Flight path planning for infrastructure inspection based on artificial potential fields

doctoral thesis

• Title (croatian): Planiranje putanje leta za inspekciju infrastrukture zasnovano na umjetnim potencijalnim poljima
• Author: Jurica Goričanec
• Mentor: Stjepan Bogdan (mentor)
• Committee member: Matko Orsag (predsjednik povjerenstva)
• Committee member: Tamara Petrović (član povjerenstva)
• Committee member: Tomislav Radišić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2024-07-11, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing
• Universal decimal classification (UDC): 004 - Computer science and technology. Computing. Data processing

Abstract

This thesis develops flight path planning algorithms and trajectory tracking methods for autonomous UAV inspection of infrastructure. Using artificial potential fields (APF), these algorithms are adaptable to different scenarios, varying in level of complexity in surveyed targets and environments. The goal is to generate safe, feasible flight paths for UAVs, ensuring comprehensive data acquisition for infrastructure assessment and damage detection. The system must map the environment, detect real-time obstacles, and execute collision avoidance while adhering to the planned path in unobstructed areas. This thesis proposes multiple inspection path planning methods, depending on the complexity and available knowledge of the target object, such as 3D models, 2D layouts, or approximate locations. Inspection parameters are set based on sensor specifications and requirements, like image overlap and feature size. The planning algorithms incorporate principles like Huygens's wave propagation and 3D envelope construction. A path tracking method for LiDAR-equipped UAVs is also introduced, with real-time obstacle detection and collision avoidance using augmented APF. The system prevents UAVs from getting trapped in local minima by combining normal and rotational repulsive forces and uses attractive forces to minimize deviations. The feasibility of waypoints is evaluated to maintain adherence to the planned path. These methods were validated through simulations, laboratory experiments, and real-world inspections of different structures.

Abstract (croatian)

Glavni fokus ove doktorske disertacije je razvoj algoritama za planiranje putanje leta i metode praćenja trajektorije za autonomnu inspekciju infrastrukture pomoću bespilotnih letjelica. Korištenjem umjetnih potencijalnih polja, ovi algoritmi se mogu prilagoditi različitim scenarijima, ovisno o razini složenosti infrastrukturnih objekata i okruženja u kojem se nalaze. Cilj je generiranje sigurne i izvedive putanje leta za UAV, osiguravajući sveobuhvatno prikupljanje podataka za procjenu stanja infrastrukture i otkrivanje oštećenja. Sustav mora mapirati okruženje, detektirati prepreke u stvarnom vremenu i izbjegavati sudare, te osigurati praćenje planirane putanje u slobodnom prostoru. U disertaciji je predloženo više metoda planiranja inspekcijskih putanja, ovisno o složenosti i dostupnim informacijama o cilju, kao što su 3D modeli, 2D tlocrti ili približne lokacije. Parametri inspekcije određuju se na temelju specifikacija senzora i zahtjeva, poput preklapanja slika i veličine prepoznatljivih značajki. Planiranje putanja uključuje principe poput Huygensove propagacije valova i konstrukcije 3D ovojnice. Također je uvedena metoda praćenja putanje leta s detekcijom prepreka u stvarnom vremenu i izbjegavanjem sudara pomoću potencijalnih polja. Sustav osigurava to da bespilotna letjelica neće zapeti u točkama lokalnog minimuma kombiniranjem normalnih i rotacijskih odbojnih sila te koristi privlačne sile za minimalizaciju odstupanja. Ove metode su potvrđene simulacijama, laboratorijskim eksperimentima i stvarnim inspekcijama različitih struktura.

• Keywords: Unmanned Aerial Vehicle; Infrastructure Inspection; Artificial Potential Fields; Flight Path Planning; Obstacle Avoidance
• Keywords (croatian): bespilotna letjelica; inspekcija infrastrukture; umjetna potencijalna polja; planiranje putanje leta; izbjegavanje prepreka
• Language: english
• URN:NBN: urn:nbn:hr:168:436124
• Promotion: 2024
• Study programme: Title: Computer Science; Study programme type: university; Study level: postgraduate; Academic / professional title: Doktor znanosti (Doktor znanosti)
• Type of resource: Text
• Extent: 117 str. : graf. prikazi
• File origin: Born digital
• Access conditions: Open access
• Created on: 2024-09-03 12:52:12