Measurements with uniform aggregated weighting using boxcar filters for time-synchronised metering, power quality assessment, and control

Boxcar (rectangular) window functions are commonly used for measurements of electrical energy for revenue purposes, or for power quality assessment. This is because they can be repeated using Bartlett's method so that the time-domain sampled data is used with a uniform aggregated weighting. However, such simple window functions are not suitable for real-time protection and control measurements due to their poor frequency-domain performance. Raised-cosine or Tukey windows offer a compromise using Welch's method, but real-time implementation of these functions, using windows adaptive to varying system frequency, can be difficult due to computational load. Alternatively, this paper shows that fast-to-execute cascaded boxcar filters can be used to achieve the desirable property of uniform aggregated weighting, as well as their previously-demonstrated frequency-domain capabilities. This means that measurements based on cascaded boxcar filters could be used for formal revenue metering and power-quality assessment, at the same time as for real-time control, phasor measurement, and protection.