Xing, YangLv, ChenHuaji, WangWang, HongCao, Dongpu2017-02-132017-02-132017-03Xing Y, Lv C, Huaji W, et al., (2017) Recognizing driver braking intention with vehicle data using unsupervised learning methods. WCX17: SAE World Congress Experience, 4-6 April 2017, Detroit, USAhttps://doi.org/10.4271/2017-01-0433https://dspace.lib.cranfield.ac.uk/handle/1826/11439Recently, the development of braking assistance system has largely benefit the safety of both driver and pedestrians. A robust prediction and detection of driver braking intention will enable driving assistance system response to traffic situation correctly and improve the driving experience of intelligent vehicles. Unsupervised machine learning algorithms has been widely used in clustering and pattern mining in previous researches. In this paper, a various unsupervised clustering methods will be used to build a driver braking intention predictor which can accurately recognize the braking maneuver based on vehicle data captured with CAN bus. The braking maneuver along with other driving maneuvers such as normal driving will be clustered and the results from different methods like K-means and Gaussian mixture model will be compared. Additionally, the evaluation of features from raw data, which are important to driving maneuvers clustering will be proposed. The experiment data are collected from one hybrid electric vehicle in real world. Final results show that the proposed method can detect driver’s braking intention in a very beginning moment with a high accuracy and the most important sets of feature for driving maneuver clustering will be discussed.Recently, the development of braking assistance system has largely benefit the safety of both driver and pedestrians. A robust prediction and detection of driver braking intention will enable driving assistance system response to traffic situation correctly and improve the driving experience of intelligent vehicles. Unsupervised machine learning algorithms has been widely used in clustering and pattern mining in previous researches. In this paper, a various unsupervised clustering methods will be used to build a driver braking intention predictor which can accurately recognize the braking maneuver based on vehicle data captured with CAN bus. The braking maneuver along with other driving maneuvers such as normal driving will be clustered and the results from different methods like K-means and Gaussian mixture model will be compared. Additionally, the evaluation of features from raw data, which are important to driving maneuvers clustering will be proposed. The experiment data are collected from one hybrid electric vehicle in real world. Final results show that the proposed method can detect driver’s braking intention in a very beginning moment with a high accuracy and the most important sets of feature for driving maneuver clustering will be discussed.enPublished by SAE. This is the Author Accepted Manuscript. Please refer to any applicable publisher terms of use.Conference paper