内容：
针对在重力梯度张量正演中计算耗时过长和核矩阵内存消耗过大等制约反演实施的瓶颈问题,在L1范数的基础上,引入种植反演,用累加求和分析替换迭代求解,避免计算或存储反演核矩阵,以减少内存占用和加快反演迭代;针对种植反演容易导致相邻异常源相互侵入的问题,引入一个基于位场水平衰减特性加权函数来限制密度吸引子的作用范围,以期使密度吸引子忽略较远的异常源,抑制相邻异常源相互干扰。反演结果及分析表明重力及重力梯度张量种植反演所需计算机内存小和水平衰减特性加权函数能有效的抑制相邻异常源的侵入。 Based on the L1 norm, the inversion of the planting is introduced, and the iterative solution is solved by the summing and analysis. In this paper, the inversion of the inversion of the constraints in the gravitational tensor forward modeling and the excessive consumption of the kernel matrix is carried out. To avoid the calculation or storage of the inverse kernel matrix to reduce the memory footprint and to speed up the inversion iteration. Aiming at the problem that the planting inversion easily leads to the intrusion of adjacent anomaly sources, a weighting function based on the field level attenuation characteristic is introduced to limit the density attractor Range, in order to make the density attractors ignore the distant source, inhibit the interference of adjacent anomalies. The results of inversion and analysis show that the weighting function of the computer memory and the horizontal attenuation characteristic of the gravity and gravitational tensor inversion can effectively suppress the intrusion of adjacent anomaly sources.

文件列表（点击上边下载按钮，如果是垃圾文件请在下面评价差评或者投诉）: