預測駕駛者行為與事故之智能駕駛輔助系統

An intelligent driver assistance system to anticipate the human traits and the accidents

巴若璞(RamPrasad Palarapu)

指導教授 ： 黃經堯

國立交通大學/電機學院/電機資訊國際學位學程/碩士(2016年)

摘要

這篇論文提出了一種透過學習駕駛員個別的固定及應變行為駕駛者的駕駛行為的智能駕駛警報系統的新穎設計，現在大多數的警報系統是基於物理特性像是他們關注車輛的軌跡，但忽略了駕駛者的能力以及特性，這項工作涉及了一些議題，像是提高現有的技術卻又得使其不複雜使用，可以使一般駕駛者覺得舒適。
現今，很多的高級車都有裝備著智能駕駛預警系統，但之後這類系統的數量很有可能會急速增加，因為它有可能成為一般的轎車的標準配備，然而，將多種警報系統引入車輛裡可能增加駕駛任務的複雜性，並有許多關鍵的人為、車輛及環境因素應考慮，像是他們如何結合年齡、性別、警示方案，系統的可靠性以及干擾等因素的交互作用去影響駕駛性能及情境感知。
在這篇論文中，我將討論這些問題並使用不同的關係模型和預測演算法來預測駕駛者的行為以及事故，此外，我還將討論不同的特性如何貢獻出不同的價值觀導致不同的行為。

關鍵字

Internet of Things

並列摘要

This thesis proposes a novel design of an intelligent driver warning system that learns each individual trait of the driver both stationary and challenging and use this information to improve the driving habits of a driver. Most of the current warning systems are physics based as they look at the vehicle trajectory, but mainly ignore the abilities as well as the characteristics of the driver. A number of research issues are involved in this work, as it has to improve upon the state of the art, yet not so complicated to use that the average driver would feel comfortable using it.
Intelligent driver warning systems can be found in many high-end vehicles on the road today, which will likely rapidly increase as they become standard equipment. However, introducing multiple warning systems into vehicles could potentially add to the complexity of the driving task, and there are many critical human factors, vehicle factors and environmental factors issues that should be considered, such as how they could interact between one factor to another along with the ages and gender, warning alert schemes, system reliabilities, and distractions combine to affect driving performance and situation awareness.
In this proposal, I will discuss these issues and describe preliminary results in using different relationship models and prediction algorithm to predict the driver’s features as well as the accidents. In addition, I will also discuss how different features contribute their values to lead to the different behaviors.

並列關鍵字

Internet of Things

參考文獻

[2] C. Laugier, I. E. Apparatchik, M. Perrollaz, M. Yong, J.-D. Yoder, C. To, K. Mekhnache, and A. Negre. Probabilistic analysis of dynamic scenes and collision risks assessment to improve driving safety. ITS Magazine, IEEE, 3(4), 2011.

[4] 2014 Crash Data Key Findings, Published by NHTSA’s National Center for Statistics and Analysis 1200 New Jersey Avenue SE., Washington, DC 20590.

[5] Early Estimate of Motor Vehicle Traffic Fatalities for the First Half (Jan – Jun) of 2015, Published by NHTSA’s National Center for Statistics and Analysis 1200 New Jersey Avenue SE., Washington, DC 20590.

[11] A Real-Time Embedded Blind Spot Safety assistance system (Research Article), Bing-FeiWu, Chih-Chung Kao, Ying-Feng Li, and Min-Yu Tsai.

[13] A Real-Time Embedded Blind Spot Safety assistance system (Research Article) Bing-FeiWu, Chih-Chung Kao, Ying-Feng Li, and Min-Yu Tsai.