论文部分内容阅读
本文采用一种基于非参数回归模型的最小K近邻算法来预测四川高速公路上车辆的行驶速度。用加权欧氏距离作为距离度量标准来确定最近的K组数据序列。然后,建立一个最小K近邻非参数回归模型来预测未来6分钟以内的平均行驶速度。用浮动车数据采集(FCD)系统随机采集的几个行驶速度数据序列来验证该模型。结果表明,通过使用FCD系统,该模型预测平均行驶速度的准确性在90%以上,因此该模型是可行的,并且是有效的。
In this paper, a K-nearest neighbor algorithm based on nonparametric regression model is used to predict the speed of vehicles on Sichuan freeway. The weighted Euclidean distance is used as a measure of distance to determine the nearest K sets of data sequences. Then, a minimum K-nearest neighbor nonparametric regression model is established to predict the average speed of the next 6 minutes. The model was verified by several driving speed data sequences collected randomly by the FCD system. The results show that by using the FCD system, the model predicts that the average driving speed is more than 90% accurate, so the model is feasible and effective.