The Research About Fiber Probe Signal Denoising Based On Wavelet Analysis And EMD Transform

Gas holdup is an important parameter of fluid state in the oil wells, which is directlyrelated to the oil extraction yield. There is often some noise in the gas holdup signals, sohow to combine the characteristics of gas holdup signal to select the appropriate methodto remove the noise has the important practical significance to improve the gas holdupmeasurement accuracy. The wavelet threshold method, the modulus maximum valuemethod and Hilbert-Huang transform method are selected to the probe signal de-noisingresearch.First of all, the factors of wavelet threshold, wavelet basis and the soft threshold andhard threshold which will affect the effect of wavelet threshold denoising are used tosimulated signal to analyze and compare the effect. The effective signal and noisemodulus maxima in the scale characteristics were analyzed in order to make modulusmaxima algorithm, which achieve the work of the probe signal denoising.Thirdly, in order to solve the problem of false component which appears afterempirical mode decomposition in Hilbert Huang transform, this paper proposes ansolution basing on deviation index. Experimental results show the method has goodresults, and the method is used to the probe signal denoisingFinally, through the test probe data of gas holdup processing, the gas holduptreatment results were compared with the wavelet threshold method, the modulusmaximum value method and EMD decomposition threshold denoising basing on thedeviation index. The linear fitting effect between the denoising signal of three methodsand the standard gas rate curve was analysised.The experimental results show that, theEMD threshold method basing on the deviation index is with less calculation and thestable denoising effect .It also improve the most significantly to the gas rate comparingwith wavelet threshold method and the modulus maximum value method. Its effect of gasrate curve of the linear fitting with standard is also the best, being able to predict the gasholdup trend, so it is the most suitable method to the probe signal denoising among thethree denoising method.