The research objective of this award is to develop methods for modeling, analysis, and design of dynamical systems with stochastic delays and apply these to connected vehicle systems. Random delays arise due to intermittencies and packet drops in wireless vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications and can lead to instabilities that degrade the performance of multi-vehicle systems. Using a system-level approach the research will result in mathematical tools for control design with stochastically delayed feedback that can be executed by the vehicles in a decentralized way. This will allow the realization of Lagrangian traffic control where the temporal responses of semi-autonomous vehicles are shaped using spatial and temporal anticipation and memory so that maximal traffic throughput is achieved without compromising safety. Deliverables also include experimental realization of a connected vehicle system in a scaled experiment using ground robots which will also be used in engineering student education.
If successful, the results of this research will provide opportunities to increase the efficiency of road transportation by ensuring smooth vehicular flows. This will reduce travel times, improve fuel-efficiency, and improve traffic safety. The design principles established go beyond the specific problem of connected vehicles and allow the design of network dynamics with stoshastically varying communication delays. The results will be disseminated to allow the realization of full-size vehicles that can exploit the connected vehicle environment given by wireless communication. Graduate and undergraduate engineering students will benefit from experimental classroom demonstrations and involvement in the research. Outreach activities will be used to inspire under-represented minority students from urban school systems to learn science and engineering through challenging and exciting hands-on experiences.
