Uma técnica de diagnóstico em tempo real de distúrbios transitórios baseada na transformada walvelet para uso em registradores digitais de perturbação.

Abstract

This thesis presents a transient disturbance diagnostic method based on the real-time analysis of the wavelet coefficient energies of both the voltages and currents. The method detects the main transient disturbances in transmission power systems, such as faults, developing faults, voltage sags, and switching transients. The method also detects the beginning and end times of the disturbances, and classifies the faults and developing faults in real-time. A new wavelet-based digital fault recorder with the proposed method as triggering method was implemented in the Real Time Digital Simulator (RTDSTM) in order to detect and record only oscillographic data with transient disturbances. In this monitoring device, the disturbance diagnostic information is recorded into the digital channels of the oscillographic data. The performance of the proposed method was real-time assessed with two power system models with faults and other transient disturbances and good results were achieved. The performance of the method was also off-line evaluated with more than two thousand actual oscillographic records from digital fault recorders of the Chesf power system and good results were also obtained. The fault classification module was assessed with various fault resistances, inception angles, and locations and good results were obtained. The distinctive feature of the method is that it could be used to detect and classify the main transient disturbances in nonspecific power system topologies, with different sampling rates and rated voltages, and in both the real-time and off-line applications.