论文部分内容阅读
为实现险性驾驶行为状态的有效辨识,提出了一套驾驶行为险态辨识方法。以单位时间误操作率为依据,采用自底向上的分段算法实现了行为险态分级,采用因子方差分析,选取听觉感知、动视野、动视力、色觉、暗适应、注意力、判断能力、反应时这8项因子构成驾驶行为状态因子集,构建驾驶行为险态辨识特征向量,然后再对行为状态指标数据予以预先分级的前提下,采用单因子分析法对试验数据予以分析。并设计出BP神经网络行为险态辨识模型,最后进行了实例分析与计算。分析结果表明:反应时、注意力、判断能力3项指标在各分级间差异显著,故可作为驾驶行为险态辨识主因子,行为状态错判误差率为2.5%。
In order to realize the effective identification of dangerous driving behavior, a set of driving behavior hazard identification method is proposed. Based on the rate of misuse per unit time, the risk-based classification was implemented by bottom-up segmentation algorithm. The variance of the behavior was used to select the auditory perception, dynamic visual field, visual acuity, color vision, dark adaptation, attention, The eight factors of the reaction form the driving behavior state factor set, construct the driving behavior hazard identification eigenvector, and then pre-classify the behavioral state index data, and then use the single factor analysis method to analyze the experimental data. And design BP neural network behavioral hazard identification model, and finally carried out an example analysis and calculation. The analysis results show that the three indicators of attention, ability of judgment have significant difference among different grades, so they can be regarded as the main factor of risk identification of driving behavior, and the misclassification error rate of behavior is 2.5%.