Generating Various and Consistent Behaviors
      in Simulations
      Benoît Lacroix 1,2, Philippe Mathieu 2 and Andras K...
Context and motivation

      Renault / LIFL UMR CNRS collaboration
      Context: traffic simulation in driving simulat...
Normative description of behaviors

      Normative systems (Noriega, 1997; Esteva et al., 2001; Vazquez-Salceda et al., ...
Generation engine

        Variety
               Randomly select parameters
                from a norm
               ...
Monitoring

 Emergence of new norms
          Feedback to the users
          Improve design and calibration

 Calibra...
Application

 Application
          Driving simulation software SCANeR™ II
          Ergonomics, embedded systems, desi...
Benoit Lacroix
Renault, Technical Center for Simulation   March 26, 2009   PAAMS 2009   7
Experimental results

 Highway database
          11 km, 3000 veh/h
          Normal, aggressive and
           cautiou...
Conclusion

 Easily create various behaviors
 Manage the generation process
          Guaranty the consistency of the b...
Thank you for your attention




                                           Contact: benoit.lacroix@gmail.com




Benoit L...
Upcoming SlideShare
Loading in …5
×

Generating Various and Consistent Behaviors in Simulations

440 views

Published on

Presentation of a model designed to easily generate various and consistent agents' behaviors in simulations. The model is applied to traffic simulation in Renault's driving simulators.

This is the presentation of the paper entitled "Generating Various and Consistent Behaviors in Simulations" at the 2009 International Conference on Practical Applications of Agents and Multi-Agent Systems (PAAMS'09).

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
440
On SlideShare
0
From Embeds
0
Number of Embeds
8
Actions
Shares
0
Downloads
5
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Generating Various and Consistent Behaviors in Simulations

  1. 1. Generating Various and Consistent Behaviors in Simulations Benoît Lacroix 1,2, Philippe Mathieu 2 and Andras Kemeny 1 1 Renault, Technical Center for Simulation 2 LIFL, University of Lille March 26, 2009 PAAMS 2009
  2. 2. Context and motivation  Renault / LIFL UMR CNRS collaboration  Context: traffic simulation in driving simulators  Evaluation of ergonomics, embedded systems, design…  Needs  Various and consistent behaviors for autonomous vehicles (cautious, aggressive…)  Usable by scenario designers  Idea  Driving psychologists classify drivers depending on their behavior (Saad, 1992)  Drivers use set of norms (based on Highway Code, informal rules…)  But they do not strictly follow these norms  Generic approach to address the issue  Behaviors description using norms  Generation engine managing the determinism  Monitoring Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 2
  3. 3. Normative description of behaviors  Normative systems (Noriega, 1997; Esteva et al., 2001; Vazquez-Salceda et al., 2005)  Organizational control in multi-agent based simulations  Improve agents coordination, communication…  In our case  Institution: parameters and associated definition domains  Norms: subsets of these parameters and domains  Behaviors: instantiations of these norms  For instance, in traffic  Parameters: maximal speed, safety time…  Institution: bounds of these parameters (max speed in [0,300] km/h)  Norms: cautious, aggressive drivers (max speed in [140,160] km/h)  Behavior: a cautious, an aggressive (max speed = 156 km/h) Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 3
  4. 4. Generation engine  Variety  Randomly select parameters from a norm  Behavioral variety within a norm  Allow violations: one or more parameters outside the limits  Consistency  Guaranteed when generation within norms limits  Mechanism to reject aberrant behaviors (quantification)  Reaction to violations at runtime Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 4
  5. 5. Monitoring  Emergence of new norms  Feedback to the users  Improve design and calibration  Calibration with real data  Learning norms from real data sets  Unsupervised learning  Kohonen Neural Networks  Description of the data space  Linear component analysis Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 5
  6. 6. Application  Application  Driving simulation software SCANeR™ II  Ergonomics, embedded systems, design, headlights…  Description  Agents’ decision model: perception – decision (finite state automata) – action (vehicle dynamic model)  Institution parameters = existing vehicles parameters of traffic model  Traffic managed by the existing model  Uses  Introduction of driving styles  Generation of the “ambient” traffic Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 6
  7. 7. Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 7
  8. 8. Experimental results  Highway database  11 km, 3000 veh/h  Normal, aggressive and cautious drivers  Speed distributions  More norms increase variety  Increased dynamicity  Lane repartition  Aggressive on left lane  Cautious on right lane Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 8
  9. 9. Conclusion  Easily create various behaviors  Manage the generation process  Guaranty the consistency of the behaviors  Allow violations if wished  Wide application range  Non-intrusive  Perspectives  Norms calibration with real data Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 9
  10. 10. Thank you for your attention Contact: benoit.lacroix@gmail.com Benoit Lacroix Renault, Technical Center for Simulation March 26, 2009 PAAMS 2009 10

×