检索结果-内蒙古大学图书馆

13th ieee data driven control and learning systems conference, DDCLS 2024

作者： Lv, Lingling Hou, Bo Zhang, Shijie Huang, Quanzhen North China University of Water Resources and Electric Power Zhengzhou450045 China Henan University of Technology Zhengzhou450001 China Henan University of Engineering Zhengzhou451191 China

ISBN: (纸本)9798350361674

To solve the problem of noise interference during satellite operation, a satellite attitude control method based on a fully actuated system method and active disturbance rejection technology is proposed. Firstly, a state feedback controller is designed for the satellite attitude system based on the high order fully actuated system method, so that the closed-loop system can be transformed into a linear time-invariant system with an arbitrary configuration of characteristic structure. Then, active disturbance rejection technology is introduced to realize the satellite attitude fast-tracking control in the presence of external interference. The simulation results show that the proposed control strategy can make the satellite attitude angle and angular velocity converge quickly to the expected value, the convergence process is uniform, and the angular velocity amplitude is small. At the same time, it has the advantages of strong anti-interference ability and fast-tracking speed. The simulation results verify the effectiveness of the controller designed in this paper. © 2024 ieee.

关键词： State feedback

来源：评论

学校读者我要写书评

暂无评论

On-Policy data-driven Linear Quadratic Regulator via Model Reference Adaptive Reinforcement learning 62

On-Policy Data-Driven Linear Quadratic Regulator via Model R...

引用

62nd ieee conference on Decision and control (CDC)

作者： Borghesi, Marco Bosso, Alessandro Notarstefano, Giuseppe Alma Mater Studiorum Univ Dept Elect Elect & Informat Engn Bologna Italy

ISBN: (纸本)9798350301243

In this paper, we address a data-driven linear quadratic optimal control problem in which the regulator design is performed on-policy by resorting to approaches from reinforcement learning and model reference adaptive control. In particular, a continuous-time identifier of the value function is used to generate online a reference model for the adaptive stabilizer. By introducing a suitably selected dithering signal, the resulting policy is shown to achieve asymptotic convergence to the optimal gain while the controlled plant reaches asymptotically the behavior of the optimal closed-loop system.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Draft & Refine: Efficient Resource Management in Federated learning under Pathological Labels Skew 31

Draft & Refine: Efficient Resource Management in Federated L...

引用

31st conference on Electronics Circuits and systems

作者： Malan, Erich Peluso, Valentino Calimera, Andrea Macii, Enrico Politecn Torino Dept Control & Comp Engn Turin Italy Politecn Torino Interuniv Dept Reg & Urban Studies & Planning Turin Italy

ISBN: (纸本)9798350377217;9798350377200

In Federated learning (FL), network end-nodes use private data for the local training of classification models that are periodically synchronized with a remote server to update a global model. A key limitation in FL is the high network traffic required to achieve acceptable accuracy, particularly when each client has samples of only a few classes, a problem known as pathological labels skew. Under such constraints, local models become biased toward the available classes, and more synchronization rounds are needed to converge, saturating the communication budget. We address this issue with Draft & Refine, a new resource management strategy to optimize accuracy by dynamically controlling clients' participation based on their remaining budget. Our approach consists of two learning phases: a low-effort drafting phase with only a few clients selected for synchronization, followed by a refinement phase with increased client participation. The available data traffic is used as the control variable to set the duration of the two phases. The proposed strategy outperforms state-of-the-art FL schemes with up to 12.20% higher accuracy within the same data traffic.

关键词： Federated learning Communication Internet of Things Pathological Labels Skew

来源：评论

学校读者我要写书评

暂无评论

learning for CasADi data-driven Models in Numerical Optimization 6

Learning for CasADi Data-driven Models in Numerical Optimiza...

引用

6th Annual learning for Dynamics and control conference

作者： Salzmann, Tim Arrizabalaga, Jon Andersson, Joel Pavone, Marco Ryll, Markus Tech Univ Munich Munich Germany Stanford Univ Stanford CA USA Tech Univ Munich Munich Inst Robot & Machine Intelligence Munich Germany

While real-world problems are often challenging to analyze analytically, deep learning excels in modeling complex processes from data. Existing optimization frameworks like CasADi facilitate seamless usage of solvers but face challenges when integrating learned process models into numerical optimizations. To address this gap, we present the learning for CasADi (L4CasADi) framework, enabling the seamless integration of PyTorch-learned models with CasADi for efficient and potentially hardware-accelerated numerical optimization. The applicability of L4CasADi is demonstrated with two tutorial examples: First, we optimize a fish's trajectory in a turbulent river for energy efficiency where the turbulent flow is represented by a PyTorch model. Second, we demonstrate how an implicit Neural Radiance Field environment representation can be easily leveraged for optimal control with L4CasADi. L4CasADi, along with examples and documentation, is available under MIT license at https://***/Tim-Salzmann/l4casadi

关键词： optimization machine learning control systems data-driven control

来源：评论

学校读者我要写书评

暂无评论

A learning-based control Framework for Human-like Whip Targeting 10

A Learning-based Control Framework for Human-like Whip Targe...

