基于近似熵和统计复杂度的交通流复杂性测度

为了定量描述交通流系统的复杂性,引入非线性动力学的近似熵和统计复杂度特征量,计算混沌区域Henon和Logistic序列、周期序列、白噪声序列的近似熵和统计复杂度,并通过计算5个实测交通流时间序列数据对其近似熵和统计复杂度进行了计算分析。结果表明:这2种特征量能够有效测度系统的随机性和非线性,可以解决复杂性刻画的"短序列、可比较"难题,适于实测交通流的复杂性测度;交通流系统随机性较强,带有非线性结构,可能产生混沌,应该综合使用多种复杂性测度方法。

Complexity Measure of Traffic Flow Based on Approximate Entropy and Statistical Complexity

In order to quantitatively describe the complexity of traffic flow system,hence two eigen indexes of nonlinear dynamics,ie approximate entropy and statistical complexity were introduced.The approximate entropies and statistical complexities of periodic sequence,white noise sequence,the Henon and Logistic sequences which were situated in the chaotic region,were calculated.The approximate entropies and statistical complexities of five actual time series of traffic flow were calculated.The results show that the two eigen indexes can be applied to effectively measure the randomicity and nonlinearity of a system and resolve the difficult problems of short sequence and comparability in depicting the complexity of a system.They are suitable to measure the complexity of the actual traffic flow.The randomicity of traffic flow system is high and there are some nonlinear structures in the traffic flow system which probably can generate chaos.Manifold methods should be used to measure the complexity of a system.