DETECTION OF LOSS OF NEUTRAL OR PHASE

Abstract

A method for monitoring a polyphase network includes the steps, repeated regularly, of: acquiring voltage measurements taken by voltage sensors of an item of electrical equipment, and producing current values of electrical parameters; accessing a reference table comprising a list of connection configurations each associated with a distinct combination of reference values of said electrical parameters, the list of connection configurations comprising configurations each corresponding to a combination of connection states for each phase (P1, P2, P3) and for the neutral (N); implementing an exploratory algorithm intended to explore the reference table to identify the current connection configuration.

Claims

1. A method for monitoring a polyphase electricity network, implemented at least partially in a processing unit of an item of electrical equipment connected to the polyphase electricity network, and comprising the steps, repeated regularly, of: acquiring voltage measurements taken by voltage sensors (R.sub.1A, R.sub.1B, R.sub.2A, R.sub.2B, R.sub.3A, R.sub.3B) of the item of electrical equipment that are connected to the polyphase electricity network; producing current values of electrical parameters from the voltage measurements; accessing a reference table that comprises a list of connection configurations each associated with a distinct combination of reference values of said electrical parameters, the list of connection configurations comprising first connection configurations each corresponding to a distinct combination of connection or disconnection states for each phase (P1, P2, P3) and for the neutral (N); implementing an exploratory algorithm intended to explore the reference table to identify the connection configuration associated with the current values of the electrical parameters.

2. The monitoring method according to claim 1, wherein the list of connection configurations comprises second connection configurations each corresponding, for one of the phases (P1, P2, P3), to an inversion of the neutral (N) and said phase (P1, P2, P3).

3. The monitoring method according to claim 1, wherein the list of connection configurations comprises third connection configurations each corresponding to one of the phases that is disconnected and that is at the potential of another phase.

4. The monitoring method according to claim 1, wherein the implementation of the exploratory algorithm comprises the steps of: testing a connection configuration, by checking whether predetermined criteria associated with said connection configuration are fulfilled, the predetermined criteria being defined based on the electrical parameters; if the predetermined criteria are fulfilled, deducing therefrom that said connection configuration is that which corresponds to the current values of the electrical parameters, if not, repeating the testing step with another connection configuration.

5. The monitoring method according to claim 1, wherein the electrical parameters comprise: measured line-to-neutral voltages, each measured line-to-neutral voltage being representative of an amplitude of a line-to-neutral voltage ({right arrow over (V.sub.1)}, {right arrow over (V.sub.2)}, {right arrow over (V.sub.3)}) between one of the phases and the neutral; and/or measured line-to-line voltages, each measured line-to-line voltage being representative of an amplitude of a line-to-line voltage ({right arrow over (U.sub.1)}, {right arrow over (U.sub.2)}, {right arrow over (U.sub.3)})) between two distinct phases; and/or measured angles (Phi.sub.1, Phi.sub.2, Phi.sub.3), each measured angle being equal to an angle between a vector representation of a line-to-neutral voltage between one phase and the neutral, and a vector representation of a line-to-neutral voltage between another phase and the neutral.

6. The monitoring method according to claim 5, wherein the measured angles are obtained by implementing a zero-crossing detection method on the voltages measured between the phases and the neutral.

7. The monitoring method according to claim 5, wherein the exploratory algorithm implemented is a first exploratory algorithm that only uses the measured line-to-neutral voltages and the measured line-to-line voltages as electrical parameters.

8. The monitoring method according to claim 4, wherein the electrical parameters comprise: measured line-to-neutral voltages, each measured line-to-neutral voltage being representative of an amplitude of a line-to-neutral voltage ({right arrow over (V.sub.1)}, {right arrow over (V.sub.2)}, {right arrow over (V.sub.3)}) between one of the phases and the neutral; and/or measured line-to-line voltages, each measured line-to-line voltage being representative of an amplitude of a line-to-line voltage ({right arrow over (U.sub.1)}, {right arrow over (U.sub.2)}, {right arrow over (U.sub.3)})) between two distinct phases; and/or measured angles (Phi.sub.1, Phi.sub.2, Phi.sub.3), each measured angle being equal to an angle between a vector representation of a line-to-neutral voltage between one phase and the neutral, and a vector representation of a line-to-neutral voltage between another phase and the neutral, wherein the exploratory algorithm implemented is a first exploratory algorithm that only uses the measured line-to-neutral voltages and the measured line-to-line voltages as electrical parameters, wherein the first exploratory algorithm comprises the step of attempting to detect one or more measured line-to-line voltages that are abnormal, the first exploratory algorithm then consisting in testing sets of connection configurations, which are defined as a function of the number of abnormal measured line-to-line voltages.

9. The monitoring method according to claim 8, wherein the measured line-to-line voltage U.sub.i is considered to be abnormal if:
U.sub.i<Uprev.sub.i or if U.sub.i>Uprev.sub.i+, where: is a margin of uncertainty, and where:
Uprev.sub.i=V.sub.n.Math.SQRT(ai.sup.2+aj.sup.2+ai.Math.aj), where ai=V.sub.i/V.sub.n, aj=V.sub.j/V.sub.n, V.sub.i is the measured line-to-neutral voltage for the phase Pi, V.sub.j is the measured line-to-neutral voltage for the phase Pj, and V.sub.n is a nominal line-to-neutral voltage.

10. The monitoring method according to claim 4, wherein the electrical parameters comprise: measured line-to-neutral voltages, each measured line-to-neutral voltage being representative of an amplitude of a line-to-neutral voltage ({right arrow over (V.sub.1)}, {right arrow over (V.sub.2)}, {right arrow over (V.sub.3)}) between one of the phases and the neutral; and/or measured line-to-line voltages, each measured line-to-line voltage being representative of an amplitude of a line-to-line voltage ({right arrow over (U.sub.1)}, {right arrow over (U.sub.2)}, {right arrow over (U.sub.3)})) between two distinct phases; and/or measured angles (Phi.sub.1, Phi.sub.2, Phi.sub.3), each measured angle being equal to an angle between a vector representation of a line-to-neutral voltage between one phase and the neutral, and a vector representation of a line-to-neutral voltage between another phase and the neutral, wherein the exploratory algorithm implemented is a first exploratory algorithm that only uses the measured line-to-neutral voltages and the measured line-to-line voltages as electrical parameters, wherein the predetermined criteria comprise comparisons each made between: a resulting voltage obtained by adding: two different measured line-to-neutral voltages; or a measured line-to-neutral voltage and a measured line-to-line voltage; and a measured line-to-neutral voltage or a measured line-to-line voltage.

11. The monitoring method according to claim 5, wherein the exploratory algorithm implemented is a second exploratory algorithm that only uses the measured line-to-neutral voltages and the measured angles as electrical parameters.

