A method of mapping suitable for a plurality of electricity meters to map which terminal of each electricity meter is connected to which nominal phase of a multi-phase electrical grid

Abstract

The present invention refers to a method of mapping suitable for a plurality of electricity meters (M1–Mn) to map which terminal (T1–T3) of each electricity meter (M1–Mn) is connected to which nominal phase (L1-L3) of a multi-phase electrical grid (1). The mapping is performed for clusters of electricity meters (M1–Mn) separately, wherein the clusters are determined on the basis of a proximity matrix, wherein the electricity meters (M1–Mn) of each cluster share a common connection line (11) to a shared distribution substation (3) of the multi-phase electrical grid (1).

Description

[0001] Applicant: Kamstrup A / S

[0002] Title: A method of mapping suitable for a plurality of electricity meters to map which terminal of each electricity meter is connected to which nominal phase of a multi-phase electrical grid

[0003] Our Ref.: KAP 3795 WO

[0004] Description

[0005] TECHNICAL FIELD

[0006]

[0001] The present invention relates to a method of mapping suitable for a plurality of electricity meters to map which terminal of each electricity meter is connected to which nominal phase of a multi-phase

[0007] 5 electrical grid, based on an analysis of voltage variables measured by and collected from the electricity meters. The present invention is applicable to multi-phase electrical grids having an inhomogeneous grid topology.

[0008] BACKGROUND

[0009] 10

[0002] Multi-phase electrical power distributing grids, such as three- phase distribution systems, are used for delivering electrical power to consumers using a plurality of phases or wires connected to points of use.

[0010]

[0003] Typically, the power consumption at each point of use is metered using electricity meters connected with the different phases of the grid. Normally, each phase is connected to a separate terminal of the electricity meter and the amount of electricity delivered to the respective point of use via each of the three phases is metered by track¬

[0011] 20 ing voltage and current over time for each phase. It should be noted

[0012] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 that the current load on a multi-phase electrical power distributing grid may be different among the phases of the grid. One phase may currently suffer from a high load while the capacity of another phase is currently not fully used. For a utility provider, it is desirable and a tech¬

[0013] 5 nical challenge to balance the load among the phases of a multiphase electrical power distributing grid in reaction of changing and unpredictable consumption behaviour. The problem of load balancing becomes even more difficult as more and more distributed energy sources, e.g. photovoltaic panels, wind turbines, biomass power plants, etc., are integrated into the electrical grid during the transition process from fossil-based energy sources to renewable energies. Depending on the current sunshine and / or wind intensity, a consumer may become a supplier generating electric power and feed that generated electric power back into the multi-phase electrical power distributing grid. An

[0014] 15 electricity meter may be used to measure the amount of electric power consumed and / or provided.

[0015]

[0004] A utility provider may use the information from the electricity meters in the multi-phase electrical power distributing grid to measure the

[0016] 20 current load on each phase. However, it is a technical challenge to find out which terminals of the electricity meters are connected to which nominal phase of the grid. So, if one incorrectly assumed, for example, that the first terminal, e.g. called Tl , of each electricity meter, was connected to a first nominal phase, e.g. called Pl, one would in fact experience an unknown arbitrary mix of the phases, depending on the phases the terminal Tl of each electricity meter is actually connected to. The same applies for the other terminals, e.g. T2 and T3.

[0017]

[0005] Hence, a method of mapping is needed for a plurality of elec¬

[0018] 30 tricity meters to map which terminal of each electricity meter is connected to which nominal phase of a multi-phase electrical grid. An example of a known method is given in EP 3449267 Bl . However, it is a problem of the known method that the phase-identification may be

[0019] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 not reliable and robust enough for inhomogeneous grid topologies, e.g. grid topologies with large differences of cable lengths between the electricity meters and their associated shared distribution substation of the multi-phase electrical grid.

[0020] 5

[0021]

[0006] An object of the present invention is thus to provide a method of mapping suitable for a plurality of electricity meters to map which terminal of each electricity meter is connected to which nominal phase of a multi-phase electrical grid, wherein the method is more reliable and robust for inhomogeneous grid topologies.

[0022] SUMMARY

[0023]

[0007] The inventive solution to the above-mentioned problem is given by the subject-matter of the independent claims. Preferred embodi¬

[0024] 15 ments of the invention may be deduced from the dependent subclaims, the description and the figures.

[0025]

[0008] According to a first aspect of the present invention, a method of mapping is provided that is suitable for a plurality of electricity meters to

[0026] 20 map which terminal of each electricity meter is connected to which nominal phase of a multi-phase electrical grid, wherein the method comprises the steps of: establishing a proximity matrix to determine clusters of electricity meters, wherein the electricity meters of each cluster share a common connection line to a shared distribution substation of the multi-phase electrical grid, and mapping, for each cluster separately, which terminal of each electricity meter of the cluster is connected to which nominal phase of the multi-phase electrical grid.

[0027] 30 This has the advantage that the mapping is much more reliable and robust for inhomogeneous grid topologies. This is, because the electricity meters within a cluster show a similar “signature” of normalised voltage

[0028] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 variables. If electricity meters of different clusters were mixed in the mapping, the mapping would be much less reliable and much less robust.

[0029] 5

[0030]

[0009] Optionally, the method may comprise: a) preselecting a group of the electricity meters that are associated with one shared distribution substation of the multi-phase electrical grid;

[0031] 10 b) in each electricity meter of the preselected group of electricity meters, measuring and storing voltage variables induced on the terminals of the respective electricity meter; c) collecting the voltage variables from each electricity meter of the group of electricity meters, preferably by a data collection system; d) for each terminal of each electricity meter of the preselected group of electricity meters, calculating a set of normalised voltage variables based on the voltage variables; characterised in that the method further comprises:

[0032] 20 e) for each electricity meter of the preselected group of electricity meters, selecting said electricity meter as a base electricity meter; f) for each selected base electricity meter, defining a solution space of lower dimensionality compared to a variable space defined by the normalised voltage variables of the selected base electricity meter; g) for each solution space, projecting the normalised voltage variables of each terminal of each electricity meter of the preselected group of electricity meters onto a terminal point in the respective solution space;

[0033] 30 h) determining the proximity matrix comprising proximity values being defined by the terminal points in each solution space;

[0034] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 i) identifying the clusters of electricity meters that share a common connection line to the distribution substation based on the determined proximity matrix; and j) wherein the step of mapping comprises

[0035] 5 one or more terminals of one or more of the electricity meters of said cluster are selected as cluster base terminals, a cluster solution space is defined being of lower dimensionality compared to a variable space defined by the normalised voltage variables of the selected cluster base terminals,

[0036] 10 the normalised voltage variables of each terminal of each electricity meter of said cluster are projected onto a cluster terminal point in the cluster solution space, and mapping which terminal of each electricity meter of said cluster is connected to which nominal phase of the multiphase electrical grid based on the cluster terminal points in the cluster solution space.

