Organizational Unit:

Aerospace Systems Design Laboratory (ASDL)

Permanent Link

https://hdl.handle.net/1853/70744

Parent Organization

Organizational Unit

Daniel Guggenheim School of Aerospace Engineering

ArchiveSpace Name Record

https://finding-aids.library.gatech.edu/agents/corporate_entities/1135

Full item page

Publication Search Results

Now showing 1 - 10 of 11

A Multi-UAS Trajectory optimization Methodology for Complex Enclosed Environments

(Georgia Institute of Technology, 2019-06) Barlow, Sarah ; Choi, Youngjun ; Briceno, Simon ; Mavris, Dimitri N.

This paper explores a multi-UAV trajectory optimization methodology for confined environments. One potential application of this technology is performing warehouse inventory audits; this application is used to evaluate the methodology's impact on minimizing total mission times. This paper investigates existing algorithms and improves upon them to better address the constraints of warehouse-like environments. An existing inventory scanning algorithm generates sub-optimal, collision free paths for multi-UAV operations, which has two sequential processes: solving a vehicle routing problem, and determining optimal deployment time without any collision. To improve the sub-optimal results, this paper introduces three possible improvements on the multi-UAV inventory tracking scenario. First, a new algorithm logic which seeks to minimize the total mission time once collision avoidance has been ensured rather than having separate processes. Next, an objective function that seeks to minimize the maximum UAV mission time rather than minimizing the total of all UAV mission times. Last, an operational setup consisting of multiple deployment locations instead of only one. These algorithms are evaluated individually and in combination with one another to assess their impact on the overall mission time using a representative inventory environment. The best combination will be further analyzed through a design of experiments by varying several inputs and examining the resulting fleet size, computation time, and overall mission time.
Multi-UAS path-planning for a large-scale disjoint disaster management

(Georgia Institute of Technology, 2019-06) Choi, Younghoon ; Choi, Youngjun ; Briceno, Simon ; Mavris, Dimitri N.

A UAS-based disaster management method has been adopted to monitor the disaster impact and protect human lives since it can be rapidly deployed, execute an aerial imaging mission, and provide a cost-efficient operation. In the case of a wildfire disaster, a disaster management is highly complex because of large-scale wildfires that can occur simultaneously and disjointly in a large area. In order to effectively manage these large-scale wildfires, it requires multiple UAS with multiple ground stations. However, conventional UAS-based management methods relies on a single ground station that can have a limitation to handle the large-scale wildfire problem. This paper presents a new path-planning framework for UAS operations including a fleet of UAVs and multiple ground stations. The framework consists of two parts: creating coverage paths for each wildfire and optimizing routes for each UAV. To test the developed framework, this paper uses representative wildfire scenarios in the State of California.
Rapid and Automated Urban Modeling Techniques for UAS Applications

(Georgia Institute of Technology, 2019-06) Choi, Youngjun ; Pate, David ; Briceno, Simon ; Mavris, Dimitri N.

Urban models for testing UAV path-planning algorithms commonly apply simple representations using cuboid or cylinderical shapes which may not capture the characteristics of a urban environment. To address this limitation of existing urban models, this paper presents two urban modeling techniques for an unmanned aircraft flight simulation in an urban environment. The first proposed urban modeling technique is an airborne LiDAR source-based approach that incorporates machine learning algorithms to identify the number of buildings and characterize them from the LiDAR information. The second proposed urban modeling technique is an artificial urban modeling technique without any airborne LiDAR resources that applies an adaptive spacing method, an iterative algorithm to define an artificial urban environment. Unlike the LiDAR source-based approach that creates an approximated urban model, the adaptive spacing-based urban modeling algorithm generates an artificial urban environment that is visually different from a reference city, but has similar the characteristics to it. To demonstrate the two proposed urban modeling techniques, numerical simulations are conducted using open-source datasets to construct several realistic urban models.
Multi-UAV Trajectory Optimization Utilizing a NURBS-Based Terrain Model for an Aerial Imaging Mission

(Georgia Institute of Technology, 2019-05) Choi, Youngjun ; Chen, Mengzhen ; Choi, Younghoon ; Briceno, Simon ; Mavris, Dimitri N.

