method shows better sample efficiency on this task comparing to the model-free and model-based baseline. It also shows comparable results on several other tasks. The application of NODE to model-based reinforcement learning enables more precise modeling of robotic system dynamics and enhances the sample efficiency of learning-based control methods. The empirical evaluation on various tasks demonstrates the efficacy of our approach, offering promising prospects for improving the performance and efficiency of real-world robotic systems. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

On Some General Methods for Mathematical modeling of the Dynamics of Manipulators with Parallel Kinematics

On Some General Methods for Mathematical Modeling of the Dyn...

引用

2022 International Conference on Industrial Engineering, Applications and Manufacturing, ICIEAM 2022

作者： Krasinskiy, Aleksandr Ya Dept. of Physical and Mathem. Disciplines Moscow Russia Moscow Aviation Institute National Research University Dept. of Computer Modeling and Probability Theory Moscow Russia

ISBN: (纸本)9781665483698

The level of development of technology for obtaining and processing measurement information, power electronics and actuators allows the use of the results of mathematical control theory with incomplete information in modern mechatronic systems. These results create the possibility of minimizing interference in the behavior of the object by minimizing the amount of measurement information and the number of actuators involved. A necessary condition for the realization of this possibility is the presence of an adequate nonlinear mathematical model of the control object. The development of methods for creating such models is hampered not so much by the technical complexity of the system under study, but by the fact that fundamentally new problems often arise in the study of a specific actually functioning device, which have never been encountered before in the relevant theory. Their solution requires research by theoretical specialists on the further development of the complex application of areas of analytical mechanics, nonlinear theory of stability, mathematical control theory with incomplete information, information technologies, etc., which were previously quite distant from each other. Rigorous justification of methods for obtaining such mathematical models, is based on complex sections of mathematics, mechanics and control theory, which have never been included in the curricula of standard engineering education. However, in our opinion, a significant expansion of the practical application of the latest results in engineering practice is possible due to such a presentation of algorithms for their qualified application, which does not require a complete understanding of all the abstract-theoretical foundations of their justification. © 2022 IEEE.

关键词： Control theory

来源：评论

学校读者我要写书评

暂无评论

Multiscale Numerical Simulation for Assessing Effective Electrical Conductivity Estimations of Geological Medium 3

Multiscale Numerical Simulation for Assessing Effective Elec...

引用

3rd IEEE International Conference on Problems of Informatics, Electronics and Radio Engineering, PIERE 2024

作者： Kutishcheva, Anastasiia Yu. Markov, Sergey I. Shtanko, Ekaterina I. Trofimuk Institute of Petroleum Geology and Geophysics Sb Ras Laboratory of mathematical modeling of multiphase processes in native and artificial multiscale heterogeneous media Novosibirsk Russia Novosibirsk State Technical University Department of Geophysical Systems Novosibirsk Russia Novosibirsk State Technical University Department of Computer Technologies Novosibirsk Russia

ISBN: (纸本)9798331516321

The paper considers methods for estimating the effective electrical conductivity of a geologic medium with an internal structure represented as a set of layer-by-layer tomographic images. Based on this type of input data, a discrete hierarchical mesh model of the sample is constructed. It is used to describe the internal singularities of the investigated medium with significant accuracy. To obtain the effective electrical conductivity of the geologic medium, a direct problem of electric field potential distribution in the sample volume is solved numerically. Since the computational domain is defined as a hierarchy of mesh partitions, a multiscale modification of the scalar finite element method is applied. The averaging procedure of the properties is based on Ohm's law. The possibility of step-by-step averaging of the electrical properties of a rock sample is also investigated. To obtain the electrical effective characteristic of the entire sample, the effective properties of individual macroelements in the mesh partitioning are determined and further averaged using the algorithms proposed in the paper. The variation of the effective electrical conductivity over certain sub-domains of the sample is given. © 2024 IEEE.

关键词： Mesh generation

来源：评论

学校读者我要写书评

暂无评论

Organization of Storage and Processing of Experimental Results from the Continuous Process of Styrene-Butadiene Copolymerization Using Databases

Organization of Storage and Processing of Experimental Resul...

引用

International Russian Smart Industry Conference (SmartIndustryCon)

作者： Tatiana Mikhailova Vladimir Mikhailov Svetlana Mustafina Department of Mathematical and Computer Modeling Ufa University of Science and Technology Ufa Russia Institute of Computational Mathematics and Information Technologies Kazan Federal University Kazan Russia

ISBN: (数字)9798331511241

ISBN: (纸本)9798331511258

The article presents an approach to storing and processing the results of industrial, laboratory, and computational experiments based on databases. The approach is presented on the example of the process of copolymerization of butadiene with styrene, which proceeds continuously in a cascade of series-connected ideal mixing reactors. A conceptual and logical database model of the industrial process under study has been created. The constructed structure can be adapted for other copolymerization processes, depending on the type of process, its formulation, and the process flow conditions. The presented physical model is implemented as a database and has been tested in the Microsoft SQL Server environment.

