Multi-agent Coordination Research Articles

Cooperative output regulation of multi-agent systems with switched leader dynamics

ABSTRACTIn this paper, the cooperative output regulation problem for heterogeneous multi-agent systems is addressed by considering a switched leader dynamics. The switched leader dynamics consists of multiple linear models and a switching rule governing the switches among them. A switched leader is capable of generating various sophisticated reference signals for more complicated multi-agent coordination tasks. A novel distributed hybrid impulsive switching control scheme is proposed to achieve cooperative output regulation. Distributed switching stability is established using the average dwell-time technique with multiple Lyapunov functions. Moreover, the associated distributed control synthesis conditions are formulated in terms of linear matrix inequalities plus linear matrix equations. As a result, both switching rules for the leader and distributed switching protocols for the followers can be jointly synthesised via efficient semi-definite programming. An example has been used to demonstrate the effectiveness of the proposed approach.

A values‐driven self‐organization mechanism for automating multiagent coordination

AbstractIn distributed and open environments, MASs (multiagent systems) generally have no mechanisms for prior coordination and self‐organization has been believed to be the necessary selection to achieve the coordination of agents. This paper first presents a values‐driven model for self‐organization in which the expected emergent properties of a system are specified as the social values while the social values are realized via implicitly inducing members to regulate their individual values and adjust their behaviors to fit the expectations of the system. Based on the values‐driven self‐organization, this paper proposes an automated coordination mechanism for decentralized MASs. In this mechanism, by indirectly changing the difficulties in acquiring resources (which may be delegated to some special agents since MASs generally do not have substantial bodies), MASs can lead agents to regulate their values to be consistent with the social values of MASs so that the coordination of MASs can spontaneously emerge from the local behaviors of agents. Finally, this paper implements a simulation traffic system using the coordination mechanism based on values‐driven self‐organization to validate the emergence of coordination among multiple agents.

Multi-Agent Coordination to High-Dimensional Target Subspaces

In this paper, we consider a coordination problem to high-dimensional target subspaces for networked multi-agent systems. It is a general form of multi-agent coordination which includes a wide range of applications, e.g., rendezvous, alignment, formation and so on. This paper discusses a fundamental issue of this coordination: what kind of network topology enables us to achieve this coordination via distributed and relative control. We derive necessary and sufficient conditions of such network graphs, and show that the connectivity in terms of cliques (i.e., complete subgraphs) plays a crucial role. This result is a natural extension of the well-known result of the consensus control, and provides us with a unified control strategy for various coordination tasks. Numerical examples are given to illustrate the validity of the obtained analysis results.

COOPERATIVE MULTI-AGENT LEARNING IN SOCIAL NETWORKS: A REVIEW

Coordination in cooperative multi-agent systems is one of the important issues in multi-agent learning and has been broadly studies in the literature. Much work in this field has been carried out yet. However, there are still some coordination issues that are required to be improved. In this work, we look over the multi-agent coordination problems in cooperative environments under the networked multi-agent learning framework using some social network structures and will try to improve coordination efficiency. In our framework, we talk about two types of learners, i.e. individual action learner and joint action learner. We are considering hierarchical multi-agent learning framework to accelerate the coordination efficiency. Our research direction is to consider utilizing the characteristics of different neighboring agents (e.g., the past coordination histories of different nodes, the relative degree of different nodes in the neighborhood) while performing the multi-agent learning to improve coordination performance.

From Incident to Inpatient: How Healthcare Coalitions Can Improve Incident Response

In recent years, many urban areas have established healthcare coalitions composed of autonomous (and often competing) hospitals, with the goal of improving emergency preparedness and response. We study the role of such coalitions in the specific context of response to multiple-casualty incidents in an urban setting, where on-scene responders must determine how to send casualties to medical facilities. A key function in incident response is multi-agency coordination. When this coordination is provided by a healthcare coalition, responders can use richer information about hospital capacities to decide where to send casualties. Using bed availability data from an urban area and a suburban area in the United States, we analyze the response capability of healthcare infrastructures under different levels of coordination, and we develop a stress test to identify areas of weakness. We find that improved coordination efforts should focus on decision support using information about inpatient resources, especially in urban areas with high inter-hospital variability in resource availability. We also find that coordination has the largest benefit in small incidents, suggesting that policy makers should improve coordination in smaller incidents. This benefit is a new value proposition for healthcare coalitions, which were originally formed to improve preparedness for large disasters.

Review of Multi-Agent Algorithms for Collective Behavior: a Structural Taxonomy

In this paper, we review multi-agent collective behavior algorithms in the literature and classify them according to their underlying mathematical structure. For each mathematical technique, we identify the multi-agent coordination tasks it can be applied to, and we analyze its scalability, bandwidth use, and demonstrated maturity. We highlight how versatile techniques such as artificial potential functions can be used for applications ranging from low-level position control to high-level coordination and task allocation, we discuss possible reasons for the slow adoption of complex distributed coordination algorithms in the field, and we highlight areas for further research and development.