Trajectory optimization precisely scanning an irregular terrain is a challenging problem since the trajectory optimizer needs to handle complex geometry topology, vehicle performance, and a sensor specification. To address these problems, this paper introduces a novel framework of a multi-UAV trajectory optimization method for an aerial imaging mission in an irregular terrain environment. The proposed framework consists of terrain modeling and multi-UAV trajectory optimization. The terrain modeling process employs a Non-Uniform Rational B-Spline (NURBS) surface fitting method based on point cloud information resulting from an airborne LiDAR sensor or other sensor systems. The NURBS-based surface model represents a computationally efficient terrain topology. In the trajectory optimization method, the framework introduces a multi-UAV vehicle routing problem enabling UAV to scan an entire area of interest, and obtains feasible trajectories based on given vehicle performance characteristics, and sensor specifications, and the approximated terrain model. The proposed multi-UAV trajectory optimization algorithm is tested by representative numerical simulations in a realistic aerial imaging environment, namely, San Diego and Death Valley, California.
Energy-Constrained Multi-UAV Coverage Path Planning for an Aerial Imagery Mission Using Column Generation

(Georgia Institute of Technology, 2019-03) Choi, Younghoon ; Choi, Youngjun ; Briceno, Simon ; Mavris, Dimitri N.

This paper presents a new Coverage Path Planning (CPP) method for an aerial imaging mission with multiple Unmanned Aerial Vehicles (UAVs). In order to solve a CPP problem with multicopters, a typical mission profile can be defined with five mission segments: takeoff, cruise, hovering, turning, and landing. The traditional arc-based optimization approaches for the CPP problem cannot accurately estimate actual energy consumption to complete a given mission because they cannot account for turning phases in their model, which may cause non-feasible routes. To solve the limitation of the traditional approaches, this paper introduces a new route-based optimization model with column generation that can trace the amount of energy required for all different mission phases. This paper executes numerical simulations to demonstrate the effectiveness of the proposed method for both a single UAV and multiple UAV scenarios for CPP problems.
Framework for Multi-Asset Comparison and Rapid Down-selection for Earth Observation Missions

(Georgia Institute of Technology, 2019) Gilleron, Jerome ; Muehlberg, Marc ; Payan, Alexia P. ; Choi, Youngjun ; Briceno, Simon ; Mavris, Dimitri N.

Observing the Earth, whether it be from space or from the air, has become easier in recent years with the advent of new space-borne and airborne technologies. First, satellites constantly provide data about almost any point on the globe with varying resolutions and in various spectral bands. Second,Unmanned Aerial Vehicles (UAV) are becoming more readily accessible to the public and may be rapidly deployed to take high resolution images of ground features or areas of interest. Third, manned aircraft may be used to image large areas of land at a higher resolution than satellites and have been used regularly in disaster monitoring and surveillance missions. However, when multiple heterogeneous assets compete to perform a given aerial imaging mission, deciding which asset is better suited and/or less costly to operate in a timely manner is challenging. Every acquisition mode is different, resolution values are computed differently and there currently does not exist a common framework to compare UAV, manned aircraft and satellites. To address this challenge, this paper describes a methodology to rapidly compare various types of aerial assets (such as UAVs and manned aircraft) and space assets (such as satellites) to decide which one would be better able to perform an Earth observation mission depending on a set of requirements. To demonstrate the proposed methodology, this paper executes numerical simulations with three different representative scenarii in California.
An Extended Savings Algorithm for UAS-based Delivery Systems

(Georgia Institute of Technology, 2019) Choi, Younghoon ; Choi, Youngjun ; Briceno, Simon I. ; Mavris, Dimitri N.

