中文摘要：

针对重力梯度数据聚焦反演结果中存在的＂上漂＂现象,在经典Tikhonov正则化理论框架下,引入最小支撑泛函数对反演模型进行约束以避免反问题解的不稳定,并针对重力梯度数据聚焦反演中存在的趋肤效应,在模型目标函数中引入指数深度加权函数。通过理论模型,对部分重力梯度张量分量进行了单独以及联合聚焦反演,验证了基于深度加权的聚焦反演方法的有效性,并将该反演方法运用到涩北一号气田区的实际数据中,反演结果较好地反映出气田位置。

英文摘要：

The focusing inversion method of gravity gradient tensor data based on depth weighting was proposed to avoid ＂resting on＂phenomenon of inversion results. In the classic Tikhonov regularization theory framework,a minimum support functional was introduced to constrain inversion model,which can avoid instability of inverse problem solution. And the index depth weighting function was also added in the model objective function to overcome the accumulation of density occurring at shallow depths. Some single components of gravity gradient tensor and their joint-components of theoretical models were inverted based on regularized focusing inversion,which proves the validity of the regularized focusing inversion method with depth weighting function. The proposed method was applied to the practical data of Sebei 1 gas field,and the inversion results could well reflect the position of the gas field.