Fault Diagnosis of Power System Considering Communication System Information Transmission Error

The existing fault diagnosis systems of power grids are subject to uncertainties in fault information, such as transmission errors in communication channels, thus causing poor practical application. To address these problems, in this paper, we describe an information model for fault diagnosis based on a real communication channel that combines with an expert system using information theory. Here the `traditional' expert system based on strict rule-matching was modified to a numerical computation method, whereby the process of diagnosis is not influenced significantly by signal chronology; this allows information transmission uncertainties to be described quantitatively. The fault diagnosis system for a power grid based on the algorithm proposed was applied successfully in a practical area control center. Our results showed that this method can solve ambiguities in information and the uncertainties of real communications. Despite interference in telecontrol communications, the diagnostic results given by the proposed method were reasonable.