Multi-agent System System Research Articles

Real-Time Energy Management for DC Microgrids Using Artificial Intelligence

Microgrids are defined as an interconnection of several renewable energy sources in order to provide the load power demand at any time. Due to the intermittence of renewable energy sources, storage systems are necessary, and they are generally used as a backup system. Indeed, to manage the power flows along the entire microgrid, an energy management strategy (EMS) is necessary. This paper describes a microgrid energy management system, which is composed of solar panels and wind turbines as renewable sources, Li-ion batteries, electrical grids as backup sources, and AC/DC loads. The proposed EMS is based on the maximum extraction of energy from the renewable sources, by making them operate under Maximum Power Point Tracking (MPPT) mode; both of those MPPT algorithms are implemented with a multi-agent system (MAS). In addition, management of the stored energy is performed through the optimal control of battery charging and discharging using artificial neural network controllers (ANNCs). The main objective of this system is to maintain the power balance in the microgrid and to provide a configurable and a flexible control for the different scenarios of all kinds of variations. All the system’s components were modeled in MATLAB/Simulink, the MAS system was developed using Java Agent Development Framework (JADE), and Multi-Agent Control using Simulink with Jade extension (MACSIMJX) was used to insure the communication between Simulink and JADE.

EARCT- ENVIRONMENT FOR AUTOMATED RANK BASED CONTINUOUS AGENT TESTING

Testing MAS (Multi-agent System) is a challenging task because these systems are distributed, complex and autonomous in nature. Agents exist in an open environment having their own locus of control and they require context awareness. So due to these agent’s characteristics, testing MAS system using existing testing techniques becomes a very tedious job. Agents also pose problems regarding message communication and semantic interoperability, as well as synchronization with other agents existing in the environment. All these features are known to be hard not only to design and to code, but also to test. In this paper we will propose a unique environment EARCT to test MAS keeping in mind the essential software engineering paradigms such as effort consumed, errors revealed etc.

Suitability assessment framework of agent-based software architectures

ContextA common distributed intelligent system architecture is Multi Agent Systems (MASs). Creating systems with this architecture has been recently supported by Agent Oriented Software Engineering (AOSE) methodologies. But two questions remain: how do we determine the suitability of a MAS implementation for a particular problem? And can this be determined without AOSE expertise? ObjectiveGiven the relatively small number of software engineers that are AOSE experts, many problems that could be better solved with a MAS system are solved using more commonly known but not necessarily as suitable development approaches (e.g. object-oriented). The paper aims to empower software engineers, who are not necessarily AOSE experts, in deciding whether or not they should advocate the use of an MAS technology for a given project. MethodThe paper will construct a systematic framework to identify key criteria in a problem requirement definition to assess the suitability of a MAS solution. The criteria are first identified using an iterative process. The features are initially identified from MAS implementations, and then validated against related work. This is followed by a statistical analysis of 25 problems that characterise agent-oriented solutions previously developed to group features into key criteria. ResultsKey criteria were sufficiently prominent using factor analysis to construct a framework which provides a process that identifies within the requirements the criteria discovered. This framework is then evaluated for assessing suitability of a MAS architecture, by non-AOSE experts, on two real world problems: an electricity market simulation and a financial accounting system. ConclusionSubstituting a software engineer’s personal inclination to (or not to) use a MAS, our framework provides an objective mechanism. It can supplant current practices where the decision to use a MAS architecture for a given problem remains an informal process. It was successfully illustrated on two real world problems to assess the suitability of a MAS implementation. This paper will potentially facilitate the take up of MAS technology.

A Petri Net Based Multi-Agent System Behavioral Testing

In Multi-Agent System (MAS), developers concentrate on creating design models and evolving them, from higher level models to lower level models, in several steps. Considerable part of MAS implementations is automatically produced from the design models. If a design model contains faults, they are passed to the generated implementations. Practical model validation techniques are required to discover and delete faults in abstract design models. In this paper, we introduce a formal approach for MAS design testing. It specifies a testing process that complements Multi-agent Systems Engineering (MaSE) methodology and strengthens the mutual relationship between UML and MAS. Besides, it defines a structured and comprehensive testing process for engineering software agents at the design level by providing a systematic way of converting the MAS design models to UML design diagram. Then a Petri Net (PN) diagram is generated from the UML models to simulate the behavior of the MAS system. Finally, because Petri Nets (PNs) are formal models, their analysis techniques can be applied to automatic MAS behavioral testing.

Cooperation Design System Based on Mobile-C Agent Platform

Over the past decades, Distributed Systems (DS) have been adopted for industrial applications to improve the system efficiency because distributed architecture has advantages in resource utilization, fault toleration .etc. Multi-Agent System (MAS) arises from combination of the theories of artificial intelligence and distributed systems. One character of MAS is their self-organization, so how to implement an effective mechanism for self-organization of agents is important to a MAS system, this paper describes the design and implementation of a Mobile-C based agent management system, in which Mobile-C was adopted as the implementation platform, and this paper also described an agent-based cooperative design application using this system to manage all the agents involved.