[0037]

[0010] It should be noted that the nominal phases can be arbitrarily defined to distinguish among the physical phases of the multi-phase elec¬

[0038] 20 trical grid. It may not be important for a utility provider to know which nominal phase corresponds to which physical phase. It may be enough to know the nominal phases to distinguish among them for taking a switching action, for example.

[0039]

[0011] The clustering of the electricity meter may use a form of hierarchical clustering based on an appropriate “distance”, e.g. a Euclidian distance of the terminal points in the solution space, as a similarity parameter. An electricity meter may be arbitrarily chosen to serve as the “base electricity meter” for as long as it is temporarily used as a refer¬

[0040] 30 ence electricity meter to define the base-specific solution space. This means that the distances of the terminal points in the solution space are determined with reference to that electricity meter which is temporarily considered to be the base electricity meter. In other words, the

[0041] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 origin of the base-specific solution space is defined by the terminal points of the base electricity meter. The solution space is defined by all possible terminal points of the other electricity meters with respect to the base electricity meter. So, the distance of an electricity meter to

[0042] 5 the origin of the base-specific solution space is the “distance” to the base electricity meter. A “terminal point” shall be defined herein as a point in the solution space onto which the normalised voltage variables are projected for each terminal of an electricity meter. The “proximity matrix” is a matrix of the proximity or distance values defined by the lo¬

[0043] 10 cation of the terminal points in each base-specific solution space. The “clusters” of electricity meters are defined, preferably by hierarchical clustering, to be those electricity meters that are identified to share a common connection line to the distribution substation.

[0044]

[0012] Once the clustering is done, the mapping is performed for each cluster separately to map which terminal of each electricity meter is connected to which nominal phase of a multi-phase electrical grid. The terminals of a currently selected base electricity meter are then denoted as “cluster base terminals”. The normalised voltage variables of

[0045] 20 the selected cluster base terminals then define the “cluster solution space”. The projection of the normalised voltage variables of the terminals of the other electricity meters of the cluster are then denoted as “cluster terminal points” in the cluster solution space.

[0046]

[0013] Optionally, the method may further comprise determining a load on a nominal phase of the multi-phase electrical grid based on voltage measurements at those terminals of the electricity meters that are mapped to the same said nominal phase of the multi-phase electrical grid. This is very beneficial to make further use of the more reliable and

[0047] 30 robust mapping which terminal of each electricity meter is connected to which nominal phase of a multi-phase electrical grid.

[0048] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025

[0014] Optionally, the method may further comprise: identifying a load imbalance on a nominal phase of the multiphase electrical grid based on the determined load on at least two of the nominal phases of the multi-phase electrical grid; and

[0049] 5 balancing the identified load imbalance by switching nominal phase connections of at least one power consumer and / or power source in the multi-phase electrical grid.

[0050] This is very beneficial to make further use of the more reliable and robust mapping which terminal of each electricity meter is connected to which nominal phase of a multi-phase electrical grid. The load imbalance may be continuously or regularly identified and balanced on-the- fly, so that load imbalances can be kept low before they grow to a more severe problem.

[0051]

[0015] Optionally, the method may further comprise executing the

[0052] 15 steps a) to j) for another preselected group of electricity meters that are associated with another shared distribution substation of the multiphase electrical grid. Thereby, it is possible to perform the mapping for all or many of the electricity meters installed in the multi-phase electrical grid.

[0053] 20

[0054]

[0016] Optionally, the method may further comprise identifying those electricity meters among the group of electricity meters that are incorrectly associated with the shared distribution substation of the multiphase electrical grid if a majority of non-diagonal proximity values in at least one row and / or column of the determined proximity matrix are below a pre-determined threshold, wherein said at least one row and / or column of the determined proximity matrix corresponds to those electricity meters that are incorrectly associated with the shared distribution substation of the multi-phase electrical grid. As an alternative to

[0055] 30 a pre-determined threshold, or in addition, other known methods to identify distinct outliers may be applied, e.g. based on a deviation from

[0056] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 an expectation value. Identifying distinct outliers is very beneficial to correct an incorrect database of a head-end-system (HES) that associates an electricity meter with a wrong distribution substation. The nondiagonal proximity values in at least one row and / or column of the de¬

[0057] 5 termined proximity matrix that are below the pre-determined threshold are an indication that the associated electricity meter experiencing such a different “signature” of normalised voltage variables compared to the other electricity meters is very likely connected to another distribution substation of the multi-phase electrical grid. Such incorrectly as¬

[0058] 10 sociated electricity meters may be excluded from the cluster for the mapping and included into another group that is preselected in step a) for being associated with another shared distribution substation.

[0059]

[0017] Optionally, the proximity values of the determined proximity matrix may be indicative of a Euclidian distance of the terminal points in the solution space, or of an angular distance of the terminal points in the solution space, or a weighted combination thereof. The less distance (in length and / or angle) the terminal points have in the solution space, the more likely it is that the terminals are connected to the same

[0060] 20 nominal phase, because these terminals experience a similar “signature” of normalised voltage variables. Vice versa, a large distance (in length and / or angle) between the terminal points in the solution space indicate that the terminals are connected to different nominal phases, because these terminals experience a different “signature” of normalised voltage variables.

[0061]

[0018] Optionally, each solution space may be spanned by principal components of the normalised voltage variables of the respective base electricity meter, and / or wherein the cluster solution space may be

[0062] 30 spanned by principal components of the normalised voltage variables of the respective cluster base terminals. The dimension reduction may thereby be a reduction to only two dimensions, i.e. each terminal point may be defined by two components of a vector in the solution space.