12. The monitoring method according to claim 2, wherein the implementation of the exploratory algorithm comprises the steps of: testing a connection configuration, by checking whether predetermined criteria associated with said connection configuration are fulfilled, the predetermined criteria being defined based on the electrical parameters; if the predetermined criteria are fulfilled, deducing therefrom that said connection configuration is that which corresponds to the current values of the electrical parameters, if not, repeating the testing step with another connection configuration, wherein the electrical parameters comprise: measured line-to-neutral voltages, each measured line-to-neutral voltage being representative of an amplitude of a line-to-neutral voltage ({right arrow over (V.sub.1)}, {right arrow over (V.sub.2)}, {right arrow over (V.sub.3)}) between one of the phases and the neutral; and/or measured line-to-line voltages, each measured line-to-line voltage being representative of an amplitude of a line-to-line voltage ({right arrow over (U.sub.1)}, {right arrow over (U.sub.2)}, {right arrow over (U.sub.3)})) between two distinct phases; and/or measured angles (Phi.sub.1, Phi.sub.2, Phi.sub.3), each measured angle being equal to an angle between a vector representation of a line-to-neutral voltage between one phase and the neutral, and a vector representation of a line-to-neutral voltage between another phase and the neutral, wherein the exploratory algorithm implemented is a second exploratory algorithm that only uses the measured line-to-neutral voltages and the measured angles as electrical parameters, wherein the predetermined criteria, associated with the following connection configurations, only comprise comparisons between the measured angles and angular thresholds: the neutral and all the phases are normally connected; the neutral is inverted with one of the phases; the neutral is disconnected and all the phases are connected.

13. The monitoring method according to claim 11, wherein the second exploratory algorithm tests whether the phases are wired to the item of electrical equipment in a direct or indirect order.

14. An item of electrical equipment comprising voltage sensors and a processing unit wherein the monitoring method according to claim 1 is implemented.

15. The item of electrical equipment according to claim 14, the item of electrical equipment being a meter.

16. (canceled)

17. A nontransitory computer-readable storage medium on which a computer program comprising instructions that cause the processing unit of an item of electrical equipment to perform the steps of the monitoring method according to claim 1 is stored, wherein the item of electrical equipment comprising the voltage sensors and the processing unit in which the monitoring method is implemented.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Reference is made to the appended drawings in which:

[0064] FIG. 1 shows two electricity meters connected to a network in which the neutral has been cut;

[0065] FIG. 2 shows an electricity meter in which the invention is implemented;

[0066] FIG. 3 shows a Fresnel diagram in the nominal case, without loss of neutral;

[0067] FIG. 4 comprises a first diagram showing a neutral cut-off on the network, and a second diagram showing a cut-off of the phase P1;

[0068] FIG. 5 shows the effect, downstream of the break point, of the loss of the neutral on the Fresnel diagram;

[0069] FIG. 6 shows the result, on the voltages shown in Fresnel diagrams, of different connection configurations;

[0070] FIG. 7 shows steps of the monitoring method according to the invention.

DETAILED DESCRIPTION OF THE INVENTION

[0071] In reference to FIG. 2, a polyphase (in this case three-phase) electricity meter 5 is connected to a polyphase (in this case three-phase) electricity network 6 comprising three phases P1, P2, P3 and a neutral N. The meter 5 is intended to measure the electrical power distributed by the electricity network 6 to a user's installation 7.

[0072] The electricity meter 5 comprises, for each phase Pi (i ranging from 1 to 3), a terminal 8 connected to said phase Pi and a phase conductor 9 connected to said terminal 8. The electricity meter 5 also comprises a terminal 10 connected to the neutral N and a neutral conductor 11 connected to the terminal 10.

[0073] The meter 5 further comprises voltage sensors that are connected to the electricity network 6 via the phase 9 and neutral 11 conductors, and via the terminals 8 and 10. The voltage sensors can be used to measure: [0074] the line-to-neutral voltage V1 between the phase P1 and the neutral N, which is (normally) the image, at the meter 5, of the line-to-neutral voltage V1 of the network 6; [0075] the line-to-neutral voltage V2 between the phase P2 and the neutral N, which is (normally) the image, at the meter 5, of the line-to-neutral voltage V2 of the network 6; [0076] the line-to-neutral voltage V3 between the phase P3 and the neutral N, which is (normally) the image, at the meter 5, of the line-to-neutral voltage V3 of the network 6.

[0077] The voltage sensors comprise a first resistive bridge, a second resistive bridge, a third resistive bridge and an analogue-to-digital converter 14 (ADC).

[0078] The first resistive bridge comprises two resistors R.sub.1A and R.sub.1B mounted between the phase conductor 9 connected to the phase P1 and the neutral conductor 11. The second resistive bridge comprises two resistors R.sub.2A and R.sub.2B mounted between the phase conductor 9 connected to the phase P2 and the neutral conductor 11. The third resistive bridge comprises two resistors R.sub.3A and R.sub.3B mounted between the phase conductor 9 connected to the phase P3 and the neutral conductor 11.

[0079] The ADC 14 comprises a first input connected to a point situated between the two resistors R.sub.1A and R.sub.1B, a second input connected to a point situated between the two resistors R.sub.2A and R.sub.2B, and a third input connected to a point situated between the two resistors R.sub.3A and R.sub.3B.

[0080] The resistive bridges allow the amplitude of the measured voltages to be reduced so that they are compatible with the measurement range of the ADC 14. The ADC 14 digitizes the measured voltages.

[0081] The characteristics of the ADC 14 (number of bits, sampling frequency) are compatible with the need to produce usable voltage measurements. Typically, the ADC 14 that is used is a ADC with 12 or more bits and a sampling frequency of 1 ksps.

[0082] The electricity meter 5 also comprises a processing unit 15.

[0083] The processing unit 15 comprises one or more processing components 16 and, for example, any general-purpose or specialized processor or microprocessor, (for example a DSP (Digital Signal Processor), a microcontroller or indeed a programmable logic circuit such as an FPGA (Field-Programmable Gate Array) or an ASIC (Application-Specific Integrated Circuit)).

[0084] The processing unit 15 also comprises memories 17. At least one of these memories 17 forms a computer-readable storage medium on which at least one computer program is stored comprising instructions that cause the processing unit 15 to perform at least some of the steps of the monitoring method according to the invention. One of these memories 17 may be integrated into one of the processing components 16.

[0085] In this instance, the ADC 14 is integrated into the processing unit 15, and optionally into one of the processing components 16.

[0086] The processing unit 15 therefore acquires the measurements of the line-to-neutral voltages V1, V2, V3 and produces, from these voltage measurements, measurements of: [0087] the line-to-line voltage U1 between the phase P1 and the phase P2, which is (normally) the image, at the meter 5, of the voltage U1 of the network 6; [0088] the line-to-line voltage U2 between the phase P2 and the phase P3, which is (normally) the image, at the meter 5, of the voltage U2 of the network 6; [0089] the line-to-line voltage U3 between the phase P3 and the phase P1, which is (normally) the image, at the meter 5, of the voltage U3 of the network 6.

[0090] Each of these line-to-line voltages, at each sampling time, is obtained by subtracting the two line-to-neutral voltages of the phases in question.

[0091] The voltages considered hereinafter are: [0092] measured line-to-neutral voltages, each measured line-to-neutral voltage being representative of an amplitude of a line-to-neutral voltage between one of the phases and the neutral; [0093] measured line-to-line voltages, each measured line-to-line voltage being representative of an amplitude of a line-to-line voltage between two distinct phases.

[0094] In this instance, the measured line-to-neutral voltages and the measured line-to-line voltages are root mean square voltagesbut the voltages representative of the amplitudes could be different, for example maximum amplitude values.

[0095] The voltages considered hereinafter are therefore root mean square voltages calculated (and possibly corrected by calibration parameters) on the basis of the samples of the line-to-neutral voltage measurements (for example, each second on the basis of the samples of the previous second, or over a sliding-second window, or over another time depth that is deemed appropriate).

