ARTIFICIAL NEURAL NETWORKS FOR RECOGNITION OF 3-D PARTIAL DISCHARGE PATTERNS

Artificial neural networks have been used very successfully to address many real world pattern recognition problems. The ability to learn from examples, capability of generalization and abstraction, and the ease with which it can be adapted to a given problem domain, when compared to other artificial intelligence paradigms, are the prime reasons for its popularity. Its recent availability in hardware form, as Neural Chips, has further enhanced the prospects of its real time implementation to practical problems. Partial discharge (PD) measurements have been carried out over the years to assess insulation systems in power apparatus for their integrity and design deficiencies. Digital PD recording, processing and its presentation as 3-d patterns is the recent trend in both industry and testing laboratories. Interpretation of these patterns can lead to evaluation of the cause of PD. This task requires much expertise, which is frequently not available, since experts are few, and also takes a long time to consolidate. Therefore, a need arises to look for methods in the domain of pattern recognition for automating this process. In this context, this paper presents results to demonstrate the possibility of using pattern recognition capabilities offered by a multilayer neural network to recognize 3-d PD patterns.