一种泥浆脉冲压力传感器信号的融合算法  

Algorithms of Multi-sensor Data Fusion for the Mud Pulse Signal

作　　者：李传伟[1] 慕德俊[1] 李安宗 

机构地区：[1]西北工业大学自动化学院,西安710072 [2]中国石油集团测井有限公司,西安710061

出　　处：《计算机科学》2009年第7期74-75,102,共3页Computer Science

基　　金：油气资源与勘探教育部重点实验室基金项目(2006K002)资助

摘　　要：针对泥浆脉冲信号传输复杂环境和采集信号的特点,提出多传感器探测泥浆脉冲信号的思想,研究了基于统计方法理论的信号间可靠性度量,并推导出信号间关系矩阵的确定方法。利用综合支持度最优数据融合方法,实现对实验数据的仿真处理和现场采集数据处理。分析结果表明,该方法克服了单传感器信号不确定性的困难,有效消除了信号中的干扰和噪声,能够很好地恢复原始编码数据,满足随钻测井的要求。According to the acquisition and transmitting conditions while drilling, the idea of detecting the signals by means of multi-sensors was proposed to get more accurate data, the reliability of sensors and the relation matrix between multi-sensors were also researched based on the fuzzy theory. The optimal methods of data fusion based on synthesis were presented to process both the example and real data. The processing results imply that the method has exhibited the abilities to reduce uncertainty and noises encountered in single-sensors detection, which ensures its application in the measure while drilling system in oil field.

关 键 词：泥浆脉冲信号 多传感器 数据融合 关系矩阵 仿真处理 去噪 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]