Modeling and Control of Information-Driven Smart Transportation Systems
- View All
Next generation of intelligent transportation systems (ITS) are expected to heavily rely on connected and automated vehicles. As automated components of ITS grow, the system will evolve into a mix of manned and automated systems. This proposal aims at developing some of the main engineering principles of such systems. We have identified modeling and control of these mixed manned/autonomous connected systems as an important step in that direction. Connected vehicle concepts rely on the use of communication for gaining situational awareness for the purpose of making transportation safer and more efficient. Cooperative collision avoidance and network based routing mechanisms are examples of such concepts. While these concepts are useful on their own, a full view of the system is not complete without considering how the reactions of the human drivers or automated driving systems to the available information change the system dynamics. Such changes result in a new system that in many cases may not exhibit the same dynamics that the original system controllers were designed for. For example, the availability or resolution of real-time vehicle movement information in a highway will change the way human drivers or automated systems respond. The result is a system that may not follow the currently known microscopic or macroscopic traffic laws, and may hinder the optimality of adaptive cruise control algorithms or the safety of collision avoidance systems. The effect is in particular more profound and less known in large scale systems where dynamics of vehicles are intrinsically and physically tied over large distances. This proposal aims at closing the loop around communication, control and physical system dynamics of the system by determining how small and large scale system dynamics change in response to availability or lack of information, and how such dynamics require information for sustaining a stable and functional system. We will derive the models that describe the mutual coupling of information networks and large and small scale transportation system dynamics. Such models will then allow the design of control mechanisms for safe and optimal operation of such systems. An important component of this research is a co-simulation tool for large scale simulation of connected transportation systems and communication networks. A test-bed for small scale experimentation will also be developed and used as a validation platform.