Article
Energy & Fuels
Jiawen Li, Tao Yu, Xiaoshun Zhang
Summary: A coordinated power control framework and a novel deep reinforcement learning algorithm EIC-MADDPG are proposed to achieve coordinated control and improve performance in a multi-area integrated energy system (IES). By combining imitation learning and curriculum learning, the algorithm can adaptively derive optimal coordinated control strategies for multiple LFC controllers.
Article
Automation & Control Systems
Yao Zou, Kewei Xia, Zongyu Zuo, Danyong Li, Zhengtao Ding
Summary: This paper solves the coordinated optimal consensus problem for second-order multi-agent systems without velocity information from the distributed perspective. It proposes a velocity-free distributed coordinated optimal control strategy by minimizing a global target function aggregated by a group of convex local ones. The proposed strategy achieves the coordinated optimal control objective even without velocity information.
Article
Chemistry, Physical
Jiawen Li, Tao Yu, Bo Yang
Summary: An intelligent control framework is proposed for coordinating the air and hydrogen supply systems in PEMFCs, using ensemble imitation learning and multi-trick deep deterministic policy gradient approach to enhance exploration efficiency. Multiple reinforcement learning explorers and control algorithm explorers are utilized to address sparse rewards and improve training efficiency. Multiple tricks are applied to improve the overestimated Q value, resulting in a model-free intelligent control algorithm with better global searching ability.
INTERNATIONAL JOURNAL OF HYDROGEN ENERGY
(2021)
Article
Energy & Fuels
Wei Deng, Wei Pei, Yuting Teng, Qi Wu, Yin Yi, Xinji Cao
Summary: This paper provides an overview of the concepts and functional advantages of DC distribution systems, along with an introduction to the main equipment, topologies, and application scenarios of multi-terminal DC (MTDC) distribution systems. It analyzes and summarizes the coordinated control technology of MTDC distribution systems, including control architecture, coordinated control, damping control, and operation mode. The existing MTDC distribution demonstration projects are summarized and recent typical projects are introduced. The paper concludes by highlighting the areas that can be studied for the innovation and application of MTDC distribution systems.
Article
Energy & Fuels
Valery Stennikov, Evgeny Barakhtenko, Gleb Mayorov, Dmitry Sokolov, Bin Zhou
Summary: The creation of intelligent integrated energy systems is a promising alternative to traditional energy systems, but faces challenges in balancing centralized and distributed energy generation. The multi-agent approach is used to model and optimize the interaction between the two.
Article
Automation & Control Systems
Wenlong Yang, Zongying Shi, Yisheng Zhong
Summary: This paper studies the distributed time-varying formation control problems for a class of linear systems with heterogeneous uncertainties and directed graphs. Both cases with and without an active leader having unknown control inputs are considered. Two new robust adaptive formation protocols are designed in a totally distributed fashion, incorporating a fully distributed nominal controller and an adaptive compensating signal to deal with uncertainties. The proposed protocols guarantee the convergence of formation errors to zero based on Lyapunov arguments, and the theoretical results are verified through numerical simulations.
Article
Computer Science, Information Systems
Maryam Nasri, Herbert L. Ginn, Mehrdad Moallem
Summary: This paper introduces an agent-based architecture for coordinating power electronic converters in stand-alone microgrids. By utilizing a publish-subscribe agent architecture over a distributed hash table searching overlay, it provides a scalable technology for real-time coordination of power converters in microgrids.
Article
Automation & Control Systems
He Wang, Wei Ren, Wenwu Yu, Dong Zhang
Summary: This paper considers the distributed consensus control problem for disturbed second order multi-agent systems with directed networks, investigating leaderless and tracking cases. Different methods are proposed to solve the leaderless and tracking problems, with simulation examples provided to verify the theoretical results.
Article
Automation & Control Systems
Xuxi Zhang, Ting Chen, Xianping Liu
Summary: This paper investigates the problem of bipartite output containment control in heterogeneous multi-agent systems over directed signed graphs by utilizing distributed control protocols. It proposes a fully distributed dynamic compensator to estimate the convex hull and symmetric convex hull spanned by the states of the leaders. Furthermore, a distributed observer based on neighbor information is presented to estimate the states of the followers. An output feedback control protocol is also provided to deal with the bipartite containment control problem of the linear heterogeneous multi-agent systems based on the fully distributed dynamic compensator and neighbor-based observer.
INTERNATIONAL JOURNAL OF SYSTEMS SCIENCE
(2023)
Article
Automation & Control Systems
Wenyan Tang, Jia Wu, Ning Liu, Haihong Mo
Summary: This study investigates the consensus tracking problem for unknown multi-agent systems with time-varying communication topology using data-driven control and model predictive control methods. The proposed distributed iterative protocol does not require knowledge of the dynamics of the multi-agent systems, utilizing only local input-output data for each agent. Numerical simulations demonstrate the effectiveness of the derived consensus conditions.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2021)
Article
Computer Science, Information Systems
Wenhui Ren, Xuxi Zhang
Summary: In this paper, the containment control problem of a class of second-order nonlinear multi-agent systems under directed graph is considered. The problem is solved using disturbance observers and adaptive distributed control method. Simulation results demonstrate the effectiveness of the proposed control strategy.
