Linear and Nonlinear Fault Location in Smart Distribution Network Under Line Parameter Uncertainty

• Published in: IEEE transactions on industrial informatics Vol. 17; no. 12; pp. 8308 - 8318
• Main Authors: Mirshekali, Hamid, Dashti, Rahman, Shaker, Hamid Reza, Samsami, Reza, Torabi, Amin Jahromi
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.12.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Distributed generations (DGs) operation modes; Distribution networks; Fault location; gradient descent (GD) method; Machine learning algorithms; Mathematical model; optimization problem; Parameter uncertainty; Particle swarm optimization; State estimation; Time-domain analysis; time-domain fault location (TDFL) method; Uncertain systems
• ISSN: 1551-3203; 1941-0050
• DOI: 10.1109/TII.2021.3067007

The line parameters of the distribution network (DN) may change because of the atmospheric, structural, and operational conditions. The uncertainty of line parameters can compromise the accuracy of the automatic fault location methods. Besides, arc faults (AFs) may happen in the DN. These faults are difficult to locate in the faulty section because of the nonlinear, asymmetric, and random nature of AF current. In this article, we present a new time-domain fault location (TDFL) method to determine the location of the fault in smart power DN under the line parameters uncertainty. In the suggested method, line parameters' accurate values are determined using a mixed gradient descent particle swarm optimization algorithm. The suggested method's performance is investigated with the help of an IEEE 123-node test feeder in MATLAB (R2018b). The effects of parameter uncertainty, distributed generations operation conditions and modes, different types of AF, various fault distances, resistances, and the fault inception angles are studied. For further evaluation of the proposed method's robustness, two practical tests in the laboratory are carried out. The results confirm that the proposed TDFL method has high accuracy.