FR2725277A1 - Detection of faults on public lighting network - Google Patents

Abstract

The process is for detecting faults in the public lighting network which has lamps (N) fed from sections (S) and supply points (D). Multiplexers and a microprocessor are used to measure values of voltage and current (U,I) in a section of the network (S), also to determine the corresponding values of real and reactive power (P,Q) and their ratio Q/P. The values of reactive power (Q) and the ratio of reactive to real power (Q/P) are compared to reference values (VINF) and (VSUP) to determine whether a fault exists. If a fault does exist, a signal is sent and/or recorded and the reference values are modified to allow for the faulty equipment in later measurements.

Description

 The invention relates to a method for detecting faults in a public lighting network and to a device for implementing this method.

STATE OF THE TECHNOLOGY
Nowadays, the streets and spaces of agglomerations are generally lit by lampposts equipped with mercury or sodium vapor lamps. The distribution of the energy required for the operation of the lamps is effected by means of distribution boxes supplied by the public distribution network and essentially comprising a circuit breaker, a general contactor, secondary circuit breakers, and a remote control relay capable of being actuated. by a signal sent over the network by the manager of this network.

 Lamps are normally checked in small and medium-sized towns by visual observation and, in large cities, by visual inspection during periodic rotations, in practice, at least once a week.

 In general, visual examination refers to the presence of illumination, followed, if necessary, by a control of the constituents (fuse, ballast, initiator).

 Until now, there has been little formal or global maintenance of lighting installations.

SUMMARY SUMMARY OF THE INVENTION
The applicant has now developed a fault detection method in a public lighting network consisting of an electrical distribution network, on which is connected at least one start D comprising at least one section S feeding at least one source of radiation N (S) .This process is characterized in that it comprises at least the following steps
a) the values of the voltage U and the intensity I are measured in the section S
b) the values of the voltage U and the intensity I are determined from:
the reactive power Q,
- and possibly the active power P and the ratio Q / P,
c) comparing said reactive power Q or said Q / P ratio with reference values VINF and VSUP,
d) if said reactive power Q or said Q / P ratio is, at the same time, less than or equal to VSUP and greater than or equal to VINF, return to step a )
e) in the opposite case, a signal indicating the existence of a defect is sent and / or recorded, then said reference values VSUP and
VINF, so that said defect is not detected again during a subsequent examination of the section S, and return to step a).

The applicant has also designed a device that can at least implement its method and comprising at least the following elements:
means for measuring the voltage U and the intensity I in at least one
section S,
determining means for determining, from the values of U and I, the reactive power Q and possibly the active power P and the ratio
Q / P,
processing means able at least to compare said power Q or said Q / P ratio with reference values VINF and VSUP, and to modify these values,
memory means containing at least one program governing the operation of said processing means,
and means for transmitting a signal.

 This device has the advantage of being able to control alone a very large number of radiation sources distributed by section and departures. It can also take into account the voltage variations of the electrical distribution network. Such a device also imposes very few modifications to the existing lighting network. It is more adaptable, inexpensive and reliable because it is very insensitive to disturbances or fluctuations in the network voltage.

DESCRIPITON SUMMARY OF FIGURES
Figure 1 is a schematic representation of a radiation source and components that are generally associated with it.

FIG. 2 is a schematic representation of a portion of the public lighting network, the dotted lines delimiting a conventional distribution box C.
Figure 3 illustrates the variations of the active and reactive powers, intensity and tangent a as a function of the voltage of the network.

 Figure 4 illustrates the variations of the active and reactive powers, the intensity and tan a as a function of time.

 FIG. 5 is a schematic representation of a box of the public lighting network in which a device according to the invention has been connected.

 Figure 6 is a block diagram showing the internal constitution of the device according to the invention.

 FIG. 7 represents an exemplary flow chart that can govern the operation of the microprocessor of the device according to the invention.

 FIG. 8 represents a flowchart governing the operation of the microprocessor according to an advantageous version of the device according to the invention.

 FIG. 9 represents another flowchart governing the operation of the microprocessor according to another advantageous version of the device according to the invention.

DETAILED DESCRIPTION OF THE INVENTION
Figure 1 illustrates a source of radiation generally used in public lighting networks. By radiation source is meant in the present description the assembly consisting of a discharge lamp R such as a mercury vapor lamp, an inductance L called ballast in series with the discharge lamp R, of a capacitor C in parallel with the discharge lamp assembly
R / L ballast, and an overcurrent protection device F which may be a fuse or a circuit breaker. The discharge lamp may also be a sodium vapor lamp. But in this case, the radiation source must be provided with an initiator.

 By default, in the present description is meant the non-functioning of a component of the radiation source, that is to say the absence of current flow through this component.

 In general, ballast defects are rare. And since the ballast is in series with the lamp, a ballast fault will be, in the present description, likened to a lamp fault.

 It should also be noted that any short circuit inside the ballast, capacitor or lamp results in a fuse blowing and can therefore be likened to a fuse fault.

