【摘要】 针对地面微地震有效信号特点和资料采集方式,结合微地震信号高阶累积量统计特征分析,考虑到有效信号和噪声在时空方向的不同分布特征,研究了时间和空间两个方向地面微地震信号的四阶累积量估计方法;考虑到贝叶斯估计方法对于弱信号估计的优势,研究了基于贝叶斯框架的四阶累积量的自适应算法,把信号四阶累积量的联合概率密度函数作为原信号的概率密度函数进行最大后验概率估计,建立了地面微地震资料四阶累积量贝叶斯估计方法;提取弱信号的同时不可避免会提取到弱的无用相关信号,使得弱有效信号不易识别,根据区域相关噪声在时间方向具有区域均匀分布而有效信号具有局部分布的特点,提出进一步采用自适应减法剔除贝叶斯估计结果中的这种区域性相关噪声。通过系列方法的分析研究,形成了地面微地震有效信号的有效提取方法。利用该方法对实际资料进行处理,取得了较好的效果。更多还原

【Abstract】 According to the surface microseismic effective signal characteristics and data acquisition methods,combined with microseismic signal higher-order cumulant statistical characteristic analysis,considering the different distribution characteristics of effective signal and noise in the time and space direction,we discuss the fourth-order cumulants estimation method of surface microseismic signal in time and space domain.Considering the advantages of Bayesian estimation method for week signal estimation,we study self-adaptive algorithm of fourth-order cumulant based on Bayes framework,take the joint probability density function of signal fourth-order cumulant as that of original signal to carry out maximum posterior estimation,to establish fourth-order cumulant Bayesian estimiation method for surface microseismic data.While extracting weak signals,useless relevant signals would be extracted,so the weak effective signals are difficult to recognize.As we all know,regional relevant noise is evenly distributed in time direction and effective signal is locally distributed,in terms of the characteristic,we propose adopting self-adaptive subtraction to remove the regional relevant noise in Bayesian estimation results.By analyzing the method series,effective extraction method for surface microseismic effective signals is formed.The method has obtained good result in the actual data application.更多还原