镨钕合金碳含量软测量研究  

Research on Soft Sensing of Carbon Content in Pr-Nd Alloy

作　　者：陈鑫宇 伍昕宇 刘飞飞[1] 曾波华 涂远泯 曹乐乐 CHEN Xin-yu;WU Xin-yu;LIU Fei-fei;ZENG Bo-hua;TU Yuan-min;CAO Le-le(Jiangxi University of Science and Technology,Ganzhou 341000,China;Jiangxi Ionic Rare Earth Engineering Research Co.,Led.,Ganzhou 341000,China;Ganzhou Nojiferrous Metallurgy Research Institute Co.,Lid.,Ganzhou 341000,China)

机构地区：[1]江西理工大学,江西赣州341000 [2]江西离子型稀土工程技术研究有限公司,江西赣州341000 [3]赣州有色冶金研究所有限公司,江西赣州341000

出　　处：《稀土》2022年第3期111-122,I0005,共13页Chinese Rare Earths

基　　金：江西省重点研发计划(20192BBE50010);江西省重点创新研发平台计划(20181BCD40009)。

摘　　要：为克服工业生产中应用化学分析法和人工质检法对稀土金属碳含量检测的缺陷,模拟人工钻检时依靠对受力大小的感知来判断金属产品是否合格的过程,提出利用自动钻检法检测镨钕合金的碳含量。自主研发微孔钻检实验装置,并通过动态钻削受力模型验证实验装置的可行性和支持向量机(SVM)建立镨钕合金碳含量的识别模型。结果显示,以切削力维度特征和时间维度特征组合训练模型的识别率达94.586%,识别时间在8 ms以内,接受器操作特性曲线(ROC)性能指标FP(将负样品识别为正样品)最小且曲线线下面积(AUC)最大,达到了镨钕合金生产质量在线检测的准确性和实时性的工业要求,为稀土类合金质量检测提供了参考。In order to overcome the shortcomings of the application of chemical analysis and artificial quality inspection methods to the carbon content detection of rare earth metals in industrial production, this paper simulates the process of manual drilling and inspection, and proposes the use of automatic drilling and inspection methods to detect the carbon content of praseodymium-neodymium alloy. The team independently developed the micro-hole drilling and inspection experimental device, and verified the feasibility of the experimental device through the dynamic drilling force model and established the identification model of carbon content of praseodymium and neodymium alloy by Support Vector Machines(SVM). The results show that the recognition rate of the model reaches 94.586%, the recognition time is within 8 ms, the Receiver Operating Characteristic(ROC) curve performance index FP(identifying negative sample as positive sample) is the minimum and the Area Under Curve(AUC) is the maximum, which meets the industrial requirements of accuracy and real time of online quality inspection of Pr-Nd alloy. It provides a reference for the quality inspection of rare earth alloys.

关 键 词：镨钕合金 人工智能 自动检测 动态钻削受力模型 支持向量机 在线检测 

分 类 号：O614.33[理学—无机化学]