Field Of Multi-agent Systems Research Articles

Securing Consensus in Fractional-Order Multi-Agent Systems: Algebraic Approaches Against Byzantine Attacks

This paper investigates the behavior of fractional-order nonlinear multi-agent systems subjected to Byzantine assaults, specifically focusing on the manipulations of both sensors and actuators. We employ weighted graphs, both directed and undirected, to illustrate the system's topology. Our methodology combines algebraic graph theory with fractional-order Lyapunov techniques to develop algebraic requirements for leader-following consensus, providing a robust framework for analyzing consensus dynamics in these complex systems. We present quantitative results demonstrating the effectiveness of our approach, including two numerical examples that validate the proposed requirements for consensus evaluation. Notably, our work highlights the novelty of using fractional-order systems to enhance resilience against adversarial conditions, contributing significantly to the field of multi-agent systems. By clarifying key terms and streamlining our language, we ensure accessibility for a broader audience while emphasizing the implications of our findings for real-world applications.

An overview: Attention mechanisms in multi-agent reinforcement learning

In recent years, in the field of Multi-Agent Systems (MAS), significant progress has been made in the research of algorithms that combine Reinforcement Learning (RL) with Attention Mechanism (AM). However, there is a lack of comprehensive reviews in this field. Based on this, this paper does the following work. Firstly, it reviews the classical algorithms of RL and AM; Secondly, it systematically introduces the combination of RL and AM; Thirdly, it sorts out their application in the field of single-agent and multi-agent, and pays attention to and looks forward to the challenges and future research prospects in this field; The last part offers a comprehensive analysis of the challenges encountered by research in this field and anticipates future research paths. The research offers a conceptual understanding and theoretical foundation for future applications of RL using AM and facilitates further in-depth study in this field for researchers.

Special Issue “Advances in Multi-Agent Systems”: Editorial

Multi-agent systems (MAS) are collections of autonomous computational entities (the agents) capable of pro-actively pursuing goals and re-actively adapting to environment change. Agents in MAS exploit their social abilities, interacting with peers, and their situated capabilities as well, by perceiving and acting on the world around them. From distributed computing to intelligent systems, the relevance of agents and MAS as software abstractions is steadily growing as they are extensively and increasingly used to model, simulate, and build heterogeneous systems across a huge variety of diverse application scenarios and business domains, ranging from industrial manufacturing to robotics, from social simulation to applications, and more. The recent, renewed popularity of AI techniques has further spread the adoption of MAS, focusing in particular on the cognitive capabilities of agents, so that intelligent systems can be modelled and built as MAS. Along those lines, this Special Issue gathers five contributions that well represent the many diverse advancements that are currently ongoing in the MAS field.

Distributed Differential Games for Control of Multi-Agent Systems

Motivated by the challenges arising in the field of multi-agent system (MAS) control, we consider linear heterogeneous MAS subject to local communication and investigate the problem of designing distributed controllers for such systems. We provide a game theoretic framework for systematically designing distributed controllers, taking into account individual objectives of the agents and their possibly incomplete knowledge of the MAS. Linear state-feedback control laws are obtained via the introduction of a <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">distributed differential game</i> , namely, the combination of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">local non-cooperative differential games</i> , which are solved in a decentralized fashion. Conditions for stability of the MAS are provided for the special cases of acyclic and strongly connected communication graph topologies. These results are then exploited to provide stability conditions for general graph topologies. The proposed framework is demonstrated on a tracking synchronization problem associated with the design of a distributed secondary voltage controller for microgrids and on a numerical example.

AntAlate-A Multi-Agent Autonomy Framework.

AntAlate is a software framework for Unmanned Aerial Vehicle (UAV) autonomy, designed to streamline and facilitate the work of application developers, particularly in deployment of Multi-Agent Robotic Systems (MARS). We created AntAlate in order to bring our research in the field of multi-agent systems from theoretical results to both advanced simulations and to real-life demonstrations. Creating a framework capable of catering to MARS applications requires support for distributed, decentralized, control using local sensing, performed autonomously by groups of identical anonymous agents. Though mainly interested in the emergent behavior of the system as a whole, we focused on the single agent and created a framework suitable for a system of systems approach, while minimizing the hardware requirements of the single agent. Global observers or even a centralized control can be added on top of AntAlate, but the framework does not require a global actor to finalize an application. The same applies to a human in the loop, and fully autonomous UAV applications can be written in as straightforward a way as can semi-autonomous applications. In this paper we describe the AntAlate framework and demonstrate its utility and versatility.