The average values of the various physical quantities during the operation of a radiation source comprising a 125W mercury vapor lamp are given in table A below.
TABLE A

<tb><SEP> P <SEP> (Watt) <SEP> Q <SEP> (VAR) <SEP> I <SEP> (Amp) <SEP>
<tb> Without <SEP> default <SEP> 129 <SEP> 32 <SEP> 0.62
<tb><SEP><SEP> capacitor fault <SEP> 123 <SEP> 194 <SEP> 1.02
<tb> Am defect <SEP><SEP> 7 <SEP> -167 <SEP> 0.71
<Tb>
FIG. 2 shows a public lighting network consisting of an electrical distribution network to which a plurality of feeders D1, D2, D3 are connected. Each of these feeders comprises 3 sections S1, S2 and S3. which are composed of the neutral N and the phases L1, L2 and L3. Radiation sources Nil, N12, ...., Nln, N21, N22 .... N2n, N31, N32,..., N3n are connected to sections S1, S2, S3.

PROCESS ACCORDING TO THE INVENTION
The Applicant has found that a fault in a network could easily be detected by monitoring at least one of the physical quantities that are the reactive power Q consumed in the section S considered or the ratio Q / P of the reactive power Q on the active power P, that is to say the tangent of the angle a corresponding to the phase shift of the voltage relative to the current in the section S considered.

 Indeed, as shown in the following Table B, in case of failure of one of the constituents of the radiation source, it immediately appears a significant variation of the values of the quantities Q and Q / P above compared to normal operation .

 This table B was made with 10 radiation sources including mercury vapor lamps of the same nominal power. The columns F, A, C respectively indicate the state of the fuse, the bulb and the capacitor. The value 0 corresponds to a normal operation and the value 1 to a fault, that is to say a non-functioning.

TABLE B

<tb><SEP> F <SEP> A <SEP> C <SEP> I <SEP> P <SEP> Q <SEP> Difference <SEP> Q <SEP> tana <SEP> Difference
<tb><SEP> (Amps) <SEP> (Watts) <SEP> (VARs) <SEP> in <SEP>% <SEP> = Q / P <SEP> tan <SEP> a
<tb><SEP> in <SEP>%
<tb><SEP> 0 <SEP> 0 <SEP> 0 <SEP> 7.0 <SEP> 1392 <SEP> 518 <SEP> 0 <SEP> 0.37 <SEP> 0
<tb><SEP> 0 <SEP> 0 <SEP> 1 <SEP> 7.3 <SE> 1380 <SE> 652 <SEP> +26 <SEP> 0.47 <SEP> +27
<tb><SEP> 0 <SEP> 1 <SEP> 0 <SEP> 6.0 <SEP> 1256 <SEP> 304 <SEP> -41 <SEP> ~ <SEP><SEP> 0.24 <SEP> -35
<tb><SEP> 1 <SEP> 0 <SEP> 0 <SEP> 6.3 <SEQ> 1247 <SEP> 440 <SEP> -15 <SEP> 0.35 <SEP> -5.4 (1) <September>
<tb><SEP> 0 <SEP> 1 <SEP> 1 (2) <SEP> 6.3 <SEP> 1254 <SEP> 457 <SEP> -12 <SEP> 0.36 <SEP> -2
<tb><SEP> 1 <SEP> 0 <SEP> 1 <SEP> 6.6 <SEQ> 1258 <SEP> 611 <SEP> +18 <SEP> 0.48 <SEP> +30
<tb><SEP> 1 <SEP> 1 (2) <SEP> 0 <SEP> 54 <SEQ> 1140 <SEP> 258 <SEP> -50 <SEW> 022 <SEP> -41
<tb> i <SEP> 1 (2) <SEP> 1 (2) <SEP> 5.6 <SEP> 1120 <SEP> ~ <SEP><SEP> 401 <SEP> -23 <SEP> 036 <SEP > -2
<tb> (1): the value should be 0 in theory because Q / P is constant. The difference is due to measurement uncertainties.

(2): defects are located on different radiation sources.

It thus appears that as soon as a fault occurs, a large difference with respect to the fault-free operation is manifested in the active power level.
Q or the tangent of the phase shift angle a.

 The monitoring of the variations of Q or tan a thus makes it possible to detect any fault occurring in the network then to transmit and / or record a signal informing of the existence of this defect and preferably also indicating the section in which this defect manifests itself. It is not necessary for the process to know more precisely the location of this defect because each section comprises a small number of radiation sources and extends only a few hundred meters. The team of agents in charge of the surveillance of the lighting network, can then, while going in the section in question, quickly determine which is the source of defective radiation.

 Indeed, if it is a defect fuse, lamp or possibly ballast, its location is easily because visually, the lamp of the radiation source concerned having extinguished.

 If against all lamps work, it means that a capacitor is defective because only a capacitor fault does not result in the extinction of a lamp. The monitoring team must then individually control the capacitors until they find the one that is defective.

 The appearance of another fault later in the same section can also be detected, if the reference values are modified so that they take into account the previous fault.

Thus, the following Table C shows, in general, the differences occurring after appearance of a capacitor fault following a fault of the bulb or vice versa:
TABLE C

<tb> nth <SEP> Default <SEP> n + <SEP> l <SEP> Default <SEP> Difference <SEP> I <SEP> Difference <SEP> P <SEP> Difference <SEP> Q <SEP> Difference <SEP > tan <SEP> a <SEP>
<tb><SEP> in <SEP>% <SEP> in <SEP>% <SEP> in <SEP>% <SEP> in <SEP>% <SEP>
<tb> Capacitor <SEP> Bulb <SEP> - <SEP> 15 <SEP> - <SEP> 10 <SEP> -30 <SEP> -30
<tb> Bulb <SEP> Capacitor <SEP> + <SEP> 5 <SEP> 0 <SEP> ~ <SEP><SEP> +20 <SEP> ~ <SEP><SEP> +20
<Tb>
The largest differences are in reactive power Q and tan a.

