The research group Transportation Informatics (TIG) is mainly involved in MACHINE LEARNING, DATA SCIENCE and NEURO-COMPUTING based modeling, simulation, optimization and control in the frame of selected TRANSPORTATION related COMPLEX SYSTEMS, which are: intelligent traffic systems, autonomous and/or networked vehicles, intelligent mobility & logistics systems, and advanced driver assistance systems.
Selected research activities:
- Cellular neural networks based online traffic light state forecast (context: traffic-light assistance systems, a future component of driver assistance Systems)
- Online proactive road safety assessment involving a neuro-computing based driver state forecast
- Driver’s emotion detection and forecast (involving cellular neural networks) for advanced driver assistance systems
- Recurrent/cellular neural networks based local adaptive road traffic control at a road junction
- Cellular neural networks based robust forecast of complex stochastic time-series data (e.g.: traffic data, etc.)
- Solving complex graph theoretical problems with recurrent/cellular neural networks (contexts: nonlinear path costs, stochastic link weights, etc.): optimizations over graphs, traveler salesman problem (TSP), shortest path problem (SPP), vehicle routing problem (VRP), etc.
- Recurrent/cellular neural networks based ultrafast solving of differential equations (contexts, e.g.: online/real-time systems simulation, …)
- Recurrent/cellular neural networks based ultrafast matrix inversion (also of time-varying matrices; contexts, e.g.: complex systems simulation, sensor networks, real-time systems, etc.)
- Recurrent/cellular neural networks based compression of images and sensor data
- Recurrent/cellular neural networks based robust control of nonlinear systems (e.g. in adaptive traffic control, robotics, etc.)
- Solving inverse problems (e.g.: online system identification, reliability modeling & forecast, etc.) by involving cellular neural networks