关键词: 旋翼气动评估, 数据融合, 贝叶斯框架, 不确定度分析, CFD

Abstract: Due to the complexity of the aerodynamic environment, there are many uncertainties in the aerodynamic evaluation of rotor blades, which have a significant impact on their performance. The current aerodynamic evaluation methods cannot take into account the uncertainty of aerodynamic data. This study applies data fusion technology to the prediction of rotor blade aerodynamics to obtain aerodynamic force distribution and confidence intervals with higher credibility under the conditions of remembering certain uncertain factors. This lays a foundation for further uncertainty analysis and engineering applications of rotor blade aerodynamic data. Based on the hypothesis that the aerodynamic force data from different sources are all normally distributed independent random variables, the relationship between the distributed aerodynamic data and measured values is used as the fusion criterion to achieve the best matching of measured values as the target. The Bayesian estimation is used to solve the maximum posterior probability distribution of the fused data, thus constructing an aerodynamic data fusion method based on the Bayesian framework. Using the UH-60A and Caradonna-Tung rotor blades as examples, the proposed method is used to fuse the aerodynamic data from different sources, and the estimation variance and prediction error of the fused results are analyzed and compared. The results show that firstly, the proposed method does not have special requirements for the source of input data, can provide confidence intervals for fused results, and reduce the estimation variance of data from different sources, which can be followed by uncertainty analysis and research. Secondly, the fused results are not limited to data from different sources and broaden the coverage of data. Finally, the fused results are more in line with physical laws than data from a single source.

Key words: Rotor aerodynamic evaluation, Data fusion, Bayesian estimation, Uncertainty analysis, CFD