Cellular Space and Agentbased Modeling and Simulation

1
Cellular Space and Agent-based Modeling and
Simulation

• Xiaolin Hu
• Computer Science Department
• Georgia State University

2
Wildland Fire Spread and Containment Cellular
Space Modeling and Simulation
3
Forest Cell and Cell Space

• Rothermels semi-empirical model to compute fire
  spread speed and fire line intensity based on
• Cell forest fuel model
• Cell field topography (mainly slope)
• Wind speed and wind direction
• Cell conditions, wet, dry, humidity and ambient
  temperature.

4
Fire Spread Simulation
5
Firefighting Simulation
Direct attack
Parallel attack
Indirect attack
6
Firefighting Simulation
Direct tail attack
Direct head attack
7
Firefighting Simulation
One agent
Two agents
Three agents
8
Fire Spread Simulation with GIS Data(Huntsville
area, Texas)
9
Fire Suppress Simulation with GIS Data
(Huntsville area, Texas)
10
Cellular Space Modeling and Simulation

• Good for simulating complex dynamical systems
  where space plays important roles
• Cell-to-cell connections are predefined and fixed
  different from agent-based MS
• Some applications
• Game of life
• Ecological systems forest fire, geo-science
• Social systems city development, social behavior
• Computational epidemiology disease spread
  simulation

11
Adaptive Behavior of Autonomous Agent
Agent-based Modeling and Simulation
12
Simulating Crayfishs Dominance Hierarchy
Formation
13
Behaviors and Behavioral Context

• The behavioral context refers to the
  psycho-socio-physical state that sets the context
  of an agents behavior.
• It can represent
• A social environment (the crayfish dominance
  hierarchy)
• A task environment (the team formation example)
• A psychological state (the crowd behavior
  simulation)
• Different behavior contexts make an agent display
  different behavior patterns

14
Dynamic Team Formation
15
Social Crowd Behavior Simulation
16
Agent-Based Simulation

• Some of the simulated entities are agents
• Explicitly represents specific behaviors of
  specific individuals contrast with traditional
  macro-level aggregated representations
• Facilitates simulation of group behavior in
  highly dynamic situations. Allows study of
  "emergent behavior"
• Parallel and distributed simulation for e.g., the
  crowd simulation
• Well-suited to populations of heterogeneous
  individuals
• vehicles (and pedestrians) in traffic situations
• actors in financial markets or social network
• consumer behavior
• humans and machines in battle fields
• people in crowds
• animals and/or plants in eco-systems
• artificial creatures in computer games
• enzymes, DNA, and mRNA in a bio cell

17

Spectrum of Agent Properties
goal management capability
intention management capability
domain knowledge
belief management capability
domains
language skills
agent model
decision making abilities
communication capabilities
perception abilities
manipulation skills
mobility skills
navigation skills
18

Discussion

• What is the problem (application) that we want to
  model?
• Do we have the theory behind the model to be
  developed?
• The goal what do we want to gain from the
  model?
• What modeling and simulation environment to use?
• The schedule whats for next step?