Based Window Function Combination Explore Inhibition Of Hht Transform Endpoint Effect

In recent years, Hilbert-Huang transformation as a new signal processing theory was proposed. It is suitable for processing the non-stationary signal. But there are still some shortcomings in some ways. Among them, the endpoint effects problem of the EMD decomposition is particularly notable. The endpoint effects will'pollution'the whole data sequence with the shifting number's increasing. Ultimately, it could make the results of EMD decomposition meaningless.HHT transformation theory is studies deeply in this paper, it shows the superiority of this theoretical, application field and its existing problems. And, it specifically analyses the'endpoint effect'. According to the cause of this problem this paper gives an improvement measures. It discusses a method of the endpoint processing problem:extreme value point continuation method, the advantages and disadvantages of the rules is analyzes in this paper. Then, the window functions are discussed, and one kind window combination is constructed according to the signal itself requirements. And it combines the advantage of the extremum continuation to improve the method of the inhibition endpoint effect. The main features of the method are ensuring signal itself characteristic, attenuating the continuation part, reducing man-made factors and making them more flexibility. In this paper, the effectiveness is proved by the simulation signal experiment.It still researches the shifting stopping standard threshold of the EMD. In a lot of experiments, by changing the threshold, the different decomposition results are showed. It explains stopping standard is an important factor to influence the decomposition effect.