The authors use model-checking techniques to automatically verify multiagent systems written in a logic-based agent-oriented programming language. The approach is to translate the language into the input notations of existing model checkers. Furthermore, property-based slicing can reduce a multiagent system's state space, thus improving the efficiency of model checking. The analysis of a typical scenario of an autonomous Mars rover illustrates this approach.
Index Terms:
Intelligent agents, multiagent systems, model checking
Citation:
Rafael H. Bordini, Michael Fisher, Willem Visser, Michael Wooldridge, "Model Checking Rational Agents," IEEE Intelligent Systems, vol. 19, no. 5, pp. 46-52, Sep./Oct. 2004, doi:10.1109/MIS.2004.47