Analysis of Ultrasound-based Collision Warning System for Low Speed Vehicle

• Main Authors: Teng-Chieh Hsu, 許登傑
• Other Authors: Yi-Yu Liu
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/93941389828181484197

碩士 === 元智大學 === 資訊工程學系 === 102 === Low speed vehicle collision is one of the most common traffic accidents in urban area. To prevent vehicle collision, ISO 15623:2013 was established to point out necessary requirements for forward collision warning. Many previous works adopted infrared and/or millimeter wave to measure distance away from close neighboring vehicles. However, millimeter wave-based forward collision warning systems are not widely deployed owing to its high cost. In this thesis, we propose a relatively low-cost collision warning system based on ultrasound. We exploit parking sensors to measure the distance between vehicle driver and adjacent obstacles/vehicles/pedestrians. Furthermore, Doppler Effect is applied to estimate relative speed and collision potential. Mathematical formulation is quantitatively derived to analyze the hazardous conditions for vehicle drivers. A lookup table is constructed to activate collision warning alarm and/or vehicle active safety. Our proposed methodology can be utilized in either forward collision warning or backward collision warning.