FR2685779A1 - Fault detector for a multiphase overhead electricity distribution network - Google Patents

Abstract

A fault detector comprising a set of independent sensors (9) each associated with one of the phases (I, II, III) of a line (4) of the network, and a receiving system (10), which are connected together by radio channels (A, B, C). Each sensor (9), mounted on one phase conductor (3, 8) and supplied by the latter, measures the amplitude (intensity) of the corresponding phase current and converts it into sampled digital data, transmitted instantaneously via the associated radio channel. The receiver (10) resynchronises the collected digital data, by an interpolation calculation, in such a way as to compensate for the phase shift related to the asynchronous character of the data transmission over the various channels (A, B, C). The homopolar current is thus reconstituted accurately in real time by summing the various phase currents, and this current is monitored in order to detect homopolar faults.

Description

DETECTOR OF DEFECTS FOR
AERIAL POLYPHASE ELECTRICAL DISTRIBUTION NETWORK
The present invention relates to a defect detector affecting a multiphase overhead electrical distribution network, more particularly a three-phase "high voltage" overhead network, that is to say in particular with voltages in the range of 10 to 40 KV, with a maximum current of 400 A. This detector measures the currents of fault currents, for the purpose
- to make a fault passage detection, with monitoring of homopolar and / or polyphase currents
to send a significant measurement of the current state of the network at a given moment
- To make a fault detection of wattmetric type, by association with a measuring element of the voltage.

 Currently, two general principles are used for the detection of defects on networks of the kind considered.

 The first existing principle is the use of a magnetic field detector, located under the distribution line. The detector thus ensures the detection of homopolar currents and its advantages are its ease of implementation and its low cost. However, it has the drawbacks of being imprecise, of being unreliable because of its sensitivity to parasitic magnetic fields, and of allowing only the measurement of the homopolar current, excluding phase currents.

 The second known principle consists, for a three-phase network, in a detector with three magnetic measuring cores, integrated in a switchgear. The advantages of such a detector are its accuracy and reliability. On the other hand, it retains the disadvantages of being able to be mounted only in the factory, and to have a high cost due in particular to the insulation of the cores.

 Furthermore, European Patent No. 0 125 796 discloses a system for monitoring a "high voltage" air transport network, with a set of sensors arranged on the conductors of a line and capable of providing a detailed report. network (phase current, phase voltage, temperature, etc ...). The quantities measured by the sensors are integrated over several sampling periods and are transmitted by radio waves, in the form of Fourier transforms, to a receiver placed for example at the base of a tower.

The data thus transmitted can not be used in real time in the context of a fault monitoring and only concern the characteristics of the phase currents, the measurement of the homopolar current being excluded here.

In addition, the installation of the sensors on the conductors from the ground, imposes a very sophisticated mechanical structure in the case of the aforementioned document.

 The aim of the invention is to eliminate all the disadvantages mentioned above, by providing a fault detector which, while applying a radio transmission between sensors and a receiver, allows a complete measurement in real time, encompassing the three currents of phase and the homopolar current, so that the measurement can be exploited for the location of defects affecting a network line, polyphase and / or homopolar, the proposed device is accurate, reliable, low cost and easy installation.

To this end, the subject of the invention is a defect detector for a multi-phase aerial overhead network, which essentially comprises, in combination
a set of sensors independent of each other, each associated with one of the phases of a line of the overhead network, each sensor being mounted on a phase conductor of the line, being electrically powered by the current of the line, measuring the intensity of the corresponding phase current, converting the analog measured value of this intensity into digital data and instantaneously transmitting this digital data via a radio channel, the data transmissions on the different radio channels associated with the network phases being asynchronous, and
a receiver assembly collecting the digital data transmitted by the different radio channels associated with the sensors, the receiver unit exploiting these digital data by calculating by calculation the homopolar current from the sum of the different phase currents, with compensation for the phase shift related to the asynchronous character of the data transmissions on the different radio channels.

 Thus, the set of sensors transmits in real time sampled values of the three phase currents (considering the usual case of a three-phase network), and these values are exploited almost instantaneously in the receiver which comprises means for rapid calculation. . As will be explained below, these calculation means have the particularity of performing a numerical interpolation resynchronizing the data which are derived from the three independent sensors and which thus arrive at the receiver asynchronously; the phase shift associated with the asynchronous nature of the transmission of the three data is thus eliminated, and the summation of the resynchronized data provides the precise instantaneous value of the zero sequence current indicative of a fault if this value is not zero.

