search

godisreal / group-social-force

Multi-Agent Simulation of Collective Behavior (Pedestrian Crowd)
24 stars 6 forks source link

A computational model of unconscious mind in colletive behavior #2

Open godisreal opened 6 years ago

godisreal commented 6 years ago

Should unconsciousness be taken into account? This an interesting issue.
The unconscious mind of an agent can be considered as a self-replusion.
This is in contrast to the replusion from others, and it could be modeled.
Any comments are welcome.

godisreal commented 5 years ago

In current model the effect of unconsciousness is computed by self-repulsion.

Very interestingly, the self-driven force is understood as generated by conscious mind of an individual, and results in one's motivation of behavior. The self-repulsion refers to the unconscious mind, and it may be against the conscious motive that we are aware of. Thus, we assume that the direction of the self-repulsion is contrary to the desired velocity. This model is useful to characterize certain crowd behavior depicted in Le Bon, 1895. That is, when an individual immerses oneself in the crowd, he may lose part of his individual feature such as his original motivation and simply follow the collective pattern of the crowd.

chraibi commented 5 years ago

Instead of introducing a new force, would it make more sense to set the desired speed (v0) to zero?

godisreal commented 5 years ago

Set the de to be zero? So you mean the desired velocity?

chraibi commented 5 years ago

sorry typo. I edited the comment.

godisreal commented 5 years ago

Setting desired velocity to be zero means that agents want to stand still and they do not want to move. In this situation agents may move around by the group social force or other interactive forces, but the motive force will drive agents to stand at where they stay. As a result the motive force becomes a resistence to any movement if desired velocity is set to be zero. Such a resistence could be large if we specify a small relaxation time. This may be useful to model behavior in pre-evacuation stage, but it is not related to unconscious mind.

chraibi commented 5 years ago

Well, then I would suggest clarifying exactly what "unconscious mind" means and how it could be defined.

godisreal commented 5 years ago

This is the explanation of unconsciousness and group dynamics in wikipeida. If you are interested, you may take a look.

https://en.wikipedia.org/wiki/Unconscious_mind https://en.wikipedia.org/wiki/Group_dynamics

It is not easy to model the things exactly in the manner of psychology. It may be complex to model unconsciousness deep inside, but it is an interesting topic, and I find a method to characterize how an agent may lose some of his individualistic feature when he or she becomes a part of crowd. As a result, the unconscious mind is related to the group dynamics in the computational model.

chraibi commented 5 years ago

That's interesting. Would you mind to share more insights about this method you developed?

godisreal commented 5 years ago

In brief the model of unconsciousness is developed based on the model of group dynamics. So the unconscious mind is computed not at the individual level, but in a collective sense, and it is effective when an individual interacts with others in the crowd. Thus, the first thing to build up is group dynamics and then collective unconsciousness. Any comments? @fij

I got some results for group dynamics, but not yet for unconsciousness. It is much more complicated. The group dynamics has been tested in FDS+EVAC. Anyway, it is an interesting study topic and thus I would like to explore it further.

The group dynamics are listed there (Please unzip GroupDynamics.7z and check the result) https://github.com/godisreal/test-crowd-dynamics