A computational model for distributed knowledge systems with learning mechanisms

This paper addresses the issues of machine learning in distributed knowledge systems, which will consist of distributed software agents with problem solving, communication and learning functions. To develop such systems, we must analyze the roles of problem-solving and communication capabilities among knowledge systems. To facilitate the analyses, we propose a computational model: LPC. The model consists of a set of agents with (a) a knowledge base for learned concepts, (b) a knowledge base for problem solving, (c) prolog-based inference mechanisms and (d) a set of beliefs on the reliability of the other agents. Each agent can improve its own problem-solving capabilities by deductive learning from the given problems, by memory-based learning from communications between the agents and by reinforcement learning from the reliability of communications between the other agents. An experimental system of the model has been implemented in Prolog language on a Window-based personal computer. Intensive experiments have been carried out to examine the feasibility of the machine learning mechanisms of agents for problem-solving and communication capabilities. The experimental results have shown that the multiagent system improves the performance of the whole system in problem solving, when each agent has a higher learning ability or when an agent with a very high ability for problem solving joins the organization to cooperate with the other agents in problem solving. These results suggest that the proposed model is useful in analyzing the learning mechanisms applicable to distributed knowledge systems.     

Reflections on the Nature of Multi-Agent Coordination and Its Implications for an Agent Architecture

The development of enabling infrastructure for the next generation of multi-agent systems consisting of large numbers of agents and operating in open environments is one of the key challenges for the multi-agent community.Current infrastructure support does not materially assist in the development of sophisticated agent coordination strategies. It is the need for and the development of such a high-level support structure that will be the focus of this paper. A domain-independent (generic) agent architecture is proposed that wraps around an agent's problem-solving component in order to make problem solving responsive to real-time constraints, available network resources, and the need to coordinate—both in the large and small—with problem-solving activities of other agents. This architecture contains five components, local agent scheduling, multi-agent coordination, organizational design, detection and diagnosis, and on-line learning, that are designed to interact so that a range of different situation-specific coordination strategies can be implemented and adapted as the situation evolves. The presentation of this architecture is followed by a more detailed discussion on the interaction among these components and the research questions that need to be answered to understand the appropriateness of this architecture for the next generation of multi-agent systems.     

Wolves, bees, and football: Enhancing coordination in sociotechnological problem solving systems through the study of human and animal groups

This paper describes how sociotechnological systems comprising human and technological agents can be considered problem solving systems. Problem solving systems typically comprise many agents, each characterized by at least partial autonomy. A challenge for problem solving systems is to coordinate system agent operations during problem solving. This paper explores how competence models of human–human and animal–animal coordination might be used to inform the design of problem solving systems so that the potential for agent coordination is enhanced. System design principles are identified based on a review of competent coordination in human groups, such as work and sport teams, and animal groups, such wolf packs and bee colonies. These principles are then discussed in relation to agent coordination in the domains of E-Science, future combat systems, and medicine, which typify real-world environments comprising problem solving systems.     

Negotiation as a metaphor for distributed problem solving

We describe the concept of distributed problem solving and define it as the cooperative solution of problems by a decentralized and loosely coupled collection of problem solvers. This approach to problem solving offers the promise of increased performance and provides a useful medium for exploring and developing new problem-solving techniques.We present a framework called the contract net that specifies communication and control in a distributed problem solver. Task distribution is viewed as an interactive process, a discussion carried on between a node with a task to be executed and a group of nodes that may be able to execute the task. We describe the kinds of information that must be passed between nodes during the discussion in order to obtain effective problem-solving behavior. This discussion is the origin of the negotiation metaphor: Task distribution is viewed as a form of contract negotiation.We emphasize that protocols for distributed problem solving should help determine the content of the information transmitted, rather than simply provide a means of sending bits from one node to another.The use of the contract net framework is demonstrated in the solution of a simulated problem in area surveillance, of the sort encountered in ship or air traffic control. We discuss the mode of operation of a distributed sensing system, a network of nodes extending throughout a relatively large geographic area, whose primary aim is the formation of a dynamic map of traffic in the area.From the results of this preliminary study we abstract features of the framework applicable to problem solving in general, examining in particular transfer of control. Comparisons with planner, conniver, hearsay-ii, and pup6 are used to demonstrate that negotiation—the two-way transfer of information—is a natural extension to the transfer of control mechanisms used in earlier problem-solving systems.     

