一种在线实时快速地判定交通流混沌的组合算法

交通控制的实时性要求高，需要在线实时快速地判定交通流混沌，才可能实现交通流的混沌控制。计算时间序列的最大Lyapunov指数是判定混沌的主要方法。本文提出一种在线实时快速地判定交通流混沌的组合算法。该算法先用关联积分法（C—C方法）确定重构相空间的两个重要参数——嵌入维m和延迟时间τ，再用小数据量方法计算时间序列的最大Lyapunov指数。为检验算法的有效性，首先将算法用于几个最大Lyapunov指数已知的经典混沌系统，比较计算结果；同时对皮埃莱（Bierley）跟驰模型产生的理论交通流时间序列做了仿真试验，计算了其最大Lyapunov指数。实验结果表明这种算法可以用于数据少的交通流时间序列，并且抗噪性好。

A Combined Algorithm for Real-time On-line Rapid Identification of the Chaos in Traffic Flow

On-line traffic control has high real time requirement.It needs to identify the chaos in traffic flow on-line and rapidly realize chaotic control of traffic flow.Computing the maximal Lyapunov exponent from the observed time series is(a main) method to identify chaos.In this paper,a combined algorithm is put forward for real-time online rapid identification of chaos in traffic flow.The algorithm first uses correlation integral method(C-C method) to estimate two important(variances) of phase space reconstruction: embedding dimension m and delay time,then,uses small data sets to calculate the maximal Lyapunov exponent from the time series.To test the validity of the algorithm,it is employed to calculate the(maximal) Lyapunov exponents of several known chaotic system.Furthermore,the maximal Lyapunov exponent of the traffic flow time series which are generated by Bierley Car-following method is calculated.The result shows that the proposed algorithm can be used for traffic flow time series with shorter data and has the anti-noise ability.