Increasing Electric Vehicles Utilization in Transit Fleets using Learning, Predictions, Optimization, and Automation

Jacopo Guanetti,Yeojun Kim,Xu Shen,Joel Donham,Santosh Alexander,Bruce Wootton,Francesco Borrelli

2023-05-24

Abstract:This work presents a novel hierarchical approach to increase Battery Electric Buses (BEBs) utilization in transit fleets. The proposed approach relies on three key components. A learning-based BEB digital twin cloud platform is used to accurately predict BEB charge consumption on a per vehicle, per driver, and per route basis, and accurately predict the time-to-charge BEB batteries to any level. These predictions are then used by a Predictive Block Assignment module to maximize the BEB fleet utilization. This module computes the optimal BEB daily assignment and charge management strategy. A Depot Parking and Charging Queue Management module is used to autonomously park and charge the vehicles based on their charging demands. The paper discusses the technical approach and benefits of each level in architecture and concludes with a realistic simulations study. The study shows that if our approach is employed, BEB fleet utilization can increase by 50% compared to state-of-the-art methods.

Systems and Control

What problem does this paper attempt to address?

The problem this paper attempts to address is improving the utilization of Battery Electric Buses (BEBs) in bus fleets. Specifically, the researchers propose a hierarchical approach based on learning, prediction, optimization, and automation to tackle the following challenges:

1. **Uncertainty in charging time and energy consumption**: The range and charging time of electric buses are affected by various factors such as temperature, road conditions, driving behavior, etc. These uncertainties and variabilities make it difficult for bus companies to plan and operate BEBs.
2. **Efficient utilization of BEBs**: Existing methods cannot accurately predict the energy consumption and charging time of each BEB under different routes, different drivers, and different conditions, leading to low utilization rates of BEBs and high operating costs.
3. **Optimized management of charging infrastructure**: How to reasonably schedule the charging time and order of BEBs under limited charging facilities to ensure the efficient operation of the fleet.

To address these issues, the paper proposes a hierarchical architecture comprising three key components:

1. **ZEV Digital Twin Module**: By constructing high-precision prediction models through learning algorithms, it can accurately predict the energy consumption and charging time of each BEB under specific routes, specific drivers, and specific conditions.
2. **Predictive Block Allocation Module**: Utilizing the above prediction results, it optimizes the daily task allocation and charging management strategies of BEBs to maximize their utilization.
3. **Parking and Charging Queue Management Module**: Based on the charging needs of the vehicles, it autonomously parks and charges the vehicles to ensure they can complete their charging tasks on time.

Through this series of technical means, the researchers hope to significantly improve the utilization of BEBs in practical applications, reduce operating costs, and promote the transition of public transportation to zero-emission vehicles.