时频谱图像分割的飞行器时变结构模态频率辨识与跟踪方法OA

Novel aerospace vehicles encounter severe structural elastic deformation problems dur-ing flight,with their structural modal frequencies exhibiting significant time-varying characteristics.To address the challenges of high difficulty and low efficiency in identifying and tracking the time-varying structural modal frequencies of aerospace vehicles,this paper proposes a method for time-varying structural modal frequency identification and tracking based on time-frequency representation image segmentation.Firstly,time-frequency analysis is performed on the measured structural vibra-tion responses,and the U-Net network model is introduced to automatically extract ridges from the time-frequency representation of responses.Subsequently,the structural modal frequencies at each discrete moment are obtained through a skeletonization algorithm.Finally,the ridge merging algo-rithm is employed to eliminate spurious modal frequencies and achieve time-varying modal tracking throughout the entire time period.Numerical simulation and telemetry data analysis results demon-strate that the proposed method can effectively identify and track the time-varying structural modal fre-quencies of aerospace vehicles.Compared to the traditional time-series model-based FS-TARMA method,it offers higher efficiency and exhibits better adaptability to telemetry data.