Challenge and objective

  • With power electronic equipment and DG, the power system has become more prone to harmonic pollution.
  • As isolated systems categorized by low inertia such as microgrids are more common, the presence of nonlinear distortion is becoming an increasing problem.
  • Commonly used methods for surveillance are thus not suited for the rising non-linearity caused by the harmonics, and there is a need for alternative surveillance methods.

Work performed

  • This thesis has explored the use of adaptive data analysis as an alternative surveillance method for harmonic detection in the power system.
  • Empirical Mode Decomposition (EMD) and its real-time extension, Online EMD, have been used in conjunction with Hilbert-Transform (HT) and Fast Fourier Transform (FFT) for instantaneous frequency and ampl. identification.

Significant results

  • The methods proved to be powerful tools for harmonic detection when supported with techniques to handle mode mixing on more complex signals.
  • The use of masking signals turned out to be a highly effective mode mixing separation technique

Impact for distribution system innovation

  • Measurement based techniques can identify special characteristics of equipment or systems when modelbased techniques may fail due to parameter uncertainty.
  • Possible to identify the connected equipment and sources based on their response to disturbances.


Oddbjørn Gjerde

WP2 Lead
+47 99 730 027
Oddbjørn Gjerde
WP2 Lead


Reference in CINELDI