Sistema de detecção de falhas de manobras em seccionadores de alta tensão baseado em processamento digital de sinais e RNA

Abstract

The purpose of this dissertation is to present the development of a system capable of detecting incipient malfunction and also verify the correct operation of opening and closing of high voltage disconnectors thru digital analysis of the torque curve supplied by the drive engine during the maneuvers of the disconnector, surpassing the difficulties involved with the direct monitoring of the contacts. This system will be based on the use of dedicated devices to the Digital Signals Processing (DSP's) associated to the algorithms of Artificial Neural Networks, for pattern recognitions of recurrent malfunction during the maneuvers of the high voltage disconnectors. The signals to be processed by the DSP will be the values of the torque of the drive engine, calculated by the immediate values of the electric power supplied. The real-time analysis of the curve of the torque supplied by the engine during the operations of opening and closing will make possible for an indirect monitoring of the position of the main contacts of the high power switch, increasing the trustworthiness of the electrical system’s operation. The information supplied by the detection system of malfunction will allow a more efficient preventive maintenance of the disconnectors, based on the indication of misaligned or broken components in the mechanical system of power switch.