A Refinement Framework for Autonomous Agents
An autonomous agent is one that is not only directed by its environment, but is also driven by internal motivation to achieve certain goals. The popular Belief-Desire-Intention (BDI) design paradigm allows such agents to adapt to environmental changes by calculating a new execution path to their current goal, or when necessary turning to another goal. In this paper we present an approach to modelling autonomous agents using an extension to Object-Z. This extension supports both data and action refinement, and includes the use of LTL formulas to describe an agent’s desire as a sequence of prioritised goals. It turns out, however, that the introduction of desire-driven behaviour is not monotonic with respect to refinement. We therefore introduce an additional refinement proof obligation to enable the use of simulation rules when checking refinement.
KeywordsAutonomous agents BDI agents Refinement Object-Z Temporal logic
Unable to display preview. Download preview PDF.
- 1.Wooldridge, M.: An Introduction to MultiAgent Systems, 2nd edn. Wiley (2009)Google Scholar
- 2.Rao, A.S., Georgeff, M.P.: BDI agents: From theory to practice. In: 1st International Conference of Multi-agent Systems (ICMAS 1995), pp. 312–319. MIT Press (1995)Google Scholar
- 4.Smith, G.: The Object-Z Specification Language. Kluwer Academic Publishers (2000)Google Scholar
- 5.Emerson, E.: Temporal and modal logic. In: van Leeuwen, J. (ed.) Handbook of Theoretical Computer Science, vol. B, pp. 996–1072. Elsevier (1990)Google Scholar
- 6.Abrial, J.R.: Modelling in Event-B. Cambridge University Press (2010)Google Scholar
- 7.Derrick, J., Boiten, E.: Refinement in Z and Object-Z, Foundations and Advanced Applications. Springer (2001)Google Scholar
- 11.Aştefănoaei, L., de Boer, F.: The refinement of multi-agent systems. In: Dastani, M., Hindriks, K., Meyer, J.J. (eds.) Specification and Verification of Multi-agent Systems, pp. 35–65. Springer (2010)Google Scholar