A. D. Germano

Characterization of polymeric insulators using thermal and UV imaging under laboratory conditions

This paper presents an efficient monitoring methodology for polymeric insulators. The experiments allow that the correlation of internal and external temperatures of the insulators be implemented. Clean and polluted insulators were also inspected with infrared and ultraviolet cameras, and the influence of well dimensioned corona rings under polluted conditions was evaluated. The electrical behavior of the insulators and corona rings was estimated in simulations using the Finite Element Method.

Ultrasound and Artificial Intelligence Applied to the Pollution Estimation in Insulations

This paper presents an electrical insulator pollution estimation technique based on the ultrasonic noise emitted by these insulators when connected to energized electrodes. The spectral subband centroid energy vectors (SSCEV) algorithm was employed in the signal processing. This algorithm can be understood as being a spectral compression, capable of selecting the most significant frequency bands. The processed audio, changed into SSCEV, constituted a database which was fed into an artificial neural network (ANN), capable of distinguishing with remarkable precision a SSCEV from a polluted insulator from a SSCEV from a less polluted insulator. Finally, in order to validate the technique in the field, measurement campaigns were performed at the Campina Grande 2 substation, belonging to the São Francisco Hydroelectric Company. During these campaigns, ultrasonic noise from several electrical equipments, exposed to different natural pollution degrees, was obtained, and the processing, based on SSCEV and the artificial neural network, was once again applied. As a result, the success rates of more than 80% were generally obtained by the ANN.

Intelligent acoustic detection of defective porcelain station post insulators

This work presents an intelligent acoustic methodology for detection of defective porcelain station post insulators, which are widely used in substations in the form of column presenting several sheds. The acoustic emission inspection aims to detect cracks or fissures in a particular shed, which will have its insulating capacity severely decreased, if cracked. The test is done by gently striking the shed with an appropriate instrument, connected to the tip of an insulated pole. The resulting acoustic emissions are recorded at the substation. A database is created with these audio files and two approaches are considered in order to emphasize the important attributes and to compact the information: Wavelet Energy Coefficients and Spectral Subband Centroid Energy Vectors. Finally, to add reliability, automation and ability to generalize and to adapt to new situations, an Artificial Neural Network is employed. The average classification accuracy is above 62% when using Wavelet Energy Coefficients and above 98% when using Spectral Subband Centroid Energy Vectors.

Methodology to Evaluate the Electromechanical Effects of Electromagnetic Forces on Conductive Materials in Transformer Windings Using the Von Mises and Fatigue Criteria

A methodology based on computer simulations and aiming to estimate the lifespan of transformer windings after the occurrence of short circuits is presented in this paper. The Von Mises and fatigue criteria are employed to determine the failure proximity of the windings. The obtained results contribute to the evaluation of the influence of electromechanical effects in transformer windings and to the determination of the number of short circuits that a transformer winding can support. The simulation results may also be employed in the design of transformers and failure diagnosis, to increase the security margin against mechanical failures.

Ultra-sound and artificial intelligence applied to the diagnostic of insulations in the field

This work studies the feasibility of implementing a system for diagnosis in the field of electrical insulation based on ultrasonic noise and artificial neural networks. Such system, proved functional under laboratory conditions, extracts spectral information from the ultrasonic noise emitted by the corona discharges that occur in electric equipment and correlates it with degrees of pollution previously defined. To achieve this classification, artificial neural networks are employed. The results show the viability of the method in the field, but they also show that its reliability is proportional to the size and diversity of the available database.