 Thus, after a nth fault has been detected, a signal is recorded and / or transmitted, and then the reference values of the faulty section are modified to take account of a fault. During the next examination of the section, we then proceed as if this section no longer has any defect.

 The examination of a section is carried out through the comparison of Q or tan with a higher reference value VSUP and a lower reference value VINF.

The VSUP value can be determined from a reference value
VNORM to which is added a gap ESUP
VSUP = VNORM + ESUP
The value VINF can be determined from said reference value
VNORM to which we deduct an EINF gap
VINF = VNORM - EINF
VNORM is a reference value derived for example from Table B.

Advantageously, VNORM is the value that was determined during the previous review of the section.

 After detection of a defect, the VSUP and VINF values are modified so that the following examination takes into account the existence of this defect. The modifications can be made by giving VNORM the current value of Q or tan a, then recalculating VSUP and VINF from this new value of VNORM according to the equations given above.

 Thus, the detection of a fault causes the sending and / or recording of a signal, then the device is found, through the modification of VSUP and VINF, in a state analogous to normal operation. Although the defect already detected is still existing, it is no longer detected a second time during subsequent examinations of the section.

 The process can then be applied successively to several sections.

For each section, we then have a VNORM value from which the VINF and VSUP values of this section can be calculated.

 It is advantageous to apply the method according to the invention so that all the sections of all the departures are examined successively, and that all these examinations begin again cyclically.

 The differences ESUP and EINF can be obtained from Table C and advantageously, after multiplication by a safety coefficient.

 Preferably, ESUP and EINF are dependent on the number N (S) of radiation sources present in the section S examined.

 The number N (S) of radiation sources present in the section S examined can be determined using the value of the active power P consumed in said section S examined or by reading in a memory.

 The differences ESUP and EINF can also be a function of the nominal characteristics of the lamps in the considered section.

It is also desirable to correlate the differences ESUP and EINF with the voltage of the electrical distribution network, because the physical quantities can vary a lot according to this voltage as illustrated by the following table D which shows the evolution of the values of the quantities I, P, Q, tan a depending on the voltage
U of the network. Figure 3 shows the corresponding curve.

 For the establishment of Table D, 10 lamps were tested.

 During the measurements, it was found that a minimum voltage of about 190 volts was required for all lamps to remain on.

TABLE D

<tb> U <SEP> (Volt) <SEP> | <SEP> I <SEP> (Ampere) <SEP> P <SEP> (Watt) <SEP> Q <SEP> (VAR) <SEP> TAN <SEP> a
<tb><SEP> 190 <SEP> 5.02 <SEP> 916 <SEP> - <SEP> 41 <SEP> -41 <SEP> - <SEP> 0.04 <SEP>
<tb><SEP> 200 <SEP> 5.66 <SEP> 1092 <SEQ> 148 <SEP> 0.13
<tb><SEP> 210 <SEP> 6.36 <SEP> C <SEP><SEP> 1242 <SEP> 315 <SEP> 0.25
<tb><SEP> 220 <SEP> 7.00 <SE> 1392 <SE> 515 <SEP> 0.37
<tb><SEP> 230 <SEP> 7.60 <SEP> 1530 <SEP> 687 <SEP> 0.44
<tb><SEP> 240 <SEP> 8.19 <SEP> 1670 <SEP> 861 <SEP> 0.51
<tb><SEP> 250 <SEP> 8.76 <SEP> 1816 <SEP> 1070 <SEP> 0.58
<Tb>
The evolution of 1, P, Q and tan at these values as a function of U is approximately linear, as shown in Figure 3.

 In addition, measurements of voltage U or intensity I must be carried out only a certain time after ignition of the radiation sources because it is necessary to wait for the stabilization of the different physical quantities to avoid detecting faults. and transmit or record erroneous signals.

 The following tables E and F show the evolution of these quantities as a function of the time flowing from the moment when the ignition is controlled. Figure 4 shows the corresponding curve.

TABLE E

<tb><SEP> t <SEP> 1.5 <SEP> 2 <SEP> 2.5 <SEP> 3 <SEP> 3.5 <SEP> 4 <SEP> 4.5
<tb><SEP> (mn)
<tb><SEP> I <SEP> 9.87 <SEP> 9.00 <SEP> 7.7 <SEP> 7.05 <SEP> 7.03 <SEP> 7.00 <SEP> 7.04
<tb> (Ampere)
<tb><SEP> P <SEP> 651 <SEP> 905 <SE> 1260 <SE> 1365 <SE> 1379 <SE> 1380 <SE> 1391
<tb><SEP> (Watt)
<tb><SEP> Q <SEP> 2170 <SEQ> 1789 <SEQ> 1039 <SEQ> 577 <SEQ> 531 <SEQ> 518 <SEP> 519
<tb><SEP> (VAR)
<tb><SEP> Tana <SEP> 333 <SEP> 1.98 <SEP> ~ <SEP><SEP> 0.82 <SEP> 0.42 <SEP> 0.38 <SEP> 0.37 <SEP > 0.37
<Tb>
TABLE F