[0063] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025

[0019] Optionally, the normalised voltage variables may be mean free voltage change, MFVC, variables based on time derivatives of the voltage variables. The MFVC variables are particularly well suited to estab¬

[0064] 5 lish a “signature” of normalised voltage variables.

[0065]

[0020] Optionally, the method may further comprise identifying which terminal of each electricity meter is connected to which physical phase of the multi-phase electrical grid based on a known order in which the terminals of at least one reference electricity meter are connected to physical phases of the multi-phase electrical grid.

[0066]

[0021] According to another aspect of the present invention, a computer program is provided comprising instructions which, when the

[0067] 15 computer program is executed by at least one computer, cause the at least one computer to carry out the steps of the method described above.

[0068]

[0022] Optionally, the computer program may be configured to re¬

[0069] 20 ceiver user input for carrying out one or more steps of the method described above.

[0070]

[0023] According to another aspect of the present invention, an advanced metering infrastructure, AMI, is provided for reading electricity consumption from a plurality of electricity meters in a multi-phase electrical grid, the AMI comprising: a plurality of electricity meters having terminals being connected to nominal phases of a multi-phase electrical grid; a data collection system being configured to collect voltage vari¬

[0071] 30 ables from each of the electricity meters via a communication connection; and at least one computer being part of or in communication connection with the data collection system and / or the plurality of

[0072] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 electricity meters, wherein the at least one computer is configured to carry out the steps of the method described above.

[0073]

[0024] The inventive method, computer program and advanced me¬

[0074] 5 tering infrastructure, AMI, provides a terminal-phase-association for a plurality of electricity meters installed in a multi-phase electrical grid, wherein voltage measurements of the electricity meters are used that are preferably taken and collected anyway for billing purposes. The inventive method, computer program and AMI does not require additional expensive measuring equipment. In an advantageous embodiment, the inventive method, computer program and AMI may be implemented as a retrofit in form of a firmware or software update to existing equipment, e.g. electricity meters and / or a head-end-system (HES). A further advantage of the inventive method, computer program

[0075] 15 and AMI is that a model of the topology of an electricity grind can be created and continuously updated based on voltage variables collected according to routine.

[0076] BRIEF DESCRIPTION OF THE FIGURES

[0077] 20

[0078]

[0025] The method, computer program and AMI according to the invention will now be described in more detail with regard to the accompanying figures. The figures show one way of implementing the present invention and is not to be construed as being limiting to other possible embodiments falling within the scope of the attached claim set. Preferred embodiments of the invention may be deduced from the dependent subclaims, the description and the figures, of which:

[0079] Fig. 1 illustrates schematically a three-phase electrical grid and

[0080] 30 an AMI according to the present invention;

[0081] Figs. 2a, b shows schematically different grid topologies the present invention is able to cope with;

[0082] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 Fig. 3 shows a flow chart of an example of the method according to the present invention;

[0083] Figs. 4a-c show diagrams of measured voltage variables over time

[0084] 5 or the three terminals of three different electricity meters;

[0085] Figs. 5a-c show diagrams of calculated normalised voltage variables based the voltage variables shown in Figs. 3a-c;

[0086] 10 Figs. 6a-c show terminal points in form of projections of the normalised variables shown in Figs. 4a-c onto a two-dimensional solution space spanned by two principle components;

[0087] Figs. 7a, b show the terminal points of Figs. 5a-c in one diagram, wherein distances to distribution centres are shown in Fig. 6b;

[0088] Fig. 8 shows a proximity matrix with proximity values being indicative of a normalised distances between terminal points in

[0089] 20 Figs. 6a, b, wherein each terminal of each electricity meter acts as a base terminal for spanning the two-dimensional solution space;

[0090] Fig. 9 shows a proximity matrix with proximity values being indicative of a normalised distance between terminal points in Figs. 6a, b, wherein each electricity meter acts as a base electricity meter for spanning the two-dimensional solution space;

[0091] 30 Fig. 10 shows a proximity matrix with proximity values being indicative of a normalised distance between terminal points for

[0092] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 each electricity meter in Figs. 6a, b, wherein each electricity meter acts as a base electricity meter for spanning the two-dimensional solution space;

[0093] 5 Fig. 1 1 shows a clustering of electricity meters that share a common connection line to the distribution substation based on the proximity matrix shown in Fig. 9; and

[0094] Fig. 12 shows a clustering of electricity meters that share a com¬

[0095] 10 mon connection line to the distribution substation based on a proximity matrix of the electricity meters shown in Fig. 2b.

[0096] DETAILED DESCRIPTION

[0097]

[0026] The invention can be implemented by means of hardware, software, firmware or any combination of these. The invention or some of the features thereof can also be implemented as software running on one or more data processors and / or digital signal processors.

[0098] 20

[0099]

[0027] The individual elements of an embodiment of the invention may be physically, functionally and logically implemented in any suitable way such as in a single unit, in a plurality of units or as part of separate functional units. The invention may be implemented in a single unit, or be both physically and functionally distributed between different units and processors.

[0100]

[0028] Fig 1 shows a three-phase electricity grid 1 having multiple phases, such as a three-phase grid, and a plurality of smart electricity

[0101] 30 meters Ml -M3 installed at respective points of use, e.g. households. Each of three physical phases Pl - P3 of the grid 1 are connected with one of three terminals Tl - T3 of the respective electricity meters Ml -M3 and each electricity meter Ml -M3 measures the amount of electricity

[0102] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 delivered via a shared distribution substation 3 to the respective point of use using each of the three phases Pl - P3. For the purpose of metering electricity consumption, each electricity meter Ml -M3 measures and stores voltage variables in a memory of the electricity meter MI ¬

[0103] 5 MS.