Cooperative Multi-Agent Systems Using Distributed Reinforcement Learning Techniques

In this paper, the fully cooperative multi-agent system is studied, in which all of the agents share the same common goal. The main difficulty in such systems is the coordination problem: how to ensure that the individual decisions of the agents lead to jointly optimal decisions for the group? Firstly, a multi-agent reinforcement learning algorithm combining traditional Q-learning with observation-based teammate modeling techniques, called TM_Qlearning, is presented and evaluated. Several new cooperative action selection strategies are then suggested to improve the multi-agent coordination and accelerate learning, especially in the case of unknown and temporary dynamic environments. The effectiveness of combining TM_Qlearning with the new proposals is demonstrated using the hunting game.

Distributed coordination for a class of non‐linear multi‐agent systems with regulation constraints

In this paper, a multi-agent coordination problem with steady-state regulation constraints is investigated for a class of nonlinear systems. Unlike existing leader-following coordination formulations, the reference signal is not given by a dynamic autonomous leader but determined as the optimal solution of a distributed optimization problem. Furthermore, we consider a global constraint having noisy data observations for the optimization problem, which implies that reference signal is not trivially available with existing optimization algorithms. To handle those challenges, we present a passivity-based analysis and design approach by using only local objective function, local data observation and exchanged information from their neighbors. The proposed distributed algorithms are shown to achieve the optimal steady-state regulation by rejecting the unknown observation disturbances for passive nonlinear agents, which are persuasive in various practical problems. Applications and simulation examples are then given to verify the effectiveness of our design.

Exponential Stabilization of Fully Actuated Planar Bipedal Robotic Walking With Global Position Tracking Capabilities

This paper focuses on the development of a model-based feedback controller to realize high versatility of fully actuated planar bipedal robotic walking. To conveniently define both symmetric and asymmetric walking patterns, we propose to use the left and the right legs for gait characterization. In addition to walking pattern tracking error, a biped's position tracking error in Cartesian space is included in the output function in order to enable high-level task planning and control such as multi-agent coordination. A feedback controller based on input–output linearization and proportional–derivative control is then synthesized to realize exponential tracking of the desired walking pattern as well as the desired global position trajectory. Sufficient stability conditions of the hybrid time-varying closed-loop system are developed based on the construction of multiple Lyapunov functions. In motion planning, a new method of walking pattern design is introduced, which decouples the planning of global motion and walking pattern. Finally, simulation results on a fully actuated planar biped show the effectiveness of the proposed walking strategy.

Distributed Multi-agent Bidding-Based Approach for the Collaborative Mapping of Unknown Indoor Environments by a Homogeneous Mobile Robot Team

Abstract This paper deals with the problem of the collaborative mapping of unknown indoor environments by a homogeneous mobile robot team. For this aim, a distributed multi-agent coordination approach is proposed for the mapping process to offer a global view of the entire environment. First, the scheme starts by assigning the most suitable robots to the different zones of the environment to be mapped based on a bidding strategy. Then, while a Robot agent of the group explores its local surroundings and collects information about its neighborhood, it sends mapping data to the Human/Machine Interface agent to integrate them into a single global map. Furthermore, a geometric map representation and an algorithm based on obstacles and environment limits detection are used to provide an explicitly geometric representation of the workspace. For validation purposes, Player/Stage simulator is used to show the effectiveness of the proposed distributed approach and algorithms without needing a real multi-robot system and environment. Finally, various scenarios have been carried out and results are compared in terms of (i) required mapping time, (ii) accuracy of the global generated map, and (iii) number of exchanged messages between the agents.

Design of Complex Engineered Systems Using Multi-Agent Coordination

In complex engineering systems, complexity may arise by design, or as a by-product of the system's operation. In either case, the cause of complexity is the same: the unpredictable manner in which interactions among components modify system behavior. Traditionally, two different approaches are used to handle such complexity: (i) a centralized design approach where the impacts of all potential system states and behaviors resulting from design decisions must be accurately modeled and (ii) an approach based on externally legislating design decisions, which avoid such difficulties, but at the cost of expensive external mechanisms to determine trade-offs among competing design decisions. Our approach is a hybrid of the two approaches, providing a method in which decisions can be reconciled without the need for either detailed interaction models or external mechanisms. A key insight of this approach is that complex system design, undertaken with respect to a variety of design objectives, is fundamentally similar to the multi-agent coordination problem, where component decisions and their interactions lead to global behavior. The results of this paper demonstrate that a team of autonomous agents using a cooperative coevolutionary algorithm (CCEA) can effectively design a complex engineered system. This paper uses a system model of a Formula SAE racing vehicle to illustrate and simulate the methods and potential results. By designing complex systems with a multi-agent coordination approach, a design methodology can be developed to reduce design uncertainty and provide mechanisms through which the system level impact of decisions can be estimated without explicitly modeling such interactions.

