Nonlinear dynamical study of partial discharge phenomena

Some studies have been conducted in recent years that investigate partial discharge phenomena from the viewpoint of deterministic chaos, by analyzing time series data on discharge magnitude. Here we have analyzed the chaotic nature of partial discharges from another viewpoint, namely, interdischarge intervals. The observed time series data have been analyzed with such statistical indicators of deterministic chaos as the maximum Lyapunov exponents, the fractal dimension and so on. Data of discharge intervals were measured with an IEC (b) electrode system. The maximum Lyapunov exponents calculated from these data were estimated to be positive values. Additionally, the calculated correlation dimension showed non-integer value, and also autocorrelation rapidly converged on a near zero value. This result implies chaotic characteristics in partial discharge phenomena in an IEC (b) electrode system