Behavior Composition Meets Supervisory Control

Barati, Masoud; St-Denis, Richard

Abstract

With the evolution of software engineering since the advent of structured programming until now, software engineers are faced with tremendous challenges mostly due to the development of large software programs that behave as open systems. Multi-agent systems, which consist of multiple cooperating intelligent agents within an environment, form a particular class of such systems. This sort of software program, which carries out operations by repeatedly interacting with dynamic environments, has become more and more complex with the emergence of ubiquitous communication and computing technologies that constantly grow and evolve. Agent-oriented computing constitutes an appealing solution for coping with this level of complexity because systems can be built by combining agents. The automated composition of software artifacts to generate new ones may rest on recent progress in artificial intelligence and automatic control that has its roots in the tradition of program synthesis. The goal is to provide software engineers with effective methods in which the planning and control of software actions are integral parts of composition operators, together with synthesis procedures that automatically generate an execution strategy that governs the discrete dynamics of the composed artifact in order to satisfy given requirements. This approach ensures a higher degree of safety because it relies on formal methods. This paper shows how the behavior composition problem issued from the artificial intelligence community can be solved within the framework of the supervisory control theory with the aim to benefit from all of its rich facets.

• Presentation Conference Type: Conference Paper (Published)
• Conference Name: 2015 IEEE International Conference on Systems, Man, and Cybernetics (SMC)
• Start Date: Oct 9, 2015
• End Date: Oct 12, 2015
• Online Publication Date: Jan 14, 2016
• Publication Date: 2015
• Deposit Date: May 4, 2021
• Publisher: Institute of Electrical and Electronics Engineers
• Book Title: 2015 IEEE International Conference on Systems, Man, and Cybernetics (SMC)
• DOI: https://doi.org/10.1109/smc.2015.33
• Keywords: Behavior composition problem, artificial intelligence, multi-agent systems, supervisory control theory, formal methods, delegation, control
• Public URL: http://researchrepository.napier.ac.uk/Output/2767185