Fault Detection and Classification of Time Series Data Using Support Vector Regression and Inception-v3
2018
It is important to detect and classify the cause of abnormalities by monitoring the abnormal status of the periodic signal generated by the equipment of the process system in real time.We propose research methods, support vector regression (SVR), and Inception-v3 to effectively detect and classify outlier data of periodic signals measured at process system facilities.Therefore, in this study, outliers are detected as SVR method and apply the models classified as normal and abnormal to the Inception-v3 model.The real-time process monitoring result graph is predicted as normal or Abnormal through Inception model.