Optimizing hybrid microgrids based battery swapping and charging station for UAV operations in remote areas with considering multiple weather uncertainties

Leveraging agility, maneuverability, and aerial accessibility, unmanned aerial vehicles (UAVs) demonstrate tremendous potential in remote-area applications like patrol, remote sensing, and logistics. Since remote areas lack adequate grid coverage that supports autonomous operation of UAVs, microgrids (MGs) based battery swapping and charging station (MGBSCS) emerge as a technically appealing solution. Yet, the operations of both UAVs and microgrids are under the influence of microclimate variations, which introduce uncertainties that pose notable challenges to the energy system configuration (ESC) for MGBSCS. To address ESC of UAVs-oriented MGBSCS under multiple uncertainties, this paper proposes a copula scenario tree based two-stage robust optimization (TRO) method. Specifically, multiple weather uncertainties (i.e., wind fields, solar irradiance, and temperature) are incorporated into UAVs energy consumption and MGs power generation modeling, with their correlations captured through a 3D Gaussian copula. Furthermore, a UAV task network is established to simulate energy demand distribution, complemented by SOC-based wireless fast charging (WFC) and battery-swapping (BS) strategies that link UAVs operations with MGBSCS. The system-coupled model is transformed into a mixed-integer linear programming (MILP) problem and subsequently extended into a TRO problem, which is solved by a scenario clustering generation tree-based column-and-constraint generation (SCGTC&CG) algorithm. Based on real-world meteorological data, which cover five remote areas with heterogeneous climates (i.e., North Dakota, Atacama Desert, Utqiaġvik, Sahara Desert, and Kolkata), multi-scenario comparative experiments are conducted to validate the effectiveness and robustness of the proposed method. The experimental results demonstrate the proposed method achieves a modeling error under 5.71 % in uncertainty correlation while yielding over 4 % cost reduction compared to the baselines, 49.46 % computation time savings compared to direct nonlinear programming solvers and achieves annual carbon emission reduction of 251.7 tons. This study demonstrates that UAVs-oriented MGBSCS offers substantial environmental benefits, with the TRO effectively enhancing the cost-effectiveness by modeling the correlations of weather uncertainties.