 According to one embodiment of the invention, each sensor comprises a magnetic core core and winding for its power supply, a magnetic core toroid and winding for measuring the phase current, an electronic power supply circuit in relation with the first toroid, an electronic circuit for converting the analog signal, supplied by the second toroid, into digital data of the "series" type, an electrical circuit constituting a radio transmitter, and an envelope containing all of the preceding components and intended to be mounted on the phase conductor concerned so that the conductor passes through the two aforementioned tori. According to an advantageous arrangement, each sensor produced as specified above is integral with a connection ramp between an overhead line and an apparatus such as an air switch; the integration of the sensor with such a strap makes it easy to mount, from the ground and under tension, by exchange with a standard connecting strap. According to a complementary feature, contributing to the ease of installation, each sensor is "potential", that is to say galvanically linked to the phase conductor on which it is mounted.

 The receiver assembly comprises at least one radio reception antenna, a plurality of demodulators tuned to the radio channels respectively associated with the different sensors, a plurality of parallel serial / digital digital conversion stages, a central computing unit taking into account the data received from the various sensors, in particular by digital interpolation for their resynchronization and by summation providing an instantaneous value of the homopolar current, and a box containing the previously named components.

 More particularly, the processing of the phase current measurements comprises, on the side of the set of sensors, the sampling and the radio transmission of the digital data at a predefined sampling frequency, the sampling frequencies being different from one sensor to another with relatively small deviations and, on the side of the receiver assembly, a linear interpolation taking as a temporal reference the times of reception of the sampled data which correspond to a particular phase of the network and realizing, from the sampled data received at other times and corresponding to the other phases, a calculation of fictitious samples returned to the moments of reception of the reference phase. Thanks to such a resynchronization process, the phase error (due to the random differences between the reception instants for the different phases) is rendered negligible and the calculation of the amplitude of the homopolar current can be precise and significant, even if it is a residual homopolar current of very low intensity in front of the phase currents, the amplitude of the homopolar current being for example less than 10 A for phase currents of 400 A. Of course, the unit of The receiver calculation can simultaneously perform other functions, such as fault detection by comparing the calculated instantaneous value of the zero sequence current and / or instantaneous values collected from each phase current with threshold values.

In any case, the invention will be better understood with the aid of the description which follows, with reference to the appended schematic drawing showing, by way of nonlimiting example, one embodiment of this defect detector for polyphase overhead network. electrical distribution
FIG. 1 shows a fault detector according to the present invention, for a three-phase network, associated with a remotely controlled air switch, in a general perspective representation.
Figure 2 is a general block diagram of the defect detector object of the invention;
Figure 3 is a sectional view of a sensor belonging to this fault detector
Figures 4 and 5 are diagrams showing, generally and in detail respectively, the sampled data from the three sensors and reaching the receiver of the device
Figure 6 is a diagram illustrating the problem of the relative "slip" between the samples from the reference phase and another phase.

 Figure 1 shows a pole 1 which supports, via insulators 2, the conductors 3 of an overhead line 4 belonging to a three-phase power distribution network. In known manner, the pole 1 supports, at its apex, a remote control switch 5 controlled remotely, this switch 5 being actuated from a control / control box 6 carried by the lower part of the pole 1. Phase conductors 3 are connected to the air switch 5 by connecting straps 7 and 8, on either side of this switch 5.

 The three connecting straps 7 located on one side of the air switch 5 are standard shoulder straps, while the three connecting straps 8 located on the other side of this switch 5 each carry a current sensor 9 belonging to to the defect detector object of the invention. The fault detector further comprises a receiver 10, with a radio antenna 11 and a box 12, carried by the pole 1. A radio wave link is made between each of the three sensors 9 and the antenna 11 of the receiver 10, the three radio transmission channels are indicated respectively at A, B and C. The three sensors 9 are supplied with electrical energy from the corresponding conductors 3 of the overhead line 4, while the receiver 10 can be electrically powered, via a connecting cable 13, from the control / command box 6.

 The three sensors 9 and the receiver 10, connected by the radio channels A, B and C, are shown in the form of a block diagram in FIG. 2, which thus indicates their internal structures in connection with their functions.

The structures of the three sensors 9 being identical, we will describe a single sensor in the following.

 Each sensor 9 comprises a first core 14 or magnetic core feed toroid, with a winding 15, core which is crossed by the associated phase conductor 3. The winding 15 of the core 14 is connected to an electronic supply circuit 16, which delivers the power supplies necessary for the operation of all the components of the sensor 9, the operating range of the supply circuit 16 being situated for example between 5 A (minimum line current) and 400 A (maximum line current).

