【摘要】 基于对不同驾驶状态下的方向盘操作特性及其频域特性进行分析,采用方向盘角度百分比功率谱密度及角速度百分比功率谱密度作为疲劳判别指标,将判别指标变量及驾驶状态作为模型的输入和输出对BP神经网络进行训练,构建相应的网络模型,并以疲劳状态分类准确率最大为优化目标确定疲劳检测取样窗口大小,最后利用测试样本数据验证模型的准确率。结果表明,该方法对驾驶员驾驶状态的分类准确率可达到91. 5%。更多还原

【Abstract】 Based on the analysis of steering wheel operating characteristics and its frequency domain characteristics under different driving conditions,the steering wheel angular percentage power spectral density and angular speed percentage power spectral density are used as fatigue discriminant indexes,and the discriminant index variables and driving state are used as input and output of the model to train BP neural network.The corresponding network model is used to determine the size of the sampling window for fatigue detection with the aim of maximizing the accuracy of fatigue state classification. Finally,the accuracy of the model is verified by the test sample data. The results show that the classification accuracy of the method for driver’s driving state can reach 91. 5%.更多还原