Une procédure d'architecture coopérative pour les systèmes de systèmes basée sur une approche par système multi-agent auto-adaptatif

A Cooperative Architecting Procedure for Systems of Systems Based on Self-Adaptive Multi-Agent Systems

• Auteur / Autrice : Teddy Bouziat
• Direction : Pierre Glize
• Type : Thèse de doctorat
• Discipline(s) : Informatique
• Date : Soutenance le 20/11/2017
• Etablissement(s) : Toulouse 1
• Ecole(s) doctorale(s) : École doctorale Droit et Science Politique (Toulouse)
• Partenaire(s) de recherche : Equipe de recherche : Institut de Recherche en Informatique de Toulouse (1995-....)

Mots clés

Mots clés contrôlés

Systèmes informatiques -- Mesures de sûreté

Mots clés libres

Système de Systèmes

SdS

SApHESIA

SoS

Résumé

Depuis la seconde guerre mondiale, l’ingénierie des systèmes a permis le développement de méthodologies pour contrôler le développement de systèmes et de projets de plus en plus complexes. En 1990, la chute de l’URSS a provoqué un changement de doctrine militaire aux Etats-Unis en passant d’une confrontation bipolaire à une mondialisation des conflits comportant une grande variété de menaces. Sa nouvelle doctrine était de faire collaborer ses systèmes de défense existants pour produire un système de défense de haut niveau, décentralisé, adaptable et composé de systèmes indépendants. C’est l’apparition du concept de Système de Systèmes (SdS).Cette thèse de doctorat propose un nouveau modèle de SdS appelé SApHESIA (SoS ArchitectingHEuriStIc based on Agent), ainsi qu’une nouvelle méthodologie d’architecture. Cette nouvelle méthodologie est basée sur une coopération complète entre tous les composants du SdS, lui permettant d’évoluer de lui-même afin de faire face à des événements inattendus de son environnement tels que des menaces. Enfin, ce travail est testé à travers 4exemples issus de différents domaines (militaire, logistique et exploratoire).

Since the World War II, researchers have tended to develop methodologies and tools tobuild and control the development of more and more complex systems and projects. Thisinter-disciplinary research area has been called Systems Engineering (SE) and continues tobe developed nowadays. In 1990, the fall of USSR led the US Department of Defense (DoD)to re-think its defense doctrine and to switch from a one opponent confrontation to a globalizationof conflicts with a huge variety of scenarios. Its idea was to re-use and join itsdefense systems by producing a huge, decentralized and adaptive defense system that iscomposed of existing and independents (complex) systems. This is the apparition of theSystem of Systems (SoS) concept. After 2000’s, this concept spreads in civil domains suchas crisis management or logistic systems. More precisely, a SoS is a complex system characterizedby the particular nature of its components: these latter, which are systems, tend tobe managerially and operationally independent as well as geographically distributed. Thisspecific characterization led to re-think research areas of classic SE such as definition, taxonomy,modeling, architecting and so on. SoS architecting focuses on the way independentcomponents of a SoS can be dynamically structured and can change autonomously theirinteractions in an efficient manner to fulfill the goal of the SoS and to cope with the highdynamics of the environment. This PhD thesis mainly focuses on two SoS research areas: 1)SoS modeling and 2) SoS architecting. To achieve the first point, we propose a new modelcalled SApHESIA (SoS Architecting HEuriStIc based on Agent). We have used set theoryand ABM (Agent-Based Model) paradigm to define this model that takes into account thecharacteristics of SoS. Secondly, we propose a new SoS architecting procedure based on theAdaptive Multi-Agent System (AMAS) approach that advocates full cooperation betweenall the components of the SoS through the concept of criticality. This criticality is a metricthat represents the distance between the current state of a component and its goals. In thisprocedure, the SoS architecture evolves over time to self-adapt to the dynamics of the environmentin which it is plunged, while taking into account the respective local goals of itscomponents. Finally we instantiate this model and this procedure through 4 examples fromdifferent domains (military, logistics and exploratory missions) and validate the feasibility,the efficiency, the effectiveness and the robustness of the SoS architecting procedure we havedeveloped and proposed.