Database-Mediated Preview of Roadway Friction and Model Predictive path Tracking Control for Connected Vehicles
Restricted (Penn State Only)
- Author:
- Gao, Liming
- Graduate Program:
- Mechanical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- September 13, 2022
- Committee Members:
- Daniel Haworth, Professor in Charge/Director of Graduate Studies
Minghui Zhu, Outside Unit & Field Member
Bo Cheng, Major Field Member
Sean Brennan, Chair & Dissertation Advisor
Herschel Pangborn, Major Field Member
Craig Beal, Special Member - Keywords:
- Connected vehicle simulation
Database
Vehicle dynamics
Vehicle path tracking control
Model predictive control
Road-tire friction - Abstract:
- The primary focus of this dissertation is to preview roadway friction via database-mediated connected and autonomous vehicles (CAVs) and develop a method to incorporate this previewed information into vehicle path tracking control for improved performance. Tracking a target-planned path – a lane centerline of a highway, a lane-changing maneuver, or a trajectory for obstacle avoidance – is one of the most challenging tasks of vehicle driving. Due to the lack of information, current vehicle control systems generally assume that the road friction conditions ahead of a vehicle are unchanged relative to the conditions at the vehicle’s current position. This can result in dangerous situations if the friction is suddenly decreasing from the current situation or overly conservative driving styles if the friction of the current situation is worse than the roadway ahead. Future driving systems must go further so that they are capable of maneuvering even on unfavorable road conditions, for example, tracking sharp turning paths on the road with a sudden decrease in friction. This may be enabled by using new technologies, for example, the connectivity of CAVs, that can provide information about the environment, particularly the friction between vehicle tires and the road surface. Therefore, the challenge is to find a way to aggregate the data from CAVs for roadway friction preview and incorporate previewed friction information to improve vehicle path tracking performance. Specifically, the challenge in the creation of road friction preview maps is the very large quantity of data involved, and the measurements populating the map are generated by vehicle trajectories that do not uniformly cover the grid in situations of varying road surface friction. Furthermore, even if road conditions are known, incorporating the information into the path tracking control system is a challenge on its own. To incorporate previewed roadway friction information into the vehicle path tracking control, a systematic approach to the analysis and development of controllers is needed. The key contributions of this dissertation are: (1) a micro-simulation framework for studying the CAVs control and road friction preview based on a database-mediated data sharing system; (2) a road friction map generation strategy that aggregates the measured road-tire friction coefficients along the individual trajectories of CAVs into a road surface grid; (3) a vehicle longitudinal speed planning algorithm according to the previewed roadway friction and path geometry constraints; (4) a model predictive path tracking control structure that utilizes preview friction to achieve tracking accuracy and stability near the vehicle dynamic limits.