位场数据处理算法的数据挖掘试验与应用

Data Mining Experiment and Application Based on Potential Field Data Processing Algorithms

以规则形体重力异常场为正演模型,基于Matlab平台运用位场的梯度算法、解析延拓算法、边界提取算法进行模型试算,总结了这3种位场数据处理算法的特点。梯度算法能识别和分离不同场源的异常;解析延拓算法能够压制和突出深部或浅部场源信息;边界提取算法能以阶跃或梯度带的形式反映场源体边界。最后运用位场数据处理算法结合数据挖掘的视角对塔里木盆地的布格重力异常进行处理及解释,得到研究区域主要断裂系统的分布为平行于盆地,走向为NE,NW,NEE。

Taking the gravity anomalies of regular shape bodies as conversion model,this paper summarizes the features three potential field data processing algorithms,and operates model calculation using gradient algorithm,analytical continuation algorithm,boundary extraction algorithm,respectively,based on Matlab.The gradient algorithm is able to recognize and separate the anomalies coming from different sources.The analytical continuation algorithm contributes to suppress and highlight the information of deep or shallow sources.The boundary extraction algorithm is skilled at reflecting the borders of sources in the form of gradient zone and phase step.Finally,the methods are applied to the processing and interpretation of the actual Bouger gravity anomalies data in Tarim Basin from the perspective of data mining.It is concluded that the major fracture systems that orientate NE,NW and NEE are parallel to the borders of the basin.