Two Modal Decomposition Method Based On Ar Modeling In The Dither Signal Processing Research

Mode decomposition and parameter extraction based on structural random response is an important research work in flutter testing of aircraft. According to the features and analysis requirements of turbulence response two kinds of new analysis method so called AR Auto-Correlation Matching Pursuit (AR-MP) and Empirical Mode Decomposition & Matching Pursuit (EMD-MP) is proposed in this thesis. Dictionaries of AR waves based on MP algorithm are construacted. An orthogonal EMD that is more efficient is introduced from MP. The theory analysis and method principle are presented. The characteristics of the two methods are studied by computer numerical simulation. Low-speed wind tunnel test data of aeroelastic mode are used to examine the two methods. The results show that both the AR-MP and the EMD-MP are reasonable and feasible.The software is developed on Labview, Matlab and Visual C++ and success fully used in processing of wind-tunnel tesing datas of one homemade plane model.