[0104]

[0029] As explained above, during installation of an electricity meter Ml -M3, the phases Pl - P3 may be connected in an arbitrary order to the terminals T1 - T3. Therefore, no or insufficient information exists about

[0105] 10 which of the phases Pl - P3 is connected to which of the terminals Tl - T3. It may be enough for a utility provider to identify the nominal phases LI LS of the grid 1 that are connected to the terminals Tl - T3. The nominal phases LI - L3 could map to the physical phases Pl - P3 in a known or unknown way, depending on the wiring in the shared distribution substation 3. The present invention thus provides an improved method for identifying which of the nominal phases LI - L3 of the multi-phase grid 1 is connected to which terminal Tl - T3 of each of a plurality of electricity meters Ml -M3 installed in the grid 1 . This means, the electricity meters Ml -M3 can be distinguished or grouped depending on how they are

[0106] 20 connected to the nominal phases LI - L3 of the multi-phase grid 1 .

[0107]

[0030] Further, Fig. 1 illustrates an example of an advanced metering infrastructure, AMI, for collecting electric power consumption data, including voltage variables from the electricity meters Ml -M3. The electricity meters Ml -M3 are wirelessly connected to a communication network 9 including various receiver devices in the form of repeaters 7 and collectors 8. Consumption data is transmitted from the electricity meters Ml -M3 either directly to a collector 8 or indirectly to a collector 8 via a repeater 7. The number of repeaters 7 and collectors 8 required for the

[0108] 30 AMI depends amongst others on the size and topology of the AML From the collector s, consumption data from the plurality of electricity meters Ml -M3 is transmitted to a head-end system 5 with a connected server s for storing, processing and analysing consumption data. Data

[0109] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 may also be transmitted in the opposite direction in the AMI, from the head-end system 5 to the respective electricity meters Ml -M3. For instance, as will be further described below, data related to phase identification may be transmitted to the electricity meters Ml -M3 by means

[0110] 5 of the AMI.

[0111]

[0031] It is understood by the skilled person that other network devices, such as router devices or smart meters, equipped with different communication modules may also be included in the AMI. The total communication network is advantageously a radio network, wherein at least the network between the collector 8 and / or repeater 7 and the electricity meters Ml -M3 is a radio network, whereas the link between the collector 8 and the repeater 7 and the link to the head-end system 5 may be of any suitable type, such as wired or wireless connection.

[0112] 15 Further, it is understood by the skilled person that the shown AMI only includes a limited number of network components for illustrative purposes. Still further, the above-described AMI may have the functionality of an automatic reading system ARS and / or of a fixed network.

[0113] 20

[0032] The voltage variables stored in each meter are collected via the above-described AMI in regular intervals and / or on demand. In one embodiment, the subsequent data processing is carried out remote from the electricity meters Ml -M3 at a system level of the AMI, using a computer having the necessary processing power to handle the amount of data used for the phase identification. In other embodiments, at least part of the data processing may take place at meter level, i.e. one or more method steps may be performed by the electricity meters Ml -M3. For example, it is possible that part of the data processing may take place at the meter level, before meter data is trans¬

[0114] 30 mitted to the head-end system 5. Some kind of preprocessing of the voltage variables may take place at meter level to reduce the amount of data needed to be collected. Further, part of the data processing may also take place at an intermediate level between electricity meter

[0115] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 Ml -M3 and the head-end system 5, such as at a collector 8 or another gateway used for collecting metering data from a plurality of electricity meters Ml -M3 and forwarding the data to a server 6 at system level of the AMI.

[0116] 5

[0117]

[0033] The phase identification process is based on electricity meter data, including voltage variables collected from each terminal Tl - T3 of each electricity meter Ml -M3 of a group of electricity meters Ml -M3 that are supposed to be connected to the same shared distribution substation 3. An accessible database may contain correct or incorrect information about which electricity meter Ml -M3 is connected to which shared distribution substation 3. The inventive method may preferably be used to correct such incorrect information in such a database if applicable. To start off with, the database can be assumed to contain

[0118] 15 correct information to preselect a group of the electricity meters Ml -M3 that are associated with the same one shared distribution substation 3 of the multi-phase electrical grid 1 .

[0119]

[0034] The problem to be solved by the represent invention is schemati¬

[0120] 20 cally illustrated in Figs. 2a, b which show different examples of grid topologies of how the electricity meters Ml -M3 are connected to the shared distribution substation 3 of the multi-phase electrical grid 1 . Fig. 2a shows a first electricity meter Ml , a second electricity meter M2 and a third electricity meter M3 connected to the same shared distribution substation 3 of the multi-phase electrical grid 1 . The first electricity meter Ml and the second electricity meter M2 share a relatively long common connection line 1 1 via a cable box 13 to the shared distribution substation 3. The third electricity meter M3, however, has its own relatively short connection line 15 to the shared distribution substation 3.

[0121] 30 Known phase identification methods are less reliable and not robust enough for such inhomogeneous grid topologies, e.g. grid topologies with large differences of cable lengths between the electricity meters Ml -M3 and their associated shared distribution substation 3 of the multi-

[0122] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 phase electrical grid 1 , because the cable lengths may have a significant effect on the measured voltage variables. Fig. 2b shows schematically a more realistic grid topology with 18 electricity meters Ml -Ml 8 being arranged in clusters, wherein the electricity meters Ml -Ml 8 of a

[0123] 5 single cluster share a common connection line 1 1 to the shared distribution substation 3. Usually, there is no or insufficient database information available about the cable lengths, or about which of the electricity meters Ml -Ml 8 share a common connection line 1 1 to the shared distribution substation 3.

[0124] 10

[0125]

[0035] In Fig. 2b, all 18 electricity meters Ml -Ml 8 are, according to incorrect database information, supposed to be connected to the shared distribution substation 3. In fact, however, the four electricity meters M1 -M4 have no common connection line 1 1 to the shared distribution substation 3, but rather a common connection line 1 1 ’ via a cable box 13’ to another shared distribution substation 3’. As will be shown below, the inventive method will reveal the correct grid topology and identifies a hierarchical structure of clusters of the electricity meters MI MI 8 that share a common connection line 1 1 to the shared distribution

[0126] 20 substation 3, so that the phases can be identified more reliably for each cluster separately. As a further beneficial effect, the inventive method will reveal the incorrect database information about the four electricity meters M1 -M4 and allows to correct and update the database in this respect.

