On the reliability of real measurement data for assessing power quality disturbances

This paper is primarily focused on assessing the reliability of real recorded data, which is used for several applications in analyzing power quality disturbances. The prime goal becomes, for that reason, to develop a set of reliability criteria to improve the data selection. Some of the identified reliability indices are such as quantization noises, frequency resolution, energy levels, data dispersion, frequency-domain coherence and time-domain correlation. As a motivation for the presented theory, two major applications are described in this paper: the estimation of power system parameters at harmonic frequencies, and the identification of interharmonic polluting loads. These criteria are explained in detail as a guide for rejecting unreliable data used when analyzing power quality disturbances.