The Hht Endpoint Problem Suppression Method

The Hilbert-Huang transform (HHT) is a new theory of time-frequency signal analysis, it breaks through the traditional limitations of the analysis methods. It generated with different characteristics scale of intrinsic mode function(Intrinsic Mode Function, IMF)through the empirical mode decomposition(Empirical Mode Decomposition, EMD),so instantaneous frequency is reasonable in physics. HHT proposed for the signal processing has opened up a new road, there are problems because of the lack of complete theoretical basis, and the endpoint problem is one of the key issues to be resolved.This paper is mainly to do research on the endpoint of EMD. The main tasks of this paper include:(1) It has analyzed the causes on the end effects of EMD. Summarizes, the influences of the signal data envelope use different transect interpolation method.(2) The performance is compared with methods which are mirror image method, slope-based method and waveform matching method of dealing with endpoint in different interpolations. Summarized the interpolation which is more suitable for such methods, can effectively inhibit the endpoint problem of EMD.(3) This paper has introduced the new method which one kind of waveform matching method and the mirror image method combining, based on understood to each kind of processing vertex question method's research. Through the simulation experiment, we have confirmed this method relative validity.(4) Though analysis a city for many years by the average monthly air temperature data, shows that the new method which one kind of waveform matching method and the mirror image method combining is on the relative effectiveness of real data.