INFORMATION SCIENCES
(2023)
Article
Automation & Control Systems
Dario Giuseppe Lui, Alberto Petrillo, Stefania Santini
Summary: In this paper, a novel fully-Distributed Model Reference Adaptive Control (DMRAC) approach is proposed to address the containment control problem of heterogeneous uncertain high-order linear Multi-Agent Systems (MASs). The analytical derivation of the asymptotic stability of the whole closed-loop network and the boundedness of the adaptive gains validates the effectiveness of the proposed approach in achieving convergence towards the convex hull spanned by leaders.
JOURNAL OF THE FRANKLIN INSTITUTE-ENGINEERING AND APPLIED MATHEMATICS
(2021)
Article
Chemistry, Multidisciplinary
Zongcheng Liu, Hanqiao Huang, Sheng Luo, Wenxing Fu, Qiuni Li
Summary: A novel global consensus method is proposed to address the control of uncertain multi-agent systems with completely unknown system nonlinearities and unknown control coefficients. By constructing novel filters and barrier function-based distributed controllers, the method achieves global consensus of the MAS while guaranteeing prescribed tracking-error performance, as rigorously proved and demonstrated through simulation results in two examples.
APPLIED SCIENCES-BASEL
(2021)
Article
Computer Science, Information Systems
Xiongtao Shi, Yanjie Li, Qi Liu, Ke Lin, Shiyu Chen
Summary: This paper addresses the cooperative output regulation problem of heterogeneous linear multi-agent systems using an event-triggered control strategy based on a directed graph. A novel distributed event-triggered control protocol is proposed to eliminate the need for continuous update of the control rule and ensure the absence of Zeno behavior. By utilizing adaptive control technology, the cooperative output regulation problem is solved in a fully distributed manner, without relying on global information such as the number of agents or eigenvalues of the Laplacian matrix. The considered communication network is a more general directed network, resulting in a reduction of the minimum communication channel compared to the undirected case. The effectiveness of the proposed control strategy is validated on a group of heterogeneous agents.
INFORMATION SCIENCES
(2023)
Article
Automation & Control Systems
Shuyi Xiao, Jiuxiang Dong
Summary: This paper investigates distributed fault-tolerant containment control (FTCC) problem of nonlinear multi-agent systems (MASs) under a directed network topology. The proposed control framework, independent of global information about communication topology, consists of two layers. Through a hierarchical control strategy, the FTCC problem with a directed graph can be simplified into distributed containment control of the upper layer and fault-tolerant tracking control of the lower layer.
IEEE-CAA JOURNAL OF AUTOMATICA SINICA
(2021)
Article
Automation & Control Systems
Ivano Notarnicola, Ying Sun, Gesualdo Scutari, Giuseppe Notarstefano
Summary: This study focuses on distributed big-data nonconvex optimization in multiagent networks. A novel distributed solution method is proposed, where agents update one block of the entire decision vector in an uncoordinated fashion to address nonconvexity and reduce communication overhead in large-scale problems. Numerical results demonstrate the effectiveness of the algorithm and highlight the impact of block dimension on communication overhead and convergence speed.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2021)
Article
Automation & Control Systems
Andrea Camisa, Ivano Notarnicola, Giuseppe Notarstefano
Summary: This article addresses the distributed mixed-integer linear programming setup in control applications, proposing a fully distributed algorithm that can provide accurate feasible solutions in finite time with low suboptimality bounds.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2022)
Article
Automation & Control Systems
Ivano Notarnicola, Andrea Simonetto, Francesco Farina, Giuseppe Notarstefano
Summary: This paper presents a distributed optimization algorithm for solving online personalized optimization problems over a network of computing and communicating nodes. The algorithm consists of two components: a dynamic gradient tracking scheme for finding local solution estimates and a recursive least squares scheme for estimating unknown parameters. The algorithm exhibits a bounded regret under suitable assumptions. A numerical example is provided to corroborate the theoretical analysis.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2023)
Article
Robotics
Andrea Testa, Giuseppe Notarstefano
Summary: The article presents a branch-and-price algorithm for self-assignment of tasks in a network of agents, where each agent locally solves small problems and communicates with neighbors to converge to the optimal solution. By implementing the proposed algorithm in a robot operating system testbed, the team of heterogeneous robots successfully solved the task assignment problem.
