Low voltage customer phase identification methods based on smart meter data

Combined unbalanced distribution system state and line impedance matrix estimation

“…The proposed methods assume that the phase connectivity of all users is known, as well as the general topology, i.e., the network tree. Reliable methods exist that perform phase [32] and topology estimation [6] with SM only, i.e., without relying on line impedance data. Furthermore, we develop a case study in which we recover the impedance of all lines from scratch, using historical measurements.…”

Phase Identification of Distribution System Users Through a MILP Extension of State Estimation

“…For underdetermined systems, other methods than the proposed one might be preferable. In particular, voltage-based methods exist that can determine the connectivity of a few SMs when most of the system is not monitored [24]. However, PI cannot be performed for non-monitored connections, so a full picture of the system cannot be obtained, which has a negative impact on any calculation reliant on network data.…”

“…These methods are not affected by errors in the network impedances, and are also used as a benchmark in [30]. The first relies on voltage measurements only: Pearson correlation is calculated between the voltage measurement time series of each user and each transformer phase [20], [24]. The user is then assigned to the transformer phase with the highest correlation.…”

“…In general, voltage-based methods are less accurate for short and balanced feeders [18]. Thorough comparisons of different statistical and clustering methods for PI can be found in [23], [24].…”

Impact of line length, EV and PV on LV Grid topology identification based on voltage correlation

“…Methods based on smart meters data typically use voltage measurement with clustering algorithms [3][4][5][6]. This set of methods is scalable and does not require the installation of any additional equipment, but it was only applied to phase clustering so far.…”