航空兵器2026，Vol.33Issue(1)：54-64,11.DOI:10.12132/ISSN.1673-5048.2025.0166
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
Modal Frequencies Identification and Tracking Method for Time-Varying Structure of Flight Vehicles Based on Time-Frequency Representation Image Segmentation
摘要
Abstract
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.关键词
模态参数识别/时变结构/图像分割/U-Net/脊线提取Key words
modal parameter identification/time-varying structure/image segmentation/U-Net/ridge extraction分类
军事科技引用本文复制引用
刘乐翔,苏怀一,康杰,王寅,孙嘉宝..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[J].航空兵器,2026,33(1):54-64,11.基金项目
国家自然科学基金青年科学基金项目(12102178) （12102178）
航空科学基金项目(2024Z005052003) （2024Z005052003）
南京航空航天大学研究生科研与实践创新计划(xcxjh20251506) （xcxjh20251506）
