Repository hosted by TU Delft Library

Home · Contact · About · Disclaimer ·
 

Using Agent-based Simulation in a Decision Support System for Military Command & Control

Author: Luik, T.T.
Type:Thesis
Date:2012
Publisher: Vrije Universiteit Amsterdam
Place: Amsterdam
Identifier: 462578
Keywords: Command and Control C2 · Agent based simulation · Computerized Simulation · Computer generated Forces · Staff training · Military operations · Decision support systems · Defence Research · Defence, Safety and Security · Organisation · MSG - Modelling Simulation & Gaming · BSS - Behavioural and Societal Sciences

Abstract

In the military, command & control (C2) systems are used by the commanders to provide orders to their troops and to monitor the execution of these orders. By connecting such systems to a simulator, the simulator's computer generated forces can replace actual human forces, providing cheap, quick and reusable units for purposes such as training, mission preparation or even support during mission execution. To allow this connection, TNO created a multi-agent system to understand the orders from the military commander and subsequently provide the simulator with the correct tasks to execute the given order. Building on this foundation, this Master's project's task was to take it one step further by finding and creating a prototype system (DSS). For a decision support system based on simulation to actually provide support to the commander, its interface needs to be usable in the C2 process and the simulation's results need to be realistic. To solve these problems, requirements from related research and military experts on both these matters were consulted. Based on these examples, our decision support system now provides a table in which multiple orders from the C2 system can be mapped to an analysis of their execution in the simulator. During, prior or after simulation, these analysis criteria can be shown or hidden, allowing the user to compare the orders on only the most interesting criteria. To increase the realism of the simulation's results, military doctrine hasbeen introduced to the multi-agent system. Now the agents can break up an assault in the required phases, move in the correct formations and respond to unexpected enemy behaviour with a blocking position, based on doctrinal documents provided by the Royal Netherlands Army. By doing this, we have shown a glimpse of the future where the computer generated forces realistically mimic soldiers' behaviour like teamwork, coordination, reactive capabilities and the deliberate doctrinal planning, thereby allowing the simulator to realistically foresee the results of the commander's orders. When used in conjuction with an expanded version of our decision support system, commanders will be able to realistically evaluate and compare their courses of action and choose the best one for accomplishing their current mission.