Connected Vehicles at Signalized Intersections: Traffic Signal Timing Estimation and Optimization

Summary: While traffic signals ensure safety of conflicting movements at intersections, they also cause much delay, wasted fuel, and tailpipe emissions. Frequent stops and goes induced by a series of traffic lights often frustrates passengers. However, the connectivity provided by connected vehicles applications can improve this situation. A uni-directional traffic signal to vehicle communication can be used to guide the connected vehicles to arrive at green which increases their energy efficiency; and in the first part of the dissertation, we propose a traffic signal phase and timing estimator as a complementary solution in situations where timing information is not available directly from traffic signals or a city’s Traffic Management Center. Another approach for improving the intersection flow is optimizing the timing of traditional traffic signals informed by uni-directional communication from connected vehicles. Nevertheless, one can expect further increase in energy efficiency and intersection flow with bi-directional vehicle-signal communication where signals adjust their timings and vehicles their speeds. Autonomous vehicles can further benefit from traffic signal information because they not only process the incoming information rather effortlessly but also can precisely control their speed and arrival time at a green light. The situation can get even better with 100% penetration of autonomous vehicles since a physical traffic light is not needed anymore. However, the optimal scheduling of the autonomous vehicle arrivals at such intersections remains an open problem. The second part of the dissertation attempts to address the scheduling problem formulation and to show its benefits in microsimulation as well as experiments. Intellectual Merit: In the first part of this research, we study the statistical patterns hidden in the connected vehicle historical data stream in order to estimate a signal’s phase and timing (SPaT). The estimated SPaT data communicated in real-time to connected vehicles can help drivers plan over time the best vehicle velocity profile and route of travel. We use low-frequency probe data streams to show what the minimum achievable is in estimating SPaT. We use a public feed of bus location and velocity data in the city of San Francisco as an example data source. We show it is possible to estimate, fairly accurately, cycle times and duration of reds for pre-timed traffic lights traversed by buses using a few days worth of aggregated bus data. Furthermore, we also estimate the start of greens in real-time by monitoring movement of buses across intersections. The results are encouraging, given that each bus sends an update only sporadically (≈ every 200 meters) and that bus passages are infrequent (every 5-10 minutes). The accuracy of the SPaT estimations are ensured even in