中文摘要：

为解决卫星重力梯度大尺度密度反演存在的平面模型误差、病态矩阵正则化、大型矩阵求逆、先验数据融合等问题,依托Tesseroid单元体模型,分别采用广义岭估计、代数重建（ART）、遗传算法完成了附加高斯白噪声的模拟异常体密度反演实验,并将反演结果与原始设定模型进行比较分析。

英文摘要：

In this paper we examine the problems in the density inversion from satellite gravity gradient data which including plan model error,the regulation of ill-conditional matrix,the inversion of big matrix,the combination of nonlinear prior data,etc.Using the Tesseroid model we complete the simulation model inversion adding the white Gauss noise based on methods which include general ridge estimation,ART（algebra reconstruction technique）and genetic algorithm.Finally,all the results are compared and analyzed with the prior model.