油管壁厚测量数据的一致性加权融合估计算法  

Consensus Weighted Fusion Algorithm of Multisensor Data for Wall Thickness Measurement in Oil Well Tubing

作　　者：金建华[1] 杨叔子[2] 

机构地区：[1]上海交通大学信息检测技术与仪器系,上海200030 [2]华中科技大学,湖北武汉430074

出　　处：《仪表技术与传感器》2002年第11期43-45,共3页Instrument Technique and Sensor

摘　　要：在随机扰动环境中采用漏磁场法在线测量油管壁厚时 ,每个磁性传感器只能得到环境的部分信息。为了对各传感器的局部测量值进行稳健融合估计 ,提出了油管壁厚测量数据的一致性加权融合估计算法。利用改进的数据探测技术中的分布图方法对各个传感器的一致性进行检验 ,并采用简单加权平均法求得一致性传感器测量数据的稳健融合估计值。融合实例表明 ,该算法能有效地提高系统测量的稳健性 ,且具有精度高、运算简单等特点。In an on-line wall thickness measurement system of oil well tubing using magnetic flux leakage technique,each magnetic sensor can only offer partial information with stochastic noise about it's environment.The consensus weighted fusion algorithm for the robust fusion of multisensor data is developed.The improved distribution display method in exploratory data analysis is used to test various sensor's consensus.Meanwhile,the simple weighted mean method is introduced to fuse consensus sensor data,and the robust fused data of multisensor data is obtained as a result.The fusion results show that the robustness of the measurement system is evidently enhanced,and the algorithm has specific advantages over some conventional algorithms such as high-accuracy,and high-efficiency.It is predicted that the algorithm is available for real-time,accurate fusion of multisensor data.

关 键 词：多传感器 数据融合 信号处理 油管 无损检测 厚度测量 漏磁检验 

分 类 号：TE931.2[石油与天然气工程—石油机械设备] TG11[金属学及工艺—物理冶金]