检索结果-内蒙古大学图书馆

Data-driven cooperative output regulation of multi-agent systems under distributed denial of service attacks

Science China(Information Sciences) 2023年第9期66卷 5-20页

作者： Weinan GAO Zhong-Ping JIANG State Key Laboratory of Synthetical Automation for Process Industries Northeastern University Department of Electrical and Computer Engineering Tandon School of Engineering New York University

This paper addresses an optimal, cooperative output regulation problem for multi-agent systems with distributed denial of service attacks and unknown system dynamics. Unlike existing studies, the proposed solution is essentially a learning-based control strategy such that one can obtain a distributed control policy with internal models through online data and analyze the resilience of closed-loop systems, both without the precise knowledge of system dynamics in the state-space model. The efficiency of the proposed methodology is validated using computer simulations.

关键词： cooperative output regulation adaptive dynamic programming learning-based control multi-agent systems cybersecurity

来源：评论

学校读者我要写书评

暂无评论

A Sensitivity-based Data Augmentation Framework for Model Predictive control Policy Approximation

引用

IEEE TRANSACTIONS ON AUTOMATIC control 2022年第11期67卷 6090-6097页

作者： Krishnamoorthy, Dinesh Harvard Univ Harvard John A Paulson Sch Engn & Appl Sci Cambridge MA 02138 USA

Approximating model predictive control (MPC) policy using expert-based supervised learning techniques requires labeled training datasets sampled from the MPC policy. This is typically obtained by sampling the feasible state space and evaluating the control law by solving the numerical optimization problem offline for each sample. Although the resulting approximate policy can be cheaply evaluated online, generating large training samples to learn the MPC policy can be time-consuming and prohibitively expensive. This is one of the fundamental bottlenecks that limit the design and implementation of MPC policy approximation. This technical article aims to address this challenge, and proposes a novel sensitivity-based data augmentation scheme for direct policy approximation. The proposed approach is based on exploiting the parametric sensitivities to cheaply generate additional training samples in the neighborhood of the existing samples.

关键词： Training data Training Optimization Optimal control Sensitivity Deep learning Approximation algorithms Predictive control learning-based control data augmentation parametric optimization

来源：评论

学校读者我要写书评

暂无评论

On Tuning Parameterized control Policies Online for Safety-Critical Systems - Applied to Biomedical Systems 22

On Tuning Parameterized Control Policies Online for Safety-C...

引用

22nd World Congress of the International Federation of Automatic control (IFAC)

作者： Krishnamoorthy, Dinesh Eindhoven Univ Technol Dept Mech Engn NL-5600 MB Eindhoven Netherlands

ISBN: (纸本)9781713872344

There are different approaches to tune a control policy that would result in a desired closed-loop performance. The typical design flow involves tuning the control policies offline using (high-fidelity) simulators until satisfactory performance is achieved. This paper on the other hand, considers the problem of tuning parameterized control policies directly by interacting with the real system that also has safety-critical constraints. We use safe Bayesian optimization using interior-point methods to tune the parameterized control policy online that guarantees constraint satisfaction with high probability. The proposed framework is applied to a personalized artificial pancreas system for type 1 diabetes. The paper shows that the parameterized control policy used for blood glucose regulation can be safely tuned to personalize the controller for each individual patient using our approach and thus improve its performance. Copyright (c) 2023 The Authors. This is an open access article under the CC BY-NC-ND license (https://***/licenses/by-nc-nd/4.0/)

关键词： learning-based control Bayesian Optimization biomedical control

来源：评论

学校读者我要写书评

暂无评论

learning-based Stochastic Model Predictive control with State-Dependent Uncertainty 2

Learning-based Stochastic Model Predictive Control with Stat...

引用

2nd Annual Conference on learning for Dynamics and control (L4DC)

作者： Bonzanini, Angelo D. Mesbah, Ali Univ Calif Berkeley Dept Chem & Biomol Engn Berkeley CA 94720 USA

The increasing complexity of modern engineering systems can introduce a great deal of uncertainty in our knowledge of system dynamics, which can, in turn, pose a major challenge to safe model-based control. This paper presents a learning-based stochastic model predictive control (LB-SMPC) strategy with chance constraints for tracking. The LB-SMPC strategy systematically handles mismatch between the actual system dynamics and a system model via a state-dependent uncertainty term that is intended to correct model predictions at each sampling time. A chance constraint handling method is presented to ensure state constraint satisfaction to a desired level for the case of state-dependent model uncertainty. Closed-loop simulations demonstrate the usefulness of LB-SMPC for control of a safety-critical plasma system for processing of (bio)materials with hard-to-model and time-varying dynamics.

