Real-time multi-objective charging scheduling unlocks efficient renewable integration and grid-friendly electric vehicle operations at public stations

The global surge in electric vehicle adoption stands as a cornerstone of efforts to combat climate change and reduce reliance on fossil fuels, with EVs poised to play a transformative role when paired with renewable energy sources. Public charging stations, as critical nodes in this ecosystem, face mounting pressure from uncoordinated large-scale charging that risks overloading local grids, driving up peak demand, and hindering the seamless incorporation of intermittent solar and wind power. While grid upgrades offer one path forward, their prohibitive costs make smarter management of existing infrastructure far more practical. The extended dwell times typical of most charging sessions—often far exceeding the actual energy transfer needed—create valuable flexibility for optimized scheduling, enabling EVs to act as distributed energy resources through Vehicle-to-Grid interactions that stabilize grids and maximize clean energy utilization.

Building on real-world insights, researchers have developed a sophisticated real-time charging scheduling framework tailored for public charging stations equipped with micro-grids and renewable generation. The approach begins with a comprehensive EV charging behavior database derived from 329,632 actual charging records collected across 1,268 stations in Beijing from January to mid-February 2023. This dataset captures the nuanced temporal uncertainties and distinct patterns of fast- and slow-charging on weekdays versus weekends, providing a robust foundation far more representative than idealized or statistically simplified models. A novel charging pile allocation mechanism then dynamically assigns available chargers to arriving vehicles, aiming to maximize pile utilization, minimize wait times, reduce the likelihood of drivers abandoning sessions, and determine an optimal vehicle-to-charger ratio under realistic capacity constraints—addressing a frequent oversight in prior studies that assumed unlimited infrastructure availability.

At the core of the scheme lies a multi-objective optimization model that simultaneously pursues grid stability, economic advantages for stakeholders, and enhanced renewable energy absorption. The framework incorporates efficient V2G capabilities, allowing parked EVs to discharge power back to the micro-grid during periods of high renewable output or grid stress. To handle the dynamic, uncertain nature of real operations—fluctuating renewable generation, variable arrival patterns, and shifting grid conditions—a sliding window mechanism continuously updates charging and discharging plans in real time, bridging individual vehicle behaviors with station-wide goals. The Non-dominated Sorting Genetic Algorithm II generates a Pareto front of trade-off solutions, while the Entropy-TOPSIS method selects the most balanced compromise, ensuring equitable consideration of conflicting priorities without arbitrary weighting.

Simulation results affirm the scheme’s strong all-around performance. Compared to uncoordinated or benchmark approaches, the proposed method markedly reduces distribution network load fluctuations, lowers average charging costs for users, and shrinks real-time mismatches between energy consumption and renewable availability—outcomes that translate directly into smoother grid operations, reduced peak infrastructure strain, and higher effective utilization of clean energy. By cutting waiting times and abandonment rates, the charging pile allocation component also improves user experience and station throughput, fostering greater confidence in public EV infrastructure.

The implications reach well beyond individual stations. This data-driven, adaptive scheduling approach offers a scalable pathway to integrate growing EV fleets with expanding renewable capacity, supporting carbon neutrality targets while deferring expensive grid reinforcements. It empowers charging operators to balance financial viability with environmental responsibility and grid reliability, creating win-win scenarios across utilities, station owners, drivers, and society.

Looking ahead, the framework holds considerable promise for broader deployment and enhancement. Future extensions could incorporate real-time traffic flow predictions to anticipate arrival surges, account for distribution network capacity limits, or integrate advanced forecasting of renewable output using machine learning. Scaling to city-wide networks of interconnected stations or coupling with demand-response programs could further amplify benefits. Real-world pilot implementations will be key to validating transferability and refining user acceptance.

In summary, this innovative real-time multi-objective charging scheduling strategy represents a meaningful advance in managing the complex interplay between EV growth, renewable integration, and grid stability. Grounded in extensive real-world data and designed for practical dynamism, it delivers measurable improvements in efficiency, cost, and sustainability—paving the way for public charging ecosystems that not only accommodate the EV revolution but actively accelerate the transition to a cleaner, more resilient energy future.

Reference

Author: Lei Zhang ^a b, Yingjun Ji ^a b, Xiaohui Li ^a b, Zhijia Huang ^a b, Dingsong Cui ^c, Haibo Chen ^c, Jingyu Gong ^d, Fabian Breer ^d, Mark Junker ^d, Dirk Uwe Sauer ^d

Title of original paper: Multi-objective charging scheduling for electric vehicles at charging stations with renewable energy generation

Article link: https://www.sciencedirect.com/science/article/pii/S2773153725000337

Journal: Green Energy and Intelligent Transportation

DOI: 10.1016/j.geits.2025.100283

Affiliations: ^a Collaborative Innovation Center for Electric Vehicles in Beijing, Beijing Institute of Technology, Beijing 100081, China

^b National Engineering Research Center for Electric Vehicles, Beijing Institute of Technology, Beijing 100081, China

^c Institute for Transport Studies, University of Leeds, 34-40 University Road, Leeds LS2 9JT, UK

^d Institute for Power Electronics and Electrical Drives, RWTH Aachen University, 52056 Aachen, Germany