Effect of Aircraft Model Fidelity in an Agent-Based Simulation on System of Systems Level Metrics

Abstract

Agent-Based Simulations are a valuable tool in the System of Systems engineer’s belt, as it allows the modelling of the stakeholders in the System of Systems including their respective characteristics and behaviour. The fidelity level with which to model the stakeholder characteristics and behaviour in the agent-based simulation is a key decision which needs to be made early on. However, arriving at such a conclusion can prove difficult as the implications of the chosen fidelity level is not always clear. This paper aims to use the case study of aerial wildfire fighting, to investigate the effect of fidelity level of the aircraft performance modelling on the optimal TLARs of the aircraft obtained from the System of Systems analysis. In order to perform the study, aircraft performance models of varying fidelity levels are integrated into the Agent-Based Simulation with other parameters kept consistent. A simple value function representing the key stakeholders’ interest is to be used as the Measure of Effectiveness of the System of Systems. The optimal vehicle level TLARs are derived from the SOS level Measure of Effectiveness for each aircraft model in the Agent-Based Simulation. A comparative analysis of the TLARs sensitivity is to be presented. The research questions posed by this study are as follows: • What is the effect of aircraft model fidelity on the SoS Measures of Effectiveness, when evaluated in an Agent-Based Simulation? • Is there a diminishing-returns effect on the SoS Measures of Effectiveness as we increase the aircraft model fidelity?