Thesis and Doctoral Defense Session No. 65 of the PPGEEC - Application of EMD and VMD Associated with the Teager Energy Operator in Fault Location in Transmission Lines Based on Traveling Wave Theory

 

STUDENT: ANDRESSA PEREIRA OLIVEIRA

 

DATE: August 19, 2026

 

TIME: 9:00 AM

 

LOCATION: https://conferenciaweb.rnp.br/users/fernando-augusto-moreira/personal_room

 

TITLE: Application of EMD and VMD Associated with the Teager Energy Operator in Fault Location in Transmission Lines Based on Traveling Wave Theory

 

KEYWORDS: Fault location; Variational decomposition of modes; Teager energy operator; Traveling wave theory; Transmission lines.

 

ABSTRACT: Locating faults in electrical power systems is a challenging problem, as is the accurate identification of their occurrence and the rapid restoration of power supply. Different digital signal processing techniques (conventional and combined with artificial intelligence) have been used in recent decades to better characterize faults and improve the systems studied. This work proposes a fault location methodology based on traveling wave theory, using the analysis of characteristic data from voltage or current signals obtained from one or two reading terminals of the electrical power transmission system. For processing the read signals, the variational mode decomposition (VMD) algorithm associated with the Teager energy operator (VMDT) is used to improve the identification of peaks equivalent to the arrival times of traveling waves to estimate the location of the occurrence. The power system simulation was performed using the Alternative Transients Program (ATP) software (for simulating the studied test systems) and MATLAB/Simulink (for processing the fault location algorithm), using the Bergeron and JMarti models, which are implemented assuming an ideally transposed transmission line. The two test systems consist of a transmission line between two Thevenin equivalents, with lengths of 200 km, 210 km, and 360 km. The effects of the main parameters that can affect the accuracy of fault location are analyzed in this study based on their variations for all types of faults: fault resistance, angle of incidence, sampling rate (20 kHz, 200 kHz, 1 MHz, 2 MHz, and 5 MHz), and fault occurrence locations (every 10% of the line length or at other specific points). To analyze the behavior of the applied algorithm in the face of the influence of noisy signals affecting electrical signals, the variation of the signal-to-noise ratio is applied using different values. The results obtained demonstrate 100% accuracy for fault detection and classification, and that the proposed methodology (VMDT) can extract characteristic data from faulty voltage/current signals and estimate the fault location with high precision (errors less than 1% for the two-terminal analysis for most fault types, especially single-phase faults), regardless of the variables studied. The case studies performed are compared with results obtained using the wavelet transform, the DS filter, the standard EMD (empirical mode decomposition) model (the precursor technique to VMD), EMD associated with the Teager energy operator, and standard VMD, to demonstrate the efficiency of the proposed methodology and the influence of the Teager energy operator on accuracy.

 

EXAMINING COMMITTEE MEMBERS:

 

FERNANDO AUGUSTO MOREIRA (ADVISOR) - UFBA (President)

 

EDUARDO FURTADO DE SIMAS FILHO - UFBA

 

FREDERICO OLIVEIRA PASSOS - UNIFEI

 

RICARDO CANELOI DOS SANTOS - UFABC

 

ALESSANDRA FREITAS PICANÇO - IFBA

 

ALMIR LARANJEIRA NERI JUNIOR - SENAI/CIMATEC

Em 18/08/2026