Årsrapport 2012, Syntetiska aktörer

Authors:

  • Linus Luotsinen
  • Fredrik Johansson
  • Mikael Lundin
  • Petter Ögren
  • Emil Salling

Publish date: 2012-12-17

Report number: FOI-R--3557--SE

Pages: 64

Written in: Swedish

Keywords:

  • Intelligent agents
  • multi-agent systems
  • multi-agent simulation
  • agent-based modeling
  • population simulation
  • computer generated
  • forces
  • behavior imitation
  • behavior optimization
  • behavior recognition

Abstract

This report describes the work that was conducted within the project Synthetic actors (FoT nr 922:01.03) in 2012. The purpose of the project is:  to provide the Swedish ground combat school with tools targeting the modeling and simulation of populations to improve the training of military personnel in military operations in urban terrain,  to acquire knowledge, develop tools, methods and algorithms targeting the modeling of synthetic actors for use in today's as well as the future simulator systems of the Swedish Armed Forces. The project has since 2009 developed a prototype, POPSIM, which is a tool developed to simulate population behaviors in urban areas on an individual level of abstraction. The development of POPSIM has been the main activity of 2012. In addition to the population behavior model, POPSIMalso include models to represent environment and infrastructure (buildings, road-network and mobile communication), irregular forces and military forces. Some of the highlights of this year's work has been the addition of relationships within the population and the development of opinion dynamics to model, for instance, the population's attitude towards the military forces. The project has, in addition to the POPSIM activity, also conducted several smaller studies focusing on behavior modeling. Specifically, these studies explored the possibilities to imitate, optimize and recognize behaviors. The purpose of these studies has primarily been to support the development of POPSIM but are also considered relevant in other similar simulation tools. Finally, the project has in 2012 co-financed and participated in the European Urban Simulation for Asymmetric Scenarios (EUSAS) project.