Framework of knowledge-based systems : Multiple meta-level architecture for representing problems and problem-solving processes

Currently available expert systems have a performance limit because of the lack of capability to describe problems and problem-solving methods. It is closely related with knowledge representation language, but this is not the only concern with this issue. Real world problems and problem-solving methods are not so simple as to be represented always in the same way by the same language. Their representations must be different depending on various factors involved in the problems themselves and the situations these problems are surrounded with. In this paper, the author discusses first the intrinsic nature of problem representation and problem-solving process representation. The requirements for and the conceptual framework of a knowledge-based system that is suited for dealing with various problems then become apparent quite naturally. The author asserts that a multiple meta-level architecture is necessary as well as a knowledge-representation language that can describe complex data structures as the basic framework of knowledge-based systems.     

Coordination Algorithm for Distributed Testing

With the emergence of new models, architectures and middleware such as ODP, TINA and CORBA, for developing open distributed systems, testing technology requires adaptation for use within conformance assessment in such systems. All these frameworks are object-based and aim at creating open distributed environments supporting interworking, interoperability, and portability, in spite of heterogeneity and autonomy of the related systems. In this context, an open distributed system may be viewed as a system providing standardized distributed interfaces for interacting with other systems. Conformance of such a system can be assessed by attaching a related tester at each provided interface. However, many problems of controllability and observability influencing fault detection during the testing process, arise if there is no coordination between the testers. In this paper, we show how to cope with these problems by using test coordination procedures in a distributed testing architecture.     

Distributed Constraint Optimization Problems: Review and perspectives

Intelligent agents is a research area of the Artificial Intelligence intensely studied since the 1980s. Multi-agent systems represent a powerful paradigm of analyzing, projecting, and developing complex systems. One of the main difficulties in modeling a multi-agent system is defining the coordination model, due to the autonomous behavior of the agents. Distributed Constraint Optimization Problems (DCOP) have emerged as one of most important formalisms for coordination and distributed problem solving in multi-agent systems and are capable of modeling a large class of real world problems naturally. This work aims to provide an overview and critical review of DCOP, addressing the most popular methods and techniques, the evolution and comparison of algorithms, and future perspectives on this promising research area.     

Learning to Improve Coordinated Actions in Cooperative Distributed Problem-Solving Environments

Coordination is an essential technique in cooperative, distributed multiagent systems. However, sophisticated coordination strategies are not always cost-effective in all problem-solving situations. This paper presents a learning method to identify what information will improve coordination in specific problem-solving situations. Learning is accomplished by recording and analyzing traces of inferences after problem solving. The analysis identifies situations where inappropriate coordination strategies caused redundant activities, or the lack of timely execution of important activities, thus degrading system performance. To remedy this problem, situation-specific control rules are created which acquire additional nonlocal information about activities in the agent networks and then select another plan or another scheduling strategy. Examples from a real distributed problem-solving application involving diagnosis of a local area network are described.     

Distributed control of multi-agent systems over unknown communication networks using extremum seeking

In this paper, the solution of large-scale real-time optimization problems of multi-agent systems (MAS) is tackled in a distributed and a cooperative manner without the requirement of exact knowledge of network connectivity. Each agent in the communication network measures a local disagreement cost in addition to its local cost. The agents must work collaboratively to ensure that the system's unknown overall cost (i.e., the sum of the local cost of all the agents) is minimized. In order to minimize this cost, the local disagreement cost of all the agents must first be minimized. This minimization requires the solution of a consensus estimation problem and ensures that the agents reach agreement on their decision variables. To address this challenging problem, a distributed proportional-integral extremum seeking control technique is proposed, one that solves both problems simultaneously. Three simulation examples are included, they demonstrate the effectiveness and robustness of the proposed technique.     

