激进驾驶状态识别算法研究

A Study on the Recognition Algorithm of Aggressive Driving State

为识别和量化激进驾驶这一极易导致交通事故的驾驶状态,建立了激进驾驶状态识别算法。首先根据驾驶的操作特点提出驾驶状态-驾驶行为-驾驶模式3层结构并建立其数学描述。然后基于Stevens幂定律提出驾驶模式激进程度的量化指标——激进指数,并在此基础上建立了激进驾驶状态识别算法以区分不同驾驶状态的激进程度。最后通过不同驾驶状态的实车实验样本完成算法参数确定和算法验证,结果显示该算法能有效识别激进驾驶状态。

For identifying and quantifying aggressive driving, an accident-prone driving state, a recognition algorithm of aggressive driving state is established. Firstly a driving state-driving behavior-driving mode three layer structure is proposed with its math description given based on the manipulation feature of driving. Then with Stevens power law, a quantitative indicator, aggressive index, is put forward to describe the aggressive level of driving modes, based on which a recognition algorithm is set up to distinguish different aggressive levels of driving state. Fi-nally based on the data of different driving states collected in real vehicle experiments, algorithm parameters are determined and the algorithm is verified. The results show that the proposed algorithm can identify aggressive driving state effectively.