Kobbert, Jonas (2019)
Optimization of Automotive Light Distributions for Different Real Life Traffic Situations.
Technische Universität Darmstadt
Ph.D. Thesis, Primary publication
|
Text
2019-01-21_Kobbert_Jonas.pdf Copyright Information: CC BY-NC-SA 4.0 International - Creative Commons, Attribution NonCommercial, ShareAlike. Download (36MB) | Preview |
Item Type: | Ph.D. Thesis | ||||
---|---|---|---|---|---|
Type of entry: | Primary publication | ||||
Title: | Optimization of Automotive Light Distributions for Different Real Life Traffic Situations | ||||
Language: | English | ||||
Referees: | Khanh, Prof. Dr. Tran Quoc ; Neumann, Prof. Dr. Cornelius | ||||
Date: | 21 January 2019 | ||||
Place of Publication: | Darmstadt | ||||
Date of oral examination: | 7 December 2018 | ||||
Abstract: | The major goal of this thesis is to find a way to optimize current automotive headlamps in order to provide safer nighttime driving. While this has already been done in the past with the works by Damasky and Huhn, the current approach combines methods previously not used in one single study. In the first steps, the influence of different headlamp parameters on viewing distance of the driver is evaluated in field tests. In the second step, the current German traffic space is analysed before in the third step, the gaze behaviour of drivers is recorded and investigated for different situations. The combination of these studies is then used to propose new light distributions. In the first part, field tests are conducted in order to investigate detection distances with different lighting conditions. The gained data is used to provide recommended luminous intensity values for certain detection distances. Furthermore, the data is used to extract luminous intensity recommendations for different angular positions relative to the hot spot. These investigations show, that the current limits set by the ECE for high beam headlamps are sufficient to provide safe detection distances for nearly all situations. However, the data also shows, that low beam should be disregarded for any situation and only be used if high beam cannot be used at all. The traffic space analysis in the second part of this thesis shows, that there are significant differences between different road categories in terms of object location and frequency. For these situations, optimized segment distributions are proposed, leading to significant benefits over the conventional high beam setup. The difference between the proposed segment partitioning and the standard setup is, that the segments are not set equal in size. The segments at the centre of the distribution are set to be smaller in order to better mask out traffic that is further away. Furthermore, it is shown, that the benefit of additional segments is limited at around 280 segments, where a performance identical to a 10000 pixel headlamp is achieved. In the last section, regarding the gaze analysis a large driving test, including 54 test subjects is performed. Here the findings by Diem, Damasky, Brückmann and Weber are confirmed. New approaches regarding the correlation between the driver’s gaze and objects in the traffic space are tested. On a general level, no correlation between the object distribution and the gaze is found. However, a large databank containing object positions as well as driver’s gaze, speed, lighting condition and position in the world is set up for further, more detailed information. The data from all presented studies is then used to propose new, optimized light distributions. |
||||
Alternative Abstract: |
|
||||
URN: | urn:nbn:de:tuda-tuprints-83828 | ||||
Classification DDC: | 600 Technology, medicine, applied sciences > 620 Engineering and machine engineering | ||||
Divisions: | 18 Department of Electrical Engineering and Information Technology 18 Department of Electrical Engineering and Information Technology > Light Technology (from Oct. 2021 renamed "Adaptive Lighting Systems and Visual Processing") |
||||
Date Deposited: | 25 Jan 2019 12:29 | ||||
Last Modified: | 09 Jul 2020 02:29 | ||||
URI: | https://tuprints.ulb.tu-darmstadt.de/id/eprint/8382 | ||||
PPN: | 441510620 | ||||
Export: |
View Item |