[0127]

[0036] Fig. 3 shows a flow chart of an example of the inventive method to improve the phase identification. In a first step SI , a group of n electricity meters Ml -Mn is preselected which are associated with one shared distribution substation 3 of the multi-phase electrical grid 1 . If the

[0128] 30 multi-phase electrical grid 1 comprises more than one distribution substation 3, the inventive method is preferably performed separately for

[0129] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 each preselected group of n electricity meters Ml -Mn that are associated with a respective shared distribution substation 3. In a second step S2, voltage variables U are measured and stored at each electricity meter Ml -Mn of the preselected group of n electricity meters Ml -Mn,

[0130] 5 wherein the voltage variables U are induced on each connected terminal T1 -T3 of the respective electricity meter Ml -Mn. The voltage variables U are collected in step S3 from each electricity meter Ml -Mn of the group of n electricity meters Ml -Mn by a data collection system (5, 6, 7, 8, 9). It should be noted that the steps S2 and S3 are preferably per¬

[0131] 10 formed during regular consumption metering. In other words, the steps S2 and S3 may be part of the ordinary consumption metering anyway. In a fourth step S4, a set of normalised voltage variables are calculated based on the voltage variables U collected from the respective electricity meter Ml -Mn for each terminal T1 -T3 of each electricity meter Ml -Mn of the preselected group of n electricity meters Ml -Mn. Alternatively, the step S4 of calculating normalised voltage variables may be performed at each electricity meter Ml -Mn. In that case, the voltage variables U collected in step S3 are already normalised. The normalised voltage variables may for example be mean free voltage change vari¬

[0132] 20 ables (MFVC variables) as described in EP 3449 267 Bl .

[0133]

[0037] In step S5, for each electricity meter Mi, where i e {1, ... , n} of the preselected group of n electricity meters Ml -Mn, said electricity meter Mi is selected as a base electricity meter Mi. In other words, the method loops over the n preselected electricity meters Ml -Mn. Within the loop, for each selected base electricity meter Mi, a solution space of lower dimensionality compared to a variable space defined by the normalised voltage variables of the selected base electricity meter Mi is defined in step S6. For example, a principal component analysis (PCA)

[0134] 30 may be performed on the normalised voltage variables, i.e. MFVC variables, to define a two-dimensional solution space spanned by principal

[0135] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 components PCI , PC2 of the distribution of the normalised voltage variables of the selected base electricity meter Mi. In step S7, the normalised voltage variables of each terminal T1-T3 of each electricity meter Ml -Mn of the preselected group of n electricity meters Ml -Mn are pro¬

[0136] 5 jected onto a terminal point TP in the respective solution space i e {1, for each of the n solution spaces. In the following step S8, a proximity matrix is determined comprising proximity values being defined by the terminal points TP in each of the n solution spaces. For example, the proximity values of the determined proximity matrix may be

[0137] 10 indicative of a Euclidian distance of the terminal points TP in the respective solution space. Alternatively, or in addition, the proximity values may be indicative of an angular distance of the terminal points TP in the respective solution space. Also, a weighted or unweighted combination of these distances is possible. Fig. 7b shows an example of a terminal point TP31 of the terminal T1 of the electricity meter M3 having three distances to distribution centres in a solution space. The distribution centres may be defined by centres of clusters of terminal points TP of terminals T1 -T3 that are likely to be connected to the same nominal phase of the grid 1 . In the following step S9, clusters of electricity meters

[0138] 20 are identified that share a common connection line 1 1 to the distribution substation 3 based on the determined proximity matrix. In other words, the proximity matrix reveals a topology of the grid 1 for the preselected electricity meters Ml -Mn, e.g. as shown in Figs. 2a, b. Finally, it is mapped in step S10 for each cluster separately which terminal T1 -T3 of each electricity meter Ml -Mn is connected to which nominal phase of the multi-phase electrical grid 1 . Step SI 0 comprises four sub-steps S101-S104 for mapping which terminal T1 -T3 of each electricity meter of a cluster is connected to which nominal phase of the multi-phase electrical grid 1 . In the first sub-step S 101 , one or more terminals T1 -T3 of one

[0139] 30 or more of the electricity meters of the cluster are selected as cluster base terminal. The choice of the one or more cluster base terminal(s) should in principle not significantly affect the result. This is, because the

[0140] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 voltage signature of electricity meters within a cluster is supposed to be similar enough. If more than one terminal T1 -T3 of one or more electricity meter(s) of the cluster is used as a cluster base terminal, an average or another combination of them can be used as cluster base termi-

[0141] 5 nal(s). In sub-step S102, a cluster solution space is defined being of lower dimensionality compared to a variable space defined by the normalised voltage variables of the selected cluster base terminal(s). The cluster solution space may be spanned by principle components PCI , PC2 derived from the normalised voltage variables, e.g. MFVC varia¬

[0142] 10 bles, of the selected cluster base terminal(s). In sub-step SI 03, the normalised voltage variables, e.g. MFVC variables, of each terminal T1 -T3 of each electricity meter of said cluster are projected onto a cluster terminal point TP in the cluster solution space. The cluster terminal points TP have a clear distance from each other in the cluster solution space so that the phase identification is very robust and reliable. In the last substep S104, it is mapped which terminal T1 -T3 of each electricity meter of said cluster is connected to which nominal phase of the multi-phase electrical grid 1 based on the cluster terminal points TP in the cluster solution space.

[0143] 20

[0144]

[0038] The advantage of the present invention is that the separation of the cluster terminal points TP in the cluster solution space is much clearer and distinguished than for electricity meters of different clusters. Therefore, the mapping which terminal T1 -t3 of each electricity meter is connected to which nominal phase of the grid 1 is much more robust and reliable for each cluster separately.

[0145]

[0039] Figs. 4a-c to Fig. 1 1 show diagrams, solutions spaces and proximity matrices for a grid topology shown in Fig. 2a, comprising three elec¬

[0146] 30 tricity meters Ml -M3, wherein electricity meter M3 has its own connection line 15 to the distribution substation 3 and electricity meters Ml and M2 have a longer shared connection line 1 1 via a cable box 13 for connection with the distribution substation 3.

