论文部分内容阅读
应用相关性理论,研究了交通流数据中缺失值与其他数据的相关性,对与缺失值不同相关性的数据给予不同的权重值,提出了基于交通流时空相关权重的重构算法,并以北京市二环快速路为研究对象,运用VISSIM仿真软件建立仿真模型,利用仿真数据对新算法和现有算法进行了对比分析。研究结果表明:在连续缺失1~10个数据时,模型1的重构值与仿真值平均相对误差最大仅为1.8766%,一般情况下,平均相对误差均在1.0000%以下,可见,模型1算法优于现有的重构算法。
Based on the relevance theory, the correlation between missing values and other data in traffic flow data is studied. Different weight values are given to the data with different correlations with missing values. A reconstruction algorithm based on spatio-temporal correlation weight of traffic flow is proposed. Beijing Second Ring Expressway as the research object, the use of VISSIM simulation software to establish a simulation model, the use of simulation data on the new algorithm and the existing algorithms were analyzed. The results show that the average relative error between the reconstructed value and the simulated value of Model 1 is only 1.8766%, while the average relative error is below 1.0000% in the case of continuous deletion of 1 to 10 data. It can be seen that Model 1 Better than the existing reconstruction algorithm.