Integrated human decision behavior modeling using extended decision field theory and soar under BDI framework

An integrated Belief-Desire-Intention framework is proposed for the human decision behavior modeling, which involves extended Decision-Field-Theory, Bayesian belief network, and Soar. To demonstrate the proposed framework, crowd evacuation behaviors under the terrorist bomb attack are considered. Three particular scenarios are bomb explosion, bomb threatening, and threatening after explosion. For each scenario, various agent behaviors are characterized based on familiarity with the area, risk-taking behavior, confidence index, and guidance by police. The environment and agents are implemented in AnyLogic® agent-based simulation software, where Netica® and Soar® are called to perform perceptual processing, real-time planning, and decision-execution in a distributed computing environment.