基于自组织竞争网络的黄河和长江物源判别

利用黄河和长江在地质历史时期的沉积物与现代黄河长江沉积物应具有一定的相似性这一原理,＂将今论古＂,以现代黄河和长江入海沉积物中部分常量元素的百分含量作为基准数据建立自组织竞争网络,对来自不同时代的黄河和长江的沉积物进行判别和验证,可靠性达到94.4%,并规定了其相应置信度下的置信区间.以此为基础,对南黄海NT2孔的物质来源进行了判别,结果表明钻孔中0～19.36m、28.07～52.88m深度范围内的沉积物为长江沉积物,19.36～28.07m、52.88～70.28m深度范围内以及表层沉积物为黄河沉积物.

A new method based on self-organizing ANN to distinguish between sediments from the Yellow River and the Yangtze River

The Yellow River and the Yangtze River are the most main sources of sediment accumulation in Chinese eastern sea,and the sediment provenance identification for the Yellow River and the Yangtze River is one of the hot research problems.In this picture,based on the principle that the geologic historical period＇s sediment of the Yellow River and the Yangtze River should have being similar to modern times＇,we established a self-organizing competitive neural network depending on part of constant element percentage content,and discriminated and verified sediments from different times of the Yellow River and the Yangtze River,the reliability of which reached 94.4% in the corresponding confidence interval.On this base,we also discriminated the sources of core NT2 sediments,and the results showed that the sediment on the scope within 0—19.36 m,28.07—52.88 m depth came from the Yangtze River,while 19.36—28.07 m,52.88—70.28 m depth and surface deposit came from the Yellow River.