Noise indicators for corona acoustic signal from power lines - estimation in intensified interference conditions

In the spectrum of corona acoustic signal two
characteristic components can be distinguished: (1) tonal components - higher
harmonics of the network frequency and (2) noise component, observed in the
frequency band extending above 1 kHz. On the other hand the total intensity of
that signal is relatively low - up to 50dB at 15m distance from the outmost
conductor - and it is strongly dependent on the environmental conditions. Thus
the nature of that signal is essentially random. Because of the above-mentioned
features, in practice the best evaluation method of the negative environmental
impact is a continuous signal monitoring. Unfortunately for such a procedure
some problems are encountered with automated extraction of the parameters
(indicators) normally used for evaluation of the signal's acoustic annoyance,
especially in intensified interference conditions. One of the more efficient
methods of filtering out the random interference signals is the statistic
spectrum measurement method. In this work a practical realization is described
for estimation of corona noise indicators, making use of the statistic spectrum
measurement. High efficiency of such an approach has been shown, especially in
favorable weather conditions, when the acoustic signal of corona process is
relatively low, while the random interference level may be intensified.