A Multi-Objective Design Optimization Approach for sUAS Fleet Design & Allocation in Meteorological Sampling Operations
Abstract
Many potential small Unmanned Aerial Systems (sUAS) applications require or would benefit from deploying a fleet of aircraft to perform a mission. Fleet operations present advantages to maximize mission success by using additional sUAS; however, an additional advantage exists in the ability to design a bespoke sUAS and capitalize on the innovation in rapid prototyping and collaborative control. If this fleet includes new, yet-to-be-designed sUAS, the problem of both designing the new sUAS and allocating the fleet becomes a challenging optimization problem. Starting from promising work in similar applications, this work will investigate how a decomposition approach will allow a top-level multi-objective problem to optimize fleet-level performance metrics, while lower-level problems will perform the aircraft design and sUAS allocation tasks. The allocation problem uses a fleet operations simulation environment to evaluate fleet level metrics outputted from the aircraft sizing problem. The aircraft sizing problem, driven by a SQP algorithm with legacy sUAS sizing predictions and a Breguet endurance equation-based sizing code, predicts the geometry, weight and performance of future sUAS. The study incorporates objectives to minimizes the lifecycle cost of a five-year operation and maximize data collection opportunities for sUAS profiling missions within the atmospheric boundary layer (ABL). The newly found Pareto sUAS design solutions are then used to construct a response surface to describe the Pareto set and guide decision-makers through the selection process for best sUAS architecture and fleet solution. Traditional ABL profiling operations use weather towers, weather balloons, and manned fixed-wing aircraft to acquire data; however, leveraging the economic, modular, and simultaneous spatial coverage advantages, a sUAS fleet can provide an alternative to using manned aircraft. In this research, the goal is to determine how to formulate a multi-objective design optimization problem that guides decision-makers on what sUAS platforms and the number of sUAS needed to profile the ABL.
Degree
M.S.A.A.
Advisors
Crossley, Purdue University.
Subject Area
Design|Aerospace engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.