<tb><SEP> t <SEP> 5 <SEP> 5.5 <SEP> 6 <SEP> 6.5 <SEP> 7 <SEP> 7.5
<tb><SEP> (mn)
<tb><SEP> 7.01 <SEP> 7.05 <SEP> 7.05
<tb><SEP> I <SEP> 7.02 <SEP> 7.07 <SEP> 6.99
<tb> (Ampere)
<tb><SEP> P <SEP> 1391 <SE> 1392 <SE> 1382 <SE> 1382 <SE> 1390 <SE> 1396
<tb><SEP> (Watt)
<tb><SEP> Q <SEP> 519 <SEQ> 534 <SEQ> 534 <SEQ> 512 <SEQ> 513 <SEQ> 528
<tb><SEP> Tan <SEP> a <SEP> 0.37 <SEP> 0.38 <SEP> 0.38 <SEP> 0.37 <SEP> 0.37 <SEP> 0.37
<Tb>
It appears that after about 3 minutes and a half, we can estimate that the stability is reached.

 Similarly, when examining a section, it is advantageous to repeatedly carry out measurements of U and I until the measured values are at least approximately constant, so as to avoid the recording and / or transmission of erroneous messages.

 The experiments carried out by the applicant have shown that it is preferable to base the detection of defects on the surveillance of tan, because it is the quantities which present the least important fluctuations.

 According to an advantageous version of the method according to the invention, the signal recorded and / or sent further includes indications as to the nature of the detected defect. The process therefore makes a distinction between bulb, capacitor or fuse faults.

For this, the process can be based
- on the monitoring of the variation of the reactive power Q
- or on that of the variation of the tangent of a completed by the monitoring of either the active power P or the reactive power Q.

If the process is based on the monitoring of the variations of the reactive power Q in the section S considered, it is necessary to define two lower deviations EINF1 and EINF2 and, respectively, two lower reference values.
VINF1 and VINF2.

 Indeed, as it appears through the third and fourth rows of Table B, the decrease of Q may be more or less important depending on whether it is a defect bulb or fuse. The largest decrease (approximately greater than 40%) corresponding to a bulb defect and the smallest decrease (approximately less than 20%) to a fuse fault.

By defining: VINF1 = VNORM - EINF1,
VINF2 = VNORM - EINF2,
and VSUP = VNORM + ESUP,
- if in the section under consideration, Q is lower than VINFI, but greater than
VINF2, we will be in the presence of a fuse fault and
- if Q is lower than VINF2, we will be in the presence of a bulb defect.

 On the other hand, if Q increases and becomes greater than VSUP, it can only be a capacitor fault.

If the process is based on the monitoring of tan variations completed by that of either P or Q, defining
VSUP = VNORM + ESUP
VINF = VNORM - EINF,
if tan a becomes greater than VSUP, there is a capacitor fault,
- if tan a becomes lower than VINF, we are in the presence of a bulb defect
On the other hand, if tan a varies little and remains both greater than VINF and lower than VSUP, P or Q should be examined to see if there is not a fuse fault. According to the fourth row of Table B, if there is a fault in the fuse, P drops by about 10% and Q by about 15%.

 Thus, in the absence of variation or low variation of tan, a fuse fault is detected if P decreases by more than 8% and Q by more than about 10%. By setting a PINF value, compare P (or Q) to PINF to see if there is a fuse fault. Preferably, the monitoring of the active power P. is used in addition to the monitoring of tan a.

DEVICE ACCORDING TO THE INVENTION
In order to be able to implement its method, the applicant has also designed a device that is intended to be at least partially disposed in the box of a public lighting network. This device comprises at least:
means for measuring the voltage U and the intensity I in at least one
section S,
determining means for determining, from the values of U and I, the reactive power Q and possibly the active power P and the ratio
Q / P,
means for comparing said Q power or said Q / P ratio with reference values VINF and VSUP,
processing means able at least to compare said power Q or said Q / P ratio with reference values VINF and VSUP, and to modify these values,
memory means containing at least one program governing the operation of said processing means,
and means for transmitting a signal.

 Of course, by comparing Q, Q / P (tan a) P or Q with reference values, those skilled in the art will hear that it is in practice a comparison of the image voltages of these magnitudes with said reference values.

 The connection of the device according to the invention to the network is visible in Figure 5 which shows a public lighting network box in which was connected a portion of the device according to the invention.

In parallel on the network is connected a block 1 of sensing voltages. This block 1 is composed of three resistors 2 and a current sensor preferably of the Hall effect type. The output of block 1 thus supplies the voltage currents of the voltages
V1, V2 and V3 phases L1, L2 and L3 respectively. These currents V1, V2, and V3 are applied to block 3 which contains at least one current / voltage adapter.

 Block 4 is connected to surge arresters 5, pulsadis relay KA1, general circuit breaker QG, and general contactor KMG. Its role is to provide logical information to block 6 on the state of these components.

 Block 6 comprises at least one voltage / voltage interface.

