Interpretation of Intelligent Transport Systems Data in Moscow for Bernoulli Approximation of Deterministic-Stochastic Model

The BigData technology development allows to measure traffic flows on highways in detail. The flow dynamics depends on many parameters, in particular social and economical parameters. Based on the results of measurements on highways in Deutschland, B.S. Kerner has concluded that the known mathematical traffic models are not always adequate. The congested flows are very unstable, and, therefore, it is important to know the real function of state, which shows the dependence of the flow intensity on the density. In Moscow, an intellectual transport system has been created. This system collects the data on the flow characteristics. In this paper, an algorithm has been developed such that this algorithm calibrates a traffic mathematical model. This model is based on the deterministic-stochastic approach and concepts of synchronous or asynchronous exclusion processes. The model parameters are set based on data of measurements made on a section of Leningradsky prospect in Moscow with the intellectual transport system SS125 - Traffic Sensor Smartsensor Wavetronix. Data of year 2011 are not very accurate. However, these data contain information on the intensity, density, velocity and vehicle types for any minute segment during a day. The data of measurements made in 2019 are more accurate but only total traffic intensity in each direction were measured. The algorithm, developed in this paper, is used for the estimation of the traffic density and velocity dynamics based on the data of 2019 year. We consider a version of homogeneous flow model and two versions of heterogeneous flow model. One of these versions is a new model.