Distributed Learning Applications in Power Systems: A Review of Methods, Gaps, and Challenges

In recent years, machine learning methods have found numerous applications in power systems for load forecasting, voltage control, power quality monitoring, anomaly detection, etc. Distributed learning is a subfield of machine learning and a descendant of the multi-agent systems field. Distributed learning is a collaboratively decentralized machine learning algorithm designed to handle large data sizes, solve complex learning problems, and increase privacy. Moreover, it can reduce the risk of a single point of failure compared to fully centralized approaches and lower the bandwidth and central storage requirements. This paper introduces three existing distributed learning frameworks and reviews the applications that have been proposed for them in power systems so far. It summarizes the methods, benefits, and challenges of distributed learning frameworks in power systems and identifies the gaps in the literature for future studies.

Scalable Consensus in Networks of Multiagent Systems Using High-Gain Observers

Consensus algorithms are popular in the field of multiagent systems due to their wide application in formation control, distributed estimation, sensor networks, etc. Generally, for certain classes of undirected graphs, with an increase in the network size, the second smallest eigenvalue of the graph Laplacian decreases toward zero, which leads to a slow convergence rate. We present a scalable consensus algorithm using proportional derivative (PD) control where the eigenvalues of the closed-loop Laplacian matrix are invariant with respect to the size of the network for general directed graphs. The PD controller is realized using a high-gain observer. We show that the trajectories of the closed-loop system when the high-gain observer is used can be brought arbitrarily close to the trajectories under the PD controller. Simulation results are presented to demonstrate the efficacy of the proposed algorithm.

Special Issue “Multi-Agent Systems”: Editorial

Multi-agent systems (MAS) are built around the central notions of agents, interaction, and environment. Agents are autonomous computational entities able to pro-actively pursue goals, and re-actively adapt to environment change. In doing so, they leverage on their social and situated capabilities: interacting with peers, and perceiving/acting on the environment. The relevance of MAS is steadily growing as they are extensively and increasingly used to model, simulate, and build heterogeneous systems across many different application scenarios and business domains, ranging from logistics to social sciences, from robotics to supply chain, and more. The reason behind such a widespread and diverse adoption lies in MAS great expressive power in modeling and actually supporting operational execution of a variety of systems demanding decentralized computations, reasoning skills, and adaptiveness to change, which are a perfect fit for MAS central notions introduced above. This special issue gathers 11 contributions sampling the many diverse advancements that are currently ongoing in the MAS field.

BASTA: BDI-based architecture of simulated traffic agents

ABSTRACT It is expected that connected and autonomous vehicles (CAVs) will soon become a regular means of transportation. To fully leverage the potential of this new technology, it is necessary to equip such cars with efficient algorithms permitting them to drive in a safe and optimal manner. Thereby we aim to design and implement tools for convenient evaluation of strategies for driving and interactions in various settings. In this paper, we present the results of the first stage of our bigger research program on a simulation framework of CAVs. A search for a balance between complexity and comprehensibility of the solution led us to the field of multiagent systems. Beliefs-Desires-Intentions (BDI) systems offer useful abstractions for activities of a single self-driving car and collective intelligence of such vehicles. Indeed, the BDI framework helps to combine two distinct natures of a self-driving car: its reactiveness and proactiveness. Moreover, the modularity of the resulting architectures for an individual CAV and urban traffic induced by these cars makes the design intelligible and flexible. Our prototype verifies the feasibility of this concept.

A survey of modelling and smart management tools for power grids with prolific distributed generation

Abstract Over the past few decades, the proliferation of distributed generation (DG), fuelled chiefly by environmental concerns, has brought about fundamental changes to the structure and behaviour of the power grid. These changes have spawned new challenges in the way with which the power grid should be managed. As a result, the development of new modelling approaches, as well as management techniques deemed “smart”, is called for to better monitor and control this increasingly complex system by taking into account the specificities of distributed generators (DGs) and their impact on the grid. This paper explores the evolution of power grids into “smart grids” in the light of the penetration of distributed generation and the features that ensue. Then, it gives an overview of notable works in the literature aimed at the modelling and control of power grids within this framework. The reader is provided with an in-depth analysis of a variety of methods through the lens of the applications for which they are most suited. Though well-established methods that have been in use for infrastructure planning for decades, such as optimal power flow (OPF), remain reliable tools, cutting-edge research in the field of multi-agent systems (MAS) opens the door to putting a myriad of complex numerical algorithms in the service of power grid monitoring and control. In addition, a special interest is paid to demand-side management (DSM) techniques.

