作　　者： ; ; ; ;

机构地区： 北京理工大学自动控制系

出　　处： 《北京理工大学学报》 2004年第4期331-334,共4页

摘　　要： 针对由于驾驶行为的不确定性,难以建立精确的车辆跟驰模型的问题,应用径向基函数神经网络建立了跟驰模型,改进了基于最近邻聚类的网络学习算法,并利用跟驰数据对模型进行了验证.结果表明,该网络模型与多层前馈网络模型相比,结构简单,训练时间短,精度高,适宜在线进行实时预测. It is hard to establish a precise car-following model because of the uncertainty in driver's behavior. A car-following model is developed based on the radial basis function (RBF) network. With this the nearest neighbor-clustering algorithm (NNCA) is improved, and the results of modeling are examined by the car-following data. The simulation results show that the proposed RBF network has a higher precision and requires shorter training in the prediction of the car-following model compared with the multiplayer neural network.

关 键 词： 车辆跟驰 人工神经网络 最近邻聚类学习算法 径向基网络

领　　域： [自动化与计算机技术] [自动化与计算机技术] [交通运输工程] [交通运输工程]