The paper considers the problem of implementing a practical approach to predictive maintenance (maintenance based on the actual technical condition). With this type of service, the state of the system is analyzed continuously or periodically. Based on the data obtained, a forecast of the technical condition of the equipment for a certain period of time is carried out, programs and maintenance plans are formed and, if necessary, adjusted. The aim of the work is to develop a generalized approach to building a predictive service system based on data of some complex technical object, collected by the SCADA system, with their further processing using computer modeling and machine learning methods. Implementation of this approach minimizes the likelihood of an unplanned system shutdown. As a result, it should be noted that to predict the remaining useful lifetime of the unit (RUL), both linear and nonlinear models are studied, including parametric and nonparametric types. Various transformations for the out-put data are tested in order to select the best prediction form. The best form is chosen based on the predictive characteristics of the models.
Technical Diagnostics of Equipment Using Data Mining Technologies
Lect. Notes in Networks, Syst.
International Scientific Siberian Transport Forum ; 2021 May 11, 2021 - May 14, 2021
International Scientific Siberian Transport Forum TransSiberia - 2021 ; Chapter : 178 ; 1613-1622
2022-03-19
10 pages
Article/Chapter (Book)
Electronic Resource
English
Technical Diagnostics of Equipment Using Data Mining Technologies
Springer Verlag | 2021
|A Data Mining Based Approach for Gear Fault Diagnostics Using Vibration Sensors
British Library Conference Proceedings | 2012
|