Resilient Multiscale Coordination Control against Adversarial Nodes

Multiscale consensus has been studied recently as a new concept in the field of multi-agent systems, which is able to accommodate many complicated coordination control tasks where values are measured in different scales due to, e.g., the constraints of physical environment. In this paper, we investigate the problem of resilient multiscale coordination control against a set of adversarial or non-cooperative nodes in directed networks. We design a multiscale filtering algorithm based upon local information which can withstand both faulty and Byzantine nodes. Building on the concept of network robustness, we establish necessary and sufficient conditions guaranteeing multiscale consensus with general time varying scales in the presence of globally bounded as well as locally bounded threats. In particular, for a network containing at most R faulty nodes, multiscale consensus is achieved if and only if the network is (R+1,R+1)-robust. The counterpart when having at most R Byzantine nodes instead is that the induced subnetwork of cooperative nodes is R+1-robust. Conditions guaranteeing resilient consensus for time-dependent networks are developed. Moreover, multiscale formation generation problems are introduced and solved as the generalizations. Finally, some numerical examples including applications in modular microgrids and power systems are worked out to demonstrate the availability of our theoretical results.

Distributed Constraint Optimization Problems and Applications: A Survey

The field of multi-agent system (MAS) is an active area of research within artificial intelligence, with an increasingly important impact in industrial and other real-world applications. In a MAS, autonomous agents interact to pursue personal interests and/or to achieve common objectives. Distributed Constraint Optimization Problems (DCOPs) have emerged as a prominent agent model to govern the agents' autonomous behavior, where both algorithms and communication models are driven by the structure of the specific problem. During the last decade, several extensions to the DCOP model have been proposed to enable support of MAS in complex, real-time, and uncertain environments. This survey provides an overview of the DCOP model, offering a classification of its multiple extensions and addressing both resolution methods and applications that find a natural mapping within each class of DCOPs. The proposed classification suggests several future perspectives for DCOP extensions and identifies challenges in the design of efficient resolution algorithms, possibly through the adaptation of strategies from different areas.

From roles to standards: a dynamic maintenance approach using incentives

Social coordination has been addressed in multi-agent systems, making use of concepts such as institutions, norms, commitments, conventions, roles, or trust. In this paper, we argue the need to tackle open and dynamic environments with yet another concept: the notion of a standard, seen as a measurable and non-committing expectation. Not much work has been done in the field of multi-agent systems addressing the evolving nature of roles, especially in open systems, in which changes in the population bring about changes in the expectations generated from roles. Using standards measured from roles as the focus of attention, we propose an incentive-based mechanism to maintain roles over time. This approach is put in contrast with reorganization, which is needed when incentives are not cost-effective. Different search algorithms are proposed to illustrate incentive-based maintenance. Some empirical results are shown based on the principal-agent model from economics.

A Formal Model of Communication and Context Awareness in Multiagent Systems

Awareness is a concept that has been frequently studied in the context of Computer Supported Cooperative Work. However, other fields of computer science can benefit from this concept. Recent research in the multi-agent systems field has highlighted the relevance of complex interaction models such as multi-party communication and context awareness for simulation and adaptive systems. In this article, we present a generic interaction model that enables to use these different models in a standardized way. Emerging as a first-order abstraction, the environment, in the sense of a common medium for the agents, is a suitable paradigm to support the agents' awareness. We present an operational model, called Environment as Active Support of Interaction, to take into account all the agents that can be interested in a communication. This model is then extended for the regulation of multiagent systems interactions. Priority policies are given to manage the rules governing the context (un-)awareness of the agents. We also present a new AUML connector to create protocols that take into account the agent awareness to implement proactive behaviour, and several communication scenarios are proposed to show practical applications of this model.

Consensus Control Design for 360 MN Extrusion Machine Producing Process

This paper mainly addresses the issue of 360 MN extrusion machine and focuses on the stabilization control of main table attitude. We will first introduce the problem and then model the extrusion machine. As the machine is a multi-input multioutput (MIMO) and strong coupling system, it is challenging to apply existing control theory to design a controller to stabilize the main table attitude. Motivated by recent research in the field of multiagent systems, we design a consensus control protocol for our system and derive our convergence conditions based directly on Routh stability criterion. The advantages of the design are also demonstrated by numerical simulation.

