Synergistic perception and control simplex for verifiable safe vertical landing

Autonomous driving systems play an essential role in modern robotic systems. However, advanced machine learning-based perception and control algorithms may fail to avoid collisions occasionally. In this thesis, we integrate and evaluate a holistic safety driving system. Our work and evaluation mainly focus on an air taxi system, a prospective future commute tool that will significantly enhance efficiency and air mobility. The innovative safety driving framework contains verifiable algorithms. We adopt the Perception Simplex system for reliable obstacle detection to avoid collisions. In addition, we integrate L1 adaptive control to improve the system’s robustness. To further reduce the landing time and enhance efficiency, we update the safety envelope by considering real-time dynamic confirmation of the control capability instead of the static worst-case control capability. This work demonstrates the success and reliability of the safety driving framework, which can significantly improve landing efficiency while ensuring safety and robustness. This framework can also be integrated into other autonomous systems to improve safety in many fields further.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Mechanical Engineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2024

Author dc:creator

• Zhao, Yang

Contributors dc:contributor

• Hovakimyan, Naira

• Vertical Take-off And Landing
• Air Taxi
• Adaptive Control
• Safe Driving Framework

Statement dc:rights

• Copyright 2024 Yang Zhao

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/124610

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F124610