[0147] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025

[0040] Fig. 4a shows measured and stored voltage variables U of the electricity meter Ml over the daytime t for each terminal T1 -T3 of the electricity meter Ml . The voltage variables U can be measured and

[0148] 5 stored continuously or regularly. The voltage variables U may be averaged over a certain time period, e.g. 5, 10 or 15 minutes. The data collection system 5, 6, 7, 8 collects the voltage variables U as they have been measured and / or in a processed form, e.g. averaged and / or normalised. Figs. 4b and 4c show analogously the voltage variables U of the electricity meters M2 and M3 at each of their respective terminals T1-T3. Figs. 4a and 4b look very similar, so that it is likely that electricity meters Ml and M2 belong to the same cluster. Fig. 4c, however, looks significantly different, so that it is likely that electricity meter M3 belongs to a different cluster than electricity meters Ml and M2.

[0149] 15

[0150]

[0041] Figs. 5a-c show normalised voltage variables, here in form of MFVC variables, derived from the voltage variables U shown in Figs. 4a- c for the respective electricity meters Ml , M2 and M3. It becomes even clearer from Fig. 5c that electricity meter M3 has a significantly different

[0151] 20 voltage signature than the electricity meters Ml and M2 shown in Figs. 5a and 5b.

[0152]

[0042] Figs. 6a-c show a two-dimensional solution space spanned by principal components PCI , PC2 of the distributions of MFVC variables shown in Figs. 5a-c, respectively, for the electricity meters Ml , M2 and M3. The MFVC variables at terminal T1 of electricity meter Ml are projected onto a terminal point TP1 1 in the solution space shown in Fig. 6a. Similarly, the MFVC variables at terminal T2 of electricity meter Ml are projected onto a terminal point TP12 and the MFVC variables of termi¬

[0153] 30 nal T3 of electricity meter Ml are projected onto terminal point TP13. As can be seen, the terminal points TP1 1 , TP12 and TP13 have a distinct separation in the solution space and can in principle be used to identify

[0154] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 the nominal phase of the grid 1 that the respective terminal T1 -T3 is connected to. Fig. 6b shows a very similar solution space for the electricity meter M2 with terminal points TP21 , TP22 and TP23. Fig. 6c shows the solution space for electricity meter M3 with terminal points TP31 , TP32 and

[0155] 5 TP33. As can be seen, the overall distance between the terminal points TP31 , TP32 and TP33 of the electricity meter M3 is smaller than for electricity meters Ml and M2 shown in Figs. 6a, b. This means that the phase identification for electricity meter M3 is somewhat less reliable and robust.

[0156] 10

[0157]

[0043] In the shown example, it appears likely that terminal T1 of electricity meter M3 is connected to the same nominal phase of the grid 1 as terminal T3 of the electricity meters Ml and M2. Analogously, terminal T3 of electricity meter M3 appears to be connected to the same nominal phase of the grid 1 as terminal T1 of the electricity meters Ml and M2. So, the order how the terminals T1 and T3 of electricity meter M3 are connected to the nominal phases of the grid 1 is switched compared to the order in which the terminals T1 and T3 of the electricity meters Ml and M2 are connected to the grid 1 .

[0158] 20

[0159]

[0044] Figs. 7a, b show all nine terminal points TP1 1 , TP12, TP13, TP21 , TP22, TP23, TP31 , TP32 and TP33 in one solution space. In Fig. 7b, it is shown as an example how distances of terminal point TP31 to distribution centres may be determined. The distances of the terminal points TP to the distribution centres may be used to map which terminal T1-T3 of each electricity meter M1-M3 is connected to which nominal phase of the multi-phase electrical grid. When it comes to determining the proximity values of the proximity matrix, the distances between the terminal points may be used. It should be noted that distances between the ter¬

[0160] 30 minal points TP may be measured in many ways. For example, the distance of each terminal point TP to the other terminal points TP and / or

[0161] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 distribution centres of terminal points TP may be determined on an arbitrary scale ranging from 0 to 1 . Such a scale value may serve as a proximity value of a terminal point TP.

[0162] 5

[0045] Fig. 8 shows a proximity matrix that comprises proximity values ranging from 0 to 1 , wherein each proximity value is defined by the terminal points TP shown in Fig. 7b. In the shown example, the proximity value is a normalised distance of each terminal point TP to the other terminal points TP. The proximity matrix shown in Fig. 8 is a 9x9-matrix

[0163] 10 with 81 proximity values, wherein each column represents a proximity of each terminal point TP to one of the nine terminal points TP in the solution space spanned by the principle components PCI , PC2 of the MFVC variables. In the first three columns 1 -3, the normalised voltage variables of the terminals Tl , T2 and T3 of electricity meter Ml are used as a basis for spanning the solution space. The values in the first column represent a proximity of all the terminal points TP to the first terminal point TP1 1 of electricity meter Ml . The values in the second column represent a proximity of all the terminal points TP to the second terminal point TP12 of electricity meter Ml . The values in the third column repre¬

[0164] 20 sent a proximity of all the terminal points TP to the third terminal point TP 13 of electricity meter Ml . In the next three columns 4-6, the normalised voltage variables of the terminals Tl , T2 and T3 of electricity meter M2 are used as a basis for spanning the solution space. The values in the fourth column represent a proximity of all the terminal points TP to the first terminal point TP21 of electricity meter M2. The values in the fifth column represent a proximity of all the terminal points TP to the second terminal point TP22 of electricity meter M2. The values in the sixth column represent a proximity of all the terminal points TP to the third terminal point TP23 of electricity meter M2. In the last three columns 7-9,

[0165] 30 the normalised voltage variables of the terminals Tl , T2 and T3 of electricity meter M3 are used as a basis for spanning the solution space. The values in the seventh column represent a proximity of all the terminal

[0166] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 points TP to the first terminal point TP31 of electricity meter M3. The values in the eighth column represent a proximity of all the terminal points TP to the second terminal point TP32 of electricity meter M3. The values in the ninth column represent a proximity of all the terminal points TP to

[0167] 5 the third terminal point TP33 of electricity meter M3. In Fig. 9, the three columns belonging to the same base electricity meter are merged. In Fig. 10, the proximity matrix is symmetrised by merging the rows of terminals belonging to the same electricity meter, so that the resulting symmetrised proximity matrix is a 3x3-matrix. The symmetrised 3X3 proximity

[0168] 10 matrix shown in Fig. 10 serves as a basis to identify clusters of electricity meters that share a common connection line 1 1 to the distribution substation 3. Fig. 1 1 shows how the proximity matrix is used to determine a hierarchical clustering of the electricity meters Ml , M2 and M3 to resemble the grid topology shown in Fig. 2a. In the shown example, the electricity meters Ml and M2 belong to the same cluster and electricity meter M3 belongs to another cluster. Once the clusters are established, the phase identification is done for the clusters separately, so that it is irrelevant which one or more terminal(s) of one or more of the electricity meters of a cluster are selected as cluster base terminals. The phase

[0169] 20 identification is much more robust and reliable for each cluster separately.