关键词： learning-based control model predictive control chance constraints

来源：评论

学校读者我要写书评

暂无评论

Robust pose tracking control for a fully-actuated hexarotor UAV based on Gaussian processes

引用

SICE JOURNAL OF control MEASUREMENT AND SYSTEM INTEGRATION 2022年第2期15卷 201-210页

作者： Ibuki, Tatsuya Yoshioka, Hiroto Sampei, Mitsuji Meiji Univ Sch Sci & Technol Kawasaki Kanagawa Japan Esol Co Ltd Tokyo Japan Tokyo Inst Technol Sch Engn Tokyo Japan

This paper presents a robust position/attitude tracking control method for a fully-actuated hexarotor unmanned aerial vehicle (UAV) based on Gaussian processes. Multirotor UAVs suffer from modelling errors due to their structure complexity and aerodynamical disturbances whose perfect mathematical formulation is intractable. To handle this issue, this paper incorporates a data-based learning technique with model-based control. The hexarotor UAV dynamical model, considering modelling errors and aerodynamic disturbances as unknown dynamics, is first derived. Gaussian process regression is next introduced as a learning method for the unknown dynamics, which provides probabilistic distributions of the predicted values. The predicted means are regarded as deterministic information and cancelled out by feedforward control inputs. The predicted variances are considered as the bounds of the model uncertainties with high probability, and a robust control method to ensure ultimate boundedness of the tracking control error is proposed for the uncertain system. The effectiveness of the proposed method is demonstrated via experiments with a self-developed hexarotor UAV testbed.

关键词： Gaussian process multirotor UAV learning-based control robust control nonlinear control

来源：评论

学校读者我要写书评

暂无评论

Behavioral Optimization in a Robotic Serial Reaching Task Using Predictive Information

引用

IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS 2023年第3期15卷 1012-1019页

作者： Sen, Deniz de Kleijn, Roy Kachergis, George Leiden Univ Math Inst NL-2333 CA Leiden Netherlands Leiden Univ Leiden Inst Brain & Cognit NL-2333 CA Leiden Netherlands Stanford Univ Dept Psychol Stanford CA 94305 USA

Prediction is a powerful approach to minimize errors and control problems in familiar environments and tasks. In human motor execution of sequential action, context effects can be observed, such as anticipation of or predictive movement toward target objects, where later subactions are affected by the execution of earlier subactions. In this article, we present a simulation framework for a serial reaching task using a 4-DoF robotic arm to examine the learning of context effects in simulated robotic reinforcement learning agents. As we demonstrate, giving robotic agents access to predictive information about a future target object's identity results in motion optimization, where the identity of the next target modulates earlier subactions. Specifically, agents learn to anticipate and predict the location of the next target object, and move toward it before it appears, thus achieving higher rewards than agents that were not given predictive information.

关键词： learning-based control motion optimization prediction sequential action learning serial response time (SRT) task

来源：评论

学校读者我要写书评

暂无评论

learning for control: An Inverse Optimization Approach

引用

IEEE control SYSTEMS LETTERS 2022年 6卷 187-192页

作者： Akhtar, Syed Adnan Kolarijani, Arman Sharifi Esfahani, Peyman Mohajerin Delft Univ Technol Delft Ctr Syst & Control NL-2628 CD Delft Netherlands

We present a learning method to learn the mapping from an input space to an action space, which is particularly suitable when the action is an optimal decision with respect to a certain unknown cost function. We use an inverse optimization approach to retrieve the cost function by introducing a new loss function and a new hypothesis class of mappings. A tractable convex reformulation of the learning problem is also presented. The method is effective for learning input-action mapping in continuous input-action space with input-output constraints, typically present in control systems. The learning approach can be effectively transformed to learn a Model Predictive control (MPC) behaviour and a case study to mimic an MPC is presented, which is a rather computationally heavy control strategy. Simulation and experimental results show the effectiveness of the proposed approach.

