High-Speed Computation and Efficient Storage in 3-DGravity and Magnetic Inversion Based on Genetic Algorithms
重磁遗传算法三维反演中高速计算及有效存储方法技术

A currently prevailing in geophysical interpretation, especially in 3-D inversion case is regularly dividing the underground source region into a set of juxtaposed cuboids, estimating the physical properties by using potential field inversion and deducing the image of the geological source. Nonlinear global methods, such as Genetic Algorithm, gradually play a leading role in the inversion. It is shown in the paper that there exists such an excessive computation that has bottlenecked the efficaciousness of such method as the Genetic Algorithms. Several special tricks have been taken to break the choke point of overwhelming computation, one of which is to separate from formulas of the geometric trellis which is invariable during the whole process of inversion, and to store the trellis as a table which would make the following reiterative forward computation as just a simple reading of the table and multiplying it with corresponding properties of the geometric cells. Also, the caused huge storage of geometric trellis is reduced tremendously by a very skill of equivalent storing. The techniques revive the nonlinear methods in their applications to the inversion of gravity and magnetic anomalies for 3-D physical properties.