引用

10th ieee RAS/EMBS International conference for Biomedical Robotics and Biomechatronics (BioRob)

作者： Wang, Junyi Xiong, Xiaofeng Tech Univ Denmark Dept Elect & Photon Engn Automat & Control Grp Odense Denmark Univ Southern Denmark SDU Maersk Mc Kinney Moller Inst SDU Biorobot Odense Denmark

Whip targeting (to hit) is a challenging motor control task owing to movement rapidness (72% reductions of joint motion tracking control errors, compared to traditional learning and control algorithms. This integratio... 详细信息

ISBN: (纸本)9798350386523;9798350386530

Whip targeting (to hit) is a challenging motor control task owing to movement rapidness (< 1.5 s) and accuracy, as well as tool flexibility. A learning-based arm control framework is proposed for manipulating a whip to hit targets in this paper. The framework combines a joint motion planner learned by the algorithm soft actor-critic (SAC) with online impedance adaptation control (OIAC). It was validated on a simulated arm and whip system with 54 degrees-of-freedom (DOFs), whose objective is to minimize the distance between the whip tip and a target in a 3D space. As a result, our integration of data-driven and physics-based algorithms allows for 89% improvement of hitting effectiveness and >72% reductions of joint motion tracking control errors, compared to traditional learning and control algorithms. This integration provides a novel proof-of-concept solution to fast targeting control in flexible object manipulation(1).

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

Efficient Collaborations through Weight-driven Coalition Dynamics in Federated learning systems 59

Efficient Collaborations through Weight-Driven Coalition Dyn...

引用

59th Annual ieee International conference on Communications (ieee ICC)

作者： El Hanjri, Mohammed Reguieg, Hamza Attiaoui, Adil Abouaomar, Amine Kobbane, Abdellatif El Kamili, Mohamed Mohammed V Univ Rabat ENSIAS Rabat Morocco Hassan II Univ Casablanca Higher Sch Technol Casablanca Morocco Al Akhawayn Univ Ifrane Sch Sci & Engn Ifrane Morocco

ISBN: (纸本)9781728190549

In the era of the Internet of Things (IoT), decentralized paradigms for machine learning are gaining prominence. In this paper, we introduce a federated learning model that capitalizes on the Euclidean distance between device model weights to assess their similarity and disparity. This is foundational for our system, directing the formation of coalitions among devices based on the closeness of their model weights. Furthermore, the concept of a barycenter, representing the average of model weights, helps in the aggregation of updates from multiple devices. We evaluate our approach using homogeneous and heterogeneous data distribution, comparing it against traditional federated learning averaging algorithm. Numerical results demonstrate its potential in offering structured, outperformed and communication-efficient model for IoT-based machine learning.

关键词： Federated learning IoT Devices Euclidean Distance Heterogeneity Coalition Formation

来源：评论

学校读者我要写书评

暂无评论

Iterative learning control With data-driven-Based Compensation

引用

ieee TRANSACTIONS ON CYBERNETICS 2022年第8期52卷 7492-7503页

作者： He, Shaoying Chen, Wenbo Li, Dewei Xi, Yugeng Xu, Yunwen Zheng, Pengyuan Shanghai Jiao Tong Univ Dept Automat Shanghai 200240 Peoples R China Shanghai Univ Elect Power Coll Automat Engn Shanghai 200090 Peoples R China

The robust iterative learning control (RILC) can deal with the systems with unknown time-varying uncertainty to track a repeated reference signal. However, the existing robust designs consider all the possibilities of uncertainty, which makes the design conservative and causes the controlled process converging to the reference trajectory slowly. To eliminate this weakness, a data-driven method is proposed. The new design intends to employ more information from the past input-output data to compensate for the robust control law and then to improve performance. The proposed control law is proved to guarantee convergence and accelerate the convergence rate. Ultimately, the experiments on a robot manipulator have been conducted to verify the good convergence of the trajectory errors under the control of the proposed method.

