Dealing with Unforeseen Situations in the Context of Self-Adaptive Urban Traffic Control: How to Bridge the Gap
Anthony Stein, Sven Tomforde, Dominik Rauh, Joerg Haehner · 2016
Autonomously adapting signalling strategies to changing traffic demands in urban areas have been frequently used as application scenario for self-organising systems in general as well as for Autonomic or Organic Computing systems in particular. The Organic Traffic Control (OTC) system is one of the most prominent representatives in this domain. OTC implements a multi-layered Observer/Controller (O/C) architecture and utilises a strongly modified eXtended Classifier System (XCSO/C) for the task of self-adaptation. In this paper, we extend the algorithmic structure of XCS-O/C by a novel interpolation-based approach that incorporates existing knowledge beyond the traditional means. We demonstrate the benefit of the developed approach in terms of reduced delay times for near-to-reality simulations of realistic traffic conditions from Hamburg, Germany.