Analysis Of The Complexities Of Time Series In Two-lane Mixed Traffic Flow

With the rapid development of social economy, the number of traffic infrastructures has been greatly increased. However, due to the growth of traffic demand more rapid than the road network capacity, the traffic congestion becomes more and more serious, and the number of traffic accidents is increasing. Nowadays, the traffic problems have been considerd widely as hot and hard issues. It is essential to solve the traffic congest problems by analysis the complexity of traffic flow. The development of the cellular automaton model points out a new direction to study on the traffic flow, and provides a available tool to analysis the global behavior and complex phenomena of traffic flow dynamics theory in the traffic system.The traffic flow is an interdisciplinary in which physics mathematics, computer science, engineering, environmental, science and many other science are used. In this paper, complex networks, detrended fluctuation analysis and cross-correlation coefficient approach are introduced to investigate the complexities of the two-lane traffic flow time series. The results are helpful for us to analysis the traffic flow characteristics and carry out reasonable and effective control measures to alliviate the problems caused by traffic congestion.First, the visibility algorithm is introuduced to map the two-lane traffic flow time series into complex networks. We found that the networks constructed by different vehicle density display the typical characteristics of regular graphs and small-world network. The different ratio of slow vehicle and asymmetry lane-change has no great influence on the characteristics of constructed networks. Additionally, the networks constructed by the blockage traffic time series display the characteristics of small-world.Secondly, the detrended fluctuation analysis method is used to investigate properties of the two-lane traffic flow time series.In addition, the scale index is obtained. According to the changes of the scale index, we find that the traffic flow time series displays long-range correlation characteristics, short-range correlation characteristics and non-power-law relation in the long-range correlation characteristics, which provides a new method to study the complexity of traffic flow space-time evolution.Finally, we further analyze the cross-correlation coefficient of the two-lane traffic flow time series. It is found that the cross-correlation coefficient increase at the first and then decrease with the increase of the vehicle density is increase. It is common even for the difference of ratio of the slow vehicle. The results show that the interrelation of two lanes is determined by the vehicle density. In addition, we also found that that asymmetric lane-change almost have no great influence to the cross-correlation coefficient.