基于K均值聚类算法与RBF神经网络的交通流预测方法

Traffic flow prediction method based on K-Means clustering algorithm and radial basis function

随着汽车保有量的增加,交通路网问题愈发严峻。针对现阶段城市交通流预测问题,在分析道路交通车流量、平均速度的周期性波动规律及变化趋势的基础上,融合K均值聚类算法(K-Means Clustering Algorithm)与RBF神经网络算法,构建了一种基于多维时间序列的交通流预测模型。实验结论表明,该模型预测结果与实际交通流数据拟合度高,精准度达到96.7%。该模型在理论研究与实际应用方面,对提升交通控制效果具有重要意义。

With the increase of car ownership,the problem of transportation network is more and more serious.For urban traffic flow prediction problem at this stage,in the analysis of road traffic flow and average velocity of periodic fluctuation and the change trend,a traffic flow prediction model based on multidimensional time series it introduced with the fusion of k-means Clustering Algorithm(K Means Clustering Algorithm)and RBF neural network Algorithm.The experimental results show that the model prediction fits the actual traffic flow data well,where the accuracy is 96.7%.The model is of great significance for improving traffic control effect in theoretical research and practical application.