Towards a dynamic system optimum based on the simulated traffic data in the microscopic traffic simulation

Kurzfassung

Microscopic traffic simulation has been applied since decades in order to better describing both drivers’ behaviors and interactive effects among network infrastructure, drivers and traffic control applications. Furthermore, it is also used as an evaluation tool for analyzing influences of proposed management strategies and traffic-related telematics technologies on network performances, such as efficiency and safety. Achieving a system optimum in a road network is the main concern of traffic managers at all times, although most road users tend to make the route choice decision which is best suitable for their journeys in practice. The main difference between system optimum and user equilibrium is the marginal total travel costs, i.e. travel times, which are the costs that an additional road user causes to the other road users already in the network during the analysis period. Generally, travel times can be determined, i.e. approximated, by given link travel time functions, which are functions of link flows. The functional forms and respective parameters of the most travel time functions are derived from empirical data. This approach has been extensively applied in the macroscopic traffic modeling and the dynamic traffic assignment modeling. Therefore, respective marginal costs can be obtained by calculating the corresponding derivatives. However, such travel time functions and their derivates are not required and also not applied in a microscopic simulation, since the travel time, travel flows and other parameters are directly measured in a simulation. In this study, how to define and calculate marginal costs with use of the simulated data is investigated.