基于Lempel-Ziv算法与TOPSIS的水上交通流复杂性测度

发布时间：2019-01-21 16:18

【摘要】：为了定量分析并划分水上交通流的复杂性等级,提出了一种基于Lempel-Ziv算法与TOPSIS的水上交通流复杂性测度方法。运用Lempel-Ziv算法求得实测水上交通流时间序列和对比序列(周期序列、Logistic序列、Henon序列、随机序列)的复杂性特征值,采用TOPSIS得到各个序列的贴近度,根据对比序列的贴近度划分复杂性等级区间,按照水上交通流时间序列的贴近度和所在的复杂性等级区间来表征各个序列的复杂程度,并对长江口南槽航道的实测水上交通流进行复杂性测度。计算结果表明:船舶交通事故数量和下行标准船舶数量与船舶交通流时序复杂性贴近度的相关系数分别为0.698 1、0.769 2,变化趋势基本一致,表明贴近度的计算结果可以反映水域船舶交通流的复杂性;周期序列的贴近度为0.000 1,随机序列的贴近度为0.999 9,Logistic序列和Henon序列的贴近度分别为0.449 2、0.537 7,其值大于周期序列的贴近度,小于随机序列的贴近度;2013年7~11月水上交通流序列的贴近度分别为0.828 0、0.852 7、0.856 5、0.823 7、0.810 7,说明序列的复杂性基本一致;5个月的水上交通流序列的贴近度远大于周期序列的贴近度,处于随机序列和Henon序列的贴近度之间,更接近随机序列的贴近度,说明水上交通流系统不是周期与完全随机的动力学系统;5个月水上交通流复杂性的整体等级为1级,表现出高复杂性的特点。
[Abstract]:In order to quantitatively analyze and classify the complexity of water traffic flow, a method of measuring complexity of water traffic flow based on Lempel-Ziv algorithm and TOPSIS is proposed. The complexity eigenvalues of time series and contrast sequences (periodic sequence, Logistic sequence, Henon sequence, random sequence) are obtained by using Lempel-Ziv algorithm. The closeness degree of each sequence is obtained by TOPSIS. According to the closeness degree of contrast sequence, the complexity degree of each sequence is represented by the closeness degree of water traffic flow time series and the complexity grade interval of each sequence. The complexity of the measured waterway traffic flow in the south channel of the Yangtze Estuary is also measured. The results show that the correlation coefficients between the number of vessel traffic accidents and the number of downlink standard vessels and the complexity closeness of ship traffic flow are 0.698 ~ 0.769 2 respectively, and the variation trend is basically the same. It is shown that the calculation results of the closeness degree can reflect the complexity of the vessel traffic flow in the water area. The closeness of periodic sequence is 0.000, the closeness of random sequence is 0.999 9 and the closeness of Henon sequence is 0.449 2 ~ 0. 537 7, which is larger than that of periodic sequence and less than that of random sequence. The closeness degree of water traffic flow sequence from July to November 2013 is 0.828 / 0 / 0.852 / 7 / 0.856 / 5 / 0.823 / 7 / 0.810 7 respectively, indicating that the complexity of the sequence is basically the same; The closeness of 5-month waterborne traffic flow sequence is much greater than that of periodic sequence, which is between random sequence and Henon sequence, and is closer to the closeness of random sequence. It is shown that the water traffic flow system is not a periodic and completely random dynamic system. The overall complexity of water traffic flow in 5 months is 1, showing the characteristics of high complexity.
【作者单位】：武汉理工大学智能交通系统研究中心;武汉理工大学国家水运安全工程技术研究中心;江汉大学数学与计算机科学学院;代尔夫特理工大学技术政策与管理学院;
【基金】：国家科技支撑计划项目(2015BAG20B00) 国家自然科学基金项目(51579203) 武汉理工大学研究生优秀学位论文培育项目(2016-YS-041)
【分类号】：U692

【相似文献】