多Agent系统问题求解的广义粒子模型方法

提出一种多Agent系统分布式问题求解的新的广义粒子模型,将复杂环境下多Agent系统资源分配和任务规划的优化问题转变为广义粒子模型中的粒子运动学和动力学问题.广义粒子模型可以描述和处理的复杂环境包括多Agent系统中的Agent之间存在的随机、并发、多类型的交互行为.各Agent有不同的个性、自治性、生命周期、拥塞程度和故障几率等.本文讨论了广义粒子模型和多Agent系统分布式问题求解的关系,提出了广义粒子模型的数学物理模型和多Agent系统分布式问题求解算法,并且证明了它们的正确性、收敛性、稳定平衡性等基本性质.通过复杂环境下多Agent系统资源分配和任务规划问题的实验和比较,证实了广义粒子模型方法的有效性及其特点.     

The coordination generalized particle model—An evolutionary approach to multi-sensor fusion

The rising popularity of multi-source, multi-sensor networks supporting real-life applications calls for an efficient and intelligent approach to information fusion. Traditional optimization techniques often fail to meet the demands. The evolutionary approach provides a valuable alternative due to its inherent parallel nature and its ability to deal with difficult problems. We present a new evolutionary approach based on the coordination generalized particle model (C-GPM) which is founded on the laws of physics. C-GPM treats sensors in the network as distributed intelligent agents with various degrees of autonomy. Existing approaches based on intelligent agents cannot completely answer the question of how their agents could coordinate their decisions in a complex environment. The proposed C-GPM approach can model the autonomy of as well as the social coordinations and interactive behaviors among sensors in a decentralized paradigm. Although the other existing evolutionary algorithms have their respective advantages, they may not be able to capture the entire dynamics inherent in the problem, especially those that are high-dimensional, highly nonlinear, and random. The C-GPM approach can overcome such limitations. We develop the C-GPM approach as a physics-based evolutionary approach that can describe such complex behaviors and dynamics of multiple sensors.     

A multi-agent approach for integrated emergency vehicle dispatching and covering problem

The most important decisions that should be made by emergency vehicle managers are related to the allocation and the covering problems. The allocation (or dispatching) problem consists of deciding which vehicle must be assigned to assist an emergency in the best times. The covering problem aims at keeping the region under surveillance well-covered by relocating available vehicles. As components are geographically distributed, decentralized solution approaches may present several advantages. This paper develops a decentralized distributed solution approach based on multi-agent systems (MAS) to manage the emergency vehicles. The proposed system integrates the dispatching of vehicles to calls with zone coverage issues. This integration means that allocation and covering decisions are considered jointly. The idea of MAS has been applied in many others real-world contexts, and has been proven to provide more flexibility, reliability, adaptability and reconfigurability. To our knowledge, there is no existing work that uses MAS for real-time emergency vehicle allocation problem while accounting for the coverage requirements for future demands. We propose a multi-agent architecture that fit the real emergency systems, and that aims at keeping good performance compared to the centralized solution. The objective is to coordinate agents to reach good quality solutions in a distributed way. For this purpose two approaches are examined. The first one is used to show the impact of distributing data and control on the solution quality, since the dispatching decisions are based only on local evaluations of the fitness. The second approach is based on implicit agents' coordination using a more refined and efficient auction mechanism. The performance of each approach is compared to the centralized solution obtained by solving the proposed model with ILOG CPLEX solver. The obtained results show the importance of the coordination method to keep a good quality of service while distributing data and decision making, and prove the performance of the second approach.     

Distributed artificial intelligence for runtime feature-interactionresolution

The feature-interaction problem has many different instances. It is argued that some instances lend themselves to a distributed artificial intelligence (DAI) approach. The use of DAI techniques in current telecommunications systems appears quite natural in light of two trends in the way these systems are designed: the distribution of functionality and the incorporation of intelligence. The author illustrates the relevance of DAI techniques to the feature-interaction problem by discussing existing work (lodes, team-CPS, multistage negotiation, and negotiating agents) that address one or more instances of the problem. He further identifies the kind of cooperation and coordination that the feature-interaction problem requires and the interesting research problems it poses to distributed artificial intelligence     