[0096] The processing unit 15 therefore estimates the measured line-to-neutral voltages V.sub.1 (which is the root mean square voltage of the voltage {right arrow over (V.sub.1)}), V.sub.2 and V.sub.3, and the measured line-to-line voltages U.sub.1 (which is the root mean square voltage of the voltage {right arrow over (U.sub.1)}), U.sub.2 and U.sub.3.

[0097] The processing unit also estimates measured angles Phi.sub.1, Phi.sub.2 and Phi.sub.3. Each measured angle is equal to an angle between a vector representation of a line-to-neutral voltage between one phase and the neutral, and a vector representation of a line-to-neutral voltage between another phase and the neutral.

[0098] Therefore, Phi.sub.1 is the angle between {right arrow over (V.sub.1)} and {right arrow over (V.sub.2)}, Phi.sub.2 is the angle between {right arrow over (V.sub.2)} and {right arrow over (V.sub.3)}, Phi.sub.3 is the angle between {right arrow over (V.sub.3)} and {right arrow over (V.sub.1)}.

[0099] The measured angles are estimated by implementing a zero-crossing detection method and appropriate filtering on the voltage measurementsand in particular on the measured line-to-neutral voltages.

[0100] The invention aims to detect a loss of neutral and/or of one or more phases, based solely on the analysis of the voltage measurements taken by the electricity meter 5, and without using reference measurements prior to the time a break occurs in one or more conductors from the network.

[0101] FIG. 3 shows the Fresnel diagram in the nominal case, when the neutral is present and with the line-to-neutral voltages {right arrow over (V.sub.1)}, {right arrow over (V.sub.2)}, {right arrow over (V.sub.3)} balanced. The vectors {right arrow over (U.sub.1)}, {right arrow over (U.sub.2)}, {right arrow over (U.sub.3)} and the angles Phi.sub.1, Phi.sub.2 and Phi.sub.3 can also be seen.

[0102] The top diagram in FIG. 4 shows and models the situation occurring after a loss of neutral (cutting off of the neutral). The break point 18 of the neutral N is assumed to be located upstream of the meter 5. The neutral line situated downstream of the break point 18 becomes a floating line and is established according to the loads situated downstream of the break point connected to it (Z.sub.1, Z.sub.2, Z.sub.3, Z.sub.12, Z.sub.12, Z.sub.23). In this example, the term upstream means on the network 6 side, and the term downstream means on the installation 7 side.

[0103] The bottom diagram in FIG. 4 shows and models the situation occurring after a loss of the phase P1 (cutting off of the phase). The break point 19 of the phase P1 is assumed to be located upstream of the meter 5. The phase line situated downstream of the break point 19 becomes a floating line and is established according to the loads situated downstream of the break point connected to it. In this specific case, it can be seen that the measured voltages V.sub.2, V.sub.3 and U.sub.2 correspond to the actual voltages of the network (V.sub.2, V.sub.3 and U.sub.2). However, the measured voltages V.sub.1, U.sub.1 and U.sub.3 are no longer those of the network (V1, U1 and U3) except in the specific case of perfectly balanced downstream loads.

[0104] Although this situation is not shown in FIG. 4, two break points may also be considered (on one phase and the neutral, or indeed on two phases). Similarly, the voltage on each of the two lines that have become floating lines is established according to the loads situated downstream of the break points that are respectively connected to them.

[0105] In reference to FIG. 5, following the loss of the neutral N, for example, the voltages naturally balance according to the downstream loads: in the Fresnel diagram, the position of the neutral that has become a floating neutral downstream of the break point is located somewhere within the triangle formed by the three phases, depending on the values of the downstream loads.

[0106] Therefore, it can be seen in FIG. 5 that, in the event that neutral is lost, the measured line-to-line voltages correspond to the actual line-to-line voltages of the network (U.sub.x=U.sub.x). However, the measured line-to-neutral voltages are no longer those of the network (V.sub.xV.sub.x), except in the specific case of perfectly balanced downstream loads. It should be noted that FIG. 5 depicts this situation with a hypothetical imbalance of the downstream loads.

[0107] The monitoring method according to the invention consists in detecting a cutting off of the neutral and/or one or several phases, based only on the voltage measurements.

[0108] The invention first comprises a preliminary step that consists in systematically and exhaustively listing and studying the different possible cases, and defining, for each of the connection configurations considered, the expected values of electrical parameters obtained from the voltage measurements.

[0109] A reference table 20 is therefore created (see FIG. 2), which comprises a list of connection configurations each associated with a distinct combination of reference values of said electrical parameters.

[0110] The list of connection configurations comprises first connection configurations each corresponding to a distinct combination of connection (normal) or disconnection (cut-off) states for each phase Pi and for the neutral N.

[0111] For example, a first connection configuration corresponds to the combination: [0112] N: OFF-P1: ON-P2: ON-P3: ON, [0113] where OFF means that the line in question (phase or neutral) is disconnected and therefore has a cut-off, and where ON on the contrary means that the connection is normal.

[0114] Other first connection combinations are, for example: [0115] N: ON-P1: OFF-P2: ON-P3: ON, [0116] N: ON-P1: OFF-P2: OFF-P3: ON, [0117] etc.

[0118] The list of connection configurations also comprises second connection configurations each corresponding, for one of the phases, to an inversion of the neutral and said phase at the terminals 8 and 10 of the meter 5. The inversion results, for example, from incorrect workmanship when installing the meter 5.

[0119] For example, a second connection configuration corresponds to the combination: [0120] N: ON (invert)-P1: ON (invert)-P2: ON-P3: ON.

[0121] In this second connection configuration, the neutral and the phase P1 are not cut but are inverted.

[0122] The list of connection configurations also comprises third connection configurations each corresponding to one of the phases which is disconnected and which, owing to the conditions of downstream loads, is brought to the potential of another phase.

[0123] For example, a third connection configuration corresponds to the combination: [0124] N: ON-P1: OFF (ON)-P2: ON-P3: ON.

[0125] In this third connection configuration, the phase P1 is cut but is brought to the potential of one of the phases P2 or P3.

[0126] FIG. 6 shows a certain number of connection configurations: [0127] in diagram D1, all of the lines N, P1, P2, P3 are connected (nominal conditions), or indeed the neutral N is disconnected but the loads are balanced; [0128] in diagram D2, the neutral N is disconnected and the loads are unbalanced; [0129] in diagram D3, the phase P1 is disconnected; [0130] in diagram D4, the phase P1 and the phase P2 are disconnected; [0131] in diagram D5, the neutral N and the phase P1 are disconnected; [0132] in diagram D6, the neutral N and the phase P1 are inverted.

[0133] In the reference table 20 mentioned above, the electrical parameters used comprise: [0134] the measured line-to-neutral voltages, each measured line-to-neutral voltage being representative (as already seen) of an amplitude of a line-to-neutral voltage between one of the phases and the neutral; and/or [0135] measured line-to-line voltages, each measured line-to-line voltage being representative (as already seen) of an amplitude of a line-to-line voltage between two distinct phases; and/or [0136] measured angles, each measured angle being equal (as already seen) to an angle between a vector representation of a line-to-neutral voltage between one phase and the neutral, and a vector representation of a line-to-neutral voltage between another phase and the neutral.

