Comparative Analysis of Emergent Behaviors of Three Drone Swarm System Models for Targeting Using Agent-Based Modeling and Simulation

This dissertation introduces a novel computational simulation framework for evaluating the emergent behaviors of three swarm drone models using Agent-Based Modeling and Simulation (ABMS). The three swarm models are a Leader-Follower swarm model based on Bruckstein's antline theory, a Flocking model based on a simplified Reynolds 'Boids’ model, and a Stigmergic model with pheromone-based coordination. The primary objective of the simulation is to evaluate the performance of these models in delivering a user-defined number of drones of each type to a target area of interest in four separate scenarios, resulting in 50,000 separate simulation trials. Each scenario was structured to systematically assess how the swarm model's performance responds to changes in agent-level parameters and external environmental factors. The simulation results reveal statistically significant differences among the models. Numerical analysis and visualization reveal the complex behaviors exhibited by each swarm model as it navigates an environment populated with randomly placed obstacles. Additionally, the degree of adaptive behavior exhibited by each model is quantified using spatial and behavioral entropy calculations. This innovative use of entropy provides a quantitative means of characterizing emergent behavior and stability across swarm types. The Flocking swarm model achieved the highest success rate, displaying robustness across all threat levels, but was sensitive to a higher number of drones required for mission success. The Leader-Follower model was challenged in environments with higher threat densities but demonstrated improved success rates when higher drone counts are required for mission success. The simplified Stigmergic model performed poorly, with pheromone evaporation rate having minimal effect. By integrating statistical analysis and entropy-based metrics, this research provides a reusable ABMS framework for analyzing swarm performance, supporting scenario-based decision-making and system optimization. The findings advance the understanding of swarm behavior, contributing to the growing body of knowledge on swarm intelligence. The findings from this research also highlight practical pathways for designing and deploying drone swarm systems.