Using Multiscale Entropy Method to Analyze the Complexity of Traffic Flow

Multiscale entropy (MSE) analysis is a method of measuring the complexity of time series across multiple time scales. The complexity of traffic flow time series generated by NS model is analyzed with MSE. The results show that both of the randomization probability and vehicle density affect the complexity of time headway. And when we focus on large time scale, we found that with the same randomization probabilities, the complexity is at a stable and high level in higher vehicle density scenarios. However, with the same randomization probabilities, the complexity will decrease as time scale increases in lower vehicle density scenarios.