面向驾驶行为预警的换道意图辨识模型研究

发布时间：2018-04-12 08:15

本文选题：驾驶意图 + 先进驾驶辅助系统(ADAS)　；参考：《中国安全科学学报》2016年02期

【摘要】：为更准确地对驾驶行为进行预警,进一步提高驾驶人换道意图的辨识准确率,借助驾驶模拟器采集数据,建立基于支持向量机(SVM)理论的换道意图辨识模型。对比分析不同人-车-路系统参数组合在换道意图和车道保持期间的差异性,选取最佳特征参数组合,运用网格和遗传算法-支持向量机(GA-SVM)寻优方法优化模型参数,并对优化模型进行验证。结果表明,以纵向加速度、方向盘转角、车辆偏离车道中心线的距离、驾驶人头部运动横坐标变化值作为表征换道意图的人-车-路系统特征参数,优化模型惩罚参数c为58.642 3、核函数参数g为222.732 6时,该模型对驾驶人换道意图的辨识准确率为90%,误警率为5%,基本实现准确辨识换道意图。
[Abstract]:In order to warn the driving behavior more accurately and improve the identification accuracy of the driver's change intention, the identification model of the change intention based on the support vector machine (SVM) theory was established with the help of the data collected by the driving simulator.By comparing and analyzing the differences between different man-vehicle-road system parameter combinations during the course of lane change and lane keeping, the optimal feature parameter combination is selected, and the model parameters are optimized by using the grid and genetic algorithm-support vector machine (GA-SVM) optimization method.The optimization model is verified.The results show that the longitudinal acceleration, steering wheel angle, the distance from the center of the lane, the change value of the transverse coordinate of the driver's head motion are taken as the characteristic parameters of the human-vehicle-road system to represent the change intention.The optimized model has a penalty parameter c of 58.642 3 and a kernel function parameter g of 222.732 6. The identification accuracy of the model is 90 and the false alarm rate is 5.
【作者单位】：江苏省交通技师学院车辆工程系;江苏大学汽车与交通工程学院;
【基金】：江苏省交通运输科技项目(2014Y17)
【分类号】：U463.6

【相似文献】