sTC-net – Subsidiary Traffic Control Network
Traffic Control using Example-based Modelling
The sTC-net team works on a new approach for an “optimal” traffic control solution. Yet it is very difficult to find a solution, which satisfies the needs of all the relevant stakeholders and participants. The reasons are manifold: First of all the objects to be controlled (the car drivers) are human beings, who are acting with a high degree of anticipation with respect to their individual wishes and needs. This leads to an inherent stochastic control problem, which cannot be handled easily by conventional analytical methods and which requires a high flexibility and adaptability. Furthermore it is hard to quantify the control requirements and to measure or identify the quality respectively the performance of regional traffic in a general way. Nevertheless this is crucial for the application of optimization methods and strategies.
To face these complexities, Andata and Salzburg Research are developing a conceptual new approach in “intelligent” traffic control. The proposed solution is a skilled combination of selected techniques from the field of “Computational Intelligence”, a hierarchical, subsidiary structure of modular, locally intelligent control elements. These elements are based on object oriented design and intensive use of stochastic simulation, as well as novel sensors and data fusion techniques. The modular control objects can be arranged vertically (across multiple hierarchical levels) and horizontally (across the regional extension) to an overall traffic control network, which constitutes to a distributed, partially self organized intelligent system.