Mission-Oriented System Architecture and Optimal Aerial Terrain Guarding for a Collaborative Team of UAVs

Abstract

Thanks to rapid technological development, multi-Unmanned Aerial Vehicle (UAV) systems have been actively adopted in the military to accomplish various missions, including intelligence, surveillance, and reconnaissance (ISR), target acquisition, and border patrol. Although multi-UAV systems have the potential to increase the efficiency of military operations, the majority of current military system architectures that utilize multiple UAVs are not well suited for tactical-level operations because most of the current military systems have been designed in an ad-hoc manner to achieve specific purposes. In addition, the developments have only focused on behaviors of individual UAVs without mission-level perspectives as a primary engineering factor. Consequently, existing ad-hoc systems are limited in their reusability across missions unless the human operators have specialized engineering skills.Given this reality, this dissertation aims to formalize the systems engineering process of military multi-UAV systems by incorporating mission-oriented features that enhance architecture reusability and composability. The proposed systems engineering approach, called mission-oriented iterative systems engineering process, is a top-down, mission-centric methodology that consists of four main steps: 1) specification of mission requirements, 2) creation of normative models, 3) simulation of models, and 4) real implementation. Based on model-based systems engineering (MBSE) and mission engineering (ME) foundations, the outputs of each step are formalized in SysML language to facilitate understanding of the system. Military doctrines, such as Mission Command and Operations Process, are regarded as primary factors for engineering the system. Throughout this dissertation, the search and attack mission, a military offensive operation used to establish or regain contact with the enemy, is utilized as the target mission to demonstrate the details of the engineering process proposed. In the search and attack operation using UAVs, it is critical to determine how to maximize the visible regions on the terrain surface by deploying multiple UAVs. Therefore, terrain visibility from the aerial observation points and related constraints are deeply examined in this research. Based on the examination, a new optimization problem, the Aerial Terrain Guarding Problem with Visibility Constraints (ATGP-VC), and its solution approach are presented to determine an optimal set of aerial points to maximize the collective visibility on terrain surface. A task assignment problem to assign aerial points to UAVs is also discussed. The developed system with the proposed engineering process is tested and validated using a hardware-in-the-loop (HITL) simulation testbed, which involves various hardware components (i.e., a real UAV equipped with onboard computational units and sensors) and software components (i.e., flight dynamics simulator and hardware interfaces). Detailed architectures of each component are designed and presented.