[0137] The reference table 20 has therefore been created in a design department, laboratory or factory, and has been downloaded to one of the memories 17 of the processing unit 15. Alternatively, the reference table 20 may be stored remotely (in a data concentrator, a server of the computer system, in the Cloud, etc.), and the processing unit 15 of the meter 5 may access it via suitable communication means. The reference table 20 may be updated while the meter 5 is in operation.

[0138] The inventors have therefore analysed the reference table 20, and have deduced from it the exploratory path and discriminating criteria for reliably determining, during operation, the present connection configurationand, in particular, the presence or absence of the neutral N and the phases Pi.

[0139] An example of a reference table is provided in Annex 1 of the present description.

[0140] The reference table 20 lists, in a virtually exhaustive manner, the different practical cases, and indicates the characteristics of each of these cases (position of each line in the Fresnel reference diagram, values of the line-to-neutral voltages and line-to-line voltages, angles between phases). However, the table in the Annex is limited, for the sake of simplicity and readability, to equal and nominal line-to-neutral voltages (V.sub.1=V.sub.2=V.sub.3=V.sub.n), and to phases wired in a direct order.

[0141] The table in the Annex shows, in bold print, generic cases for a given connection case, and, in standard print, specific cases (typically when the conditions of downstream loads are such that a disconnected line is brought to the same potential as another line).

[0142] Careful analysis and systematic study have shown that, based solely on the measured line-to-neutral voltages, and/or the measured line-to-line voltages and/or the differences in phase angles, it is possible to distinguish between each of these cases, and therefore to deduce with certainty which is the current case and finally which lines (neutral and/or phase(s)) are present or lost.

[0143] A suitable algorithm which iteratively explores the different possible configurations and checks at each step whether the criteria (line-to-neutral voltages and line-to-line voltages and/or line-to-neutral voltages and angles) corresponding to the particular configuration concerned are fulfilled is then used.

[0144] In reference to FIG. 7, the monitoring method according to the invention therefore consists, for the processing unit 15, in acquiring voltage measurements taken by the voltage sensors of the meter 5: step E1.

[0145] The voltage measurements are then processed to produce the current values of the electrical parameters: calculation of the measured line-to-neutral voltages, the measured line-to-line voltages, the differences in phase angles, etc.: step E2.

[0146] The processing unit 15 then accesses the reference table 20 and implements an exploratory algorithm, intended to search the reference table to identify the connection configuration associated with the current values of the electrical parameters: step E3.

[0147] The exploratory algorithm allows the current state to be classified on the basis of the current values of the electrical parameters, from among multiple and exhaustive study cases. The exploratory algorithm allows a systematic and exhaustive exploration of the reference table and therefore of all of the expected study cases according to the different possible connection configuration cases.

[0148] The input information of the exploratory algorithm therefore consists of the current values of the electrical parameters. At the output of this exploratory algorithm, it is reported which line(s) (neutral and/or phase(s)) are connected or disconnected.

[0149] This mechanism for detecting loss of neutral and/or of one or more phases is implemented continuously (for example in a loop, every second, when new voltage measurement values are available).

[0150] The exploratory algorithm then comprises the steps of: [0151] testing a connection configuration, by checking whether predetermined criteria associated with said connection configuration are fulfilled, the predetermined criteria being defined based on the electrical parameters; [0152] if the predetermined criteria are fulfilled, deducing therefrom that said connection configuration is that which corresponds to the current values of the electrical parameters, [0153] if not, repeating the testing step with another connection configuration.

[0154] The exploratory algorithm acquires first predetermined criteria, associated with a connection configuration number 1, corresponding to a first hypothesis of the state of the lines (phases and neutral): step E4.

[0155] The exploratory algorithm tests whether these first predetermined criteria are fulfilled: step E5. If this is the case, the exploratory algorithm deduces from this that the current connection configuration corresponds to connection configuration number 1.

[0156] If not, the exploratory algorithm acquires second predetermined criteria, associated with a connection configuration number 2, corresponding to a second hypothesis of the state of the lines (phases and neutral): step E6.

[0157] The exploratory algorithm tests whether these second predetermined criteria are fulfilled: step E7. If this is the case, the exploratory algorithm deduces from this that the current connection configuration corresponds to connection configuration number 2.

[0158] These steps are repeated a certain number of times. At the Nth iteration, the exploratory algorithm acquires Nth predetermined criteria (step EN) and tests a connection configuration number N, corresponding to an Nth hypothesis of the state of the lines: step EN+1.

[0159] If the Nth predetermined criteria are fulfilled, the exploratory algorithm deduces from this that the current connection configuration corresponds to connection configuration number N. If not, the exploratory algorithm deduces from this that the connection configuration corresponds to an unknown connection case.

[0160] The number N is the number of different connection configurations listed in the reference table 20.

[0161] The exploratory algorithm will now be described in greater detail.

[0162] First of all, a first exploratory algorithm will be described. The first exploratory algorithm uses only the measured line-to-neutral voltages and the measured line-to-line voltages as electrical parameters.

[0163] The first exploratory algorithm, written in pseudo-language and therefore easily transcribed to all major computer languages, is provided in Annex 2.

[0164] The first exploratory algorithm first comprises a preliminary part defining a plurality of parameters and, in particular, a very low voltage threshold Voff. This threshold, with which the measured line-to-neutral voltages are compared, corresponds to a voltage so low that said voltage may be considered to an absence of phase.

[0165] The first exploratory algorithm then comprises a looping instruction every second. Indeed, as already seen, the steps of the monitoring method are repeated regularly: at regular intervals (of 1 second in this instance), the processing unit 15 acquires new voltage measurements and produces new current values of the electrical parameters. The first exploratory algorithm is therefore repeated each second in order to identify the new connection configuration associated with the new current values of the electrical parameters.

[0166] The first exploratory algorithm then calculates the expected line-to-line voltages from the measured line-to-neutral voltages, when no line is cut.

[0167] The first exploratory algorithm then attempts to detect one or more measured line-to-line voltages that are abnormal.

[0168] For the phases Pi and Pj, the measured line-to-line voltage U.sub.i is considered to be abnormal if:

U.sub.i<Uprev.sub.i or if U.sub.i>Uprev.sub.i+, [0169] where: [0170] is a margin of uncertainty, [0171] and where:

Uprev.sub.i=V.sub.n.Math.SQRT(ai.sup.2+aj.sup.2+ai.Math.aj), [0172] where ai=V.sub.i/V.sub.n, aj=V.sub.j/V.sub.n, V.sub.i is the measured line-to-neutral voltage for the phase Pi, V.sub.j is the measured line-to-neutral voltage for the phase Pj, and V.sub.n is a nominal line-to-neutral voltage.

[0173] The above formula corresponds to:

U.sub.n=SQRT(3).Math.V.sub.n [0174] which is obtained in the nominal case of voltages that are balanced and at the nominal value. It should be noted that this relationship is exact if, and only if, the angle relationship between phases Pi and Pj is 120. Non-compliance of the measurement of U.sub.i with the calculation given above and obtained with the measurements V.sub.i and V.sub.j is therefore sufficient to prove that the angle relationship between the phases i and j is not 120.

[0175] The first exploratory algorithm then consists in testing sets of connection configurations, by checking whether predetermined criteria associated with the connection configurations are fulfilled.

[0176] In this instance, the predetermined criteria comprise comparisons each made between: [0177] a resulting voltage obtained by adding: [0178] two different measured line-to-neutral voltages; or [0179] a measured line-to-neutral voltage and a measured line-to-line voltage, [0180] and a measured line-to-neutral voltage or a measured line-to-line voltage.

