Open AccessDiscovery of Incomplete Diagnostic Model based on Learning
Author(s) -
Xiaoyu Wang,
Chuang Li,
Y. T. Liang
Publication year - 2020
Publication title -
international journal of circuits, systems and signal processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.156
H-Index - 13
ISSN - 1998-4464
DOI - 10.46300/9106.2020.14.21
Subject(s) - premise , computer science , process (computing) , artificial intelligence , machine learning , complete information , data mining , mathematics , philosophy , linguistics , mathematical economics , operating system
The model-based diagnosis uses the common reasoning of offline model and online observation to obtain whether and why faults occur. However, the diagnosis is based on the premise of complete model. Once there are unknown behaviors in the diagnosis process, the diagnosis results will not be obtained. In this paper, a method of incomplete model discovery based on online diagnosis process is proposed: In the online diagnosis process, the data of the complete model are learned and the model is trained and adjusted. When the incomplete behavior is found, the nature of the incomplete behavior is determined according to the historical diagnostic data and online observation data, and the corresponding transition/state/event is generated and added to the model to further obtain the definite diagnosis results.