 Block 7 is intended for capturing currents. As shown in FIG. 5, it comprises Hall effect type current sensors arranged in series with each of the sections of the feeders D1, D2, D3,... Dn managed by the device according to the invention. The outputs of said current sensors are connected to the block 3 in order to supply it with the values of the intensities I11, I12, I13,... Ik1 and the resultants Ia1, IA2 and IA3, ... IAn.

 I11 denotes the intensity of the current flowing through the section 1 of the start 1, I12 that of the current flowing through the section 2 of the start 1, I13 the intensity for the section 3 of the departure 1, ... Ikl denotes the image of the intensity of current flowing through section 1 of departure k.

 are IA2, 1A3, ... IAn are the resultants of the currents respectively traversing the departures 1, 2, 3 .... n.

 Block 3 comprises at least one current / voltage adapter.

 The currents coming from blocks 1 and 7 are transformed into image voltages by block 3 and then directed to block 8 which is a calculation module preferably comprising at least one multiplexer. Preferably, the image voltages relating to the block 1 are applied to a first multiplexer and the image voltages relating to the block 7 to a second multiplexer. Thus, said first multiplexer allows the selective distribution of an image voltage of the voltage of one or other of the phases of the network. Said second microprocessor allows the selective distribution of an image voltage of the current flowing through one or other of the sections of one or other of the feeders managed by the device according to the invention.

 The outputs of block 8 are connected to block 9 which is a microprocessor processing module coordinating the commands between the different blocks and managing the flow of data.

 The entire device according to the invention is represented in more detail by its block diagram represented by FIG.

 Inside the block 1, a voltage capture circuit board powered by 15 volts supplies the images of the voltages V1, V2 and V3 to an input 11 of the block 3, which includes a current-voltage conversion card also supplied with 15 volts. .

 In block 7, currents Ik1 of sections 1 of the feeders k are picked up by a current capture board equipped, as already indicated, with current sensors of the Hall effect type and supplied with 15 volts. The output "line currents" 14 of this plate is connected to the input 15 of said current conversion card voltages of the block 3. The two outputs 16 and 17 of this block 3 are connected to the block 8 powered at a time in 15 volts and in 5 volts.

 The output "differential currents" 19 of the current capture board of the block 7 is connected to an input 20 of the block 8.

 The function of block 8 is to determine or calculate the value of the active P and reactive powers Q and the Q / P ratio, that is to say the tangent a of the phase shift between the voltage and the current in the current section. 'examination. The determined analog values are applied via the output 21 of the block 8, to the block 9 which comprises an analog / digital conversion card 22 powered at both 5 volts and 15 volts, and a microprocessor card 23 supplied with 12 volts 5 volts.

The digital values provided by the card 22 are transmitted to the microprocessor card 23 via the input 24. Said microprocessor card 23 comprises processing means such as at least one microprocessor and memory means such as a memory erasable (for example a RAM,
EPROM or EEPROM) containing at least one software managing the operation of the microprocessor.

 Preferably, the voltages coming from the output of said first and second multiplexers of the block 8 are processed, adapted in a manner known to those skilled in the art, then applied to a third multiplexer advantageously also located in the block 8 and connected to the card 22 , and intended to provide, alternatively, the quantities U, P, Q and Q / P.

 Said multiplexers are preferably controlled by addressing by the microprocessor of the card 23.

 Thus, for each section examined, the multiplexer relating to block 1 provides an image voltage of the network voltage, while the multiplexer relating to block 7 provides an image voltage of the intensity traversing the section. These image voltages are processed, adapted, transformed in known manner, so as to provide the input of the third multiplexer of the block 8 the quantities U, P, Q and Q / P. The multiplexers relating to the blocks 1 and 7 therefore remain fixed on the same section for a time long enough for the third multiplexer of the block 8 to supply the microprocessor of the card 23, alternatively the values of U, P, Q and Q / P .

A second input 25 of said microprocessor card 23 is connected to the output of the voltage / voltage interface of the block 6 so as to allow the introduction of logic information relating to the state of the pulsadis KA1, the general circuit breaker
HQ, of the KMG main contactor in the microprocessor board 23.

 An AC voltage supply card 26 connected to one of the phases of the electrical distribution network transforms and rectifies the 240 volts AC into 18 DC volts to supply the three inputs 27, 28 and 29 of a power supply card. As an alternative, said inputs 28 and 29 can also be supplied with 12 volts from a battery card 31. The 5 volts, 12 volts and 15 volts outputs of the DC voltage supply board 30 supply the power supplies. components of the device and are connected to the inputs 32, 33 and 34 of a DC voltage threshold detection card 35 whose role is to provide the microprocessor card 23 through the input 36, a logical information relating to the exceeding of the thresholds of the different voltages.

 The microprocessor card preferably has a serial input / output port 37 for possible communication with an external computer 38 for a modification of the software and / or data contained in the memory or memories.

 An output 39 of the microprocessor card 23 is connected to means for transmitting a signal such as a modem 40 powered by 12 volts for transmission of various information such as a signal in the event of a fault, values voltages or currents, etc., at a control station 41 via a modem 42. The control station 41 comprises at least one computer 43.

 An output 44 of the microprocessor board 23 is connected to an output board 45 powered by 12 volts and including interfaces for control of the main contactor or circuit breakers.

