Modeling and Simulation of Pilot Behavior Interfaced with Avionics System

To represent the complex behavior of a human pilot operating a fighter aircraft, an agent-based modeling and simulation method is proposed in this paper. First a live virtual constructive simulation platform including an agent-based avionics system model is introduced as a test-bed. Through refining the interface information flow between a pilot and avionics system, employing fuzzy reasoning to mimic the information processing of human, and applying Bayesian network to situation assessment, then using fuzzy Petri-net to make decisions and plan actions, the hybrid agent-based pilot model will sense, think, learn and act as a human pilot do. The ongoing effort indicates its feasibility and validity.