关键词： Supervised learning Aerospace electronics Training Linear programming Cost function Symmetric matrices Least mean squares methods learning-based control supervised learning inverse optimization convex reformulation

来源：评论

学校读者我要写书评

暂无评论

A Deep-learning-based Approach to Eco-Driving-based Energy Management of Hybrid Electric Vehicles

引用

IEEE TRANSACTIONS ON TRANSPORTATION ELECTRIFICATION 2023年第3期9卷 3742-3752页

作者： Sotoudeh, Seyedeh Mahsa HomChaudhuri, Baisravan IIT Dept Mech Mat & Aerosp Engn Chicago IL 60616 USA

This article proposes a deep-learning-based hierarchical control framework for eco-driving-based energy management of connected and automated hybrid electric vehicles (HEVs). The article focuses on a computationally efficient solution for jointly optimizing the HEV's driving cycle (velocity profile) and its powertrain's power split (powertrain energy management). The proposed framework harnesses the benefits of long- and short-term decision-making and is highly computationally efficient because it learns the control law directly with a deep neural network (DNN). At the high level, pseudospectral optimal controller (PSOC) solves the powertrain's energy management problem over driving cycle previews of the entire trip, approximated via data from vehicle-to-infrastructure (V2I) communications. At the low level, a DNN-based model predictive controller uses the high-level's solution and jointly optimizes the driving cycle and powertrain's energy management over short horizons. To satisfy the constraints, a quadratic programming (QP)-based control law modification algorithm has been developed in this article. Eco-driving-based energy management results of our framework, evaluated over standard, urban, and highway driving cycles, indicate its efficacy in generating real-time applicable energy-efficient solutions for unseen driving cycles of the same traffic pattern as the DNN's training data.

关键词： Deep learning energy management strategy (EMS) hybrid electric vehicle (HEV) learning-based control model predictive control (MPC)

来源：评论

学校读者我要写书评

暂无评论

Near-Optimal Design of Safe Output-Feedback controllers From Noisy Data

引用

IEEE TRANSACTIONS ON AUTOMATIC control 2023年第5期68卷 2699-2714页

作者： Furieri, Luca Guo, Baiwei Martin, Andrea Ferrari-Trecate, Giancarlo Ecole Polytech Fed Lausanne EPFL Inst Mech Engn CH-1015 Lausanne Switzerland

As we transition toward the deployment of data-driven controllers for black-box cyberphysical systems, complying with hard safety constraints becomes a primary concern. Two key aspects should be addressed when input-output data are corrupted by noise: how much uncertainty can one tolerate without compromising safety, and to what extent is the control performance affected? By focusing on finite-horizon constrained linear- quadratic problems, we provide an answer to these questions in terms of the model mismatch incurred during a preliminary identification phase. We propose a control design procedure based on a quasiconvex relaxation of the original robust problem and we prove that, if the uncertainty is sufficiently small, the synthesized controller is safe and near-optimal, in the sense that the suboptimality gap increases linearly with the model mismatch level. Since the proposed method is independent of the specific identification procedure, our analysis holds in combination with state-of-the-art behavioral estimators beyond standard least squares. The main theoretical results are validated by numerical experiments.

关键词： Safety Noise measurement Behavioral sciences Trajectory Biological system modeling Numerical models Data models Data-driven control learning-based control linear systems optimal control robust control

来源：评论

学校读者我要写书评

暂无评论

Online Adversarial Stabilization of Unknown Networked Systems

引用

PROCEEDINGS OF THE ACM ON MEASUREMENT AND ANALYSIS OF COMPUTING SYSTEMS 2023年第1期7卷 1-43页

作者： Yu, Jing Ho, Dimitar Wierman, Adam CALTECH Pasadena CA 91125 USA

We investigate the problem of stabilizing an unknown networked linear system under communication constraints and adversarial disturbances. We propose the first provably stabilizing algorithm for the problem. The algorithm uses a distributed version of nested convex body chasing to maintain a consistent estimate of the network dynamics and applies system level synthesis to determine a distributed controller based on this estimated model. Our approach avoids the need for system identification and accommodates a broad class of communication delay while being fully distributed and scaling favorably with the number of subsystems.

关键词： online control learning-based control adversarial control distributed control stability communication delay

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：