ORGANIZATIONAL
FIG. 7 is an example of a flowchart of software that can correspond to software stored in a memory of the microprocessor card 23 and manages the operation of the microprocessor.

 The detection of a defect can be done by examining Q or tan a. Of course, the choice of one or the other of these magnitudes is done for the whole of the organization chart (and concerns all its stages).

 The beginning 1 can be determined by a predefined time, by the pulsadis signal, that is to say the signal sent by EDF, by an external brightness sensor, etc ...

In step 2, the microprocessor orders the measurement of the network voltage and the intensity traversing the examined section. In step 3, it determines the reactive power, and possibly P and Q / P (tan a), if the fault detection is based on Q / P monitoring. Then in step 4, he compares Q or tan a to VSUP. If Q or tan a is less than or equal to VSUP, he goes to step 5 and compares Q or a to
VINF. If Q or tan a is greater than or equal to VINF, this means that no defect has been detected. The microprocessor then returns to step 2 and resumes its examination.

If the comparison of step 4 shows that Q or tan a is greater than VSUP, this means that the section under examination has a defect. Then the microprocessor goes to step 6 where it orders the sending and / or recording of a signal, then it modifies VSUP and VINF to take into account the existence of a defect in the section. This modification can be done by recalculating new values for VSUP and VINF according to the equations already mentioned, and taking for
VNORM the value of Q or tan determined in step 3
- VNORM = Q or tan a
- VSUP = VNORM + ESUP
- and VINF = VNORM - EINF.

 Then, the microprocessor returns to step 2.

 If the comparison of step 4 showed that Q or tan a is less than or equal to VSUP and if step 5 reveals that Q or tan a is also lower than VINF, it also means that the section being examined has a defect. The microprocessor then moves, as previously to steps 6 and 7, then to step 2, that is to say at the beginning of the cycle. The microprocessor continuously carries out these cycles, until the power off of the network, to a predefined time, or until receipt of an order or stop signal sent, for example, by a brightness sensor.

 FIG. 8 represents an exemplary flowchart of software that can govern the operation of the microprocessor according to an improved version of the device. The microprocessor is, according to this version, able to determine the nature of the detected fault and to send a particular message for each type of fault. The determination of the nature of the fault is based on this flowchart on the monitoring of the variations of the reactive power Q. It is therefore not necessary to determine the active power P or the phase shift (or tan a).

 The beginning 1 corresponds for example to the powering up of the network (therefore departures, therefore sections and sources of radiation). The microprocessor then waits, in step 2, for example by making known a number of times a loop, that at least one of the values P, I, Q or tan has stabilized, then passes to the Step 3. This step 3 marks the beginning of loops of the type called FOR / NEXT loop in the FORTRAN programming language. This means that the following steps will be repeated for each of the departures D on which the device according to the invention is connected, inside the box, and that it manages.

 Step 4 is similar to step 3, but this time the loops are 3 in number or the number of S sections belonging to the departure D examined. The following steps are therefore generally performed for each section S of each departure D (unless it is expected that certain sections or departures are not examined).

In step 5, the microprocessor examines the values of U and I in the considered section, then in step 6, it reads in a memory the reference value
VNORM (S) of section S. This value VNORM (S) is the value of the reactive power Q corresponding to the previous operation. The microprocessor also preferably reads the number N (S) of radiation sources present in the section S considered and one or more nominal characteristics, such as their nominal power PNOM (S).

 Instead of reading in a memory the number N (S), it is possible to envisage that the microprocessor calculates N (S) from a measurement of the active power P, by division by the nominal power PNOM (S) radiation sources.

 Then the microprocessor calculates the upper deviation ESUP (S) of the section S. Preferably, ESUP (S) is a function of N (S), U and PNOM (S). n also calculates the lower deviations EINF1 (S) and EINF2 (S) which are also preferably a function of N (S), U and PNOM (S).

 Then, it calculates the upper reference value VSUP (S) and the lower reference values VINF1 (S) and VINF2 (S) of section S.

 In step 7, the microprocessor determines Q.

 In step 8, he compares Q to VSUP (S). If Q is greater than VSUP (S) it means that since the previous examination of section S, a capacitor has become defective. The microprocessor then orders, according to step 9, the recording in a memory of a message and / or the transmission of this message to a central station (PC) via, or a modem connected to a line telephone or a special line, either of a transmitter / receiver for example of the radio type.

 Preferably the message comprises, in addition to the type of defect, the numbers of the departure D and of the section S concerned and / or information making it possible to locate the section S.

 Then in step 10 the microprocessor modifies the reference value VNORM (S) of the section S, and records its new value, so that the capacitor fault is not detected again at the next examination. Then he returns to Step 4 to proceed to the next S section of the same D start or to proceed to the S sections of the following start if he has completed the examination of all sections of the start. D. Once all sections of all departures are examined, the microprocessor resumes its exams and so on, until a pulsadis signal is sent, cutting off the power supply to the network or receiving an order or signal. 'stop.

If the comparison of step 8 showed that Q is less than or equal to
VSUP (S), the microprocessor compares, according to step 11, Q to VINF1 (S). If Q is not less than VINF1 (S), it means that no new defects have appeared in section S since the previous examination. The microprocessor then returns to step 4 to examine the next section.

