摘要
用自组织人工神经网络分析碳酸盐岩中岩溶洞规模蔡煜东，杨兵，孙虹中国科学院上海冶金研究所，上海，２０００５０关键词：碳酸盐岩，岩溶洞穴，自组织人工神经网络，水文地质，北京西山ＡｎａｌｙｓｉｓｏｆｔｈｅＫａｒｓｔＣａｖｅＳｃａｌｅＩｎＣａｒｂｏｎａｔｅＲ...
The artificial neural network(ANN),a nonlinear science raised in the mid-80's,attempts to simulate some basic features of brains such as self-organization,self-adaptability and fault-tolerant ability and so on. It has been initially used in the pattern recognition,data handling and autocontrol with perfect effect. The network is used to analyze the scale of karst caves in carbonate rock taking Ordovician strata in Yangtuo of Xishan,Beijing for example.Firstly,the recognition model of neural network was set up by 18 samples(a cave for a sample).Then four natural samples acted as unknown sample were used in prediction according to the neural network with unknown formation and the results are similar to the practical performance.The network is characteristics of higher classification and recognizing abilities,suitable for use in the nonlinear prediction with multiple-factors and multiple-object.Comparing with general multiple-factor discriminated method it has higher fault-tolerant ability,quick predicting speed.It is considered that the analyses method can treat a new way of which predict karst cave scale.as the further development of the neural network theory.
出处
《江苏地质》
1994年第1期42-45,共4页
Jiangsu Geology
