Development Of Fault Diagnosis Techniques Research Articles

Model‐based fault diagnosis of networked systems: A survey

AbstractThis paper reviews the latest development of fault diagnosis techniques of networked systems. Considering various challenges raised by unreliable local sensors, signal digitization, data transmission, as well as distributed architecture, recent progress on model‐based fault diagnosis of networked systems are overviewed. We summarize the typical application examples and the typical treatment methods related to scientific problems for better understanding the framework of this area. Some promising future fault diagnosis techniques in various scenarios, such as distributed systems, with communication protocols, with cyber security issues, with intelligent algorithm challenges, as well as within active fault diagnosis framework are discussed.

Improved few-shot learning method for transformer fault diagnosis based on approximation space and belief functions

Incomplete and uncertain information is frequently observed in the data analysis processes, which has become one of main challenges for the development of fault diagnosis techniques of transformers. To address few fault cases and deficient monitoring information in diagnostic tasks, this paper provides an improved few-shot learning method based on approximation space and belief functions to accomplish fault diagnosis of transformers. The decision-making table, as an efficient structure to map the weakly correlated attributes, is extracted from transformer cases and maintenance experience. Then the approximation space is used to describe attribute correlations between diagnostic rules and the diagnostic task. We employ the 0.5-approximation set strategy to obtain the diagnostic results when the information is sufficient. Furthermore, we propose a modified basic probability assignment (BPA) calculation method to build belief functions for diagnosis when information is scanty. The modified method is verified capable of improving the decision-making reliability. The overall recognition accuracy of fault diagnosis by our improved few-shot learning algorithm is over 87% which is higher than other four peer methods. This method also shows a potential for good expandability when new diagnostic rules of transformers are discovered.

Deductive fault diagnosis in digital circuits: a survey

Fault diagnosis in digital circuits is normally based on prior computation of fault symptoms using explicit fault models and simulation followed by matching of the observed symptoms of a faulty circuit with one of the sets of precomputed symptoms. Some attempts have been made, however, at diagnosis without fault simulation by deducing the location of a fault or faults from the observed symptoms and from fault-free models of circuit behaviour only. The paper traces the early development of fault diagnosis techniques which avoided fault simulation, and then reviews the published literature on deductive techniques. Three main methods of deductive fault diagnosis are reviewed and all are shown to have limitations, some due to the use of the stuck-fault model and others due to the combinatoric problems associated with multiple faults and the resulting multiplicity of choices which an algorithm must pursue. The problem of test selection is discussed together with the choice, in future work, of processing the results of one test at a time and then combining the diagnoses from each test, as in two of the methods reviewed, or of processing a multiple set of tests together as in the other main method reviewed.