Modélisation de l’arc électrique dans les disjoncteurs SF6 à haute tension

Abstract

This work presents a new arc model to describe the behavior of the electric arc in high-voltage circuit breakers during the interruption process. Compared to existing models such as HABEDANK, SCHWARZ, KEMA, TP KEMA, and the modified MAYR arc models, the proposed model is simpler and provides a more accurate of measured arc voltage and current waveforms. Accurate modeling of nonlinear arc requires both a reliable arc model and precise parameter estimation. To achieve this, several local and global methods were applied to determine the parameters of the SCHAVEMAKER arc model (taken as example of the black box arc models), some of which have been previously used in literature. Simulation results show that heuristic optimization methods —genetic algorithm, simulated annealing, and particle swarm optimization— are the most effective. For the SCHAVEMAKER arc model specially, the genetic algorithm is best suited for 245kV and 123kV circuit breakers, while simulated annealing is more appropriate for 145kV circuit breakers. AMB (Arc Model Blockset), an extension of MATLAB/Simulink Power Systems Blockset, was updated and improved. AMB is used to analyze the arc-circuit interaction during short-line fault interruption. Previously, AMB was only compatible with MATLAB 6 and earlier versions. AMB was first updated and then improved by modifying the test-circuit and adding four artificial lines and two newly incorporated arc models. These improvements result in faster simulation times, more precise current and voltage simulations, and a more accurate transient recovery voltage (TRV) representation. This work provides a more efficient and reliable tool for studying arc behavior in high-voltage circuit breakers.