The sensor 9 comprises a second toroid 17 with a magnetic core, with a coil 18, toroid which is also traversed by the phase conductor 3, but is intended for measuring the intensity of the current in this conductor 3, the intensity being designated I1, I2 or I3 depending on the phase concerned. The winding 18 of the measuring core 17 is connected to an electronic circuit 19 of analog / digital conversion, which converts the analog signal proportional to the intensity provided by the torus 17, into a digital signal type "series". The sensor 9 further comprises a modulator 20 and an electronic circuit constituting a radio transmitter 21, emitting on the channel A,
B or C as appropriate.

 FIG. 3 shows a practical embodiment of the sensor 9. This figure again shows the supply toroid 14, the measuring toroid 17, the supply circuit 16, the analog / digital conversion circuit 19, and the circuit 21 constituting radio transmitter, which is assumed here to encompass the modulator 20. All these components are housed inside a common envelope 22, which is traversed axially by the phase conductor 3, and which is immobilized on this conductor 3. It should be noted that, by such an assembly, the sensor 9 is galvanically connected to the phase conductor 3.

 In use, the measuring toroid 17 permanently delivers an analog signal representative of the intensity I1, I2 or I3 in the phase conductor 3, which signal is used by the analog / digital conversion circuit 19. The conversion is carried out continuously at a predefined sampling frequency of, for example, between 1 KHz and 3 KHz. The converter circuit 19 thus provides, periodically and according to this frequency, "serial" data in the form of a pulse train comprising, for example, chronologically a start pulse, followed by ten data pulses corresponding to the instantaneous measured value of current, and finally four pulses to test the proper functioning of the link, each pulse being a logic level 1 or o.

 The logic signal thus formed drives the circuit 21, constituting a radio transmitter, via the modulator 20 whose center frequency represents one of the three channels A, B and C used. The high frequency signal is then amplified to achieve sufficient radiated power, less than 100 FW.

 Referring again to FIG. 2, the receiver 10 comprises a set of three demodulators 23, connected to the antenna 11 and tuned to the respective channels A, B and C. The output stages of the demodulators 23, delivering data "Series" digital corresponding respectively to those transmitted by the channels A, B and C, are connected to a set of three stages 24 of digital serial / digital parallel conversion. The outputs of the conversion stages 24 are connected to inputs of a central processing unit 25, based for example on an eight-bit microcontroller.

All the components of the receiver 10, previously defined, is housed inside the cabinet 12 already mentioned.

 In the central computing unit 25, there are several functional blocks, namely: a block 26 for calculating interpolation and reconstruction of the zero sequence current, a block 27 for monitoring the zero sequence current, and a block 28 for monitoring the current polyphase. Blocks 27 and 28 have respective outputs 29 and 30 for the indication of zero sequence faults and phase faults.

 The three radio transmitters 21, belonging respectively to the three sensors 9, are perfectly independent of each other, so that the data transmitted on the channels A, B and C arrive at the receiver 10 asynchronously, the times of reception of the relative data. at a phase that does not correspond to the instants of reception of the data relating to the other two phases, but having random deviations. The principle of computation of the homopolar current consists of adding the samples E1, E2 and E3 coming respectively from the three phases I, II and III in a "window" F of width equal to a sampling period, as illustrated in FIG. but the simple addition of three digital current amplitude values, taken at actually different times, can not express the precise instantaneous value of the homopolar current and has, on the contrary, an error. The correction of this error, due to the phase shift of the instants of reception of the data coming from the three sensors 9, is carried out by the block 26 of the central unit 25 of the receiver 10, as described hereinafter with reference to FIGS. 6.

 The principle of this correction is to use a phase, such as phase I, as a time reference, and to resynchronize the data from the other phases, and therefore phases II and III, with the data of the reference phase.

 The sampling frequency f1 of phase I is predefined and fixed for example at: fl = 2.00 KHz. I1 corresponds to a sampling period T1 = 1 / fl.

The sampling frequencies f2 and f3 of phases II and III are chosen to be different from the sampling frequency f1 of phase I, with differences e2 and e3 which are not defined but relatively small, but which are sufficient to avoid any overlap between the sampling ranges. tolerance of the three frequencies. In this example, we have
f2 = 2 KHZ + e2 T2 = 1 / f2
f3 = 2 KHZ + e3 T3 = 1 / f3 the differences e2 and e3 can be of the order of 0.20 KHz.

 Figure 5 shows three consecutive sampling windows, with a time scale t greatly dilated compared to that of Figure 4.

