Research on Data-Driven Method for Measurement Station Selection of Key Sections in Airplane Engines

In order to improve the performance of aircraft engines, research on their complex thermodynamic and aerodynamic processes is necessary. One essential experiment is the testing of key section parameters, which provides detailed information on the average flow field parameters at section boundaries and the overall characteristics of the engine. This paper discusses the algorithm and application of temperature field measurement point layout for key sections of aircraft engines, in order to obtain the characteristics of the temperature field and represent the average parameters of the section for overall engine performance evaluation. The innovation of this paper lies in addressing the issue of sparse data by utilizing simulation data from roctor37. Firstly, a single channel interpolation algorithm based on biharmonic spline interpolation is proposed to obtain regular grid node temperature data. This method transforms irregular mesh nodes and corresponding temperature values from CFD software onto a regular sectorial grid, and evaluates the gridded data using kernel probability density estimation and JS divergence. Secondly, a sampling point selection strategy based on proxy model is introduced to determine the coordinates of measurement stations in the temperature field through kriging proxy model and maximum expected prediction error. This approach yields measurement point coordinates that reflect temperature information in the flow field well. Finally, kriging interpolation algorithm is utilized to reconstruct the temperature values of the grid nodes using the few obtained measurement stations. Using the aforementioned methods, the temperature fields for different states were reconstructed with good results. This sampling strategy and interpolation algorithm could be used to reconstruct the temperature parameter information of the entire flow field where sensors are insufficient at key sections of aircraft engines.