Clipping Data Recovery Based On The Predicted Vibration Test

Vibration test is the important measure to improve the design and the capability of a plane, but the problem of the clipped data often can be found. This paper use the information provided by the right data to restore the clipped data.This dissertation researches in the vibrant clipped data of a plane, proposes regression analysis, BP learning algorithm and cubic spline interpolation to forecast the clipped data. Firstly, the paper studies the arithmetic and theories about these three methods, compiles the program based on MATLAB, and then restores the multiple clipped data. Finally, the paper compares and analyzes the accuracy. The results show these three methods have good forecasting and the cubic spline interpolation has higher precision, simple operation than others.The methods have theoretic and engineering value in vibration test.