Entropy, Field Theory and Pedestrian Dynamics: Prediction and Forensics

摘要

The aim of this paper is to develop a path or trajectory designing tool for pedestrian dynamics, as a function of time, based on Lewin's concept of psychological field and applying Shannon's concept of entropy as a measure of uncertainty. The geometrical objects we use are Bézier cubics. From the point of view of cybernetics, developing a model of human behaviour should include a discussion of the developing process itself. We present a short discussion based on ideas of von Forster. The use of time variables suggests the analysis in two directions: prediction and forensics. A case study will illustrate the method.