This paper presents an extended savings algorithm for a package delivery system using unmanned aircraft systems (UAS). The savings algorithm as a heuristic method solves a vehicle routing problem (VRP) that is commonly formulated by an operational plan for each vehicle. In general, package delivery systems need to establish an operational plan based on demand and preferred time to be visited for each customer. In UAS-based delivery systems, however, capacity and traveling time constraints must be additionally considered to create their operational schedules because of limited payload capacity and short endurance of unmanned aerial vehicles (UAVs). Because of these limitations, UAVs should be reused during operation hours to reduce acquisition costs. Thus, a recharging strategy should be included in the operational planning process. However, conventional savings algorithms cannot capture those properties at once because they have mainly focused on delivery systems with conventional vehicles such as trucks and passenger/cargo aircraft that have different vehicle features and operational characteristics, such as the endurance/speed of a vehicle and recharging strategy. To overcome the limitations of the conventional approaches, this paper proposes the extended savings algorithm, which can concurrently reflect the characteristics of both delivery systems and UAVs. To demonstrate the proposed extended savings algorithm this paper preforms numerical simulations with two representative scenarios in Annapolis, MD and Macon, GA.
Multi-UAV Trajectory Optimization and Deep Learning-Based Imagery Analysis for a UAS-Based Inventory Tracking Solution

(Georgia Institute of Technology, 2019) Choi, Youngjun ; Martel, Maxime ; Briceno, Simon ; Mavris, Dimitri N.

This paper presents a multi-UAV trajectory optimization and an imagery analysis technique based on Convolutional Neural Networks (CNN) for an inventory tracking solution using a UAS platform in a large warehouse or manufacturing environment. The current inventory tracking method is a manual and time-consuming process to scan all the inventory items. Its accuracy is not consistent depending on the complexity of the scanning environment. To improve the scanning efficiency with respect to time and accuracy, this paper discusses a UAS-based inventory solution. In particular, this paper addresses two primary topics: multi-UAV trajectory optimization to scan inventory items and a multi-layer CNN architecture to identify a tag attached on the inventory item. To demonstrate the proposed multi-UAV trajectory optimization framework, numerical simulations are conducted in a representative inventory space. The proposed CNN-based imagery analysis framework is demonstrated on a flight experiment.
Coverage path planning for a UAS imagery mission using column generation with a turn penalty

(Georgia Institute of Technology, 2018-06) Choi, Younghoon ; Choi, Youngjun ; Briceno, Simon ; Mavris, Dimitri N.

This paper introduces a novel Coverage Path Planning (CCP) algorithm for a Unmanned Aerial Systems (UAS) imagery mission. The proposed CPP algorithm is a vehicle-routing-based approach using a column generation method. In general, one of the main issues of the traditional arc-based vehicle routing approaches is imposing a turn penalty in a cost function because a turning motion of vehicle requires the more amount of energy than a cruise motion. However, the conventional vehicle-routing-based approaches for the CPP cannot capture a turning motion of the vehicle. This limitation of the arc-based mathematical model comes from the property of turning motions, which should be evaluated from two arcs because a turn motion occurs at a junction of the arcs. In this paper, to mitigate the limitation, a route-based model using column generation approach with a turn penalty is proposed. To demonstrate the proposed CPP approach, numerical simulations are conducted with a conventional CPP algorithm.
Three-dimensional UAS trajectory optimization for remote sensing in an irregular terrain environment

(Georgia Institute of Technology, 2018-06) Choi, Youngjun ; Choi, Younghoon ; Briceno, Simon ; Mavris, Dimitri N. ;

This paper presents a novel algorithm for three-dimensional UAS trajectory optimization for a remote sensing mission in an irregular terrain environment. The algorithm consists of three steps: terrain modeling, the selection of scanning waypoints, and trajectory optimization. The terrain modeling process obtains a functional model using a Gaussian process from terrain information with a point cloud. The next step defines scanning waypoints based on the terrain model information, sensor specifications, and the required image resolution. For the selection of the waypoints, this paper introduces two different approaches depending on the direction of the viewing angle: a normal offset method and a vertical offset method. In the trajectory optimization, the proposed algorithm solves a distance-constraint vehicle routing problem to identify the optimum scanning route based on the waypoints and UAS constraints. Numerical simulations are conducted with two different UAS trajectory scanning methods in a realistic scenario, Point Loma in San Diego.