Study of Signal Feature Recognition Method Based on Morlet Combined Wavelets

For the reason that the traditional Hilbert transform method is not valid to recognize the feature parameters of the target signal in an active electromagnetic detection system under the condition of a high ambient noise level, based on the depiction of principle of envelope analysis and wavelet filtering, a new envelope demodulation method is put forward, which employs the complex Morlet combined wavelets. The method takes the advantage that the imaginary part of the combined wavelets is the Hilbert transform of its real part, so it can realize filtering and envelope demodulation only by one step. Moreover, by appropriately selecting wavelet parameters, a band pass filter with zero phase shift, flat topped pass-band and rapid attenuation in the transition band can be made, consequently, the accurate demodulation of signal can be realized under a low signal to noise radio condition. The simulation results of engineering application show that this method is efficient and affective for the feature recognition of the target signal in the above detection system, and has a good application prospect.