Based On The Time-frequency Analysis Of Millimeter-wave Interferometer Signal Processing Research

Projectile’s velocity plays a major role in developing a interior ballistics theory, researching new artilleries and light weapons, inspecting and accepting the weapons and so on. Meanwhile, the Millimeter-wave interferometer, with its unique advantages in measuring the projectile’s velocity in-bore, gets more and more wide range of applications.In this thesis, focused on the Millimeter-wave interferometer signal, the method for the motion parameters measurement of the projectile in-bore, such as the projectile’s velocity, is studied. And the main work includes:First of all, the measuring principle of a projectile speed in-bore with a Millimeter-wave interferometer is discussed, and a signal model of a Millimeter-wave interferometer is made up. Then the common time-frequency analysis methods are introduced in detail for the later signal processing.Secondly, de-noising algorithms for a Millimeter-wave interferometer signal based on the time-frequency domain are focused in the paper. Especially a wavelet package-kurtosis de-noising algorithm and a time-frequency peak filtering algorithm based on the time-frequency analysis are discussed, and their de-noising performance is also simulated.Thirdly, the instantaneous frequency extraction method based on time-frequency analysis is presented. This wavelet ridge extraction, which includes a phase stabilization method and a wavelet modulus maxima method, and a method combined digital image processing technology are demonstrated and simulated. And a compare is given between a wavelet modulus maxima method and a digital image processing method,and simulated results are shown in different SNR.Finally, a Millimeter-wave interferometer measured data is analyzed with a given method to extract the instantaneous frequency of a projectile in-bore. Based on these results, the projectile’s velocity and acceleration curves are estimated and the length of a cannon is also calculated, which is consistent with the actual, proving the accuracy and usefulness of the methods proposed in this paper.