Characterizing and predicting agents via multi-agent evolution

• US 20070162405A1
• Filed: 10/12/2006
• Published: 07/12/2007
• Est. Priority Date: 10/12/2005
• Status: Active Grant

• Current Assignee: Techteam Government Solutions Incorporated (Jacobs Solutions, Inc.)
• Original Assignee: Techteam Government Solutions Incorporated (Jacobs Solutions, Inc.)
• Inventors: Brueckner, Sven, Parunak, H., Matthews, Robert, Brophy, Steven, Sauter, John, Bisson, Robert
• Granted Patent: US 7,921,066 B2
• Time in Patent Office: Days
• US Class Current: 706/12
• CPC Class Codes: G06N 3/006 based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]

Abstract

A method of predicting the behavior of software agents in a simulated environment involves modeling a plurality of software agents representing entities to be analyzed, which may be human beings. Using a set of parameters that governs the behavior of the agents, the internal state of at least one of the agents is estimated by its behavior in the simulation, including its movement within the environment. This facilitates a prediction of the likely future behavior of the agent based solely upon its internal state; that is, without recourse to any intentional agent communications. In the preferred embodiment the simulated environment is based upon a digital pheromone infrastructure. The simulation integrates knowledge of threat regions, a cognitive analysis of the agent'"'"'s beliefs, desires, and intentions, a model of the agent'"'"'s emotional disposition and state, and the dynamics of interactions with the environment. By evolving agents in this rich environment, we can fit their internal state to their observed behavior. In realistic wargame scenarios, the system successfully detects deliberately played emotions and makes reasonable predictions about the entities'"'"' future behavior.

Claims

1. A method of predicting the behavior of an agent in an environment, comprising the steps of:

modeling a plurality of software agents representing entities to be analyzed;

providing a set of parameters governing the behavior of the agents;

executing a computer simulation of an environment including the agents;

estimating the internal state of at least one of the agents based upon its behavior in the simulation, including its movement within the environment; and

predicting the likely future behavior of the agent based upon the estimate of its internal state.

2. The method of claim 1, wherein the agent'"'"'s internal state is estimated by examining the evolution of the agent'"'"'s observed behavior.

3. The method of claim 1, wherein the agent'"'"'s internal state is estimated in conjunction with an agent-based model of the environment.

4. The method of claim 1, wherein the prediction of the agent'"'"'s future behavior takes into account the dynamics of the agent'"'"'s interaction with the environment.

5. The method of claim 1, wherein the entities to be analyzed include human beings.

6. The method of claim 1, wherein the simulated environment is based upon a digital pheromone infrastructure.

7. The method of claim 6, wherein the digital pheromones are scalar variables that agents can sense and which they deposit at their current location in the environment.

8. The method of claim 7, wherein the agents respond to the local concentrations of the digital pheromones tropistically through climbing or descending local gradients.

9. The method of claim 6, wherein the pheromone infrastructure runs on the nodes of a graph-structured environment.

10. The method of claim 6, wherein the graph-structured environment is a rectangular lattice.

11. The method of claim 6, wherein each agent is capable of aggregating pheromone deposits from individual agents, thereby fusing information across multiple agents over time.

12. The method of claim 6, wherein each agent is capable of evaporating pheromones over time to remove inconsistencies that result from changes in the simulation.

13. The method of claim 6, wherein each agent is capable of diffusing pheromones to nearby places, thereby disseminating information for access by nearby agents.

14. The method of claim 6, wherein the movements of the agents change their deposit patterns.

15. The method of claim 6, wherein the simulation integrates knowledge of threat regions, a cognitive analysis of the agent'"'"'s beliefs, desires, and intentions, a model of the agent'"'"'s emotional disposition and state, and the dynamics of interactions with the environment.

16. The method of claim 1, wherein the simulation involves urban warfare.

17. The method of claim 1, wherein the simulation involves a computer game.

18. The method of claim 1, wherein the simulation involves a business strategy

19. The method of claim 1, wherein the simulation involves a sensor fusion.