The global transition towards sustainable transportation marks a critical shift in addressing energy shortages and climate concerns. As electric vehicles (EVs) proliferate, with electric buses (EBs) leading the charge, they present a mosaic of opportunities and challenges for energy storage and power grid stability.
A staggering growth in EV usage intersects with the complexity of integrating these vehicles into power systems designed primarily for traditional loads. Noteworthy is the rapidly increasing role of EBs, which, despite their environmental benefits, impose substantial demands on infrastructure due to their high charging requirements and predictable energy consumption patterns.
The core challenge lies in efficiently predicting and managing the energy demands of electric buses within the power grid framework. This challenge is magnified by the predictable yet large-scale energy needs of EBs that necessitate significant scheduling efforts. Therefore, integrating a flexible energy storage model that aggregates individual charging demands into a cohesive system is pivotal. Through such models, the potential for EBs to offer services such as peak shaving and load balancing enhances the power grid’s ability to integrate renewable energy sources.
A closer look at the predictive models reveals an interesting landscape of data-driven and physical model-driven approaches. An adaptive graph convolutional network is a notable innovation, enhancing travel time predictions by incorporating spatiotemporal dynamics. Such models improve accuracy by considering variables such as traffic congestion and passenger numbers—elements that markedly affect the scheduling and energy availability of EBs. The adaptive nature of this network aligns with the dynamic, ever-changing conditions of urban transportation, as demonstrated by case studies where it achieved a mean absolute error (MAE) of 16.33 seconds in arrival predictions, outperforming traditional models.
Contrasting with the data-centric approach is the physical model that forecasts energy consumption using vehicle dynamics equations. This model, underpinned by first-order thermodynamic principles, provides insights into energy flow and consumption factors, offering a lens through which the intricate nuances of traction and air conditioning energy use are examined. This dual methodology—data-driven on one hand, and physical on the other—underscores the complexity but the necessity of understanding energy dynamics in detail.
Amidst these complexities, the need for a rigorous approach to evaluating battery degradation emerges. By designing test profiles to investigate the impact of various temperatures on battery life, researchers have laid down a pathway to understand long-term efficiency and sustainability. The aggregated energy storage’s variability—by as much as 27.65% under different temperature scenarios—signals both the potential for optimized scheduling and the risks of inefficiencies.
As the electricity market becomes more deregulated, the role of EBs transitions beyond mere transportation to active participants in energy management. This paradigm encompasses demand response programs and emphasizes the potential of EBs to act as mobile energy storage units, contributing to a more stable and renewable-energy-friendly grid. However, as robust as these models are, the complexity of centralized scheduling for numerous EBs persists, highlighting the need for continued research and optimization.
