A modular neural network approach to fault diagnosis

Abstract

Certain real-world applications present serious challenges to conventional neural-network design procedures. Blindly trying to train huge networks may lead to unsatisfactory results and wrong conclusions about the type of problems that can be tackled using that technology. In this paper a modular solution to power systems alarm handling and fault diagnosis is described that overcomes the limitations of "toy" alternatives constrained to small and fixed-topology electrical networks. In contrast to monolithic diagnosis systems, the neural-network-based approach presented here accomplishes the scalability and dynamic adaptability requirements of the application. Mapping the power grid onto a set of interconnected modules that model the functional behavior of electrical equipment provides the flexibility and speed demanded by the problem. After a preliminary generation of candidate fault locations, competition among hypotheses results in a fully justified diagnosis that may include simultaneous faults. The way in which the neural system is conceived allows for a natural parallel implementation.

Training fuzzy logic based software components by combining adaptation algorithms

“…Hence, there is a need to automatically generate test (fault) scenarios, especially under a time-limited situation. Researchers have applied adaptation algorithms in searching fault scenarios, fault identification, and fault diagnosis [18,22,24,25,27]. In this paper, a scernario is a sequence of states where each state (st i ) is a vector of environmental and vehicular attributes (simply called an environment), e.g., road grade , cruising speed y d , road friction , and vehicular mass m. In traditional cruise control, the above three environmental factors, , , and m, contribute to one factor sensed by the engine-mere ''drag''.…”

Sensor fault diagnosis for a class of time delay uncertain nonlinear systems using neural network

“…In [15], some serious challenges were presented to conventional neural-network design procedures in fault detection and diagnosis for some certain real-world applications. The problem of robust model-based diagnosis of process faults was addressed by means of artificial neural networks in [16].…”

Electronic Circuits Diagnosis Using Artificial Neural Networks

“…A production rule based concept was reported in (Pipitone et al, 1991). ANNs have previously been applied to diagnosis (Spina & Upadhyaya, 1997;Materka, 1994;Rodrigez et al, 1994;Aminian & Aminian, 2000;He et al, 2002;Andrejević & Litovski, 2004;Aminian et al, 2002;Stopjakova et al, 2004;Yu et al, 1994;Collins et al, 1994;Catelani & Gori, 1996;Maidon et al, 1997;Yang et al, 2000). As in the case with the classical concepts, however, ANNs were predominantly applied to linear analogue circuits.…”

“…In (Materka, 1994) feed-forward ANNs were used for parameter identification (soft fault diagnosis) of linear circuits. In (Rodrigez et al, 1994) linear power networks were diagnosed by feed-forward ANNs. In order to enhance the performance of the ANN applied for diagnosing of soft faults in linear active networks, in (Spina & Upadhyaya, 1997), new "criteria" -a discriminating measure based on discrepancy of the autocorrelation function of the faultfree and the correlation function of the faulty and fault-free circuit, were introduced.…”