关键词： Industries Adaptation models Databases Organizations Mathematical models Kinetic theory Recording Servers Polymers Inductors

来源：评论

学校读者我要写书评

暂无评论

Low Matchline Voltage Swing Content-Addressable Memory Cell

Low Matchline Voltage Swing Content-Addressable Memory Cell

引用

IEEE International Symposium on Circuits and Systems (ISCAS)

作者： Cristhopher Mosquera Ramiro Taco Benjamin Zambrano Luis Miguel Prócel Esteban Garzón Marco Lanuzza Department of Computer Engineering Modeling Electronics and Systems University of Calabria Rende Italy Institute of Micro and Nanoelectronics Universidad San Francisco de Quito Quito Ecuador

ISBN: (数字)9798350356830

ISBN: (纸本)9798350356847

Content-addressable memory (CAM) is a specialized memory architecture designed for fast data searches, allowing a one-clock-cycle comparison between the search input and the entire memory content. In this work, a low matchline voltage swing CAM is proposed to reduce the search power consumption while maintaining high-speed search operations. Low voltage swing in the matchline is enabled by introducing extra circuitry in the conventional CAM cell. By means of comprehensive Monte Carlo and post-layout simulations using a commercial 65nm node, we show that the proposed CAM cell design allows for robustness against process, voltage, and temperature variations without the need for dedicated matchline sense schemes. Compared to conventional precharge high NOR-type CAM, the proposed design achieves 42% higher speed and 29 . 1% less energy consumption. Post-layout results demonstrate that the proposed CAM operates reliably at 0.6V, maintaining performant and reliable search operations across a wide temperature range.

关键词： Temperature sensors Energy consumption Temperature distribution Power demand Microprocessors Simulation Voltage Robustness Sensors Standards

来源：评论

学校读者我要写书评

暂无评论

Pre-trained language models in Spanish for health insurance coverage 5

Pre-trained language models in Spanish for health insurance ...

引用

5th Workshop on Clinical Natural Language Processing, ClinicalNLP 2023. held at ACL 2023

作者： Aracena, Claudio Rodríguez, Nicolás Rocco, Victor Dunstan, Jocelyn Faculty of Physical and Mathematical Sciences University of Chile Chile Millennium Institute Foundational Research on Data Chile Chilean Safety Association Chile Department of Computer Science Catholic University of Chile Chile Institute for Mathematical and Computational Engineering Catholic University of Chile Chile Center for Mathematical Modeling University of Chile Chile

ISBN: (纸本)9781959429883

The field of clinical natural language processing (NLP) can extract useful information from clinical text. Since 2017, the NLP field has shifted towards using pre-trained language models (PLMs), improving performance in several tasks. Most of the research in this field has focused on English text, but there are some available PLMs in Spanish. In this work, we use clinical PLMs to analyze text from admission and medical reports in Spanish for an insurance and health provider to give a probability of no coverage in a labor insurance process. Our results show that fine-tuning a PLM pretrained with the provider's data leads to better results, but this process is time-consuming and computationally expensive. At least for this task, fine-tuning publicly available clinical PLM leads to comparable results to a custom PLM, but in less time and with fewer resources. Analyzing large volumes of insurance requests is burdensome for employers, and models can ease this task by pre-classifying reports that are likely not to have coverage. Our approach of entirely using clinical-related text improves the current models while reinforcing the idea of clinical support systems that simplify human labor but do not replace it. To our knowledge, the clinical corpus collected for this study is the largest one reported for the Spanish language. © 2023 Association for Computational Linguistics.

关键词： Health insurance

来源：评论

学校读者我要写书评

暂无评论

Non-Volatile Content-Addressable Memory For Energy-Efficient & High-Performance Search And Update Operations

Non-Volatile Content-Addressable Memory For Energy-Efficient...

引用

IEEE International Symposium on Circuits and Systems (ISCAS)

作者： Alessandro Bedoya Benjamin Zambrano Ramiro Taco Luis Miguel Prócel Marco Lanuzza Esteban Garzón Department of Computer Engineering Modeling Electronics and Systems University of Calabria Rende Italy Institute of Micro and Nanoelectronics Universidad San Francisco de Quito Quito Ecuador

ISBN: (数字)9798350356830

ISBN: (纸本)9798350356847

This work presents a non-volatile content-addressable memory (NV-CAM) based on double-barrier magnetic tunnel junction technology (DMTJ). Unlike state-of-the-art NV-CAM designs that present low-performance updates, our NV-CAM allows energy-efficient, high-performance search and update operations. This makes it well-suited for applications requiring a high frequency of searches/updates, such as associative processors. The NV-CAM hybrid CMOS/DMTJ was designed using a commercial 65nm CMOS technology and a Verilog-A-based DMTJ compact model. The NV-CAM evaluation was carried out by employing Monte Carlo simulations while accounting for process variations. Simulation results show that our NV-CAM presents competitive figures of merit compared to state-of-the-art design. Our NV-CAM presents energy-efficient operations and reduces the update and search delay by about 71% and 75%, respectively, compared to other NV-CAMs.