Individual learning of coordination knowledge

Abstract. Social agents, both human and computational, inhabiting a world containing multiple active agents, need to coordinate their activities. This is because agents share resources, and without proper coordination or ‘rules of the road’, everybody will be interfering with the plans of others. As such, we need coordination schemes that allow agents to effectively achieve local goals without adversely affecting the problem-solving capabilities of other agents. Researchers in the field of Distributed Artificial Intelligence (DAI) have developed a variety of coordination schemes under different assumptions about agent capabilities and relationships. Whereas some of these researchers have been motivated by human cognitive biases, others have approached it as an engineering problem of designing the most effective coordination architecture or protocol. We evaluate individual and concurrent learning by multiple, autonomous agents as a means for acquiring coordination knowledge. We show that a uniform reinforcement learning algorithm suffices as a coordination mechanism in both cooperative and adversarial situations. Using a number of multi-agent learning scenarios with both tight and loose coupling between agents and with immediate as well as delayed feedback, we demonstrate that agents can consistently develop effective policies to coordinate their actions without explicit information sharing. We demonstrate the viabilityof using both the Q-learning algorithm and genetic algorithm based classifier systems with different pay-off schemes, namely the bucket brigade algorithm (BBA) and the profit sharing plan (PSP), for developing agent coordination on two different multi-agent domains. In addition, we show that a semi-random scheme for action selection is preferable to the more traditional fitness proportionate selection scheme used in classifier systems.     

Distributed formation output regulation of switching heterogeneous multi-agent systems

In this article, the distributed formation output regulation problem of linear heterogeneous multi-agent systems with uncertainty under switching topology is considered. It is a generalised framework for multi-agent coordination problems, which contains or concerns a variety of important multi-agent problems in a quite unified way. Its background includes active leader following formation for the agents to maintain desired relative distances and orientations to the leader with a predefined trajectory, and multi-agent formation with environmental inputs. With the help of canonical internal model we design a distributed dynamic output feedback to handle the distributed formation output regulation problem.     

Separating Interaction Concerns from Distributed Feature Components

Interaction problems between heterogeneous appli-cations require consideration of the semantic issue of reliable composition. This problem has become significant and ubiquitous in distributed systems as the Internet rapidly grows as a mainstream service platform and requires increasing automatic coordination and cooperation between services at two ends. A feature must be able to adjust itself to work with other features or services - a highly relevant problem called feature interaction. In line with this, in this paper we propose a complexity controlling method that is suitable for distributed systems in which each feature has two concerns, namely a hard logic and a soft logic. The hard logic implements exactly the specification of a feature, while the soft logic deals with the adaptation aspects of a feature, i.e. resolving interaction problems and making features work together. A two level architecture, particularly designed for aspect oriented programming, is described with a meta level being used to describe interaction resolution, with features being at the base level. Through a case study of email systems, we explain the architecture and highlight the cause of resolution interaction problems and how this particular problem is solved.     

A distributed problem-solving approach to collaborative facility engineering

A framework for collaborative facility engineering is presented. The framework is based on a distributed problem-solving approach to collaborative facility engineering and employs an integration approach called Agent-Based Software Engineering as an implementation vehicle of this approach. The focal entity of this framework is a Multiagent Design Team (MDT) that comprises a collection of software agents (e.g. design software applications with a certain standard communication interface) and a design specialist, which together perform specific design tasks. Multiagent design teams are autonomous and form an organizational structure based on a federation architecture. Every multiagent design team surrenders its autonomy to a system program called facilitator, which coordinates the interaction among software agents in the federation architecture. Facilitators can be viewed as representatives of one or more teams that facilitate the exchange of design information and knowledge in support of the design tasks they perform. In the federation architecture, design specialists collaborate by exchanging design information with others via their software agents, and by identifying and resolving design conflicts by negotiation. In addition to a discussion of the framework's primary components, its realization in an integrated distributed environment for collaborative building engineering is described.     

Collaboration rules for autonomous software agents

Can autonomous software agents that are distributed over a computer network collaborate effectively? Both empirical evidence and theory suggest that they can. Moreover, there seem to be simple rules for designing problem-solving organizations in which collaboration among such agents is automatic and scale-effective (adding agents tends to improve solution-quality; adding computers tends to improve solution-speed). This paper develops some of these rules for off-line problems and argues that they can be extended for the on-line (real-time) control of power systems.