Skip to Main content Skip to Navigation
Theses

Pedestrian and Cyclist Forward Collision Warning System Effectiveness Estimation based on Simulation of Kinematic Reconstructions

Abstract : In 2016, road fatalities reached 1.35 million in the world according to the World Health Organization. 26% of these fatalities were pedestrians and cyclists. Nowadays, more and more cars are equipped with an emergency system (called FCW – Forward Collision Warning) that can detect pedestrians and cyclists in order to warn drivers of a hazardous situation. These systems can also help in collision avoidance either by assisting driver during braking or by activating an Automatic Emergency Braking (AEB). This thesis is focused on Pedestrian and Cyclist AEB and FCW assessment and has three main objectives. First, an analysis on more than 3700 accident case reconstructions (2200 cyclist cases and 1500 pedestrian cases) from two databases, one French and one German has been performed. Accident configurations have been extracted and classified into different scenarios. A simulation software has been implemented in order to replay the accident kinematics with the integration of an AEB by varying their system characteristics. This allows the identification of optimum characteristics for a pedestrian AEB and cyclist AEB in terms of road user detection. It also allows identifying FCW trigger time and the duration of an emergency braking. Secondly, based on an experimental campaign using a driving simulator, the driver’s reactions to a FCW signal have been analyzed on different accident configurations: pedestrian/cyclist cases, with/without FCW and with different FCW triggers. Two hundred volunteers participated in this experiment. The results concern the gaze analysis, the driver’s response to the FCW signal, the time reaction to trigger a braking and the different behavior depending on the driving configurations. The third objective concerns the benefits assessment of a FCW. Based on the results of the driving simulator experiment and the kinematic reconstructions of the accidents, benefits of a FCW are estimated in terms of potential avoided or mitigated accidents. Finally, some perspectives of this work are proposed.
Document type :
Theses
Complete list of metadata

https://hal-amu.archives-ouvertes.fr/tel-03058542
Contributor : François Char <>
Submitted on : Friday, December 11, 2020 - 6:49:58 PM
Last modification on : Wednesday, December 30, 2020 - 1:08:07 PM
Long-term archiving on: : Friday, March 12, 2021 - 8:22:39 PM

File

Thesis manuscrit-Francois-CHAR...
Files produced by the author(s)

Identifiers

  • HAL Id : tel-03058542, version 1

Collections

Citation

François Char. Pedestrian and Cyclist Forward Collision Warning System Effectiveness Estimation based on Simulation of Kinematic Reconstructions. Engineering Sciences [physics]. AMU - Aix Marseille Université, 2020. English. ⟨tel-03058542⟩

Share

Metrics

Record views

83

Files downloads

42