The Extraction Methods About Doppler Echo Signal Of Targets On The Ground

When the radio fuze carries on the survey to the ground object, due to the complexity of the background on the ground and the modern targets used the stealth technology,the echo signal of the targets may be neglected by the ground interference, causing the damage of the fuze on the target to be reduced significantly. Therefore, it’s the most urgent question that the characteristic of the target signal carries on extraction in the complicated background.Through using the principle of the continuous wave doppler in the radio fuze, this article gives out the simulation of the doppler echo signal, and analyzes the characteristics of time and frequency domain of the doppler signal about the measured targets on the ground. Comparing the assumption testing and wavelet transtorm, the extent of noise wave on the ground is analyzed in the paper, and the variance is used in determining the similarities of two group of fitted curves. An improved threshold function is used to denoise the wavelet threshold value, and then combined with Hilbert transform and wavelet transform envelope is carried on extraction. Finally the characteristics of frequency variation in the measurement of analogue pill rendezvous signal is carried on extraction with the wavelet decomposition and reconstruction approaching signal. A simple and effective method on judgement of threshold value is also put forward.The results showed that:1) The application of wavelet transform has overcome the shortcomings of the supposition of the background clutter obeying specific distribution in the assumption testing, and adapts to the analysis of different background clutter. 2) By improving the threshold value,the denoising performance has certain enhancement. And with the combination of Hilbert and wavelet transform, the envelope extracted is smooth, burr-free and a significant increase in range. 3) The characteristics of frequency variation in the measurement of analogue pill rendezvous signal is extracted well with the wavelet decomposition and reconstruction approaching signal.When signal-to-noise ratio is-10 dB, correct starting rate of fuze reaches 96% above 123 datas with the method on judgement of threshold value.