An Algorithm of Voice Activity Detection Based on EMD and Wavelet Entropy Ratio

A new method was proposed to identify speech-segment endpoints based on the empirical mode decomposition（EMD） and a new wavelet entropy ratio with improving the accuracy of voice activity detection.With the EMD, the noise signals can be decomposed into several intrinsic mode functions（IMFs）. Then the proposed wavelet energy entropy ratio can be used to extract the desired feature for each IMFs component. In view of the question that the method of voice endpoint detection based on the original wavelet entropy ratio cannot adapt to the low signal-to-noise ratio（SNR）condition, an appropriate positive constant was introduced to the basic wavelet energy entropy ratio with effectively improved discriminability between the speech and noise. After comparing the traditional wavelet energy entropy ratio with the proposed wavelet energy entropy ratio, the experiment results show that the proposed method is simple and fast. The speech endpoints can be accurately detected in low SNR environments.

Using wavelet multi-resolution nature to accelerate the identification of fractional order system

Because of the fractional order derivatives, the identification of the fractional order system（FOS） is more complex than that of an integral order system（IOS）. In order to avoid high time consumption in the system identification, the leastsquares method is used to find other parameters by fixing the fractional derivative order. Hereafter, the optimal parameters of a system will be found by varying the derivative order in an interval. In addition, the operational matrix of the fractional order integration combined with the multi-resolution nature of a wavelet is used to accelerate the FOS identification, which is achieved by discarding wavelet coefficients of high-frequency components of input and output signals. In the end, the identifications of some known fractional order systems and an elastic torsion system are used to verify the proposed method.