Stochastic urban EV load forecasting: integrating markov-monte carlo simulation with spatio-temporal and traffic-weather dynamics

Urban electric vehicle (EV) charging load exhibits complex stochastic characteristics influenced by multiple factors, including weather conditions, day types, and traffic radiation effects from surrounding cities. This complexity poses significant challenges for accurate load prediction and infrastructure planning. This paper proposes a novel stochastic prediction framework that synergistically integrates these multiple influencing factors through a sophisticated Markov-Monte Carlo fusion approach. The methodology first establishes a weather-aware Markov state transition matrix that captures EV user behavior patterns under varying meteorological conditions and day types. This is then enhanced by incorporating a traffic radiation model that quantifies the impact of EV flows from surrounding cities on the central urban area’s charging demand. The framework employs Monte Carlo simulation to generate probabilistic charging scenarios, enabling comprehensive prediction of charging load distributions across different functional zones. Case studies demonstrate that the proposed method achieves superior prediction accuracy, with root mean square error (RMSE) reduced by 26.3% and 3.8% compared to existing methods. The model’s ability to capture complex spatio-temporal dynamics and traffic radiation effects makes it particularly valuable for urban charging infrastructure planning and power grid operation optimization. This research advances the field of urban EV charging demand forecasting by providing a more comprehensive and accurate prediction framework that better reflects real-world charging patterns and supports sustainable urban transportation development. © 2017 Elsevier Inc. All rights reserved.