【摘要】 叠前同步反演在储层预测及烃类检测等方面得到了广泛应用。然而如何获得最佳的反演参数,以使误差最小,是应用叠前同步反演技术面临的一个重要问题。前人主要依据经验、人为观测获得反演参数,缺点是误差较大,缺少科学的参考依据。鉴于此,引入了统计学上的数据归一化方法,使不同量纲数据如纵横波速度、密度均方根误差等得到了有效归一化,解决了三者难以在同一尺度下对比分析的问题。有效融合了正反演数据信息到相关系数与误差比这一个指标中。上海某区地震数据叠前同步反演参数优选表明,应用该技术确定了最优的反演参数,使反演结果与测井实测结果达到了最佳逼近。更多还原

【Abstract】 Pre-stack simultaneous inversion is widely used in reservoir prediction,it is a key important problem for pre-stack simultaneous inversion to obtain the best inversion parameters and minimize error.Previous methods mainly depend on experience to achieve inversion parameters subjectively and the disadvantage of them is bigger error and lack of scientific evidence.Data normalization method made different dimensional data effectively normalized,such as compressional and shear velocity,the error of root mean square of density,etc.This paper introduces data normalization method in statistics to solve the analysis problems which usually is hard to compare in the same scales and effectively integrate the method of forward and inversion information into the index of correlation coefficient to error ratio.This technique is applied in some block of Shanghai and the results show that this method generates optimal inversion parameters,additionally,inversion results agree with actual logging results at the best approximation.更多还原