Differential inequalities in multi-agent coordination and opinion dynamics modeling

Many distributed algorithms for multi-agent coordination employ the simple averaging dynamics, referred to as the Laplacian flow. Besides the standard consensus protocols, examples include, but are not limited to, algorithms for aggregation and containment control, target surrounding, distributed optimization and models of opinion formation in social groups. In spite of their similarities, each of these algorithms has been studied using separate mathematical techniques. In this paper, we show that stability and convergence of many coordination algorithms involving the Laplacian flow dynamics follow from the general consensus dichotomy property of a special differential inequality. The consensus dichotomy implies that any solution to the differential inequality is either unbounded or converges to a consensus equilibrium. In this paper, we establish the dichotomy criteria for differential inequalities and illustrate their applications to multi-agent coordination and opinion dynamics modeling.

Parallel Multiagent Coordination Optimization Algorithm: Implementation, Evaluation, and Applications

In this paper, a parallel computing implementation of a multiagent coordination optimization (MCO) algorithm is introduced using $\mathtt {parfor}$ , a built-in MATLAB function. As a novel variation of particle swarm optimization (PSO), the MCO algorithm has demonstrated significant performance improvement, in terms of both accuracy and efficiency, in solving real-time optimization problems when compared with PSO. However, the ability to handle large-scale optimization problems or use more particles in the algorithm is limited by the sequential implementation of the original MCO algorithm. A numerical evaluation of the parallel MCO algorithm was conducted using the supercomputers in the High Performance Computing Center at Texas Tech University. Based on the results of this evaluation, it was determined that the performance of the parallel MCO is not only superior to that of PSO but is highly efficient as it reduces the computational time. The parallel binary MCO (BMCO) algorithm is presented here as well. By combining the parallel MCO with parallel BMCO algorithms, a new parallel mixed-binary nonlinear programming MCO solver is proposed. Finally, a load balancing coordination problem, a multiagent formation control problem, and a power system vulnerability analysis problem are solved using the corresponding parallel MCO algorithm.

A Distributed Feedback Motion Planning Protocol for Multiple Unicycle Agents of Different Classes

This paper presents a novel feedback method for the motion planning and coordination of multiple agents that belong to two classes, namely class-A and class-B. All agents are modeled via unicycle kinematics. Agents of class-B do not share information with agents of class-A and do not participate in ensuring safety, modeling thus agents with failed sensing/communication systems, agents of higher priority, or moving obstacles with known upper bounded velocity. The method is built upon a family of 2-D analytic vector fields, which under mild assumptions are proved to be safe feedback motion plans with a unique stable singular point. The conditions which ensure collision free and almost global convergence for a single agent and the analytical form of the vector fields are then utilized in the design the proposed distributed, semi-cooperative multi-agent coordination protocol. Semi-cooperative coordination has been defined in prior work as the ad hoc prioritization and conflict resolution among agents of the same class ; more specifically, participation in conflict resolution and collision avoidance for each agent is determined on-the-fly based on whether the agent's motion results in decreasing its distance with respect to its neighbor agents; based on this condition, the agent decides to either ignore its neighbors, or adjust its velocity and avoid the neighbor agent with respect to which the rate of decrease of the pairwise inter agent distance is maximal. The proposed coordination protocol builds upon this logic and addresses the case of multiple agents of distinct classes (class-A and class-B) in conflict. Guarantees on the safety of the multi-agent system and the almost global convergence of the agents to their destinations are proved. The efficacy of the proposed methodology is demonstrated via simulation results in static and dynamic environments.

Successful Hospital Evacuation After the Kumamoto Earthquakes, Japan, 2016.

Two major earthquakes struck Kumamoto Prefecture in Japan in April 2016. Disaster response was immediately provided, including disaster medical services. Many hospitals were damaged and patients needed immediate evacuation to alternative facilities. The hospital bed capacity of Kumamoto Prefecture was overwhelmed, and transportation of more than 100 patients was needed. Hospital evacuation was carried out smoothly with the coordinated efforts of multiple agencies. The overall operation was deemed a success because patients were transported in a timely manner without any significant adverse events. Upon repair of facilities in Kumamoto Prefecture, patients were returned safely to their previous facilities. The management of inpatients after this natural disaster in Kumamoto Prefecture can serve as a model for hospital evacuation with multi-agency coordination in the future. Future efforts are needed to improve interfacility communications immediately following a natural disaster. (Disaster Med Public Health Preparedness. 2017;11:517-521).

