工业储罐缺陷检测模糊分类器研究  

Research on FuzzyClassifier for Defect Detection of Industrial Storage Tank

作　　者：赵雅阁 马骏 孙晓强 李东立 李永真 时英明 ZHAO Yage;MA Jun;SUN Xiaoqiang;LI Dongli;LI Yongzhen;SHI Yingming(Henan Boiler and Pressure Vessel Safety Inspection Institute,Zhengzhou 450016,China)

机构地区：[1]河南省锅炉压力容器安全检测研究院,河南郑州450016

出　　处：《仪表技术》2023年第5期51-53,共3页Instrumentation Technology

摘　　要：目前工业储罐缺陷检测设备对传感器与测试表面的距离不太敏感,对于校准的要求也非常低,是人工操作员快速检查储罐底板缺陷的一大障碍。基于电场指纹法介绍了一种利用模糊推理系统对原始传感器数据进行分类的方法,用以检测大型液化天然气储罐的缺陷。采用最新的滤波去噪芯片,设计专用信号处理电路,配套相应计算软件,完成特征信号的提取;建立实时测量数据库,并对数据进行动态分析;根据分析结果,生成拟合图。结果表明,所设计的工业储罐缺陷检测模糊分类器达到了预期效果,其工业平台的建立将对石油、天然气行业的增值具有重大意义。At present,industrial tank defect detection equipment is not very sensitive to the distance between the sensor and the test surface,and has very low calibration requirements,which is a major obstacle for manual operators to quickly inspect the floor.A method based on FSM electric field fingerprint method is introduced to classify raw sensor da-ta using a fuzzy inference system for detecting defects in large liquefied natural gas storage tanks.Feature signals are ex-tracted by adopting the latest filtering and denoising chips and designing specialized signal processing circuits with the calculation software.A real-time measurement database is established and the data are analyzed dynamically.A fiting graph based on the analysis results is generated.The results indicate that the designed industrial tank defect detection fuzzy classifier has achieved the expected effect,and the establishment of its industrial platform will have great signifi-cance for adding value of the oil and natural gas industry.

关 键 词：电场指纹法 模糊推理系统 分类器 缺陷检测 

分 类 号：TH878[机械工程—仪器科学与技术]