Classification of Short‐Circuit Faults in High‐Voltage Transmission Lines Using Energy‐Based and Statistical Indices Extracted From Hilbert and Wavelet Transforms of Three‐Phase Terminal Voltage and Current Signals

ABSTRACT

Accurate classification of short‐circuit faults in high‐voltage transmission lines, including identification of the faulted phase(s) and determination of ground‐involved or ungrounded faults, is a fundamental component of modern protection schemes. The correctness of fault classification directly influences relay decision‐making, single‐pole tripping strategies, and overall power system reliability. This paper presents a data‐driven, decision‐orientated, single‐terminal method for comprehensive classification of short‐circuit faults, developed based on voltage and current signals measured at one end of the transmission line. The theoretical foundation of the proposed approach relies on the physical principle that the faulted phase(s) exhibit the largest relative amplitude variations, the highest transient energy, and the most pronounced statistical asymmetry in transient signals. Accordingly, by simultaneously employing the Hilbert transform to extract instantaneous amplitude variations and the discrete wavelet transform to obtain transient energy features, a set of closed‐form, normalised, and scale‐independent indices is defined. These indices are then integrated within a hierarchical, rule‐based decision‐making framework. The designed decision logic enables coherent and interpretable detection of fault occurrence, identification of the number of faulted phases, and discrimination between grounded and ungrounded faults. The performance of the proposed method is evaluated through extensive simulations conducted in the MATLAB environment, encompassing both specific case studies and a wide range of general scenarios with variations in fault type, fault location, fault resistance, and network operating conditions. In addition, multiple sensitivity analyses are carried out to assess the robustness and reliability of the algorithm under challenging conditions. The obtained results demonstrate that the proposed algorithm achieves high accuracy, satisfactory robustness, and strong generalisation capability, successfully classifying different types of faults and confirming the effectiveness and validity of the proposed decision‐making framework for protection applications in high‐voltage transmission lines.