自适应Chirplet信号展开及其在颤振信号处理中的应用  

ACSE (Adaptive Chirplet Signal Expansion) Algorithm and Its Application to Processing of Flutter Signals

作　　者：裴承鸣[1] 舒畅[1] 宋叔飚[1] 李中群[1] 谭申刚[2] 

机构地区：[1]西北工业大学数据处理中心,陕西西安710072 [2]北京航空航天大学,北京100083

出　　处：《西北工业大学学报》2004年第5期591-595,共5页Journal of Northwestern Polytechnical University

基　　金：航空科学基金 (0 1A5 30 0 1)资助

摘　　要：提出了一种新的以高斯线调频小波作为基函数的自适应信号展开算法。算法融参数的初值估计和精确估计于一体 ,自适应地将信号展开在高斯线调频小波基函数集上 ,通过展开系数和基函数参数获得信号的自适应时频能量分布。数值仿真结果表明该算法抗噪性好 ,信号可重构性高。同时 ,该算法作为一种时频域滤波技术 ,成功应用于某型飞机的颤振试验信号处理中 ,使得颤振边界预测的精度提高了 1 %～ 3%Refs.3 and 4 dealt with chirplet decomposition. In using Refs.3 and 4, we found that estimating initial values is one key problem that must be carefully dealt with. So, in this paper on ACSE algorithm, we lay particular emphasis on discussing how to estimate initial values. ACSE algorithm is a new algorithm based on chirplet elementary functions. This new algorithm obtains initial value estimation and precise resolution simultaneously and expands signal adaptively as the sum of chirplet elementary functions. According to expansion coefficients and elementary function parameters, adaptive time-frequency energy distribution is obtained. The signal is reconstructed well and noise is also effectively reduced. We tabulated simulation results for ten parameters (five design parameters and five estimated parameters), giving amplitude, standard deviation, time, frequency, and frequency modulation rate for each of these ten parameters. Comparison of simulation results for each signal parameter with those for its corresponding estimated parameter shows that estimation precision of ACSE algorithm is very high. This paper treats ACSE algorithm as a time-frequency domain technique and applies it to processing of flutter signals obtained in an aeroelastic model test in a low-speed wind tunnel. The results show preliminarily that the precision of flutter boundary prediction can be improved by 1%～3%.

关 键 词：信号处理 颤振边界预测 CHIRPLET 时频分析 

分 类 号：TN911.72[电子电信—通信与信息系统]