The provenance analysis based on self-organizing neural network improved by immune algorithm

The provenance analysis is the most important aspect in the research of basins and orogene belt, and it is also of great significance to rebuild paleogeographic environment. Determine a reasonable clustering according to the theory of sedimentary heavy mineral composition similar to its characteristic value under the same provenance. Self-organizing neural network is a useful tool for source analysis, and in order to analyze the provenance data more precisely, the immune algorithm is introduced to adjust the radius distance of the adjacent neurons and the number of neurons in competitive layers. In this article, we classify the heavy minerals data by immune self-organizing neural network, then compare the cluster centers of three members with sandstone contour chart of Kongdian Formation in Kongnan area of Huanghua depression. The results prove the effectiveness of the algorithm. © 2011 ISSN 2185-2766.