The three windows are designated F (n - 1), F (n) and
F (n + 1), and each window is referenced to a sample El (n - 1), El (n) and El (n + 1) of the current of phase I. Considering more particularly the central window F (n) , the goal is clearly illustrated it is to define, for phases II and III, two respective samples X2 (n) and X3 (n) fictitious, located on the same time line as the sample El (n). For the moment, it is assumed that there are no "collisions", that is to say that the samples E2 and E3 of phases II and III do not coincide with the time line considered; it is also admitted that these samples E2 and E3 are present one and the other inside each of the windows
F (n - 1) and F (n), the corresponding measured values being designated respectively: E2 (n - 1), E2 (n) and
E3 (n - 1), E3 (n).

The interpolation is of linear type, and is carried out according to the following operating mode
Upon acquisition of the sample El (n), the computing unit 25 initiates a measurement of the times t2 (n) and t3 (n) separating, respectively, the sample El (n) of the samples E2 (n). and E3 (n).

At the acquisition of the sample E2 (n), the calculation unit 25 which has kept the previous sample E2 (n - 1) of the phase II performs the following interpolation calculation
X2 (n) = E2 (n - 1) x t2 (n) + E2 (n) x (T1 - t2 (n)) T1
and thus determines the hypothetical sample X2 (n) synchronous with the sample El (n).

 In a similar way, at the acquisition of the sample E3 (n), the calculation unit calculates the imaginary sample X3 (n), given by a formula comparable to the previous one.

At the end of these two calculations, the calculation unit makes the sum
Xh (n) = El (n) + X2 (n) + X3 (n)
whose result Xh (n) is the fictitious sample corresponding to the instantaneous value of the homopolar current Ih.

The E1 samples being considered as time reference, the appearance of the samples E2 and
E3 is random with respect to this reference as already mentioned. Moreover, if it is assumed that the difference e2 of the sampling frequencies fl and f2 of the phases I and II, as well as the difference e3 of the sampling frequencies f1 and f3, are not zero, it is will follow a relative "sliding" effect between samples E1 and E2 as well as E1 and E3. This sliding, positive or negative according to the sign of the difference, will provoke at regular intervals a quasi simultaneity in the appearance of the samples El, E2 or El, E3. This is illustrated in Figure 6, the upper half of which corresponds to a negative slip, and the lower half to a positive slip.

 If the slip is positive (f2, f3 <fl), the samples E2 and E3 appear with a lower recurrence than the samples E1, and it may happen that no sample E2 or E3 appears in a window F synchronized with the samples. El samples.

 Conversely, if the slip is negative (f2, f3> fl), the samples E2 and E3 appear with a stronger recurrence than the samples El, and it can happen that two samples E2 and E3 appear in a synchronized window F with the sample El.

 Given the necessary calculation software and the requirements for speed of calculation, it does not seem appropriate to deal with the two previous situations, even if "software" solutions exist for each case, and it seems preferable to use the negative slip, which provides a surplus of samples. Consequently, the sampling frequencies f 2 and f 3 of phases II and III are voluntarily chosen to be greater than the frequency f 1 of phase I used as a reference, by construction of the sensors 9, for example with a relative difference of the order of 10% as already shown in the numerical examples given above.

 On the receiver side, given this choice, it becomes certain that at least one sample E2 or E3 appears in each sampling window F, which corresponds to the hypothesis made previously. However, "collisions" may occur, ie a sample E2 or E3 may occur in time coincidence with a sample E1 (in fact, within a small predefined time interval, centered on the moment of reception of the sample El used as reference). In this case, the calculation unit 25 differs the interpolation calculation until the appearance of the next sample E2 or E3, which will necessarily be located towards the other end of the window F, within it. Thus, one will take into account a time interval t2 or t3 correctly measurable and exploitable. By cons, the sample E2 or E3 coming into "collision" with the sample El is well memorized, and it is involved in the interpolation calculation as the previous sample.

 In other words, since there appears on average more samples E2 and E3 than samples El, the adjustment is made using only once the samples E2 and E3 coming in "collision", whereas in normal situation each sample E2 or E3 is used twice, successively, namely as E (n - 1) then E (n) by partially taking up the previous notations.

 In view of the foregoing, the calculating unit 25 reconstitutes in all cases, almost instantaneously, the homopolar current Ih, this in a sufficiently precise manner to be effectively exploited. In particular, the block 27 makes it possible to monitor the homopolar current Ih by comparison with a threshold value, and to issue, if necessary, on the output 29 a signal indicating a homopolar fault.

