Study On Rapid Identification Of Traffic Flow Chaos By Using Intelligent Technology

To design and study transportation system, on the view of nonlinear science and chaos, is a new kind of scientific approach in traffic control and management. The primary principle of this approach is to identify chaos in time using traffic flow chaos identification system, to achieve suppressing the phenomenon of traffic congestion and improving road capacity by adding chaos control signal. To identify rapidly traffic flow chaos is the foundation of controlling traffic flow chaos. Due to a huge number of time series samples needed, the existing methods to identify traffic flow chaos do not satisfy to the requriements of realistic programme. Therefore, doing research on the means of identifying rapidly traffic flow chaos, that is to say, determining whether traffic flow is chaos or not by adopting only a few time series observation samples, is a much more significant leading-edge in theoretical research and valuable to engineering application.A literature review on traffic flow chaos, this paper further analyzed the principle of intelligence system of rapid identification traffic flow chaos, synthesized the characterristics of engineering application and technological development and further regulated the structure of intelligence system of rapid identification traffic flow chaos, then illustrated the subsystem function of system. On the basis of that, the theoretical research were done on: time headway be chosen as the characteristic of traffic condition different from other scholars selecting traffic flow or traffic density; identication results online mode by selecting different lengths of time series samples, experiment results showed that a small quantity of time series samples could identify traffic flow chaos under the real-time demand; according to the advantages of particle swarm optimization algorithm, intelligent algorithm models optimized by particle swarm optimization algorithm were established, in order to identify traffic flow chaos online mode, the experiment results showed models validity and feasibility.Based on the theoretical research, according to the specialty of rapid identification traffic flow chaos, from the standpoint of engineering application and software development, an intelligent application software was developed. This software on the basis of Visual C++ 6.0 took MFC object-oriented programming technology as the development platform, using Microsoft Office Access 2007 as database, and collected time headway time series to be observation samples, then identified chaos offline mode by calculating the largest Lyapunov exponents and using the intelligent algorithm models optimized by particle swarm optimization algorithm online mode. Experiment results show that this software is a fully functional, clear system structure, user-friendly, object-oriented intelligent system which could identify rapid traffic flow chaos.