Inhalt des Dokuments
Traffic signal control can be seen as a backbone in the organization of traffic flows in urban areas. A signalized intersection consists of several traffic lights, which regulate the right of way for each direction at an intersection. For optimizing a system of traffic signals, one may vary different parameters like green split, cycle time, or offset. The green split specifies the fraction of green time for each approach of a signalized junction. The cycle time identifies the period of the signal, while the offset determines the coordination between different traffic lights. Since changing traffic signal parameters influences the travel times on paths passing the particular signal, traffic signal optimization directly interacts with traffic assignment and route choice of road users.
This project analyses the interaction between traffic signal control and users’ behavior from two perspectives. On the one hand it concentrates on the optimization of traffic assignment and signal control in a strict mathematical programming approach. On the other hand it studies the implications of signal control optimization strategies on travelers’ behavior using advanced traffic simulation tools. The mathematical approach uses an integrated, cyclically time-dependent optimization model, which calculates traffic signal control, red-green split and traffic assignment simultaneously, in order to design efficient fixed time signal plans for networks of inner city traffic. The quality of the solutions provided by the optimization model for non-adaptive signals are then assessed by simulation and compared to state of the art adaptive signal control approaches. An important element is to assess how demand reacts to those changes. For simulation, the microscopic, dynamic simulation MATSim is used.
Prof. Dr. Kai Nagel , TU Berlin
Dominik Grether , TU Berlin
Theresa Thunig , TU Berlin