Neural network system using the multi-layer perceptron technique for the recognition of PD pulse shapes due to cavities and electrical trees

Abstract

A neural network system, utilizing the multi-layer perceptron approach has been applied to disttnguish between partial discharge pulse shapes that are characteristic of cavities and electrical trees. The neural network was found to be capable of recognizing the differences in PD pulses produced by single cavity and electrical tree discharge sources. It also could differentiate between the discharge pulse forms emanating from electrical trees of different lengths; likewise, it was able to recognize changes in the shape of the discharge pulses with time due to aging effects. However, as these recognition capabilities relate only to comparisons of single discharge sources on a one-to-one basis, the application of neural networks to PD pulse shape recognition on actual cables, where a number of different discharge sources may be discharging simultaneously, is quite premature at this time without more detailed exploratory work on complex discharge Dattems.

Assessment of dielectric degradation by ultrawide-band PD detection

“…One of the problems that have to be overcome is the distortion of the discharge signal in real components. Several possibilities are suggested to recognize the distorted discharge signal, like the use of neural networks [66] and Fourier analysis [74]. Preliminary results of the author of the use of Fourier analysis of distorted pulses are promising.…”

Classification and separation of partial discharge signals by means of their auto-correlation function evaluation

“…Physical considerations, mathematical models and laboratory experiments, have shown that the PD-pulse shape is closely related to the PD mechanisms that can take place in an artificial defect composed of a gas cavity surrounded by a solid dielectric, [19][20][21][22]. Simultaneous signal acquisition and optical recording of PD light emission, performed during aging progression of polyethylene or epoxy resin provided with flat and spherical voids, [23][24][25][26], has shown a close correlation between the PD-signal shape and the PD mechanisms acting during aging of the samples.…”

Partial Discharge Pattern Recognition by Neuro-Fuzzy Networks in Heat-Shrinkable Joints and Terminations of XLPE Insulated Distribution Cables

“…The Multilayer Perceptron (MLP) network with Back Propagation has been applied with good results of convergence [2], [3], [11], [12]. The weak point of this network is the need of prior fixing its topology, thus precluding the possibility of modifying the complexity of the network based on the complexity of the problem under consideration.…”