Fault Diagnosis and Bad Data Detection of Power Transmission Network - A Time Domain Approach

TL;DR

This paper introduces a linear observer-based method that simultaneously detects bad data and identifies faults in power transmission networks by modeling faults as control failures, enabling improved fault diagnosis and data validation.

Contribution

It presents a novel control failure modeling approach and observer design that jointly detects bad data and locates faults in transmission lines, with fault resistance estimation as a byproduct.

Findings

Effective fault location independent of fault resistance

Simultaneous bad data detection and fault diagnosis

Residuals exhibit fixed directional characteristics

Abstract

Fault analysis and bad data are often processed in separate manners. In this paper it is proved that fault as well as bad current measurement data can be modeled as control failure for the power transmission network and any fault on the transmission line can be treated as multiple bad data. Subsequently a linear observer theory is designed in order to identify the fault type and bad data simultaneously. The state space model based observer theory allows a particular failure mode manifest itself as residual which remains in a fixed direction. Moreover coordinate transformation is performed to allow the residual for each failure mode to generate specific geometry characteristic in separate output dimensions. The design approach based on the observer theory is presented in this paper. The design allows 1) bad data detection for current measurement, and 2) fault location, and fault resistance estimation (as a byproduct) where the fault location accuracy is not affected by fault resistance. However it loses freedom in designing the eigenvalues in the excessive subspace. While the theoretical framework is general, the analysis and design are dedicated to transmission lines.