关键词： Convergence Uncertainty Trajectory Iterative learning control Manipulators Task analysis data models data-driven method iterative learning control (ILC) robot manipulator robust

来源：评论

学校读者我要写书评

暂无评论

data-driven Finite-Time Three-Dimensional Differential Game Guidance Law Against a Maneuvering Target With Input Constraints

引用

ieee TRANSACTIONS ON AEROSPACE AND ELECTRONIC systems 2024年第6期60卷 8483-8493页

作者： Liu, Hao Cheng, Ming Wang, Linbo Jiang, Yong Beihang Univ Inst Artificial Intelligence Beijing 100191 Peoples R China Beihang Univ Sch Astronaut Beijing 100191 Peoples R China Beijing Inst Elect Syst Engn Beijing 100854 Peoples R China Changfeng Electromech Technol Design Acad Beijing 100854 Peoples R China

This article addresses the problem of finite-time interception against a high-speed maneuvering target formulated as a two-player zero-sum differential game framework with input constraints. The interceptor seeks to minimize the performance index and the maneuvering target seeks to maximize it. A nonquadratic value function is constructed to incorporate the control constraints into the differential game problem for both sides of the players. A finite-time convergent algorithm is proposed to solve the optimal value function of the differential game from the nonlinear Hamilton-Jacobi-Isaacs equation within finite time. The optimal game strategies are consequently learned using collected data of the engagement states based on the reinforcement learning. Uniform and global finite-time convergent property is confirmed via Lyapunov analysis with the obtained Nash strategies. Simulation results of an interceptor against a high-speed maneuvering target are provided to show the efficacy of the proposed method.

关键词： Games Differential games Aerospace and electronic systems Navigation Mathematical models Vectors Transmission line matrix methods Differential game input constraint optimal guidance reinforcement learning (RL) nonlinear system

来源：评论

学校读者我要写书评

暂无评论

data driven Approach for User-owned Renewable Energy Sources Allocation in Community Microgrid 34

Data Driven Approach for User-owned Renewable Energy Sources...

引用

34th Australasian Universities Power Engineering conference

作者： Alam, Md Morshed Hossain, Jahangir Bin Mofidul, Raihan Univ Technol Sydney Elect & Data Engn Sydney NSW Australia

ISBN: (纸本)9798350377958;9798350377941

Massive integration of photovoltaic (PV) and energy storage systems (ESS) in the energy system increases the utilization of renewable energy and system complexity. Optimal and centralized scheduling and management of these sources can make them more effective and acceptable in the energy community. In this regard, this paper proposes a virtual scheduling of the PV-ESS system based on a deep learning-based data-driven approach. The proposed system divides the time span into two segments based on the availability of PV generation. A deep learning model is applied to predict the day-ahead PV generation and demand during the two-time segments. An optimization algorithm is then applied to schedule the PV and ESS based on the predictive outcomes. For validating the proposed system, a community microgrid framework is considered where multiple PV and ESS systems are connected. In addition, the user-owned PV-ESS systems are controlled by the microgrid. The simulation results demonstrate the proposed system's performance in terms of reliability and sustainability.

关键词： Community microgrid Deep learning Scheduling Optimization Energy storage system

来源：评论

学校读者我要写书评

暂无评论

Combined ILC and DOB for Repetitive control with Application to Fast Tool Servo System 20

Combined ILC and DOB for Repetitive Control with Application...

引用

20th ieee/ASME International conference on Mechatronic and Embedded systems and Applications (MESA)

作者： Jing, Yajie Hou, Beibei Lin, Pengbo Lu, Shuaishuai Yan, Peng Qilu Univ Technol Sch Mech Engn Shandong Acad Sci Jinan Peoples R China Shandong Inst Mech Design & Res Jinan Peoples R China Shandong Univ Sch Mech Engn Key Lab High Efficiency & Clean Mech Manufacture Minist Educ Jinan Peoples R China

ISBN: (纸本)9798331516246;9798331516239

Fast tool servo (FTS) systems using piezoelectric technology show great promise for ultra-precise machining due to their high precision and responsiveness. However, during machining, these systems often struggle with tracking accuracy due to factors like system complexity and external disturbances. To address this, we propose a robust control method specifically for piezo-driven FTS. Our approach aims to improve tracking accuracy and disturbance rejection during machining. We use a robust repetitive control scheme with a disturbance observer (DOB) to minimize the effect of interference on tracking. However, in practice, solving the inverse model of the controlled object can introduce uncertainty. To enhance the disturbance observer's performance, we add iterative learning control (ILC). This helps refine control actions based on past errors, improving overall system performance.

关键词： FTS DOB iterative learning control

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：