During the overall design phase of solar-powered unmanned aerial vehicles (UAVs), a large amount of high-fidelity (HF) propeller aerodynamic performance data is required to enhance design performance, but the acquisition cost is prohibitively expensive. To improve model accuracy and reduce modeling costs, this paper constructs a multi-fidelity aerodynamic data fusion model by associating data with different fidelity. This model utilizes a low-fidelity computational method to quickly determine the design space. The constrained Latin hypercube sampling based on the successive local enumeration (SLE-CLHS) method and the expected improvement (EI) criterion were adopted to achieve the efficient initialization and fastest convergence of the Co-Kriging surrogate model within the design space. This modeling framework was applied to acquire the aerodynamic performance of high-altitude propellers, and the model was evaluated using various performance indicators. The results demonstrate that the proposed model has excellent predictive performance. Specifically, when the surrogate model was constructed using 350 high-fidelity samples, there were improvements of 13.727%, 12.241%, and 5.484% for thrust, torque, and efficiency compared with the surrogate model constructed from low-fidelity samples.

High-altitude propellers experience diverse aerodynamic conditions, due to the extensive altitude variation and wide range of wind speeds encountered in the flight envelope of High Altitude Long Endurance (HALE) aircraft. Developing a high-precision aerodynamic model requires extensive high-fidelity data. This data is indispensable for aircraft design and control but incurs significant costs. According to the extensive altitude range of HALE aircraft, this paper proposes two efficient and cost-effective modeling methods, the Global Surrogate Model (GSM) and the Stratified Interpolation Surrogate Model (SISM), to acquire the aerodynamic data. Both models are assessed using different performance metrics, including the mean relative error (MRE) and the maximum relative error (MARE), to evaluate their predictive capabilities. The comparative results reveal that the aerodynamic characteristics of high-altitude propellers exhibit highly nonlinear trends in response to changes in altitude. In this study, the two surrogate model construction methods are employed, and it is found that SISM offers superior model accuracy. Specifically, when compared to GSM with an equal number of high-fidelity samples, SISM demonstrates a 2.16% increase in model accuracy. These findings provide a cost-effective approach for HALE propeller modeling.

The high computational cost of high-fidelity CFD simulations forms a major bottleneck in aerodynamic design. This paper introduces a multi-fidelity transfer learning framework to rapidly predict airfoil aerodynamics with high accuracy. Our approach involves pre-training a deep fully connected neural network on a large dataset of low-fidelity Euler simulations. The pre-trained model is then fine-tuned using a limited set of high-fidelity RANS data, enabling efficient knowledge transfer from low- to high-fidelity domains. A specialized logarithmic-exponential normalization method is developed to handle the scale differences between aerodynamic coefficients. The framework demonstrates exceptional performance: after fine-tuning with only 700 high-fidelity samples, the model accurately predicts pressure distributions (lowest RMSE = 0.053) and force coefficients (R² > 0.947 for lift and drag). This method successfully bridges the gap between computational efficiency and high accuracy, providing a powerful data-driven surrogate model that can significantly accelerate the aerodynamic design and optimization process.

In order to enhance the wind-resistance capability and achieve a lightweight design of high-altitude unmanned airships, this study proposes a rapid optimization method for a heterogeneous propeller propulsion system. This system integrates contra-rotating and ducted propellers to exploit their respective aerodynamic advantages. First, surrogate models of the contra-rotating propeller, contra-rotating motor, ducted propeller, and ducted motor were constructed using an optimal Latin hypercube sampling method based on the max–min criterion. Then, within the optimization framework, propeller–motor matching principles and energy balance constraints were incorporated to minimize the total weight of the propulsion and energy systems. A case study on a conventional high-altitude unmanned airship demonstrates that, under the same wind-resistance capability, the adoption of the heterogeneous propeller electric propulsion system reduces the total propulsion-and-energy system weight by 24.94%. This method integrates the advantages of contra-rotating and ducted propellers, thereby overcoming the limitations of conventional propulsion architectures. It provides a new approach for designing lightweight, efficient, and long-endurance propulsion systems for near-space high-altitude platforms.