MartaVanin
commented
4 years ago Papers present pretty standard result formats: a table with mean and max error between the results of the estimator and a power flow calculation (see Table II and III from https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8416770). The errors typically refer to the buses' voltage magnitudes and angles. If the test feeder does not have too many buses and there is room enough, sometimes the errors at each bus are plotted with a scatter+line plot (see Fig. 2 from https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8416770). The plot perhaps makes more sense if you show comparison between different methods.
Dzafic et al (https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7516595) do something slightly different: they show errors in tables but also report the absolute quantities (in kW/A,..). However, they don't show max/mean error across all the buses in the feeder but : 1) those between the measurements they gave as input and their estimated value, 2) those at the fancier components (transfos, etc..), whose modeling was kind of the selling point of the paper.
I find the first approach more the norm, while what Dzafic et al. do is a bit exotic..
Sporadically, computation time appears in tables, too.
We should thoroughly review the necessary post-processing features for v0.1, starting from the analyses regularly used in literature.