【摘要】 现有车辆跟驰模型大多建立在控制论和运动学观点之上,以至于在仿真过程中连续运用单一模型控制规则来支配驾驶员的驾驶行为,忽略了对多源高负载信息感知变量的运用。文章直接运用信息挖掘技术最大限度地榨取实测数据所携带的有关驾驶行为的个体有用信息,通过多元非参数三次样条回归模型剔除了数据中由白噪声产生的干扰,构建了一种基于三次样条非参数拟合的驾驶行为仿真模型。仿真试验表明,此模型具有可移植、高精度的特性,能很好地反映和预测多源高负载信息感知变量刺激下跟驰过程中驾驶员的驾驶行为。更多还原

【Abstract】 For the existed car-following models, most of them are developed on the basis of cybernetics and kinematics, and a single controller is continuously used to control the driver behavior and the cognitive multi-variable with high loading information is ignored. The information mining technology was used to extract the useful individual driving behavior information from the field data, and the noise was eliminated by the cubic spline regression model, and a new simulation model of driving behavior was established based on the cubic spline nonparametric fitting regression. Simulation results show that the simulation model can effectively predict the driving behavior in the car-following process.更多还原