Wavelet Theory And Its Applications In The Aerial Defense Weapon System

Target identification is an important link in the chain of information processing for air-defense. In target identification or classification, extracting effective identification features from original target signals is very important. But, for a great number of non-stationary or time varying signals, such as speech, echo, earthquake signals, ect. Identification features are often localized both in time and frequency, so thus extracting effective features from them by general transformation methods is very difficult. The wavelet analysis is a novel maths method. It keeps the advantages of the Fourier analysis, simultaneously making up the shortages that the Fourier analysis can not describe the function space beyond L2,the shortages that it can neither do the local analysis; the wavelet analysis can not only describe almost all the familiar functions by using the Wavelet coefficients, but also describe the local lubricant characters with expansion wavelet coefficient, especially in signal analysis, it is more effective than the existing methods in data compressing and margin detecting due to its predominant properties in local analysis. The effective features of target can be extracted by wavelet transforms.This paper analyses and researches into wavelet theory in all sides, introducing both definitions and characteristics of Wavelet transforms with many kinds and the decomposed features which the wavelet package acts on signals and the space decomposed methods. Introducing basic conceptions of genetic algorithm. Introducing applications of wavelet transform in aerial defense weapon system.Chapter 3 analyses and researches into wavelet approximation theory, introduces the basic definitions and characteristics of wavelet package and genetic algorithm, set forth an improved genetic algorithm. In the light of the features of radar signals in aerial defense weapon system, a chosen method of optimal wavelet package radix is brought forward on the basis of inheritance algorithm and improved genetic algorithm.Chapter 4 deals with the basic conceptions of and generic and radar signal processing methods, a method for resolving radar signals is brought forward, which has strong frequency resolving power, thus it can be used in instantaneous radar signal distinguishing. In C3I system since traditional data fusion algorithms are more computationally complicated with fusion precision being lower, the character of wavelet transform of radar data in imprecise data is analyzed, a method for aerial defense weapon system data fusion is brought forward based on wavelet transforms.In accordance with distinguishing features of target acoustic signals, using wavelet transform and genetic algorithm, a new method based on acoustic signals characteristics for identifying airplane is set forth, in this method target acoustic signal is taken as source signal to make characteristic extracted and pattern recognized. Computing wavelet transform of radar echo, using criterion and genetic algorithm, selecting optimal basis via given by training sample sets, the targets are recognized by effective features.In Chapter 5, firstly, through analysis of high resolving radar echo signals, a mathematical model of radar-target-range-profile is brought forward, a radar target identifying method in aerial defense weapon system is set forth based on wavelet transforms and genetic algorithm, with the experimentation results indicating that such method has high identifying rate. Then based on wavelet transforms of the second generation, the characteristics of infrared target image are introduced, along with a detecting method for infrared target image put forth. First, infrared image is deposited by wavelet transforms. The edge information of the infrared target image is kept in wavelet transforms domain. Four wavelet decomposition components are processed. Second, infrared image is constructed by inverse wavelet transforms. Finally, the infrared image is filtered. The target is detected by two-norm image. The multi-targets can be segmented by using the method. The experiments show that this approach can achieve quite satisfactory results.Thirdly, investigating the varying features of wavelet coefficients of radar target signals and chaff jamming signals, on basis of wavelet transforms and its correlation technique, a radar target identifying method under the foil strip jamming is brought forward by using the deference full deference between the radar target signal and foil strip jamming signal in time domain and frequency domain. First, the echo-signals are quickly decomposed by the Mallat-algorithm. Second, the correlation coefficient of near scales wavelet coefficient is calculated.The wavelet coefficient are rocessed through a threshold,and then the target signals are reconstructed and the radar target is recognized.In this dissertation the methods have been simulated and tested with feasible results that have greater theoretical significance and actual applied value in regared to information processing and target identifying in our aerial defence weapon system.