Abstract
Research and/or Engineering Questions/Objective
An electric motor driven road vehicle, such as a hybrid and/or electric vehicle, could provide ecologically friendly environments that contain a silent soundscape as well as low air pollution. However, silent driving could place some pedestrians in danger when approaching at low speed. This is because the acoustic vehicle alerting system (AVAS) should be regulated in Europe and the US, which could cause different kinds of noise pollution. This study not only aims to reduce electric vehicle driving noise, but also to keep pedestrians safe with a directional warning system.
Methodology
For the sake of both pedestrian safety and a silent soundscape, the selective warning system for pedestrian safety has two parts; one is to detect pedestrian moving position, and the other is to focus alert sounds within a narrow spatial range. Detection of pedestrians could be achieved by an active driver assistance system (ADAS) camera based on advanced image recognition technology. Through the position data of people acquired by this system, the pedestrian area is considered a focused warning zone. The focused alert sound zone could be realized by sound ball technology, which controls the spatial distribution of warning sounds for pedestrian safety. The combination of pedestrian recognition and a directional alert system can minimize noise pollution and enable pedestrians to recognize an approaching vehicle.
Results
The ADAS-based pedestrian recognition system can detect multiple positions of people who could be in danger if they do not perceive the approaching vehicle. The directional warning system can focus the alert sound distribution around pedestrians in danger. People in the selective warning zone can perceive a concentrated alert sound of about 15dB more than those who are not in the area. Sound directivity can be achieved by optimizing the speaker array pattern, which minimizes the unnecessary side lobe area of directional sound. The direction of the warning sound can be freely changed within 120 degrees of the frontal area with acceptable spatial sound concentration performance.
Limitations of this study
This study aims to check the feasibility of a selective warning system using sound ball technology, which means that it has a limitation of application to real vehicle implementation. The pedestrian detection system is a component of a real ADAS that is available in real vehicles. However, the system contains a digital signal processing module for severe calculation and a speaker array, which may not be applicable for commercial purposes because of their cost and durability.
What does the paper offer that is new in the field including in comparison to other work by the authors?
Manipulation of a focused warning sound zone can be achieved by minimizing side lobe and maximizing the steering angle of directional sound.
Conclusions
The selective warning system for pedestrian safety can guarantee a low level of noise pollution with good perceptibility performance for approaching electric vehicles. It can be possible through a directional warning system based on sound spatial distribution control technology with a pedestrian detection system based on advanced image recognition technology in ADAS.
KEYWORDS : Pedestrian safety, Sound ball technology, Selective warning system.