nach oben

2014 | Buch

Kapitel lesen Erstes Kapitel lesen

Knowledge-Based Driver Assistance Systems

Traffic Situation Description and Situation Feature Relevance

verfasst von: Michael Huelsen

Verlag: Springer Fachmedien Wiesbaden

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

The comprehension of a traffic situation plays a major role in driving a vehicle. Interpretable information forms a basis for future projection, decision making and action performing, such as navigating, maneuvering and driving control. Michael Huelsen provides an ontology-based generic traffic situation description capable of supplying various advanced driver assistance systems with relevant information about the current traffic situation of a vehicle and its environment. These systems are enabled to perform reasonable actions and approach visionary goals such as injury and accident free driving, substantial assistance in arbitrary situations up to even autonomous driving.

Inhaltsverzeichnis

Frontmatter

1. Introduction

Abstract

The declared objective of the European Commission was to halve the number of traffic fatalities in the period from 2001 to 2010 [European Commission, 2001]. With a decrease of 44% this goal could not completely be achieved (see Fig. 1). Consequently, the goal was renewed and adjusted for the period from 2011 to 2020 in [European Commission, 2010] and new strategic objects based on the current “3^rd road safety action programme” (RSAP) are derived to propose new actions for the subsequent RSAP. The aim for 2020 is to “halve the overall number of road deaths in the European Union by 2020 starting from 2010”.

Michael Huelsen

2. The Research Domain of this Thesis and its State of the Art

Abstract

Driver Assistance Systems (DAS) are additional electronic systems in vehicles to support the driver in specific driving situations. They may aim at increasing safety, higher comfort, less fuel (or resource) consumption or informing the driver about the current traffic condition, the traffic situation or the route. These systems may be informative, semi-autonomous or autonomous. They may intervene in vehicle propulsion, actuation or vehicle control or may simply provide useful additional information.

Michael Huelsen

3. Theoretical Foundations Relevant to this Thesis

Abstract

This chapter gives a brief overview of the theory behind the methods applied in this thesis.

Sections 2.2.3 and 2.2.4 in Chapter 2 motivated feature selection and feature relevance estimation to be supportive of an effective and lean situation description. Mutual information was found to be a method useful to select situation features and worth further investigation.

Michael Huelsen

4. Situation Feature Relevance on Measurement Data

Abstract

Predictive driver assistance systems require information about the future progress of a situation. This may be the upcoming action of the driver, the vehicle course or an event, such as a collision. Due to its complexity, this prediction is often hardly tangible for the human mind, so that machine learning methods to perform this kind of functions are increasing in popularity.

Michael Huelsen

5. Knowledge-Based Traffic Situation Description

Abstract

Beginning with section 5.1, this chapter describes the knowledge engineering and structure of a generic traffic situation description ontology that was motivated in section 2.2.6. Aspects relevant to a generic situation description, as pointed out in section 2.2.3, are moreover discussed on the developed ontology. The introduced ontology is designed for complex traffic situations, especially those at intersections.

Michael Huelsen

6. Relevance by Mutual Information on Ontology Features

Abstract

Chapter 1 delineates situation features and a holistic method to create semantics with these features to be necessary for a generic situation description. Mutual information based feature selection as a method to identify relevant situation features with respect to some priory known target function or class is described in section 3.1. Its application on vehicle and environmental sensor measurement data is investigated and applied in Chapter 4. Ontologies are explained in section 3.2 and found to be suitable for a holistic description of traffic situations in Chapter 5.

Michael Huelsen

7. Conclusion

Abstract

With the European Commission among other international agencies aiming to significantly reduce the number of traffic fatalities ongoing and in the future as well as to improve the performance of traffic systems and participants intensive research on Driver Assistance Systems is part of today’s science. These systems support the driver in critical situations or intervene in the driving process to avoid accidents or to reduce their severity. In addition, comfort assistance functions are increasingly making driving more convenient.

Michael Huelsen

Backmatter

Titel: Knowledge-Based Driver Assistance Systems
verfasst von: Michael Huelsen
Verlag: Springer Fachmedien Wiesbaden
Electronic ISBN: 978-3-658-05750-3
Print ISBN: 978-3-658-05749-7
DOI: https://doi.org/10.1007/978-3-658-05750-3

Premium Partner

Marktübersichten

Die im Laufe eines Jahres in der „adhäsion“ veröffentlichten Marktübersichten helfen Anwendern verschiedenster Branchen, sich einen gezielten Überblick über Lieferantenangebote zu verschaffen.

Zur Marktübersicht