[0181] The configuration sets are defined as a function of the number of abnormal measured line-to-line voltages.

[0182] Thus, if the number of abnormal measured line-to-line voltages is equal to zero, the first exploratory algorithm considers, by default, that the current connection configuration, over time T, is a nominal connection configuration, in which the neutral and the phases are connected normally, then tests connection configurations in which at least one phase is disconnected.

[0183] The first exploratory algorithm therefore first considers, by default, that the neutral and the three phases are connected normally.

[0184] The first exploratory algorithm then attempts to detect one or more disconnected phases, by comparing, for each phase Pi, the measured line-to-neutral voltage V.sub.i with the threshold Voff. If, for a phase Pi, the measured line-to-neutral voltage is strictly less than Voff, said phase Pi is considered to be disconnected.

[0185] If the number of abnormal measured line-to-line voltages is equal to one, the first exploratory algorithm considers, by default, that the current connection configuration is a nominal connection configuration, then tests connection configurations in which a single phase is disconnected, and connection configurations in which a phase is disconnected and is brought to the potential of another phase.

[0186] The first exploratory algorithm therefore considers, by default, that the neutral and the three phases are normally connected.

[0187] Then, for each phase Pi, the algorithm checks whether:

Vi+Ui=Vj, or whether Vi+Uk=Vk (in which j=i+1 and k=i+2).

[0188] If this is the case, the algorithm deduces from this that the phase Pi is disconnected.

[0189] If the algorithm has detected that the phase Pi is disconnected, the algorithm checks whether:

Vi=Vj or whether Vi=Vk.

[0190] If this is the case, it deduces from this that the phase Pi is connected to another phase.

[0191] If the number of abnormal measured line-to-line voltages is equal to two, the first exploratory algorithm considers, by default, that the current connection configuration is a nominal connection configuration, then tests connection configurations in which a single phase is disconnected (and brought back between the two other phases and the neutral).

[0192] The first exploratory algorithm therefore considers, by default, that the neutral and the three phases are normally connected.

[0193] Then, for each phase Pi, the algorithm checks whether:

Vi<Vj and Ui<Vj or whether

Vi<Vk and Uk<Vk.

[0194] If this is the case, the algorithm deduces from this that the phase Pi is disconnected.

[0195] If the number of abnormal measured line-to-line voltages is equal to three, the first exploratory algorithm considers, by default, that the current connection configuration is a connection configuration in which the neutral only is cut (and in which there is clearly a significant imbalance in the downstream loads), then tests connection configurations in which one phase and the neutral are cut (and brought back between the two other phases), then a connection configuration in which one phase is connected to another phase, then connection configurations in which two phases are cut (brought back between the neutral and the remaining phase), then a connection configuration in which three phases are connected to each other, then connection configurations in which one phase is inverted with the neutral.

[0196] The first exploratory algorithm first considers, by default, that the neutral is disconnected and the three phases are correctly connected.

[0197] Next, for each phase Pi, the algorithm checks whether:

Vj+Vk=Uj and Ui+Uk=Uj

[0198] If this is the case, the algorithm deduces from this that the phase Pi is disconnected.

[0199] If the algorithm has detected that the phase Pi is disconnected, it checks whether:

U=Uj or whether Uk=Uj.

[0200] If this is the case, it deduces from this that the phase Pi is connected to another phase.

[0201] Next, the algorithm attempts to detect the loss of two phases only (brought back between the neutral and the remaining phase).

[0202] For each phase Pi, the algorithm checks whether:

Vi+Uk=Vk and Vj+Uj=Vk

[0203] If this is the case, the algorithm deduces from this that the phase Pi and the phase Pj are disconnected.

[0204] If the algorithm has detected that the phase Pi and the phase Pj are disconnected, the algorithm checks whether:

Vi=Vj and whether Vi=Vk.

[0205] If this is the case, it deduces from this that the three phases are connected together.

[0206] Next, the algorithm attempts to detect an inversion of the neutral and one phase in the wiring of the meter 5.

[0207] For each phase Pi, the algorithm calculates the expected line-to-line voltages Uprev_inv[i], Uprev_inv[j], Uprev_inv[k] knowing the measured line-to-neutral voltages and assuming an inversion of the neutral and the phase Pi.

[0208] If the following applies:

[0209] Uprev_inv[i]=Vj and Uprev_inv[j]=Uj and Uprev_inv[k]=Vk, the algorithm deduces from this that the neutral and the phase Pi are inverted.

[0210] Next, for each phase Pi, the algorithm compares the line-to-neutral voltage Vi with predetermined thresholds in order to assign a status concerning the voltage levels coming from the network: overvoltage, undervoltage, significant undervoltage, dangerous overvoltage, etc.

[0211] The algorithm also assigns a status to the dangerousness of the voltage level on the downstream side, and possibly controls the opening of the breakers in order to protect the equipment connected downstream of the meter 5 (equipment of the installation 7).

[0212] It should be noted that all of these comparisons can be made by adding margins allowing uncertainties to be taken into account; for example, as shown in the first exploratory algorithm provided in Annex 2, rather than checking whether:

Vi+Ui=Vj or Vi+Uk=Vk, [0213] the following is checked:

Vi+Ui=Vj+/(2.Math.Y) or Vi+Uk=Vk+/(2.Math.Y), [0214] Y being a margin allowing uncertainty in the voltage measurements to be taken into account.

[0215] The first exploratory algorithm (like the second which is described below) therefore provides the margins required to take into account measurement uncertainties and initial tolerances of the electricity network.

[0216] A second exploratory algorithm will now be described.

[0217] The second exploratory algorithm uses only the measured line-to-neutral voltages and the measured angles as electrical parameters.

[0218] The second exploratory algorithm, written in pseudo-language, is provided in Annex 3.

[0219] The second exploratory algorithm first comprises a preliminary part defining a plurality of parameters and, in particular, a high angular threshold and a low angular threshold.

[0220] The second exploratory algorithm then comprises a looping instruction every second.

[0221] The second exploratory algorithm then compares each measured line-to-neutral voltage with the threshold Voff and, if all of the measured line-to-neutral voltages are strictly less than said threshold Voff, the algorithm is inhibited.

[0222] For each phase Pi, if the line-to-neutral voltage measured on the phase Pi is strictly less than the predetermined threshold Voff, the measured angles involving said phase are not considered reliable and are set to zero.

[0223] The algorithm first tests a nominal connection configuration in which the neutral and the phases are connected normally.

[0224] The algorithm compares each measured angle with a reference value Aref and, if the measured angles are all equal to Aref (possibly taking into account a margin of uncertainty), the algorithm deduces from this that the neutral and the three phases are normally connected and that the phases are wired to the meter 5 in a direct order.

[0225] If this condition is not met, the algorithm compares each measured angle with a reference value Aref and, if the measured angles are all equal to Aref, the algorithm deduces from this that the neutral and the phases are normally connected and that the phases are wired to the meter in a direct order.

[0226] The algorithm then tests connection configurations, in each of which one phase is inverted with the neutral.

[0227] For this purpose, the algorithm compares each measured angle with the high and low angular thresholds. The predetermined criteria therefore only comprise comparisons between the measured angles and angular thresholds.

[0228] Next, the algorithm tests a connection configuration in which a single phase is disconnected, the two other phases and the neutral being normally connected. The predetermined criteria comprise comparisons between the measured angles and angular thresholds, and comparisons between the measured line-to-neutral voltages and voltage thresholds.