If the comparison of step 11 reveals that Q is greater than or equal to
VINF1 (S), the microprocessor proceeds to step 12 where it compares Q to VINF2 (S). If Q is less than VINF2 (S), this means since the previous examination of section S, a bulb has gone out. The microprocessor then orders, according to step 13, the sending and / or recording of the corresponding message, then in step 14 modifies the reference value VNORM (S) of section S, and records its new value, so that the defect of the bulb is not detected again at the next examination. It can also decrement the number N (S) of radiation sources in the state of Mache in section S and record the new value of N (S). Then he goes back to step 4.

If the comparison of step 12 shows that Q is greater than or equal to
VINF2 (S), this means that a fuse has blown. The microprocessor then orders, according to step 15, the sending and / or recording of the corresponding message, then goes through step 14 and returns to step 4.

 Fig. 9 shows an exemplary flow chart of software that can govern the operation of the microprocessor according to another improved version of the device. According to this version, the microprocessor is also able to determine the nature of the detected fault and to send and / or record a particular message for each type of defect, but in this case, the determination of the nature of the defect is based on the monitoring of the defects. variations of tan completed by that of P.

 Steps 1 to 5 are similar to those of the flowchart shown in Figure 8.

In step 6, the microprocessor reads from a memory the reference value
VNORM (S) of section S and the reference value P (S) of the active power in section S. VNORM (S) and P (S) are the reference values, respectively, of tan a and the power active P corresponding to the previous operation.

 The microprocessor also preferably reads the number N (S) of radiation sources present in the section S considered and one or more nominal characteristics, such as their nominal power PNOM (S).

 Then the microprocessor calculates the upper deviation ESUP (S) of the section S. Preferably, ESUP (S) is a function of N (S), U and PNOM (S). It also calculates the lower deviation EINF (S) and the power deviation EP (S) which are also preferably a function of N (S), U and PNOM (S).

Then, it calculates the upper reference value VSUP (S), the lower reference value VINF (S) of the section S, and the lower value of the power
PINF (S).

 In step 7, the microprocessor determines tan a and P.

 In step 8, he compares tan a to VSUP (S). If tan a is greater than VSUP (S) it means that since the previous examination of section S, a capacitor has become defective. 11 then orders, according to step 9, the recording in a memory of a message and / or the transmission of this message.

 Then in step 10 the microprocessor modifies the reference value VNORM (S) of the section S, and records its new value, so that the capacitor fault is not detected again at the next examination. Then he returns to Step 4 to proceed to the next section S review.

If the comparison of step 8 showed that tan a is less than or equal to
VSUP (S), the microprocessor compares, according to step 11, tan a to VINF (S). If tan a is not less than VINF (S), the microprocessor proceeds to step 14 where it compares
P to PINF (S). If P is not less than PINF (S) this means that no new defects have appeared in section S since the previous examination. The microprocessor then retires at step 4 to examine the next section.

If the comparison of step 11 reveals that tan a is greater than or equal to
VINF (S), this means since the previous examination of section S, a bulb has gone out. The microprocessor then orders, according to step 12, the sending and / or recording of the corresponding message, then in step 13 modifies the reference values VNORM (S) and P (S) of section S, and saves their new values so that the bulb's defect is not detected again at the next test. It can also decrement the number N (S) of radiation sources in the state of Mache in section S and record the new value of N (S). Then he goes back to step 4.

If the comparison of step 14 showed that tan a is greater than or equal to
PINF (S), this means that a fuse has blown. The microprocessor then orders, according to step 15, the sending and / or recording of the corresponding message, then goes through step 13 and returns to step 4.

 It should be noted that the probability of two defects occurring simultaneously in the same section is negligible and is therefore not taken into account in the flowcharts of Figures 7, 8 and 9.

 The intensity I and voltage U measurements are preferably performed several times in succession, for example three times in succession, so as to avoid the sending and / or recording of erroneous messages.

 The entire device according to the invention may advantageously be grouped together in a shaped enclosure or a housing, preferably portable, and from which preferentially only the elements (wires ...) necessary for its connection and, where appropriate, to his power supply.

 When implementing the device according to the invention for the first time, the reference values can be determined experimentally, or from tables B and C above.

 Where the radiation sources do not all have the same nominal characteristics, it is preferable to provide for the differences between EINF and / or and / or ESUP and / or EINF1 and / or EINF2 to take this into account.

 It is possible to provide that the microprocessor performs the provisional recording of the messages and periodically sends, for example every two hours, in a grouped manner, the results of the examinations it has carried out on all the sections of all the departures it manages.

 According to another advantageous version of the device according to the invention, the software is designed to detect and also signal the existence in the section considered of defects of logic type, that is to say of unquantified defects but of which only the absence or existence is decisive. The existence of a logical defect then results, for example, in the value 1 and its absence by the value 0. These logical type defects can be for example the closure (0) or the opening (1) of the door of the cabinet, operation (0) or non-operation (1) of the circuit breaker, etc ...

 It is also possible to consider assigning a priority degree to each defect and to provide that the microprocessor records only the low priority faults and sends a signal only in the event of a high priority fault.

 The device according to the invention can provide that information such as the voltage of the network, the currents flowing through the sections, the powers relating to each of the sections, etc. are periodically or constantly transmitted to the central station or stored in a memory. view of the compilation of statistics.

