This article proposes a new approach to predicting failures of high-tech manufacturing equipment using machine learning methods. As part of the study, various modern approaches for solving similar problems were analyzed, their advantages and disadvantages relative to the proposed method were described. Model training data includes sensor readings measuring temperatures, pressures, currents, and vibrations for six exhausters over three years. In the process of data preprocessing, standardization, noise removal and extraction of new features were carried out. Since blower failure is a rare and short-term occurrence, the use of supervised learning methods is difficult. In this regard, it is necessary to introduce an additional stage of screening out periods of normal operation. Thus, the process of early diagnosis of failures consisted of two successive stages. Firstly, anomalous operating modes of equipment were identified using the isolation forest method. Next, intervals were classified using a recurrent neural network, defined as anomalous, according to the technical locations that caused the equipment failure. As a result, on the test sample it was possible to achieve an average value of the F1 metric equal to 0.74 for all analyzed technical locations.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.