[0229] If the algorithm has detected a disconnected phase, it then attempts to detect whether said phase has been brought to the potential of one of the other phases.

[0230] Next, the algorithm attempts to detect a connection configuration in which the neutral is disconnected, the three other phases still being connected.

[0231] Once again, the predetermined criteria only comprise comparisons between the measured angles and angular thresholds.

[0232] The algorithm then tests a connection configuration in which two lines are disconnected, the other lines remaining connected.

[0233] The predetermined criteria comprise comparisons between the measured angles and angular thresholds, and comparisons between the measured line-to-neutral voltages and voltage thresholds.

[0234] The algorithm then attempts to detect a connection configuration in which two phases are disconnected, the neutral and the other phase remaining connected (specific case of the above configuration).

[0235] If the algorithm has detected a disconnected phase, it then attempts to detect whether said phase has been brought to the potential of one of the other phases.

[0236] The algorithm then attempts to detect a disconnection configuration in which the neutral and one phase are disconnected, the two other phases being normally connected (specific case of the above configuration).

[0237] If the algorithm has detected a disconnected phase, it then attempts to detect whether said phase has been brought to the potential of one of the other phases.

[0238] Next, the algorithm once more tests a connection configuration in which the neutral and a phase are disconnected, the two other phases being normally connected.

[0239] Finally, like the first exploratory algorithm, for each phase Pi, the second exploratory algorithm compares the line-to-neutral voltage Vi with predetermined thresholds in order to assign a status concerning the voltage levels coming from the network: overvoltage, undervoltage, significant undervoltage, dangerous overvoltage, etc.

[0240] The algorithm also assigns a status to the dangerousness of the voltage level on the downstream side, and possibly controls the opening of the breakers in order to protect the equipment connected downstream of the meter 5 (equipment of the installation 7).

[0241] This invention is therefore based on criteria that are easily measured in the profession (voltage measurements and their processing) and highly deterministic, and it does not require any reference measurements from before the suspected loss of neutral. The voltage measurements taken at time T are sufficient on their own in order to be able to ascertain whether or not a loss of neutral and/or of one or more phases has taken place, because they are sufficient in order to determine the current connection configuration by analysing the corresponding characteristic conditions in the Fresnel diagram.

[0242] In the final implementation of the proposed solution, irrespective of the specific embodiment of the method, it is advantageous to allow a suitable margin of uncertainty, relating in particular to the accuracy of the voltage measurements and possibly of angles (typically of +/1V or +/1% and +/2 respectively, depending on the method and the acquisition means), as proposed in the specific embodiments of the exploratory algorithms provided.

[0243] Optionally, in order to minimize the risk of false positives, irrespective of the specific embodiment of the method, it may be ascertained that the predetermined detection criteria are fulfilled over several consecutive iterations before reporting an actual loss of neutral and/or of one or more phases.

[0244] Optionally, the detection of loss of neutral and/or of one or more phases generates an alarm that is timestamped in one of the memories 17 of the processing unit 15 of the meter 5 and/or, in the case of communicating meters (for example communicating via a transmission channel based on PLC technology or cellular radio technology), an alarm signal may be sent to equipment external to the meter, and, for example, to the remote information system (possibly spontaneously, depending on the high-level protocol used).

[0245] Optionally, the detection of loss of neutral and/or of one or more phases generates a specific and dedicated display on the local display of the meter or by means of an indicator light.

[0246] Optionally, the detection of loss of neutral and/or of one or more phases generates the emission of a sound signal.

[0247] Optionally, on detection of loss of neutral and when the imbalance in the line-to-neutral voltages is considered to be abnormally significant, an alarm signal is generated and timestamped in a memory of the processing unit of the meter and/or, in the case of communicating meters, sent to equipment external to the meter, and, for example, to the remote information system.

[0248] Optionally, on detection of loss of neutral and when the imbalance in the line-to-neutral voltages is considered to be too high and dangerous for the equipment downstream, in the case of meters provided with breakers, the breakers are activated to switch to the open state and thus protect the equipment downstream.

[0249] Finally, it is also possible to use the invention proposed here in a different way in order not to detect the loss of a conductor (phase(s) and/or neutral), but rather to automatically detect the wiring used for connection to the network (2 wires, 3 wires, 4 wires, etc.).

[0250] Naturally, the invention is not limited to the described embodiments, but covers any variant that falls within the scope of the invention as defined by the claims. The item of electrical equipment in which the invention is implemented (at least partially) is not necessarily a meter, but may be any type of equipment connected to a network and capable of being provided with voltage sensors (a circuit breaker, for example).

[0251] Annexes

[0252] Annex 1

[0253] An example of a reference table is provided here.

[0254] The reference table, which comprises 32 rows and 20 columns, has been split into 4 tables to improve the readability of the table.

[0255] Table 1 comprises 32 rows and the 6 left-hand columns of the reference table; table 2 comprises 32 rows and 5 columns situated to the right of the columns of table 1; table 3 comprises 32 rows and 6 columns situated to the right of the columns of table 2; table 4 comprises 32 rows and 3 columns situated to the right of the columns of table 3.

TABLE-US-00001 TABLE 1 Presence/Absence of each Line Connection configurations N Li Lj Lk All Lines Connected 3P + N ON ON ON ON (Nominal Conditions) Neutral Disconnected 3P OFF ON ON ON Neutral Disconnected 3P OFF ON ON ON Neutral Disconnected 3P OFF ON ON ON Neutral Disconnected 3P OFF ON ON ON Neutral Disconnected 3P OFF ON ON ON Line i Disconnected 2P + N ON OFF ON ON Line i Disconnected 2P + N ON OFF ON ON Line i Disconnected 2P + N ON OFF ON ON (false 3P + N) (ON) Line i Disconnected 2P + N ON OFF ON ON (false 3P + N) (ON) Line i & Line j Disconnected 1P + N ON OFF OFF ON Line i & Line j Disconnected 1P + N ON OFF OFF ON Line i & Line j Disconnected 1P + N ON OFF OFF ON Line i & Line j Disconnected 1P + N ON OFF OFF ON Line i & Line j Disconnected 1P + N ON OFF OFF ON Line i & Line j Disconnected 1P + N ON OFF OFF ON (false 2P + N) (ON) Line i & Line j Disconnected 1P + N ON OFF OFF ON (false 2P + N) (ON) Line i & Line j Disconnected 1P + N ON OFF OFF ON (false 2P + N) (ON) Line i & Line j Disconnected 1P + N ON OFF OFF ON (false 2P + N) (ON) Line i & Line j Disconnected 1P + N ON OFF OFF ON (false 3P + N) (ON) (ON) Neutral & Line i Disconnected 2P OFF OFF ON ON Neutral & Line i Disconnected 2P OFF OFF ON ON Neutral & Line i Disconnected 2P OFF OFF ON ON Neutral & Line i Disconnected 2P OFF OFF ON ON Neutral & Line i Disconnected 2P OFF OFF ON ON Neutral & Line i Disconnected 2P OFF OFF ON ON Neutral & Line i Disconnected 2P OFF OFF ON ON (ON) Neutral & Line i Disconnected 2P OFF OFF ON ON (ON) Neutral & Line i Disconnected 2P OFF OFF ON ON (ON) Neutral & Line i Disconnected 2P OFF OFF ON ON (ON) Neutral & Line i Inverted 3P + N ON ON ON ON N P invert invert invert

