A Study of Adaptive Vehicle Behavior Detection Based on OBD and Wireless Sensor Networks

碩士 === 國立高雄應用科技大學 === 電子工程系碩士班 === 101 === The innovation of the car is the great contribution for people not only led the industrial revolution, but also made economic growth. Car brings many benefits for human, and creates a lot of problems such as traffic accidents, environmental pollutions, glob...

Full description

Bibliographic Details
Main Authors: Yun-Je Tsai, 蔡昀哲
Other Authors: Jen-Yea Jan
Format: Others
Language:zh-TW
Published: 2013
Online Access:http://ndltd.ncl.edu.tw/handle/08851103555077401902
Description
Summary:碩士 === 國立高雄應用科技大學 === 電子工程系碩士班 === 101 === The innovation of the car is the great contribution for people not only led the industrial revolution, but also made economic growth. Car brings many benefits for human, and creates a lot of problems such as traffic accidents, environmental pollutions, global warming. That will be a threat to human life and property. The traffic safety and telematics have received lots of attention from industrials and academics. The more cars we have, the more traffic accidents have happened to damage peoples’ life and properties. In this study we propose a new method of detecting the dangerous driving behaviors. In this work, three-axis accelerometers are used for measuring acceleration variation in three axes; a ZigBee wireless sensor network is used for transmitting sensing data in this system to a tablet PC. For more accurate of identifying driving behavior of this system, an OBD-II is applied for capturing driving speed to adjust behavior-detection parameters accordingly. The results of experiments show this method able to identify driving behaviors with more accuracy rate. In comparison with previous work, the proposed methodology improves the accuracy of behavior detections up to 92.3%, in conditions of downtown road and highway. The study is not only using the information of speed and rpm to aid driving behavior detection but also successfully utilizing MAF (Mass Air Flow) and throttle between foregoing information to identify idling vehicles. Furthermore, the system will actively send the message to remind drivers and all of detected data can be stored to the cloud service for further application. In conclusion, we hope the system can effectively reduce environmental pollution and traffic accident.