TRaffic Interaction with Platoon (TRIP): designing controllers for safe trajectory planning |
|
|
Student(en): |
Betreuer: Ashish Kumar Cherukuri, John Lygeros |
|
Beschreibung: Introduction: Platoons of heavy-duty vehicles is going to be an essential part of future freight transportation. In a platoon, a group of vehicles move in one lane of the highway for long distances while maintaining close distance to each other. This coordinated driving provides significant fuel benefits for these vehicles. In a sense, platooning is environment friendly due to this improvement in fuel efficiency. While various control algorithms have addressed the problem of maintaining safe platoon formation, none of these have analyzed the effect of the platoon on vehicles moving around it. The long chain of vehicles certainly limits the maneuverability of vehicles running alongside, e.g., changing lanes become difficult. The overall goal is to investigate possible solutions to this problem from various viewpoints. Project aim: The objective of this particular project is to design algorithms that efficiently integrate platoons and the surrounding traffic. The designed actions for the platoon and other vehicles will ensure efficient operation of the platoon and allow "weaving" of vehicles through the platoon in order to make them reach their destination efficiently. The actions of the platoon consist of regulating the speed, opening-up or closing-down gaps with neighboring trucks, and regulating the distance to surrounding vehicles. The actions of the vehicle involve regulating the speed and changing lanes. Project outline: The first step is to familiarize with the literature on platoons and automated vehicles. Next step is to come up with a practical problem formulation. This will involve codifying the requirements of each vehicle in terms of temporal logic specifications. Given this problem formulation, the student will design control strategies that have three properties. First is tractability. This means the solution can be obtained in reasonable amount of time as decision making has to be done in real-time. Second is robustness. That is, the algorithm performs "reasonably" even in the presence of perturbations, such as error in control implementation, modelling uncertainty, etc. Third is safety. This means providing bounds on the safety margins of the trajectories generated by the controller. Finally, the controllers will be simulated to validate the theoretical guarantees. Weitere Informationen |
Professor: John Lygeros |
|
Projektcharakteristik: Typ: Art der Arbeit: 80% Theoretical, 20% simulation Voraussetzungen: Sufficient mathematical maturity and solid background on controls is required. A firm hold on concepts of linear algebra, dynamical systems, analysis, and optimization will be an advantage. Validation of the designed algorithms will require coding skills (for example, MATLAB). | |
|
Anzahl StudentInnen: Status: taken | |
|
Projektstart: February 2018 Semester: Spring 2018 | |
 |
|
|