TU Darmstadt / ULB / TUbiblio

Synthetic Generation of Radar Sensor Data for Virtual Validation of Autonomous Driving

Holder, Martin Friedrich (2021):
Synthetic Generation of Radar Sensor Data for Virtual Validation of Autonomous Driving. (Publisher's Version)
Darmstadt, Technische Universität Darmstadt,
DOI: 10.26083/tuprints-00017545,
[Ph.D. Thesis]

Abstract

The first accidents in otherwise promising deployments of autonomous driving fleets has underscored the importance of safety certification. Safety certification is expensive, especially when conducted via real-world driving. As such, high expectations are placed on virtual testing as an economic alternative. Autonomous driving functionality often relies heavily on radar sensors, but adequately modeling these radar sensors presents a particular challenge. While automotive simulation techniques have improved, there have yet to be systematic evaluations to prove that radar simulation models describe typical radar anomalies adequately such that downstream data processing algorithms behave correctly when operated with synthetic data. This dissertation builds a series of contributions to address this need. First, sensing artifacts are extracted from targeted measurements and evaluated for modeling relevance. A new method based on ray tracing is presented that generates spectral radar data from a 3D virtual environment, while addressing measurement ranges, limited resolution, and unambiguous intervals. This method, called "Fourier tracing", is of particular note in that it can model multipath related sensing artifacts. This dissertation presents a standard set of experiments for evaluating radar models and applies these to compare Fourier tracing to other modeling approaches using real radar data. The evaluation focuses on multipath propagation with respect to wave superposition, the visibility of occluded objects, on mirror targets, and on object tracking with synthetic data. The analysis makes it possible to falsify aspects of the examined radar sensor models, question their underlying assumptions, and to identify where these methods adequately simulate radar behavior.

Item Type: Ph.D. Thesis
Erschienen: 2021
Creators: Holder, Martin Friedrich
Status: Publisher's Version
Title: Synthetic Generation of Radar Sensor Data for Virtual Validation of Autonomous Driving
Language: English
Abstract:

The first accidents in otherwise promising deployments of autonomous driving fleets has underscored the importance of safety certification. Safety certification is expensive, especially when conducted via real-world driving. As such, high expectations are placed on virtual testing as an economic alternative. Autonomous driving functionality often relies heavily on radar sensors, but adequately modeling these radar sensors presents a particular challenge. While automotive simulation techniques have improved, there have yet to be systematic evaluations to prove that radar simulation models describe typical radar anomalies adequately such that downstream data processing algorithms behave correctly when operated with synthetic data. This dissertation builds a series of contributions to address this need. First, sensing artifacts are extracted from targeted measurements and evaluated for modeling relevance. A new method based on ray tracing is presented that generates spectral radar data from a 3D virtual environment, while addressing measurement ranges, limited resolution, and unambiguous intervals. This method, called "Fourier tracing", is of particular note in that it can model multipath related sensing artifacts. This dissertation presents a standard set of experiments for evaluating radar models and applies these to compare Fourier tracing to other modeling approaches using real radar data. The evaluation focuses on multipath propagation with respect to wave superposition, the visibility of occluded objects, on mirror targets, and on object tracking with synthetic data. The analysis makes it possible to falsify aspects of the examined radar sensor models, question their underlying assumptions, and to identify where these methods adequately simulate radar behavior.

Place of Publication: Darmstadt
Collation: XXIV, 204 Seiten
Divisions: 16 Department of Mechanical Engineering
16 Department of Mechanical Engineering > Institute of Automotive Engineering (FZD)
16 Department of Mechanical Engineering > Institute of Automotive Engineering (FZD) > Test Methods
Date Deposited: 07 May 2021 07:46
DOI: 10.26083/tuprints-00017545
Official URL: https://tuprints.ulb.tu-darmstadt.de/17545
URN: urn:nbn:de:tuda-tuprints-175450
Referees: Winner, Prof. Dr. Hermann ; Biebl, Prof. Dr. Erwin
Refereed / Verteidigung / mdl. Prüfung: 20 January 2021
Alternative Abstract:
Alternative abstract Language

Trotz vielversprechenden Präsentationen von Entwicklungsfahrzeugen zum autonomen Fahren rückt in Folge der ersten Unfälle mit selbstfahrenden Autos die Bedeutung ihres Sicherheitsnachweises in den Fokus. Hierbei werden an virtuelle Fahrversuche als wirtschaftliche Alternative zum Realversuch hohe Erwartungen gestellt. Mit seiner Schlüsselposition in autonomen Fahrfunktionen stellt die Modellierung des Radarsensors eine besondere Herausforderung dar. Obwohl eine zunehmenden Leistungsfähigkeit der Simulation festgestellt werden kann, fehlt ein systematischer Nachweis der Fidelität der Simulationsmodelle in Bezug auf die korrekte Abbildung radartypischer Annomalien sowie dem korrekten Verhalten von nachgelagerten Datenverabreitungsalgorithmen bei Stimulation mit synthetischen Daten. Hierauf baut diese Dissertation auf: Zunächst werden aus gezielt durchgeführten Messungen Artefakte extrahiert und für die Relevanz in der Modellierung bewertet. Zur Generierung von synthetischen Radardaten aus einer 3D-Welt wird ein neues, auf Raytracing basierendes Verfahren vorgestellt, das spektrale Radardaten aus einem virtuellen 3D Umfeld unter Berücksichtigung von Mess-, Auflösungs- und Eindeutigkeitsbereichen generiert. Das als "Fouriertracing"' bezeichnete Verfahren zeichnet sich dadurch aus, dass insbesondere mehrwegebedingte Artefakte abgebildet sind. Ein Standardsatz an Falsifikationsexperimenten zur Bewertung von Radarmodellen wurde erstellt und damit das entwickelte Fouriertracing Modell gegenüber anderen Modellierungsansätzen und Realdaten bewertet. Schwerpunkte stellen die Untersuchung der Mehrwegeausbreitung hinsichtlich Wellenüberlagerung, Sichtbarkeit verdeckter Objekte, Spiegelzielen und dem Verhalten eines Objekttrackingalgorithmus mit synthetischen Radardaten dar. Hierdurch gelingt es, Teilaspekte der untersuchten Modelle sowie der zugrundeliegenden Annahmen falsifizierbar zu machen und Gültigkeitsbereiche der Radarsimulation zu identifizieren.

German
Export:
Suche nach Titel in: TUfind oder in Google
Send an inquiry Send an inquiry

Options (only for editors)
Show editorial Details Show editorial Details