MVD and LWC prediction in icing environment based on genetic algorithm optimized neural network

Mean volumetric diameter (MVD) and liquid water content (LWC) are two important meteorological parameters that affect aircraft icing, but they are difficult to measure accurately in practice. If these two parameters can be obtained in real time and accurately, it can provide some guidance for ice accumulation prediction and the establishment of aircraft airworthiness certification standards. This paper proposes a prediction model for icing meteorological parameters based on genetic algorithm optimized neural network. With ice thickness and icing rate of different measurement point combinations, ambient temperature, flight speed and wing angle of attack as input parameters, and icing meteorological parameters MVD and LWC as output parameters, the genetic algorithm optimized icing meteorological parameter prediction model is constructed, and the prediction model is used to predict the icing meteorological parameters of the numerical calculation test group data and the icing wind tunnel experimental data. The results show that the prediction model based on genetic algorithm optimized Elman neural network predicts the icing meteorological parameters of the test group within the relative error of the test group 10%, and the relative error of the experimental data is within the relative error of the experimental data 20%. This method has certain feasibility.