Evaluation method using chaotic analysis for the model of vehiclesʹ behavior in road traffic system

We have already proposed a simulator of a road traffic system, and called it MITRAM. The MITRAM consists of microscopic models for vehicles which have the capability of their own decision-making through the application of fuzzy logic. This microscopic model is called a fuzzy model vehicle (FMV). We have simulated the driving operation of a following vehicle through the FMV. Data of the following vehicle speed, the relative speed of the following one to the leading vehicle, and so on, were obtained as the simulation data of the FMV. To calculate only the mean and the variance of the simulation data was not sufficient to describe the logic inferences of the FMV. In this paper, we propose an evaluating method by using a chaotic analysis. Correlation exponents and Lyapunov exponents were calculated for the data of the following vehicle speed, the relative speed of the following one to the leading vehicle, and the spacing distance between the following and the leading vehicles. We also calculated both the exponents of the data normalized by the standard deviation, in order to evaluate the effect of the variance. The correlation exponents and the maximum Lyapunov exponents of the relative speed were different from those of the following vehicle speed and the spacing distance. Differences between the simulation and the measurement data were found at both the correlation exponents and the Lyapunov exponents. The variance had a larger effect on the correlation exponents for the simulation data than for the measurement data.