Preference Elicitation in DCOPs for Scheduling Devices in Smart Buildings

Researchers have used Distributed Constraint Optimization Problems (DCOPs) as a powerful approach to model various multi-agent coordination problems, taking into account their preferences and constraints. A core limitation of this model is the assumption that all agents’ preferences are specified a priori. However, in a number of application domains such knowledge become available only after being elicited from users in these domains. In this abstract, we explore the effects of preference elicitation in our motivating application of scheduling smart appliances with the aim of reducing users’ electricity bill cost as well as increasing their comfort.

BD-ADOPT: a hybrid DCOP algorithm with best-first and depth-first search strategies

Distributed Constraint Optimization Problem (DCOP) is a promising framework for modeling a wide variety of multi-agent coordination problems. Best-First search (BFS) and Depth-First search (DFS) are two main search strategies used for search-based complete DCOP algorithms. Unfortunately, BFS often has to deal with a large number of solution reconstructions whereas DFS is unable to promptly prune sub-optimal branch. However, their weaknesses will be remedied if the two search strategies are combined based on agents’ positions in a pseudo-tree. Therefore, a hybrid DCOP algorithm with the combination of BFS and DFS, called BD-ADOPT, is proposed, in which a layering boundary is introduced to divide all agents into BFS-based agents and DFS-based agents. Furthermore, this paper gives a rule to find a suitable layering boundary with a new strategy for the agents near the boundary to realize the seamless joint between BFS and DFS strategies. Detailed experimental results show that BD-ADOPT outperforms some famous search-based complete DCOP algorithms on the benchmark problems.

A hierarchical clustering approach to large-scale near-optimal coalition formation with quality guarantees

Coalition formation is a fundamental approach to multi-agent coordination, and a key challenge in this context is the coalition structure generation problem, where a set of agents must be partitioned into the best set of coalitions. This problem is NP-hard and typical optimal algorithms do not scale to more than 50 agents: efficient approximate solutions are therefore needed for hundreds or thousands of agents. In this paper we propose a novel heuristic, based on ideas and tools used in the data clustering domain. In particular, we present a coalition formation algorithm inspired by the well known class of hierarchical agglomerative clustering techniques (Linkage algorithms). We present different variants of the algorithm, which we call Coalition Linkage (C-Link) and demonstrate how such algorithm can be adapted to graph restricted coalition formation problems (where an interaction graph defined among the agents restricts the set of feasible coalitions). Moreover, we discuss how we can provide an upper bound on the value of the optimal coalition structure, and we show that for specific characteristic functions we can provide such bounds while maintaining polynomial computational costs and memory requirements. We empirically evaluate the different variants of the C-Link algorithm on two synthetic benchmark data-sets, as well as in two real world scenarios, involving a collective energy purchasing and a ride-sharing application. In these settings C-Link achieves promising results providing high quality solutions and solving problem involving thousands of agents in few minutes.

Chasing viruses feverishly

A number of viral diseases have emerged and re-emerged in India and globally, in the last few years. Effective prevention and control of these diseases require, in addition to a functioning disease surveillance system, interventions both before and after disease occurrence, and a combination of personal and population services. However, the current efforts to control emerging viral diseases in India has major therapeutic focus (and attention on diagnostic and curative services) and there is limited attention on preventive and promotive components. It is proposed that for an effective and successful control, a systematic approach is adopted with an appropriate selection of personal and population health services, delivered by government through participation of private sector. This is possible through commitment and leadership of Government and other public health agencies, supplemented by multi agency coordination, sufficient funding and an accountability mechanism.

Supporting Multi-agent Coordination and Computational Collective Intelligence in Enterprise 2.0 Platform

In this paper, we propose a novel approach utilizing a professional Social network (Pro Social Network) and a new coordination protocol (CordiNet). Our motivation behind this article is to convince Small and Medium Enterprises managers that current organizations have chosen to use Enterprise 2.0 tools because these latter have demonstrated remarkable innovation as well as successful collaboration and collective intelligence. The particularity of our work is that is allows employer to share diagnosis and fault repair procedures on the basis of some modeling agents. In fact, each enterprise is represented by a container of agents to ensure a secured and confidential information exchange between intra employers, and a central main container to connect all enterprises’ containers for a social information exchange. Enterprise’s container consists of a Checker Enterprise Agent (ChEA), a Coordinator Enterprise Agent (CoEA) and a Search Enterprise Agent (SeEA). Whereas the central main container comprises its proper agents such as Selection Agent (SA), and a Supervisor Agent (SuA). JADE platform is used to allow agents to communicate and collaborate. The FIPA-ACL performatives have been extended for this purpose. We conduct some experiments to demonstrate the feasibility of our approach.