Transport flow model based on interaction of particles with action potential

The current booming development of driverless cars raises the question of assessing the impact of their penitration on the traffic flow, which makes the task of creating a mixed model relevant. The object of the research is the traffic flow. The goal is to create an adequate model based on the idea of action potential, in which cars are represented as particles attracting or repelling each other at a certain distance. Existing software products implementing various microscopic models of traffic flow are considered. There is described proposed mathematical model including two kinds of cars: driverless cars strictly obeying the rules and cars with real drivers who allow speeding and random changes of velocity. A program has been developed that implements the created model and allows changing its parameters. The macroscopic characteristics of the simulated flow have been compared with the available values of analogues and works of other researchers. The behavior of simulated cars when driving on a single-lane circular road, a two-lane circular road, and a single-lane figure-of-eight with crossing roads is considered. The results of model runs for mixed flow are presented.