摘要
针对高速公路交通流速度分布特征,选取高斯核函数进行核密度估计分布分析。首先,采用独立分布和联合分布筛选微波检测器采集数据,并对故障数据和丢失数据采用加权平均值法进行修复。然后,在对交通流速度数据样本描述性统计和K-S检验不服从正态分布结果的基础上,采用均方误差(MSE)和启发法确定最优窗宽,并用拟合检验法对核密度估计分布进行拟合检验。结果表明,核密度估计法可以较好的拟合交通流速度的概率分布。
According to the velocity distribution characteristics of traffic flow,the paper use Gauss kernel function to estimate the distribution of kernel density.Firstly,using the independent and joint distribution method to screen the microwave detector data,and the weighted average method to repair the fault and missing data.Then,on the basis of the results of the descriptive statistics and the Kolmogorov-Smirnov(K-S)test,the mean square error(MSE)and heuristic method are used to determine the optimal bandwidth,and the chi-squared test is used to fit the distribution of kernel density estimation.The results show that the kernel density estimation method can well fit the probability distribution of traffic flow velocity.
作者
郑伟
朱洪磊
符锌砂
梁中岚
ZHENG Wei;ZHU Honglei;FU Xinsha;LIANG Zhonglan(School of Civil Engineering and Transportation,South China University of Technology,Guangzhou,China,510641)
出处
《公路工程》
北大核心
2018年第2期113-117,128,共6页
Highway Engineering
基金
广东省交通运输厅科技项目(科技-2015-02-071
科技-2015-02-003
科技-2015-02-004)
关键词
交通工程
交通流
核密度估计
速度分布
traffic engineering
traffic flow
kernel density estimation
velocity distribution