Resource Sharing for Control of Wildland Fires

Wildland fires (or wildfires) occur on all continents except for Antarctica. These fires threaten communities, change ecosystems, destroy vast quantities of natural resources and the cost estimates of the damage done annually is in the billions of dollars. Controlling wildland fires is resource-intensive and there are numerous examples where the resource demand has outstripped resource availability. Trends in changing climates, fire occurrence and the expansion of the wildland-urban interface all point to increased resource shortages in the future. One approach for coping with these shortages has been the sharing of resources across different wildland-fire agencies. This introduces new issues as agencies have to balance their own needs and risk-management with their desire to help fellow agencies in need. Using ideas from the field of multiagent systems, we conduct the first analysis of strategic issues arising in resource-sharing for wildland-fire control. We also argue that the wildland-fire domain has numerous features that make it attractive to researchers in artificial intelligence and computational sustainability.

A survey of privacy in multi-agent systems

AbstractPrivacy has been a concern for humans long before the explosive growth of the Internet. The advances in information technologies have further increased these concerns. This is because the increasing power and sophistication of computer applications offers both tremendous opportunities for individuals, but also significant threats to personal privacy. Autonomous agents and multi-agent systems are examples of the level of sophistication of computer applications. Autonomous agents usually encapsulate personal information describing their principals, and therefore they play a crucial role in preserving privacy. Moreover, autonomous agents themselves can be used to increase the privacy of computer applications by taking advantage of the intrinsic features they provide, such as artificial intelligence, pro-activeness, autonomy, and the like. This article introduces the problem of preserving privacy in computer applications and its relation to autonomous agents and multi-agent systems. It also surveys privacy-related studies in the field of multi-agent systems and identifies open challenges to be addressed by future research.

Self-organization and multiagent systems: I. Models of multiagent self-organization

Nowadays, there are many problems whose complexity is much higher than the capabilities of modern information technologies. Such problems arise in economics, ecology, managing state-level infrastructures and global computer and telecommunication systems, ensuring the safety of society, and in many other fields. Even though these problems seem to be quite different, they have many common features, which imply common difficulties in their solution. These features are as follows: they are large-scale problems, they are open, have unpredictable dynamics and complex structure, include mobile components, and some others. The management in such systems is a challenging task, which requires a revision of modern views, models, architectures, and development technologies. A response to this challenge is the increasing activity in the field of principles and mechanisms of self-organization and in the software tools for their development. Although the paradigm of self-organizing control systems is not new, it is now at a new step of development, which involves, in particular, its integration with the multiagent system paradigm. The purposes of this paper are to analyze the state-of-the art in the field of multiagent self-organizing systems, to provide a critical review of the available applications, analyze development techniques, and generalize the results obtained in this field. The paper consists of two parts. In the first part, we discuss the modern interpretation of the principles of self-organization is analyzed, and the reasons for which the integration of these principles with the achievements in the field of multiagent systems provides a new impetus to the development of information technologies in the context of most complex modern applications. A systematization and description of the self-organization models and mechanisms implemented in the framework of the multiagent architecture is given, and biological self-organization mechanisms are discussed. Applications of self-organizing multiagent systems in telecommunication, grid resource management, and routing in computer networks with dynamic topology, as well as applications in distributed learning and in detecting intrusions into computer networks are described.

Cooperative agents-based decentralised and scalable approach for complex task allocation pro massive multi-agent system

A major challenge in the field of multi-agent systems is to enable autonomous agents to allocate tasks efficiently. In the context of massive multi-agents system (MMAS) which is characterised by a large number of dynamic and heterogeneous agents, traditional tasks allocations approaches based on the negotiation between agents, or a single allocator agent, proved impracticality. In this paper, we propose a decentralised and scalable approach for complex task allocation pro cooperative MMAS. Our approach is based on the idea of grouping agents according to their capacities to execute sub-tasks. Thus, the approach is based on two steps: 1) hierarchical organisation of agent groups using Galois sub-hierarchy, imminent from the formal concepts analysis approach (FCA); 2) computing the optimal allocation. These two steps are fully distributed among agents with the minimum of communication and a grant of finding the optimal allocation in a polynomial time. Further, this paper extends our last approach by distributing the global allocation process among all agents. It provides a solution based on cooperation among agents. This solution prohibits generation of conflicts. It is based on the idea that each agent has to pick out its own sub-task.

Using Agents to Improve International Maritime Transport Security

Agent-based techniques have the potential to improve the security of international maritime transport threatened by a steep rise of maritime piracy. We have demonstrated that a randomized route-selection strategy resulting from a normal two-player game formulation of the transit problem can decrease the number of attacks and the payoff accumulated by pirates. Coordinating the movement of patrol and transit vessels without a central authority requires techniques for semi-cooperative planning and coalition formation. With regard to existing applications of agent-based techniques, the maritime domain seems currently under-represented compared to other traffic and transportation domains. This work is a first step in addressing the situation and bringing this important domain into the focus of researchers in the multiagent systems field.