 Simultaneously, the sampled data from the three phases I, II and III and representing the phase currents are processed by the polyphase current monitoring block 28, which also compares this data with thresholds, and if necessary causes, on the output 30, the emission of a signal indicating a phase fault.

Of course, these signals can be transmitted remotely by any means, and cause any desired action for the protection of the air network concerned.

 It goes without saying that the invention is not limited to the sole embodiment of this defect detector for a multiphase overhead electrical distribution network which has been described above, for example; it embraces, on the contrary, all variants of implementation and application respecting the same principle.

It is thus, in particular, that one would not depart from the scope of the invention by purely constructive modifications, concerning, for example, the internal structure of the sensors in which the electronic circuits can be grouped in a different manner from that illustrated in FIG. 3. It is also conceivable to use the same defect detector on a post that does not support switchgear, and more generally to implement the invention at any point of an overhead network, not necessarily in association with a pole.

Claims (8)

 A fault detector for a multiphase overhead electrical distribution network, characterized in that it essentially comprises, in combination  a set of sensors (9) independent of each other, each associated with one of the phases (I, II, III) of a line (4) of the overhead network, each sensor (9) being mounted on a phase conductor (3,8) of the line (4), being electrically powered by the line current, measuring the intensity of the phase current (Tt, I2, corresponding, converting the analog measured value of this intensity into digital data (El, E2, E3) and instantaneously transmitting this digital data via a radio channel (A, B, C), the data transmissions on the different radio channels (A, B, C) associated with the phases (I, II, III) of the network being asynchronous, and  a receiver unit (10) collecting the digital data (E1, E2, E3) transmitted by the different radio channels (A, B, C) associated with the sensors (9), the receiver unit (10) using these digital data ( E1, E2, E3) by computationally reconstructing the homopolar current (Ih) from the sum of the different phase currents, with compensation of the phase shift due to the asynchronous nature of the data transmissions on the different radio channels (A, B, C) ).  2. Fault detector according to claim 1, characterized in that each sensor (9) comprises a core (14) magnetic core and winding (15) for its power supply, a core (17) magnetic core and winding ( 18) for the measurement of the phase current (I1, I2, I3), an electronic supply circuit (16) in relation to the first core (14), an electronic circuit (19) for converting the analog signal, provided by the second torus (17), in digital data of the "series" type (El, E2, E3), an electronic circuit (20, 21) constituting a radio transmitter, and an envelope (22) containing all of the preceding components ( 14 to 21) and intended to be mounted on the phase conductor (3, 8) concerned so that the conductor passes through the two aforementioned toroids (14, 17).  3. Fault detector according to claim 1, characterized in that each sensor (9) is integral with a connecting strap (8) between an overhead line (4) and an apparatus such as air switch (5).  4. Fault detector according to claim 2 or 3, characterized in that each sensor (9) is "at potential", that is to say galvanically connected to the phase conductor (3, 8) on which it is located. mounted.  Fault detector according to one of Claims 2 to 4, characterized in that the receiver unit (10) comprises at least one radio reception antenna (11) and a plurality of channel-tuned demodulators (23). radio (A, B, C) respectively associated with the different sensors (9), a plurality of stages (24) of serial / digital parallel digital conversion, a central computing unit (25) taking into account the data (E1, E2 , E3) received from the different sensors (9), in particular by digital interpolation for their resynchronization and by summation providing an instantaneous value of the homopolar current (Ih), and a box (12) containing the components (23 to 25) previously appointed.  Defect detector according to claim 5, characterized in that the measurement processing comprises, on the sensor assembly (9) side, the sampling and the radio transmission of the digital data (E1, E2, E3). at a predefined sampling frequency, the sampling frequencies (f1, f2, f3) being different from one sensor to another with relatively small deviations (e1, e2) and on the receiver side (10 ), a linear interpolation taking as a temporal reference the times of reception of the sampled data (El) corresponding to a particular phase (I) of the network and realizing, from sampled data (E2, E3) received at other times and corresponding in the other phases (II, III), a calculation of fictitious samples (X2, X3) returned to the reception instants of the reference phase (I).  Fault detector according to Claim 6, characterized in that the sampling frequencies (f2, f3) of the phases (II, III) other than the reference phase (I) are chosen to be greater than the sampling frequency ( fl) of the reference phase (I).  8. A fault detector according to claim 7, characterized in that, in case of "collision" of two samples (E1, E2 or E1, E3), the calculation unit (25) differs the interpolation calculation until at the onset of the next sample (E2 or E3), the sample (E2 or E3) colliding with the reference sample (E1) being stored and involved in the interpolation calculation.