IEEE TRANSACTIONS ON ROBOTICS
(2022)
Article
Automation & Control Systems
Guido Carnevale, Ivano Notarnicola, Lorenzo Marconi, Giuseppe Notarstefano
Summary: This paper proposes a distributed optimization algorithm ASYNCHRONOUS TRIGGERED GRADIENT TRACKING to solve consensus optimization over networks with asynchronous communication. It introduces two triggered versions of the algorithm, one in synchronous and the other in asynchronous way. The stability analysis and simulations show that both versions achieve exponential stability for any estimate initialization and improve the performance of inter-agent communication.
Article
Automation & Control Systems
Andrea Camisa, Giuseppe Notarstefano
Summary: This article discusses the distributed control of microgrids, taking into account the unpredictability of renewable energy sources. A distributed methodology based on neighboring communication is proposed and its effectiveness is verified through numerical experiments.
IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY
(2023)
Article
Automation & Control Systems
Guido Carnevale, Giuseppe Notarstefano
Summary: This article introduces the Gradient Tracking algorithm in the nonconvex distributed consensus optimization framework, and proves its convergence properties through system theoretical analysis.
IEEE CONTROL SYSTEMS LETTERS
(2022)
Article
Economics
Brian D. O. Anderson, Manfred Deistler, Marco Lippi
Summary: The survey discusses the formulation of modelling problems for dynamic factor models and potential algorithms for solving them, with an emphasis on understanding error handling requirements and the relevance of the proposed model application. It also considers mixed frequency problems and the identification of certain classes of processes.
Article
Automation & Control Systems
Guido Carnevale, Andrea Camisa, Giuseppe Notarstefano
Summary: This article focuses on studying an online version of the emerging distributed constrained aggregative optimization framework for applications in cooperative robotics. Inspired by an existing scheme, a distributed algorithm named projected aggregative tracking is proposed to solve the online optimization problem. The article proves the bounded dynamic regret and linear convergence rate in the static case. Numerical examples also demonstrate the effectiveness of the approach in a robotic surveillance scenario.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2023)
Article
Robotics
Andrea Camisa, Andrea Testa, Giuseppe Notarstefano
Summary: This paper discusses a large-scale instance of the Pickup-and-Delivery Vehicle Routing Problem (PDVRP) solved by a network of mobile cooperating robots. A distributed algorithm based on primal decomposition is proposed, which ensures privacy of sensitive information and exhibits good scalability. The effectiveness of the algorithm is demonstrated through Gazebo simulations and experiments on a real testbed with ground and aerial robots.
IEEE TRANSACTIONS ON ROBOTICS
(2023)
Proceedings Paper
Automation & Control Systems
Lorenzo Sforni, Andrea Camisa, Giuseppe Notarstefano
Summary: This paper discusses a linear quadratic optimal control problem in which the system dynamics is unknown and the feedback control is required to have a desired sparsity pattern. The authors propose a reinforcement learning framework based on Q-learning to address this problem. Numerical tests on a scenario with randomly generated graph and unstable dynamics show the effectiveness of the algorithm in producing stabilizing and sparse feedback control.
2022 IEEE 61ST CONFERENCE ON DECISION AND CONTROL (CDC)
(2022)
Proceedings Paper
Automation & Control Systems
Nicola Mimmo, Lorenzo Marconi, Giuseppe Notarstefano
Summary: The paper investigates a well-known extremum seeking scheme and proves its uniformity properties in terms of dither signal amplitudes and cost function. These properties are then used to demonstrate that the scheme guarantees semi-global practical stability of the global minimizer even in the presence of local saddle points. To achieve these results, the average system associated with the extremum seeking scheme is analyzed using arguments based on the Fourier series.
2022 IEEE 61ST CONFERENCE ON DECISION AND CONTROL (CDC)
(2022)
Proceedings Paper
Automation & Control Systems
Guido Carnevale, Giuseppe Notarstefano
Summary: This paper addresses the problem of distributed aggregative optimization, where agents in a network aim to minimize the sum of local objective functions. The authors propose a novel data-driven distributed algorithm that learns the parameters of unknown functions and optimization steps using users' noisy feedback. Upper bounds for dynamic regret are proven, showing that the asymptotic performance of the algorithm is not affected by initial conditions and learning errors.
2022 IEEE 61ST CONFERENCE ON DECISION AND CONTROL (CDC)
(2022)
Proceedings Paper
Automation & Control Systems
Guido Carnevale, Nicola Mimmo, Giuseppe Notarstefano
Summary: In this paper, the authors propose a novel distributed feedback optimization law called Aggregative Tracking Feedback, which guides network systems to an optimal steady state by reconstructing information in the network. The effectiveness of the method is demonstrated through system theoretical analysis and numerical simulations.
Article
Automation & Control Systems
Andrea Camisa, Ivano Notarnicola, Giuseppe Notarstefano
Summary: This letter presents a distributed stochastic optimization framework for agents in a network to cooperatively learn an optimal policy. The proposed algorithm utilizes consensus iterations and stochastic approximation to find the optimal solution without a central coordinator, showcasing scalability properties.
IEEE CONTROL SYSTEMS LETTERS
(2022)
