Date of Award

9-2024

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Department of Electrical and Computer Engineering

First Advisor

Scott L. Nykl, PhD

Abstract

This dissertation presents a novel approach to autonomous docking using machine learning for visual perception, particularly during probe and drogue aerial refueling. Autonomous vehicles have become pervasive in both civilian and defense sectors, and their ability to interact with their surroundings and each other autonomously is critical for future operations. Traditional methods relying on signals or inertial sensors face significant limitations such as interference, jamming, and drift. This research focuses on developing a computer vision-based solution to overcome these limitations. A novel pipeline, termed relative vectoring, is introduced, which utilizes dual object detection and machine learning to estimate relative positions between the receiver and the drogue during aerial refueling using imagery alone. The proposed solution leverages object detection models to detect and match 2D image points to 3D object points, enabling accurate pose estimation and vector computation without relying on extrinsic camera calibrations. The pipeline was validated through extensive simulation using the AftrBurner graphics engine, which provided realistic imagery and dynamic scenarios. Simulation results demonstrated the pipeline’s accuracy, reliability, real-time performance, and resilience to occlusions. Additionally, relative vectoring was tested in real-world scenarios using transfer learning techniques. The findings showed that scene augmentation significantly enhances model generalization and accuracy, bridging the sim-to-real gap. Real-world results confirmed that this method is reliable, accurate (within 3 cm at contact), and fast (56 fps) even under varying environmental conditions. Overall, results indicate that the proposed solution can effectively perform autonomous docking in complex and dynamic environments, thus advancing the capabilities of autonomous aerial refueling and related applications.

AFIT Designator

AFIT-ENG-DS-24-S-028

Comments

A 12-month embargo was observed for posting this work on AFIT Scholar.

Distribution Statement A, Approved for Public Release. PA case number on file.

Download

Share

COinS