Multi-agent coordination by iterative learning control: Centralized and decentralized strategies

Ahn, Hyo‐Sung; Chen, YangQuan; Moore, Kevin L.

doi:10.1109/isic.2011.6045400

Cited by 13 publications

(6 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, MBC has been the subject of research for many decades, and well‐established design and analysis techniques are available in the literature. Applications in MAS of MBC systems with adaptation include consensus/tracking via passivity‐based techniques, asymptotic and finite‐time consensus, and centralized/decentralized formation control, for example.…”

Section: A Unifying Framework For Hybrid Online Learning Control In Mmentioning

confidence: 99%

Hybrid online learning control in networked multiagent systems: A survey

Poveda

Benosman

Teel

2018

Adaptive Control & Signal

View full text Add to dashboard Cite

This survey paper studies deterministic control systems that integrate three of the most active research areas during the last years: (1) online learning control systems, (2) distributed control of networked multiagent systems, and (3) hybrid dynamical systems (HDSs). The interest for these types of systems has been motivated mainly by two reasons: First, the development of cheap massive computational power and advanced communication technologies, which allows to carry out large computations in complex networked systems, and second, the recent development of a comprehensive theory for HDSs that allows to integrate continuous-time dynamical systems and discrete-time dynamical systems in a unified manner, thus providing a unifying modeling language for complex learning-based control systems. In this paper, we aim to give a comprehensive survey of the current state of the art in the area of online learning control in multiagent systems, presenting an overview of the different types of problems that can be addressed, as well as the most representative control architectures found in the literature. These control architectures are modeled as HDSs, which include as special subsets continuous-time dynamical systems and discrete-time dynamical systems. We highlight the different advantages and limitations of the existing results as well as some interesting potential future directions and open problems. KEYWORDSadaptive control, hybrid dynamical systems, learning, multiagent systems, networked systems 228One of the fields where control systems with learning and adaptation have recently seen increased attention is in the area of networked multiagent systems (MAS), 16,17 which are systems comprised of multiple interacting subsystems, each subsystem having individual dynamics and computational capabilities as well as limited sensing, communication, and actuation. Depending on the structure of the underlying communication graph between the agents of the system, MAS can be centralized or decentralized. In a centralized MAS, there exists one central agent which has access to the state information of all other agents and which is able to compute the control action for all agents of the system. On the other hand, in a decentralized or distributed MAS, agents individually compute their own control signals, which are shared with a subset of the other agents characterized by a communication graph. Figure 1 illustrates three MASs with different communication graphs. Centralized MASs are typically not scalable, and they present the disadvantage of having a single point of failure that could potentially shut down the complete system. 16 Decentralized and distributed systems, on the other hand, are scalable and more robust, and in nonstationary environments, they allow for adaptation and self-organization of the system. 18 Recent technological advances in communication and computation have made MAS ubiquitous, and today, they can be found in several engineering systems such as in power generation and distribution systems, 19-21 teams of ...

show abstract

Section: A Unifying Framework For Hybrid Online Learning Control In Mmentioning

confidence: 99%

Hybrid online learning control in networked multiagent systems: A survey

Poveda

Benosman

Teel

2018

Adaptive Control & Signal

View full text Add to dashboard Cite

show abstract

“…In recent years, the ILC‐based coordination problems of MASs have been widely investigated due to many industry processes that require repeatable executions and coordination among a group of entities, for example, maintenance of satellite formations 18 and cooperative transport by robots 19 . Based on the ILC approach, the formation control problems were discussed in References 20‐22 for nonlinear MASs. The high‐precision formation control was considered in Reference 23 for discrete‐time nonlinear MASs.…”

Section: Introductionmentioning

confidence: 99%

“…The main contributions are as follows. First, the MASs we consider are OSL NSs that encompass the classical Lipschitz NSs, which are different from the aforementioned works 20‐31 . Second, unlike the consensus control methods in References 46‐49, the ILC algorithms proposed in this paper can enable the OSL MASs have complete tracking capability for the leader's trajectory.…”

Section: Introductionmentioning

confidence: 99%

“…First, the MASs we consider are OSL NSs that encompass the classical Lipschitz NSs, which are different from the aforementioned works. [20][21][22][23][24][25][26][27][28][29][30][31] Second, unlike the consensus control methods in References 46-49, the ILC algorithms proposed in this paper can enable the OSL MASs have complete tracking capability for the leader's trajectory. Last, difference from the work, 50 in this paper, two kinds of MAS structures are considered (i.e., regular and irregular OSL MASs), and P-type and D-type learning schemes with initial state learning are adopted, respectively.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Iterative learning consensus control for one‐sided Lipschitz multi‐agent systems

Gu,

Wang,

Chen

et al. 2023

Intl J Robust & Nonlinear

View full text Add to dashboard Cite

By applying iterative learning control approach, the consensus is studied for multi‐agent systems (MASs) with one‐sided Lipschitz (OSL) nonlinearity. Firstly, the P‐type and D‐type learning schemes with initial state learning are introduced for such MASs. Then, utilizing the OSL and the quadratically inner‐bounded constraints, the convergence conditions of the consensus algorithms are presented and analyzed under a directed communication graph. We show that both algorithms, on a fixed finite‐time interval, can achieve perfect consensus tracking. Finally, the correctness of the obtained results is illustrated with simulation examples.

show abstract

“…Most of the existing approaches are assumed that initial states are the same at all iterations. Recently, some research investigates a way to alleviate the above hypothesis by proposing the initial state learning algorithm described as follows [10], [27]. The stopping criteria for iterative learning procedure are given as follows.…”

Section: Algorithmmentioning

confidence: 99%

Application of linear multi-agent system approach to design building temperature control system

Pham

View full text Add to dashboard Cite

Building temperature control system (BTCS) is a large-scale interconnected system with a high demand of energy consumption in building. BTCS is generally described by linear time-invariant system using electric-analogous devices such as resistors and capacitors. In view of graph theory, it is modeled as linear multi-agent system (MAS) subjected to undirected communication topology. The main goal of BTCS is to track the reference temperature trajectory which is equivalent to make the tracking error go to zero in finite time. Hence, BTCS properly suits with framework of distributed control, decentralized control, and centralized control. In this thesis, we focus on comparison of three design methods, namely, distributed consensus controllers (DCC), decentralized iterative learning control (ILC), and centralized ILC. First, the DCC is applied to BTCS by solving resource allocation problem with hard control input constraints. Next, we apply decentralized ILC using the derivative of tracking error (D-type). Lastly, centralized ILC design is formulated as minimization problem with a quadratic cost function subject to control input constraints. This problem is also called Q-ILC. We apply the Alternating Direction Method of Multipliers (ADMM) approach to Q-ILC design and derive analytical solution of control input in each iteration. The convergence property of ADMM and Q-ILC algorithm are given. Numerical results are provided to illustrate the effectiveness of these algorithms.

show abstract

Multi-agent coordination by iterative learning control: Centralized and decentralized strategies

Cited by 13 publications

References 10 publications

Hybrid online learning control in networked multiagent systems: A survey

Hybrid online learning control in networked multiagent systems: A survey

Iterative learning consensus control for one‐sided Lipschitz multi‐agent systems

Application of linear multi-agent system approach to design building temperature control system

Contact Info

Product

Resources

About