[0170]

[0046] In Fig. 12, a symmetrised 18X18 proximity matrix is shown for the electricity meters M1-M18 of the topology shown in Fig. 2b. The first four electricity meters M1-M4 form clearly a cluster, which is, however, so distinct from the other clusters that it can be assumed that the data base information of these electricity meters M1-M4 being connected via a shared connection line 1 1 to the distribution station 3 is probably not correct. Instead, it can be assumed that the electricity meters Ml

[0171] 30 and M4 share another connection line 1 1 ’ to another distribution station 3’. The proximity matrix can be used to update such incorrect data

[0172] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 base information. For the other 14 electricity meters M5-M18, three clusters M1-M7, M8-M9 and M10-M18 can be identified from the proximity matrix. The cluster M5-M7 further comprises a sub-cluster M6-M7. The cluster Ml 0-M 18 comprises also three sub-clusters MI O, Ml 1-M12 and

[0173] 5 M13-M18. The phase identification is preferably performed for each sub-cluster and cluster separately, so that the phase identification is more robust and reliable.

[0174]

[0047] The information about which phase is connected to which terminal of an electricity meter may be used for various purposes. The load level of each phase of a multi-phase system may be determined based on information about power usage from individual meters. Knowing the load level of each phase can be used to better balance the phases and improve grid performance.

[0175] 15

[0176]

[0048] Further, the phase information may be distributed to each electricity meter and displayed on a display of the meter. The phase information may also be stored and / or updated in a database and distributed upon request, for example to installation or maintenance person¬

[0177] 20 nel during installation, repair or modification of existing or new electrical installations. Such distribution of phase information may be effectuated by a handheld device, such as a dedicated installation tool or general- purpose device, such as a smartphone, a tablet or computer. During installation of new heavy loads, such as a charger for an electrical car, installation personal may request the phase information from a central server using the handheld device. Based on the phase information received, the new load may be connected to the phase best suited, for example, the presently least loaded phase. The skilled person may also envisage other uses, such as active load control etc.

[0178] 30

[0179]

[0049] When it is stated that voltage variables or other variables or values are collected, processed or treated in any other way, it is implicit that it is information or data representing such values or variables that

[0180] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 are collected, processed or treated. Further, data and information are terms used interchangeably throughout the specification.

[0181]

[0050] Where, in the foregoing description, integers orelements are men¬

[0182] 5 tioned which have known, obvious or foreseeable equivalents, then such equivalents are herein incorporated as if individually set forth. Reference should be made to the claims for determining the true scope of the present disclosure, which should be construed so as to encompass any such equivalents. It will also be appreciated by the reader that integers or fea¬

[0183] 10 tures of the disclosure that are described as optional, preferable, advantageous, convenient or the like are optional and do not limit the scope of the independent claims.

[0184]

[0051] The above embodiments are to be understood as illustrative examples of the disclosure. It is to be understood that any feature described in relation to any one aspect or embodiment may be used alone, or in combination with other features described, and may also be used in combination with one or more features of any other of the aspects or embodiments, or any combination of any other of the aspects

[0185] 20 or embodiments. While at least one exemplary aspect or embodiment has been shown and described, it should be understood that other modifications, substitutions and alternatives are apparent to one of ordinary skill in the art and may be changed without departing from the scope of the subject matter described herein, and this application is intended to cover any adaptations or variations of the specific embodiments discussed herein.

[0186]

[0052] In addition, "comprising" does not exclude otherelements or steps, and "a" or "one" does not exclude a plural number. Furthermore, charac¬

[0187] 30 teristics or steps which have been described with reference to one of the above exemplary aspects or embodiments may also be used in combination with other characteristics or steps of other exemplary embodiments described above. Method steps may be applied in any order or

[0188] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025 in parallel or may constitute a part or a more detailed version of another method step. It should be understood that there should be embodied within the scope of the patent warranted hereon all such modifications as reasonably and properly come within the scope of the contribution to

[0189] 5 the art. Such modifications, substitutions and alternatives can be made without departing from the spirit and scope of the disclosure, which should be determined from the appended claims and their legal equivalents.

[0190]

[0053] List of reference numerals:

