Competitive learning is an efficient tool for Self
Organizing Maps, widely applied in variety of signal
processing problems such as classification, data
compression, etc.
In the field of data analysis two terms frequently
encountered are supervised and unsupervised clustering
methodologies. While supervised methods mostly deal
with training classifiers for known symptoms,
unsupervised clustering provides exploratory techniques
for finding hidden patterns in data. With the huge volumes
of data being generated from the different systems
everyday, what makes a system intelligent is its ability to
analyze the data for efficient decision-making based on
known or new cluster discovery. The partial discharge
(PD) is a common phenomenon which occurs in insulation
of high voltage, this definition is given in IEC 60270 [1].
In general, the partial discharges are in consequence of
local stress in the insulation or on the surface of the
insulation. This phenomenon has a damaging effect on the
equipments, for example transformers, power cables,
switchgears, and others. The first approach for a diagnosis
is to select the different features to classify measured PD
activities into underlying insulation defects or source that
generate PD’s. In particular for solid insulation like XLPE
on power cables where a complete breakdown seriously
damages the test object the partial discharge measurement
is a tool for quality assessment