Day-ahead distributed robust scheduling of active distribution networks considering source-load uncertainties and large-scale electric vehicle integration

• Establish an uncertainty model for EV response, transforming EV load uncertainty from a static probability distribution into a dynamic mechanism based on electricity price signals, while incorporating uncertainties such as charging station congestion. This enables the model to more accurately reflect user charging behavior, electricity price signals, and infrastructure operational responses. • A robust scheduling method for active distribution networks is proposed that accounts for source-load uncertainty and EV participation. Based on uncertainty sets for source-load errors and EV user response, a two-stage distributed robust optimization model is constructed. Validation demonstrates improved system economics and robustness. • By employing a method that dynamically adjusts the fuzzy set radius to fuse affine strategies with dual theory, we solve a distribution-robust scheduling model, transforming the nonlinear problem into a linear one and improves solution efficiency. The flexible large-scale development and grid integration of new energy sources, such as photovoltaic (PV) and wind power, along with electric vehicles (EVs), pose significant challenges to grid operation due to their inherent randomness and volatility. To enhance the adaptability of day-ahead dispatch decisions under uncertainty, this study proposes an active day-ahead distributed robust (DRO) dispatch method for distribution networks that integrates source-load uncertainties and incorporates EV participation. First, an active distribution network is established, along with an EV demand response model that accounts for uncertainty grounded in time-of-use (TOU) pricing and an elasticity coefficient matrix. Second, using day-ahead forecast data, a robust day-ahead distribution scheduling model for the active distribution network is developed, incorporating both source-load uncertainty and EV participation. A fuzzy set that contains the true distribution is defined employing the Wasserstein distance. Affine strategies and dual theory are employed to facilitate the linear transformation and subsequent solution of the proposed model. Finally, a comprehensive case study on an adapted IEEE 33-bus benchmark system is conducted to validate the effectiveness of the proposed solution. A cost reduction of nearly 4 % was achieved, confirming the efficacy of the proposed methodology in the case study. System robustness and economic efficiency can be balanced by adjusting the fuzzy set radius. The EV dynamic response mechanism successfully achieves peak-valley load shifting, mitigates node voltage over-limit and load loss phenomena, and enhances wind and solar power absorption rates, thereby improving overall system resource utilization.