 The device according to the invention may comprise display means such as a screen or display, a keyboard such as an alpha-numeric keypad, additional memories, additional software enabling an agent to search and display information relating to all or part of the departures and sections managed by the device according to the invention.

 The device according to the invention can also be provided with means enabling the connection of an external computer for searching and displaying the information contained in the memory.

 The method and the device according to the invention can also be used to individually manage the defects of a single radiation source. The device is then preferably installed in the cabinet of the radiation source concerned, the parameters are modified accordingly (number of starts = 1, number of sections in the start = 1, number of radiation sources in section N (S) = 1, etc ...).

 The software whose flow charts are represented by FIGS. 7, 8 or 9 can then be adapted, for example, so that the sending of a signal is done by superimposition (as described in particular in French Patent No. 2,695,286). on the carrier that constitutes the phase of the examined section, with, in addition, information or a code corresponding to the number and / or location of the radiation source concerned.

 Obviously, the invention is in no way limited by the features that have been specified in the foregoing or by the details of the embodiments or examples chosen to illustrate it. Many modifications can be made to what has just been described by way of illustration without departing from the scope of the invention. The latter therefore encompasses all the means constituting technical equivalents of the means described as well as their combination.

Claims (14)

1. Method for detecting faults in a public lighting network consisting of a  electrical distribution network, on which is connected at least one departure D  comprising at least one section S feeding at least one source of  N radiation, characterized in that it comprises at least the following steps:  a) the values of the voltage U and the intensity I are measured in the section S,  b) we determine from the values of the voltage U and the intensity I ::  the reactive power Q,  - and possibly the active power P and the ratio Q / P,  c) comparing said reactive power Q or said Q / P ratio with values  VINF and VSUP,  d) if said reactive power Q or said Q / P ratio is, at the same time, lower than  or equal to VSUP and greater than or equal to VINF, we return to step a)  e) if not, send and / or record a signal indicating  the existence of a defect, then modifying said VSUP reference values  and VINF, so that said fault is not detected again when  a subsequent review of section S, and return to step a). 2. Method according to claim 1, characterized in that VINF and / or VSUP  is / are a function of the number of radiation sources present in the section  examined. 3. Method according to claim 1 or claim 2, characterized in that  VINF and / or VSUP and / are a function of the mains voltage. 4. Method according to claim 1 or claim 2, characterized in that  VINF and / or VSUP and / are function of the nominal power of the lamps of the  section examined. 5. Method according to any one of claims 1 to 4, characterized in that  includes at least the following steps:  a) the values of the voltage U and the intensity I are measured in the section S  b) from the values of the voltage U and the intensity I, the  reactive power Q,  c) comparing said reactive power Q with VINF reference values  and VSUP,  d) if said reactive power Q is both less than or equal to VSUP and  greater than or equal to VINF, we return to step a)  e) if not, send and / or record a signal indicating  the existence of a defect, then modifying said VSUP reference values  and VINF, so that said fault is not detected again when  a subsequent review of section S, and return to step a). 6. Method according to any one of claims 1 to 4, characterized in that  includes at least the following steps  a) the values of the voltage U and the intensity I are measured in the section S,  b) the values of the voltage U and the intensity I are determined from:  the reactive power Q, the active power P and the Q / P ratio,  c) comparing said Q / P ratio with reference values VINF and VSUP,  d) if said Q / P ratio is both less than or equal to VSUP and greater than or equal to  equal to VINF, we go back to step a)  e) if not, send and / or record a signal indicating  the existence of a defect, then modifying said VSUP reference values  and VINF, so that said fault is not detected again when  a subsequent review of section S, and return to step a). 7. Device for detecting faults in a public lighting network  consisting of an electrical distribution network, to which at least one  starting point D comprising at least one section S feeding at least one source of  N radiation, characterized in that it comprises at least the elements  following ::  means for measuring the voltage U and the intensity I in at least said  section S,  determination means, for determining, from the values of U and  of I, the reactive power Q and possibly the active power P and the  Q / P ratio,  processing means capable of at least comparing said power Q or  said Q / P ratio with reference values VINF and VSUP, and to be modified  these values,  memory means containing at least one program governing the  operation of said processing means,  and means for transmitting a signal. 8. Device according to claim 7, characterized in that said means of  are able to give VINF and / or VSUP values which are  function of the number N (S) of radiation sources present in section S  examined. 9. Device according to claim 7 or claim 8, characterized in that  said processing means are adapted to give VINF and / or VSUP  values that are a function of the network voltage. 10. Device according to one of claims 7 to 9, characterized in that said means  are able to give VINF and / or VSUP values which are  function of the nominal power of the lamps of the examined section. 11. Device according to one of claims 7 to 10, characterized in that  comprises at least one multiplexer. 12. Device according to one of claims 7 to 11, characterized in that said  means of determination comprise at least three multiplexers, the outputs  first and second multiplexers being connected to the input of the third  multiplexer. 13. Device according to one of claims 7 to 12, characterized in that it is  programmed so as to be able to apply the method according to claims 1 to  6, successively to several sections. 14. Implementation of the process according to claims 1 to 6, characterized in that  that said method is successively applied to several sections.