【摘要】 在地震子波非因果、混合相位的假设下,分别应用滑动平均(MA)和自回归滑动平均(ARMA)模型对地震记录建模,并采用基于高阶累积量的线性方程法对子波提取和模型适应性进行了研究。数值仿真结果表明,ARMA模型比MA模型在描述地震记录时具有参数节俭、模型更为高效的特点;基于高阶累积量的线性方程法对加性高斯色噪声有较好的压制效果,却强烈依赖于累积量样本估计的准确性。如果累积量样本估计的误差和方差适度,结合AR-MA模型描述的累积量线性方程法适用于非因果、混合相位的子波提取,其有效性通过实际地震数据的处理得到了验证。更多还原

【Abstract】 Based on the assumption that the wavelet is non-causal and mixed phase,both the moving average(MA) and autoregressive moving average(ARMA) models were introduced to fit the seismic record.A linear equation approach based on higher-order cumulant was employed to extract the wavelet and evaluate the adaptability of each model.Numerical simulations demonstrate that the ARMA model provides a parsimonious,more efficient signal model for fitting seismic trace than the MA model does.The cumulant-based linear equations approach is not sensitive to Gaussian colored noise,but strongly relies upon the accuracy of the trace cumulant estimates.If the estimated error and variance of trace cumulants are moderate,the cumulant-based linear equations approach combined with the ARMA model description of the seismic record is appropriate for non-causal,mixed phase wavelet extraction.The real seismic data examples demonstrate possible uses of the method.更多还原