Real time collision warning system in the context of vehicle-to-vehicle data exchange based on drivings behaviours analysis
Worldwide injuries in vehicle accidents have been on the rise in recent years, mainly due to driver error regardless of technological innovations and advancements for vehicle safety. Consequently, there is a need for a reliable-real time warning system that can alert drivers of a potential col...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/4123/1/24p%20HUSSEIN%20ALI%20AMEEN.pdf http://eprints.uthm.edu.my/4123/3/HUSSEIN%20ALI%20AMEEN%20COPYRIGHT%20DECLARATION.pdf http://eprints.uthm.edu.my/4123/2/HUSSEIN%20ALI%20AMEEN%20WATERMARK.pdf http://eprints.uthm.edu.my/4123/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Worldwide injuries in vehicle accidents have been on the rise in recent years, mainly
due to driver error regardless of technological innovations and advancements for
vehicle safety. Consequently, there is a need for a reliable-real time warning system
that can alert drivers of a potential collision. Vehicle-to-Vehicle (V2V) is an extensive
area of ongoing research and development which has started to revolutionize the
driving experience. Driving behaviour is a subject of extensive research which gains
special attention due to the relationship between speeding behaviour and crashes as
drivers who engage in frequent and extreme speeding behaviour are overinvolved in
crashes. National Highway Traffic Safety Administration (NHTSA) set guidelines on
how different vehicle automation levels may reduce vehicle crashes and how the use
of on-board short-range sensors coupled with V2V technologies can help facilitate
communication among vehicles. Based on the previous works, it can be seen that the
assessment of drivers’ behaviours using their trajectory data is a fresh and open
research field. Most studies related to driving behaviours in terms of acceleration�deceleration are evaluated at the laboratory scale using experimental results from
actual vehicles. Towards this end, a five-stage methodology for a new collision
warning system in the context of V2V based on driving behaviours has been designed.
Real-time V2V hardware for data collection purposes was developed. Driving
behaviour was analyzed in different timeframes prior obtained from actual driving
behaviour in an urban environment collected from OBD-II adapter and GPS data
logger of an instrumented vehicle. By measuring the in-vehicle accelerations, it is
possible to categorize the driving behaviour into four main classes based on real-time
experiments: safe drivers, normal, aggressive, and dangerous drivers. When the
vehicle is in a risk situation, the system based on NRF24L01+PA/LNA, GPS, and
OBD-II will pass a signal to the driver using a dedicated LCD and LED light signal.
The driver can instantly decide to make the vehicle in a safe mood, effectively avoid
the happening of vehicle accidents. The proposed solution provides two main functions: (1) the detection of the dangerous vehicles involved in the road, and (2) the display of
a message informing the driver if it is safe or unsafe to pass. System performance was
evaluated to ensure that it achieved the primary objective of improving road safety in
the extreme behaviour of the driver in question either the safest (or the least aggressive)
and the most unsafe (or the most aggressive). The proposed methodology has retained
some advantages for other literature studies because of the simultaneous use of speed,
acceleration, and vehicle location. The V2V based on driving behaviour experiments
shows the effectiveness of the selected approach predicts behaviour with an accuracy
of over 87% in sixty-four real-time scenarios presented its capability to detect
behaviour and provide a warning to nearby drivers. The system failed detection only
in few times when the receiving vehicle missed data due to high speed during the test
as well as the distances between the moving vehicles, the data was not received
correctly since the power transmitted, the frequency range of the signals, the antenna
relative positions, and the number of in-range vehicles are of interest for the V2V test
scenarios. The latter result supports the conclusion that warnings that efficiently and
quickly transmit their information may be better when driver are under stress or time
pressure. |
|---|