基于变分模态分解和小波能量熵的微震信号降噪
Research on microseismic signal denoising based on variational mode decomposition and wavelet energy entropy
-
摘要: 微震监测技术被广泛应用于矿业工程、石油天然气开采、安全监测等领域。 针对微震监 测采集到的微震信号存在随机噪声的问题,本文提出了一种变分模态分解(variationalmodedecomposition,VMD)和小波能量熵(waveletenergyentropy,WEE)结合改进阈值函数的降噪算法。 对原始信号进行VMD分解,将得到的各模态分量(intrinsicmodefunction,IMF)进行多尺度小波 分解,用小波能量熵表征各尺度信号的含噪状态,并以小波能量熵最大子区间的小波系数计算各 尺度层的阈值,通过改进阈值函数进行降噪处理后得到新的IMF,重构微震信号。 对仿真信号和 实测信号进行降噪处理,结果表明,该算法优于经验模态分解(empirical mode decomposition, EMD)、集合经验模态分解(ensembleempiricalmodedecomposition,EEMD)、VMD结合小波硬阈值 和软阈值降噪方法,提高了微震信号的信噪比。Abstract: Microseismicmonitoringtechnologyiswidelyusedinsuchfieldsasminingengineering,petroleumandgasexploitation,andsafetymonitoring.Inordertosolvetheproblemofrandomnoiseinmicroseismicsignalscollectedbymicroseismicmonitoring,adenoisingalgorithmbasedonvariationalmode decomposition(VMD)combinedwithwaveletenergyentropy(WEE) andimprovedthresholdfunction isproposed.Wavelettransformationisperformedoneachintrinsicmodefunction(IMF)componentafter VMDdecompositionoftheoriginalmicroseismicsignals.Thenoisestateofeachscalesignalischaracterizedbywaveletenergyentropy.Thethresholdofeachscalelayeriscalculatedbywaveletcoefficients ofthemaximumsubintervalofwaveletenergyentropyandthenthemicroseismicsignalsarereconstructedbythenewIMFsdenoisedthroughtheimprovedthresholdfunction.Theresultsfromnumericalsimulationsignalandrealsignalshowthattheproposedalgorithmissuperiortoempiricalmodedecomposi
tion(EMD),ensembleEMD(EEMD)andVMDcombinedwithwaveletenergyentropyhardthreshold functionandsoftthresholdfunction.Thesignal-to-noiseratioofmicroseismicsignalsareimproved.
下载: