摘要
针对复杂工业过程的微小故障诊断问题,提出一种数据预处理与重构贡献图相结合的故障诊断方法;为了克服非高斯分布数据对故障检测准确性的影响,通过基于数据变化率的方法对样本原始数据进行预处理后,可以有效地检测过程变量的微小故障,以此建立故障诊断主元分析模型;检测出系统故障后,为了提高故障辨识准确度,采用一种平均残差差值重构贡献图的方法对故障进行辨识;通过正常样本数据和故障数据在残差子空间中的投影,获取两个数值的残差差值向量,计算重构贡献值来确定故障变量;以田纳西-伊斯曼(TE)过程为对象进行了故障诊断仿真实验,并与传统贡献图和重构贡献图方法的辨识准确率相比较,结果表明所提方法具有良好的故障诊断性能。
A fault diagnosis method is proposed for tiny faults in complex industrial processes,which combines data preprocessing with reconstruction contribution mapping.In order to overcome the impact of non-Gaussian distribution data on fault detection accuracy,the original data of process variables is effectively detected by preprocessing the samples based on data change rate,thus establishing a fault diagnosis principal component analysis model.In order to improve the accuracy of fault identification after the system fault detection,a method of reconstructing contribution mapping of average residual difference is used to identify the fault.Two residual difference value values are obtained by projecting the normal sample data and fault data in the residual subspace,and the reconstruction contribution value is calculated to determine the fault variable.Fault diagnosis simulation experiments are conducted by taking the Tennessee Eastman(TE)process as a object,and the identification accuracy is compared with traditional contribution mapping and reconstruction contribution mapping methods.The results show that the proposed method has a good fault diagnosis performance.
作者
赵凯阳
张家良
ZHAO Kaiyang;ZHANG Jialiang(School of Electronic Information Engineering,Xi'an Technological University,Xi'an 710021,China)
出处
《计算机测量与控制》
2023年第12期14-20,共7页
Computer Measurement &Control
基金
陕西省自然科学基础研究计划项目(2023-JC-YB-579)。
关键词
主元分析
故障诊断
贡献图
田纳西-伊斯曼过程
数据变化率
principal component analysis
fault diagnosis
contribution plot
Tennessee Eastman process
data change rate
