The Identification of Chaos in Theoritical Traffic Flow Based on Improved Surrogate-Data Technique

Abstract

It is important for more effective traffic management and control to study on identification of chaos in traffic flow. The improved surrogate-data technique can avoid the limitation of directly identifying chaos. The study on theoretical traffic flow can avoid various complex factors and easily obtain expected traffic flow states with parameters change and be easy to draw regular conclusions and can more easily verify the effectiveness of the method. Therefore, theoretical traffic flows generated by Bierley car-following model are analyzed. Based on improved implementation of the improved surrogate-data technique, with correlation dimension used as identification evidence, the chaos in the generated time series were identified, and the results were compared with those of the power spectrum method. The experimental results show new algorithm implementation improves the speed of computation, and there is chaos in theoritical traffic flow based on Car-following model, and improved surrogate-data technique can identify chaos exactly.