Bousaadia Baadji

Fault detection, classification, and location prediction are crucial for maintaining the stability and reliability of modern power systems, reducing economic losses, and enhancing system protection sensitivity. This paper presents a novel Hierarchical Deep Learning Approach (HDLA) for accurate and efficient fault diagnosis in transmission lines. HDLA leverages two-stage transformer-based classification and regression models to perform Fault Detection (FD), Fault Type Classification (FTC), and Fault Location Prediction (FLP) directly from synchronized raw three-phase current and voltage samples. By bypassing the need for feature extraction, HDLA significantly reduces computational complexity while achieving superior performance compared to existing deep learning methods. The efficacy of HDLA is validated on a comprehensive dataset encompassing various fault scenarios with diverse types, locations, resistances, inception angles, and noise levels. The results demonstrate significant improvements in accuracy, recall, precision, and F1-score metrics for classification, and Mean Absolute Errors (MAEs) and Root Mean Square Errors (RMSEs) for prediction, showcasing the effectiveness of HDLA for real-time fault diagnosis in power systems.

This article presents a novel comprehensive learning bat algorithm (CLBAT) for the optimal coordinated design of power system stabilizers (PSSs) and static VAR compensator (SVC) for damping electromechanical oscillations in multi-machine power systems considering a wide range of operating conditions. The CLBAT incorporates a new comprehensive learning strategy (CLS) to improve microbat cooperation; location updating is also improved to maintain the bats’ diversity and to prevent premature convergence through a novel adaptive search strategy based on relative travelled distance. In addition, the proposed elitist learning strategy speeds up convergence during the optimization process and drives the global best solution towards promising regions. The superiority of the CLBAT over other algorithms is demonstrated via several experiments and comparisons through benchmark functions. The developed algorithm ensures convergence speed, credibility, computational resources and optimal tuning of PSSs and SVCs of multi-machine systems under different operating conditions through eigenanalysis, nonlinear simulation and performance indices.

In conventional transmission line protection, a distance relay is used to provide the primary as well as backup protection. The voltage and current phasors measurement needed by the distance relay for determining the impedance may be affected by the power disturbances such as power swing. Consequently, this power swing may cause mal-operation of Zone three distance relays which in turn may affect on the reliability of the whole protective scheme. Many power swing blocking functions (PSB) have been developed to mitigate these effects. In this paper, a new model of Mho distance relay with dual-quadrilateral power swing detection characteristic has been developed and implemented first in PC using LabVIEW, then tested using Power System Simulink Model under different faults and power swing conditions. Finally, the relay prototype has been realized using acquisition card NI USB-6009, which acquires real-time signals of the currents and the voltages, processes them digitally and outputs tripping or blocking signal to the circuit breaker. The obtained results show that the relay provides good discrimination between a fault and power swing condition.