TABLE-US-00002 TABLE 2 Position of each Line on the Fresnel diagram Impedance Conditions between with respect to the ABCD Fresnel reference each Line (on downstream side in Points (for i = 1; j = 2; k = 3) regard to the disconnection point(s)) N P1 P2 P3 Any A B C D Balanced Zi = Zj = Zk Any Unbalanced Zi Zj Zk within BCD B C D (except A) Zi << (Zj & Zk) B B C D (e.g. terminal N shunt to terminal i) Zj << (Zi & Zk) C B C D (e.g. terminal N shunt to terminal j) Zk << (Zi & Zj) D B C D (e.g. terminal N shunt to terminal k) Any A ACD C D Zi << (Zij & Zik) A A C D Zij << (Zi & Zik) A C C D (e.g. terminal i shunt to terminal j) Zik << (Zi & Zij) A D C D (e.g. terminal i shunt to terminal k) Any A [AD] [AD] D Zij << (Zi & Zik) or Zij << (Zj & Zjk) A [AD] [AD] D Zi << Zik A A [AD] D Zj << Zjk A [AD] A D Zi << Zik and Zj << Zjk A A A D Zik << Zi A D [AD] D (e.g. terminal i shunt to terminal k) Zik << Zi and Zj << Zjk A D A D Zjk << Zj A [AD] D D (e.g. terminal j shunt to terminal k) Zi << Zik and Zjk << Zj A A D D Zik << Zi and Zjk << Zj A D D D Any [CD] [CD] C D Zi << (Zij & Zik) or Zi << (Zj & Zk) [CD] [CD] C D Zj << Zk (e.g. terminal N shunt to terminal j) C [CD] C D Zj << Zk and Zij << (Zi & Zik) C C C D (e.g. terminal N & i shunt to terminal j) Zk << Zj D [CD] C D (e.g. terminal N shunt to terminal k) Zk << Zj and Zik << (Zi & Zij) D D C D (e.g. terminal N & i shunt to terminal k) Zij << (Zi & Zik) [CD] C C D (e.g. terminal i shunt to terminal j) Zik << (Zi & Zij) [CD] D C D (e.g. terminal i shunt to terminal k) Zj << Zk and Zik << (Zi & Zij) C D C D (e.g. terminal N shunt to terminal j & terminal i shunt to terminal k) Zk << Zj and Zij << (Zi & Zik) D C C D (e.g. terminal N shunt to terminal k & terminal i shunt to terminal j) Any B A C D

TABLE-US-00003 TABLE 3 Measured Angles between phases (X: not measurable, due to low voltage condition on associated Line(s)) Measured Line-to-Neutral Voltages (i, j) (j, k) (k, i) Vi Vj Vk 120 120 120 Vn Vn Vn [60; 120[ [60; 120[ [60; 120[ [0; SQRT(3).Vn] [0; SQRT(3).Vn] [0; SQRT(3).Vn] ]120; 180] ]120; 180] ]120; 180] X 60 X 0 SQRT(3).Vn SQRT(3).Vn X X 60 SQRT(3).Vn 0 SQRT(3).Vn 60 X X SQRT(3).Vn SQRT(3).Vn 0 [120; 0] 120 [120; 0] [0; Vn] Vn Vn X 120 X 0 Vn Vn 0 120 120 Vn Vn Vn 120 120 0 Vn Vn Vn 0 0 0 [0; Vn] [0; Vn] Vn X 0 X [0; Vn] [0; Vn] Vn X 0 X 0 [0; Vn] Vn X X 0 [0; Vn] 0 Vn X X X 0 0 Vn 0 0 0 Vn [0; Vn] Vn X X 0 Vn 0 Vn 0 0 0 [0; Vn] Vn Vn X 0 X 0 Vn Vn 0 0 0 Vn Vn Vn 0 or 180 180 0 or 180 [0; SQRT(3).Vn] [0; SQRT(3).Vn] [0; SQRT(3).Vn] X 180 X 0 [0; SQRT(3).Vn] [0; SQRT(3).Vn] X X 0 [0; SQRT(3).Vn] 0 SQRT(3).Vn X X X 0 0 SQRT(3).Vn 0 X X [0; SQRT(3).Vn] SQRT(3).Vn 0 X X X 0 SQRT(3).Vn 0 0 180 180 [0; SQRT(3).Vn] [0; SQRT(3).Vn] [0; SQRT(3).Vn] 180 180 0 [0; SQRT(3).Vn] [0; SQRT(3).Vn] [0; SQRT(3).Vn] X X 0 SQRT(3).Vn 0 SQRT(3).Vn 0 X X SQRT(3).Vn SQRT(3).Vn 0 30 60 30 Vn [0; SQRT(3).Vn] [0; SQRT(3).Vn]

TABLE-US-00004 TABLE 4 Measured Line-to-line voltages Ui Uj Uk SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn [0; SQRT(3).Vn] SQRT(3).Vn [0; SQRT(3).Vn] Vn SQRT(3).Vn Vn 0 SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn 0 [0; Vn] [0; Vn] [0; Vn] 0 [0; Vn] [0; Vn] [0; Vn] [0; Vn] Vn [0; Vn] Vn [0; Vn] 0 Vn Vn [0; Vn] [0; Vn] 0 Vn Vn 0 [0; Vn] 0 [0; Vn] Vn 0 Vn 0 0 0 [0; SQRT(3).Vn] SQRT(3).Vn [0; SQRT(3).Vn] [0; SQRT(3).Vn] SQRT(3).Vn [0; SQRT(3).Vn] [0; SQRT(3).Vn] SQRT(3).Vn [0; SQRT(3).Vn] 0 SQRT(3).Vn SQRT(3).Vn [0; SQRT(3).Vn] SQRT(3).Vn [0; SQRT(3).Vn] SQRT(3).Vn SQRT(3).Vn 0 0 SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn SQRT(3).Vn 0 SQRT(3).Vn SQRT(3).Vn 0 0 SQRT(3).Vn SQRT(3).Vn Vn SQRT(3).Vn Vn

[0256] Annex 2

[0257] The first exploratory algorithm is as follows:

TABLE-US-00005 // Parameters Vn = 230 //Vrms; nominal voltage Tol_high = 10% // initial tolerance for the input voltage: Vn + Tol_high Tol_low = 10% // initial tolerance for the input voltage: Vn Tol_low Vdanger = 276 // Vrms; voltage threshold considered dangerous for the downstream equipment Vlow = 184 //Vrms; voltage considered abnormally low Voff = 30 //Vrms; voltage so low that it is considered to be an absence of phase X = 1% // to take into account, in particular, the 120 inter-phase uncertainty Y = 1 //V; to take into account, in particular, the uncertainty in the voltage measurement // Hypothesis: three-phase, four-wire network (3P+N) with fixed and constant phase shift of 120 between phases (+/(X/2)%) // V[i] = root mean square voltage measured between the phase_i & the neutral over the last second // U[i] = root mean square voltage measured between the phase_i & the phase_j over the last second // In principle, take the fundamental root mean square voltage measurement (50Hz) loop every second NeutralConnectStatus = UNKNOWN PhaseConnectStatus[ ] = [UNKNOWN;UNKNOWN;UNKNOWN] PhaseVoltageStatus[ ] = [UNKNOWN;UNKNOWN;UNKNOWN] for i = 1 to 3 a[i] = V[i]/Vref end for for i = 1 to 3 j = i+1 %3 // Calculation of the expected line-to-line voltages, knowing the line-to- neutral voltages, if all the lines are present (3P+N) Uprev[i] = Vref.SQRT(a[i].sup.2+a[j].sup.2+a[i].a[j]) // Counting of the measured line-to-line voltages with an abnormal value if ((U[i]<(Uprev[i]) or (U[i]>(Uprev[i]))) // if ((U[i]<(Uprev[i](2.Y))) or (U[i]>(Uprev[i]+(2.Y)))) abnormal[i] = 1 else abnormal[i] = 0 end if end for if (sum(abnormal[ ])=0) // default case: all the lines are present NeutralConnectStatus = CONNECTED PhaseConnectStatus[ ] = [CONNECTED;CONNECTED;CONNECTED] // specific case: one phase has an extremely low voltage, which is considered to an absence of phase for i = 1 to 3 if (V[i]<Voff) PhaseConnectStatus[i] = DISCONNECTED end if end for end if if (sum(abnormal[ ])=1) // default case: loss of the neutral only (and obviously low imbalance of the downstream loads) // NeutralConnectStatus = DISCONNECTED // default case (alternative): not yet significant enough to declare a loss of neutral, it is still considered potentially present NeutralConnectStatus = CONNECTED PhaseConnectStatus[ ] = [CONNECTED;CONNECTED;CONNECTED] // counter-proposal: loss of one phase only (brought back between one of the remaining phases and the neutral) for i = 1 to 3 j = i+1 %3 k = i+2 %3 if ((V[i]+U[i]=V[j]) or (V[i]+U[k]=V[k])) // if ((V[i]+U[i]=V[j]+/ (2.Y)) or (V[i]+U[k]=V[k]+/(2.Y))) NeutralConnectStatus = CONNECTED PhaseConnectStatus[i] = DISCONNECTED // processing of the specific case of the counter-proposal (obtained if 1 phase is connected to another phase) if ((V[i]=V[j]) or (V[i]=V[k])) // if ((V[i]=V[j]+/Y) or (V[i]=V[k]+/Y)) PhaseConnectStatus[i] = CONNECTED end if end if end for end if if (sum(abnormal[ ])=2) // default case: loss of the neutral only (and obviously low imbalance of the downstream loads) // NeutralConnectStatus = DISCONNECTED // default case (alternative): not yet significant enough to declare a loss of neutral, it is still considered potentially present NeutralConnectStatus = CONNECTED PhaseConnectStatus[ ] = [CONNECTED;CONNECTED;CONNECTED] // counter-proposal: loss of one phase only (brought back between the 2 other phases and the neutral) for i = 1 to 3 j = i+1 %3 k = i+2 %3 if(((V[i]<V[j]) and (U[i]<V[j])) or ((V[i]<V[k]) and (U[k]<V[k]))) // if (((V[i]<V[j]+Y) and (U[i]<V[j]+Y)) or ((V[i]<V[k]+Y) and (U[k]<V[k]+Y))) NeutralConnectStatus = CONNECTED PhaseConnectStatus[i] = DISCONNECTED end if end for end if if (sum(abnormal[ ])=3) // default case: loss of the neutral only (and clearly significant imbalance of the downstream loads) NeutralConnectStatus = DISCONNECTED PhaseConnectStatus[ ] = [CONNECTED;CONNECTED;CONNECTED] // counter-proposal 1: loss of the neutral and also of one phase (brought back between the 2 other phases) for i = 1 to 3 j = i+1 %3 k = i+2 %3 if ((V[j]+V[k]=U[j]) and (U[i]+U[k]=U[j])) // if ((V[j]+V[k]=U[j]+/(2.Y)) and (U[i]+U[k]=U[j]+/(2.Y))) PhaseConnectStatus[i] = DISCONNECTED // processing of the specific case of the counter-proposal 1 (obtained if 1 phase is connected to another phase) if ((U[i]=U[j]) or (U[k]=U[j])) // if ((U[i]=U[j]+/(2.Y)) or (U[k]=U[j]+/(2.Y))) PhaseConnectStatus[i] = CONNECTED end if end if end for // counter-proposal 2: loss of 2 phases only (brought back between the neutral and the remaining phase) for i = 1 to 3 j = i+1 %3 k = i+2 %3 if ((V[i]+U[k]=V[k]) and (V[j]+U[j]=V[k])) // if ((V[i]+U[k]=V[k]+/(2.Y)) and (V[j]+U[j]=V[k]+/(2.Y))) NeutralConnectStatus = CONNECTED PhaseConnectStatus[i] = DISCONNECTED PhaseConnectStatus[j] = DISCONNECTED // processing of the specific case of the counter-proposal 2 (obtained if the 3 phases are connected together) if ((V[i] = V[j]) and (V[i] = V[k])) // if ((V[i] = V[j]+/Y) and (V[i] = V[k]+/Y)) PhaseConnectStatus[ ]=[CONNECTED;CONNECTED;CONNECTED] end if end if end for // counter-proposal 3: inversion of the neutral and one phase in the wiring of the meter for i = 1 to 3 j = i+1 %3 k = i+2 %3 // Calculation of the expected line-to-line voltages, knowing the line-to-neutral voltages, if all the lines are present (3P+N) and assuming an inversion of the neutral and the phase i a[i] = V[i]/Vref a[j] = U[i]/Vref a[k] = U[k]/Vref Uprev_inv[i] = Vref.SQRT(a[i].sup.2+a[j].sup.2+a[i].a[j]) Uprev_inv[j] = Vref.SQRT(a[j].sup.2+a[k].sup.2+a[j].a[k]) Uprev_inv[k] = Vref.SQRT(a[k].sup.2+a[i].sup.2+a[k].a[i]) if (Uprev_inv[ ]=[V[j];U[j];V[k]) // if (Uprev_inv[ ]=[V[j]+/ (2.Y);U[j]+/(2.Y);V[k]+/(2.Y)) NeutralConnectStatus = INVERTED_WITH_PHASE PhaseConnectStatus[i] = INVERTED_WITH_NEUTRAL PhaseConnectStatus[j] = CONNECTED PhaseConnectStatus[k] = CONNECTED end if end for end if for i = 1 to 3 // A status is given for the voltage levels from the network if (NeutralConnectStatus = DISCONNECTED) PhaseVoltageStatus[i] = UNKNOWN elseif (PhaseConnectStatus[i] = DISCONNECTED) PhaseVoltageStatus[i] = UNKNOWN elseif (V[i]>Vdanger) PhaseVoltageStatus[i] = DANGEROUS_OVERVOLTAGE elseif (V[i]>(1+Tol_high).Vn) PhaseVoltageStatus[i] = OVERVOLTAGE elseif (V[i]<(1Tol_low).Vn) PhaseVoltageStatus[i] = UNDERVOLTAGE elseif (V[i]<Vlow) PhaseVoltageStatus[i] = DEEP_UNDERVOLTAGE else PhaseVoltageStatus[i] = NORMAL end if // A status is given for the dangerousness of the voltage level on the downstream side end for if (max(V[ ])>Vdanger) DownStreamVoltageDanger = TRUE // possibly open the breakers in order to protect the equipment connected downstream else DownStream VoltageDanger = FALSE end if end loop

[0258] Annex 3

[0259] The second exploratory algorithm is as follows: