Modeling the behavior of a hierarchy of command agents with context-based reasoning

Abstract

Context-based reasoning is a paradigm for modeling agent behavior that is based on the idea that humans only use a small portion of their knowledge at any given time. It was specially designed to represent human tactical behavior and has been successfully implemented in systems with single agents or two agents working together. In this paper, we apply this idea in a hierarchical multi-agent system of command agents, where the agents' actions are to command and coordinate subordinates, send reports to their superiors, and communicate with other agents at the same level. We focus on how contexts and actions can be defined for these higher level command agents and how the contexts and actions for the different command agents are related. The proposed methodology is implemented and tested for a hierarchy of command agents that are interpreting and planning an operational order at a battalion level and carrying it out in a computer generated forces environment.

Chain of command in autonomous cooperative agents for battles in real-time strategy games

“…There have been multiple approaches to implement a command structure. Majority of approaches (Løvlid et al 2017;Stanescu et al 2014) implement a hierarchical structure in which agents are ranked and, therefore, could have subordinates that follow their command, whilst, in turn, they would themselves be subordinated to higher-ranked agents which they receive commands from. Such approaches dispense the planning and tactical thinking into multiple levels at unit level where various units would make tactical decisions at different abstraction levels based on their rank within the hierarchy.…”

“…Swarm intelligence does not tend to have leadership or chain of commands in their implementations, yet battle management is an essential part of an RTS which is not fully implemented in swarms. The concepts of rank and chain of command are integral to realistic implementations as they would allow for the planning, coordination, and monitoring of units through leaders (Løvlid et al 2017).…”

A logic-based event controller for means-end reasoning in simulation environments

“…On the simulation platform side, software frameworks such as VR-Forces, Virtual Battle Space (VBS), Calytrix, MASA Sword and others support the declaration and development of highly realistic domain-specific object behavior and interactions (viable technologies for emulating such behavior include agent-based models, 41 multiagent systems, 42 and probabilistic action networks 43 ).…”

“…On the simulation platform side, software frameworks such as VR-Forces, Virtual Battle Space (VBS), Calytrix, MASA Sword and others support the declaration and development of highly realistic domain-specific object behavior and interactions (viable technologies for emulating such behavior include agent-based models, 41 multi-agent systems, 42 and probabilistic action networks 43 ). Although there is support in these frameworks for scripting sequences of events, and even for expressing such scripts in terms of high-level and low-level operational commands, 44 there is no inherent notion of causality and means-end relationships.…”

Krigsspill i operasjonsplanlegging: Hva kan datasimuleringer bidra med?

“…En teknologi for å få det til er modeller basert på intelligente agenter som kan oppfatte og reagere på miljøet rundt seg, som har mål som de skal oppnå og som kommuniserer med andre agenter. K2 og stridsledelse kan dermed modelleres som et hierarki av slike intelligente agenter (Løvlid, Bruvoll, Brathen & Gonzalez, 2018). Metoder fra kunstig intelligens benyttes for å modellere oppførselen til agentene, f.eks.…”