[0191] 1 multi -phase electrical grid

[0192] 3, 3’ distribution substation

[0193] 5 head-end system HES

[0194] 6 remote server

[0195] 15 7 repeater

[0196] 8 collector

[0197] 9 wireless network

[0198] 1 1, 1 1 ’ connection line

[0199] 13 cable box

[0200] 15 connection line

[0201] Ml-Mxn electricity meter

[0202] T1-T3 terminals

[0203] TP1 1-TP33 terminal points

[0204] U voltage variables

[0205] 25 PC1 , PC2 principal component

[0206] P1-P3 physical phases

[0207] L1-L3 nominal phases

[0208] Patentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025

Claims

27 / 32Claims1 . A method of mapping suitable for a plurality of electricity meters (Ml-Mn) to map which terminal (T1-T3) of each electricity meter (Ml-Mn) is connected to which nominal phase of a multi-phase5 electrical grid (1 ), wherein the method comprises the steps of: establishing a proximity matrix to determine clusters of electricity meters (Ml-Mn), wherein the electricity meters (Ml-Mn) of each cluster share a common connection line ( 1 1 ) to a shared distribution substation (3) of the multi-phase electrical grid (1 ),10 and mapping, for each cluster separately, which terminal (T1-T3) of each electricity meter (Ml-Mn) of the cluster is connected to which nominal phase (L1-L3) of the multi-phase electrical grid (1 ).15 2. The method of claim 1 , comprising: a) preselecting a group of the electricity meters (Ml-Mn) that are associated with one shared distribution substation (3) of the multi-phase electrical grid (1 ); b) in each electricity meter (Ml-Mn) of the preselected group of20 electricity meters (Ml-Mn), measuring and storing voltage variables (U) induced on the terminals (T1-T3) of the respective electricity meter (Ml-Mn); c) collecting the voltage variables (U) from each electricity meter (Ml-Mn) of the group of electricity meters (Ml-Mn);25 d) for each terminal (T1-T3) of each electricity meter (Ml-Mn) of the preselected group of electricity meters (Ml-Mn), calculating a set of normalised voltage variables based on the voltage variables (U); characterised in that the method further comprises: atentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025e) for each electricity meter (Ml-Mn) of the preselected group of electricity meters (Ml-Mn), selecting said electricity meter (Ml-Mn)as a base electricity meter (Ml-Mn); f) for each selected base electricity meter (Ml - Mn), defining a5 solution space of lower dimensionality compared to a variable space defined by the normalised voltage variables of the selected base electricity meter (Ml-Mn); g) for each solution space, projecting the normalised voltage variables of each terminal (T1-T3) of each electricity meter (Ml-10 Mn) of the preselected group of electricity meters (Ml-Mn) onto a terminal point (TP) in the respective solution space; h) determining the proximity matrix comprising proximity values being defined by the terminal points (TP) in each solution space; i) identifying the clusters of electricity meters (Ml-Mn) that share a common connection line ( 1 1 ) to the distribution substation (3) based on the determined proximity matrix; and j) wherein the step of mapping comprises one or more terminals (T1-T3) of one or more of the electricity meters (Ml-Mn) of said cluster are selected as cluster20 base terminals (T1-T3), a cluster solution space is defined being of lower dimensionality compared to a variable space defined by the normalised voltage variables of the selected cluster base terminals (T1-T3), the normalised voltage variables of each terminal (T1-T3) of each electricity meter (Ml-Mn) of said cluster are projected onto a cluster terminal point (TP) in the cluster solution space, and it is mapped which terminal (T1-T3) of each electricity me¬30 ter (Ml-Mn) of said cluster is connected to which nominal phase (LI -L3) of the multi-phase electrical grid ( 1 ) based on the cluster terminal points (TP) in the cluster solution space. atentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 20253. The method of claim 1 or 2, further comprising determining a load on a nominal phase (L1 -L3) of the multiphase electrical grid (1 ) based on voltage measurements at those terminals (T1-T3) of the electricity meters (Ml-Mn) that5 are mapped to the same said nominal phase (L1 -L3) of the multi-phase electrical grid (1 ).

4. The method of claim 3, further comprising identifying a load imbalance on a nominal phase (L1 -L3) of the multi-phase electrical grid (1 ) based on the determined load10 on at least two of the nominal phases (L1 -L3) of the multi-phase electrical grid ( 1 ); and balancing the identified load imbalance by switching nominal phase connections of at least one power consumer and / or power source in the multi-phase electrical grid (1 ).

5. The method of any of the preceding claims, further comprising executing the steps a) to j) for another preselected group of electricity meters (Ml-Mn) that are associated with another shared distribution substation (3’) of the multi-phase electrical grid (1 ).20 6. The method of any of the preceding claims, further comprising: identifying those electricity meters (Ml-Mn) among the group of electricity meters (Ml-Mn) that are incorrectly associated with the shared distribution substation (3) of the multi-phase electrical grid (1 ) if a majority of non-diagonal proximity values25 in at least one row and / or column of the determined proximity matrix are below a pre-determined threshold, wherein said at least one row and / or column of the determined proximity ma- atentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025trix corresponds to those electricity meters (Ml-Mn) that are incorrectly associated with the shared distribution substation (3) of the multi-phase electrical grid (1 ).

7. The method of any of the preceding claims, wherein the proximity5 values of the determined proximity matrix are indicative of a Euclidian distance of the terminal points (TP) in the solution space, or of an angular distance of the terminal points (TP) in the solution space, or a weighted combination thereof.

8. The method of any of the preceding claims, wherein each solution10 space is spanned by principal components (PCI, PC2) of the normalised voltage variables of the respective base electricity meter (Ml-Mn), and / or wherein the cluster solution space is spanned by principal components (PCI, PC2) of the normalised voltage variables of the respective cluster base terminals (T1-T3).15 9. The method of any of the preceding claims, wherein the normalised voltage variables are mean free voltage change, MFVC, variables based on time derivatives of the voltage variables.

10. The method of any of the preceding claims, further comprising:20 identifying which terminal (T1-T3) of each electricity meter (Ml-Mn) is connected to which physical phase of the multiphase electrical grid ( 1 ) based on a known order in which the terminals (T1-T3) of at least one reference electricity meter (Ml-Mn) are connected to physical phases of the multi-phase25 electrical grid (1 ).1 1. A computer program comprising instructions which, when the computer program is executed by at least one computer, cause atentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 202531 / 32 the at least one computer to carry out the steps of the method of any of the preceding claims.

12. The computer program product of claim 1 1 , wherein the computer program is configured to receive user input for carrying out5 one or more steps of the method of any of the claims 1 to 10.

13. An advanced metering infrastructure, AMI, for reading electricity consumption from a plurality of electricity meters (Ml-Mn) in a multi-phase electrical grid (1 ), the AMI comprising: a plurality of electricity meters (Ml-Mn) having terminals (T1-T3)10 being connected to nominal phases (L1-L3) of a multi-phase electrical grid (1 ); a data collection system (5, 6, 7, 8) being configured to collect voltage variables from each of the electricity meters (Ml-Mn) via a communication connection; and at least one computer being part of or in communication connection with the data collection system and / or the plurality of electricity meters (Ml-Mn), wherein the at least one computer is configured to carry out the steps of the method of any of the claims 1 to 10.20 atentanwdlte Hemmer Lmdfeld Frese KAP 3795 WO, 16 / 07 / 2025

Images