关键词： Semiconductor device modeling Program processors Monte Carlo methods Nonvolatile memory Simulation Energy efficiency Delays High frequency Junctions Magnetic tunneling

来源：评论

学校读者我要写书评

暂无评论

On the Possibility of Different Accuracy of Mathematical modeling of a Mechanical Component and a Control Loop in Mechatronic Systems with Electric Drives

On the Possibility of Different Accuracy of Mathematical Mod...

引用

2022 International Russian Automation Conference, RusAutoCon 2022

作者： Krasinskiy, Aleksandr Ya. Msufp Dept. of Physical and Mathem.disciplines Moscow Russia Moscow Aviation Institute National Research University Dept. of Computer Modeling and Probability Theory Moscow Russia

ISBN: (数字)9781665466554

ISBN: (纸本)9781665466554

The current level of development of the element base of signal and power electronics and information technology makes it possible to implement control algorithms of any complexity in mechatronic systems. One of the main problems of any control method is to minimize the cost of all types of resources for its implementation. However, a necessary condition for the effective use of these approaches is the availability of an adequate mathematical nonlinear model of the control object based on the complex application of the relevant sections of analytical mechanics, applications of the nonlinear stability theory and mathematical control theory with incomplete information. On the other hand, estimates of the required degree of modeling accuracy must be obtained to ensure that strictly justified results are obtained. Excessive complication of the model directly complicates the analysis of the obtained results. It is shown how the complex application of the results of analytical mechanics, mathematical control theory, nonlinear stability theory can reasonably reduce the accuracy of the model for individual components of the system while maintaining the accuracy of the general nonlinear model. © 2022 IEEE.

关键词： Electric drives

来源：评论

学校读者我要写书评

暂无评论

RLA-DDTC: A Reinforcement Learning Agent-based approach for Decision-making for Dynamic Traffic Control in C-ITS systems

引用

Procedia computer Science 2024年 246卷 3475-3484页

作者： Ameni Aloui Hela Hachicha Ezzeddine Zagrouba University of Tunis El Manar Higher Institute of Computer Science Laboratory of Informatics Modeling and Information and Knowledge Processing (LIMTIC) 2 Rue Abou Rayhane Bayrouni 2080 Ariana Tunisia College of Computer Science and Engineering - University of Jeddah Jeddah Saudi Arabia

The Cooperative Intelligent Transport System (C-ITS) aims to address traffic efficiency, road safety, and environmental sustain-ability by facilitating communication among autonomous entities with complex behaviors. Traffic congestion is a significant issue in modern urban areas, influenced by factors like accidents, road works, weather and peak hours, leading to wasted time, fuel, and increased pollution. To address these challenges, multi-agent systems have proven effective in managing dynamic traffic in C-ITS environments. In this paper, we propose a novel Reinforcement Learning Agent (RLA) method for decision-making in C-ITS systems to optimize traffic flow and control. Our method utilizes the Q-learning algorithm, enabling effective problem-solving. We conducted experiments to demonstrate the versatility and effectiveness of our approach in complex C-ITS environments. Our method surpasses traditional metrics like mean travel time and mean speed, validated through comparisons with Actor-Critic methods and against the Original Traffic Trace (OTT). This research contributes to advancing decision-making capabilities within C-ITS systems by leveraging RLA for traffic optimization and efficiency.

关键词： Reinforcement learning Q-learning Decision-making C-its system multi-agent-system

来源：评论

学校读者我要写书评

暂无评论

A Client Selection Method Based on Loss Function Optimization for Federated Learning

引用

computer modeling in Engineering & Sciences 2023年第10期137卷 1047-1064页

作者： Yan Zeng Siyuan Teng Tian Xiang Jilin Zhang Yuankai Mu Yongjian Ren Jian Wan School of Computer Science and Technology Hangzhou Dianzi UniversityHangzhou310018China Key Laboratory of Complex System Modeling and Simulation Ministry of EducationHangzhou310018China Zhejiang Engineering Research Center of Data Security Governance Hangzhou310018China Intelligent Robotics Research Center Zhejiang LabHangzhou311100China HDU-ITMO Joint Institute Hangzhou Dianzi UniversityHangzhou310018China

Federated learning is a distributedmachine learningmethod that can solve the increasingly serious problemof data islands and user data privacy,as it allows training data to be kept locally and not shared with other *** trains a globalmodel by aggregating locally-computedmodels of clients rather than their ***,the divergence of local models caused by data heterogeneity of different clients may lead to slow convergence of the global *** this problem,we focus on the client selection with federated learning,which can affect the convergence performance of the global model with the selected local *** propose FedChoice,a client selection method based on loss function optimization,to select appropriate local models to improve the convergence of the global *** firstly sets selected probability for clients with the value of loss function,and the client with high loss will be set higher selected probability,which can make them more likely to participate in ***,it introduces a local control vector and a global control vector to predict the local gradient direction and global gradient direction,respectively,and calculates the gradient correction vector to correct the gradient direction to reduce the cumulative deviationof the local gradient causedby *** experiments to verify the validity of FedChoice on CIFAR-10,CINIC-10,MNIST,EMNITS,and FEMNIST datasets,and the results show that the convergence of FedChoice is significantly improved,compared with FedAvg,FedProx,and FedNova.

关键词： Federated learning model aggregation Non-IID

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：