Few-Shot Learning for Dynamic Operations of Automated Electric Taxi Fleets under Evolving Charging Infrastructure: A Meta-Deep Reinforcement Learning Approach
This addresses a critical problem for urban transportation systems by enabling more efficient and adaptive fleet operations in evolving charging environments, representing a novel method for a known bottleneck rather than a foundational advancement.
The paper tackles the challenge of managing autonomous electric taxi fleets under dynamic and uncertain charging infrastructure by proposing GAT-PEARL, a meta-reinforcement learning framework that integrates graph attention networks and probabilistic embeddings, demonstrating superior generalization to unseen layouts and higher operational efficiency in simulations on real-world data from Chengdu, China.
With the rapid expansion of electric vehicles (EVs) and charging infrastructure, the effective management of Autonomous Electric Taxi (AET) fleets faces a critical challenge in environments with dynamic and uncertain charging availability. While most existing research assumes a static charging network, this simplification creates a significant gap between theoretical models and real-world operations. To bridge this gap, we propose GAT-PEARL, a novel meta-reinforcement learning framework that learns an adaptive operational policy. Our approach integrates a graph attention network (GAT) to effectively extract robust spatial representations under infrastructure layouts and model the complex spatiotemporal relationships of the urban environment, and employs probabilistic embeddings for actor-critic reinforcement learning (PEARL) to enable rapid, inference-based adaptation to changes in charging network layouts without retraining. Through extensive simulations on real-world data in Chengdu, China, we demonstrate that GAT-PEARL significantly outperforms conventional reinforcement learning baselines, showing superior generalization to unseen infrastructure layouts and achieving higher overall operational efficiency in dynamic settings.