Hierarchical Control for Vehicle Repositioning

We introduce a novel hierarchical structure attempting to balance empty vehicle supply and passenger demand efficiently. We cluster the city area into multiple regions initially. The framework consists of multiple control levels: the upper level manages aggregated traffic elements in a centralized manner, e.g. how many vehicles should be relocated to other regions; while the lower level provides position guidance for individual vehicles in each region, e.g. which intersection one empty vehicle should go. This hierarchical structure bridges the gap between macroscopic and microscopic perspectives.

This adaptability enables the testing of alternative algorithms at each layer of the proposed structure, further refining and optimizing our solution for managing autonomous Mobility-on-Demand systems.

More specifically,
• Model-based upper-layer design
We have worked on an upper-layer control employing an aggregated Macroscopic Fundamental Diagram (MFD) model and model predictive control (MPC) to determine optimal vehicle relocation flows between neighboring regions.

We have a published paper on this topic: 
C. Beojone, P. Zhu, I. I. Sirmatel, and N. Geroliminis, "A Two-layer Approach for Rebalancing Ride-hailing Vehicles", 26th IEEE International Conference on Intelligent Transportation Systems (ITSC 2023), Bilbao, Bizkaia, Spain.

• Data-driven upper-layer design:
We implement a novel data-enabled predictive control algorithm. Constructed by collected historical data from the considered unknown system, a non-parametric representation is used to predict future behavior and obtain optimal control actions, circumventing the costly system modeling process.

We have a published paper on this topic:
P. Zhu, G. Ferrari-Trecate and N. Geroliminis, "Data-enabled Predictive Control for Empty Vehicle Rebalancing," 2023 European Control Conference (ECC), Bucharest, Romania, 2023, pp. 1-6, doi: 10.23919/ECC57647.2023.10178140.

Idle-vehicle Rebalancing Coverage Control

The ride-sourcing system can provide passengers with fast and efficient service with a fleet of vehicles, while asymmetry between origin and destination distributions of trips leads to the imbalance between passenger demand and vehicle supply. Thus proactively relocating idle vehicles to the high-demand regions, also known as vehicle rebalancing, is an emerging problem that can significantly improve urban transportation efficiency. We formulate this problem as a Voronoi-based coverage control problem for the coordination and deployment of multiple mobile agents in city scenarios, which vehicles can benefit from by allocating them according to the different demand densities of different city districts.

We have one published paper on this topic:
P. Zhu, I. I. Sirmatel, G. F. Trecate and N. Geroliminis, "Idle-vehicle Rebalancing Coverage Control for Ride-sourcing systems," 2022 European Control Conference (ECC), London, United Kingdom, 2022, pp. 1970-1975, doi: 10.23919/ECC55457.2022.9838069.