Ambiguity groups detection in analog systems diagnostics using Self-Organizing Maps

Piotr Bilski
Abstract:
The paper presents the application of Self- Organizing Maps (SOM) to the ambiguity groups detection in the analog system. This type of neural network is able to find dependencies in data, indicating groups of similar examples in the data set used for training the classifier or the regression machine. Various configurations of the network were implemented and compared. The ability to detect ambiguity groups was verified on the model of the induction machine. Results show the efficiency of the approach, able to identify examples difficult to distinguish by the fault detection and location scheme.
Download:
IMEKO-TC10-2016-057.pdf
DOI:
-
Event details
IMEKO TC:
TC10
Event name:
TC10 Workshop on Technical Diagnostics 2016
Title:

14th IMEKO TC10 Workshop “New Perspectives in Measurements, Tools and Techniques for system’s reliability, maintainability and safety”

Place:
Milano, ITALY
Time:
27 June 2016 - 28 June 2016