RU2249827C2 - Electric arc tracking method and device - Google Patents

Abstract

FIELD: electrical engineering.

SUBSTANCE: method can be used tracking arcs in cable, in particular, in cable of aircraft board system. AC registered signal I(t) is read out discretely in time and angular function I(k) is determined by means of interpolation of definite number of read-out values y(k). Angular function I (k) models ac characteristic and is used for derivation of ac real-valued frequency ω. By comparing real-valued frequency ω with preset frequency ω' it is possible to judge on origin of alarm electric arc I_SA, I_GA. Then the warning signal S_arc can be generated if necessary. Device for arc tracking is integrated mainly in protective switch for aircraft board system, so it can be used for detecting and switching off electric arcs induced in cable of aircraft board system.

EFFECT: improved reliability of protection.

17 cl, 6 dwg

Description

The invention relates to a method for detecting emergency electric arcs in a cable, in particular in an on-board network cable of an air or spacecraft. It also relates to a device for implementing the method. In this case, an emergency electric arc is understood, in particular, the so-called Arc Tracking.

The problem of the so-called "Cable Arc Tracking" has long been known, in particular, in the on-board networks of air and / or spacecraft. In this case, Arc Tracking in the cable bundle between two cables passing next to each other due to, for example, earlier damage and / or pollution (Wet Arc Tracking) or due to friction along the metal edge (Dry Arc Tracking), a conductive connection occurs. This compound can be first high resistance (Wet Arc Tracking) or directly low resistance (Dry Arc Tracking). In this case, Wet Arc Tracking refers to a conductive connection between two insulation faults created by an electrolyte, such as cleaning fluid, and acting on the cable insulation. The result of the formation of an electric arc is often the thermal destruction of the entire cable bundle at a certain length. In contrast, with Dry Arc Tracking, a metal conductor creates a relatively low-impedance conductive connection, and an arc occurs almost immediately. Also, this can lead to thermal destruction of the cable bundle.

The problem with this is not only the destruction of the cable bundle itself. The subsequent damage resulting from this resulting from the burning of the surrounding material is critical. Explosion-like subsequent damage resulting from such cable fires can, in turn, cause a catastrophe in an air or spacecraft.

When studying the causes of emergency electric arcs or Arc Tracking in cables of such on-board networks, it was possible to prove that, on the one hand, the resistance of cable insulation to Arc Tracking depends solely on the design of the materials used in the cable sheath. On the other hand, the polyimide material used so far is characterized by very high water absorption, which contributes to preliminary failure due to cracking in the insulation, especially with frequent temperature changes that occur during flight mode. In addition, with thermal destruction, starting at about 800 ° C, a carbonizing coating forms, which contributes to the occurrence, i.e. moving the arc along the cable cut. In contrast, materials such as polytetrafluoroethylene (PTFE) are relatively resistant to Arc Tracking.

Since the insulation used for aeronautical cable types consists of a polyimide film, the goal of all efforts made in this field is to develop tests to classify cables as uniquely as possible according to their resistance to Arc Tracking. At the same time, polyimide as an integral part of cable sheaths has not yet been replaced due to its excellent dielectric properties and its high temperature strength. However, despite the optimization of processes and the utmost care in the manufacture and installation of cable bundles, it was not possible to completely prevent accidents due to Cable Arc Tracking. In addition, it was found that in plug connectors, under the influence of moisture and contamination, breakdowns can occur, which in the worst case lead to ignition of neighboring parts.

When installing protective devices, an even more difficult circumstance is that the amplitude of the current flowing during the Arc Tracking process, in particular during Wet Arc Tracking, can lie significantly lower than the rated current of the normally installed overload protection devices, as a result of which they do not work. In contrast, the amplitude values for Dry Arc Tracking are known to be higher. They arise, however, usually periodically, so the converted energy is often not enough to operate the installed protection devices.

To solve the described problem, it is therefore desirable to provide, in addition to the existing traditional protection against overload current in the on-board network of such an aircraft, protection from emergency electric arcs in order to protect the available on-board equipment as reliably as possible.

The invention is therefore based on the task of developing a particularly suitable method for detecting emergency electric arcs in a cable, in particular an onboard network cable of an air or spacecraft. In addition, a device suitable for implementing the method must be created, which in a simple and in particular compact way would provide particularly reliable protection of the cable or cable bundle and / or the consumer fed through it also from emergency electric arcs (Arc Tracking).

In relation to the method, this problem is solved according to the invention by the features of claim 1 of the formula. Preferred improvements are the subject of subordinate dependent clauses.

For this purpose, the obtained ac signal is recorded discretely in time, and by interpolating a predominantly small number of read values, an angular function is determined that models the ac characteristic and from which the actual ac frequency is derived. By comparing the actual frequency of the alternating current with a reference or predetermined frequency, a conclusion is then made about the occurrence of an emergency electric arc and, if necessary, an alert signal is generated. The reference or predetermined frequency is the generated frequency of the alternating current, i.e. in the preferred case, the use of the frequency of the onboard network or half the frequency of the onboard network. You can also set the critical frequency derived from a given frequency. It is, for example, more than 10% excess of the generated AC frequency.

The invention is based on the fact that when an electric arc or Arc Tracking occurs, the corresponding interference signal is superimposed on the signal passing through the supply wire inside the cable bundle, i.e. to a sine or cosine AC signal, and due to interruptions of the signal due to continuous quenching and reignition of the electric arc, each half-wave of the signal has a characteristic frequency component, higher than the rated frequency with which the on-board network operates. The Arc Tracking signal is contained in the received measurement signal in the form of signal peaks or peaks that can occur in different places of the positive and / or negative half-wave of the measurement signal, described below as a cosine signal. Since such interference or interference due to Arc Tracking is detected mainly in the current signal and to a lesser extent in the voltage signal, it is advisable to record the current signal by measurement.

The invention proceeds from the fact that a continuous measuring signal, the characteristic of which is recorded discretely in time by means of an analog-to-digital converter, is modeled only by a few values or read values according to the formulas:

and on this basis, you can accurately evaluate the actual frequency of the actual current characteristics according to the ratio:

After that, it is preferable only for the three read values to carry out the interpolation of the cosine function accepted as the preferred angular function, namely, the maximum value (max) and the read values (y (t _{l, r} )) lying to the left and right next to it. Based on only a relatively small processing amount of data, a particularly short processing time is required. The method is thus distinguished, in particular, by a high processing speed. Instead of the cosine function, it is possible, as a condition, to involve the sine function as a condition.

критерий или соответствующее

значение этих обоих значений отсчета.Both read values lying in time in front of or behind the maximum value preferably have the same ordinate value or amplitude value. This is ensured, expediently, by using a predetermined threshold value: during reading, the corresponding (left) read value is just above it, and the corresponding (right) read value is just below it. For further processing, the abscissa values related to these two read values are used, and thus the corresponding time steps, i.e. the threshold value is given the appropriate

criterion or appropriate

the value of both of these reference values.

With regard to the device, the named task is solved according to the invention by means of the features of clause 12. Preferred options are subject to subordinate dependent items.

The advantages achieved by the invention consist, in particular, in that, by means of an appropriate algorithm, by interpolating, it is advisable that only the three read values of the AC signal taken discretely in time evaluate the actual AC frequency, and if it deviates from the given frequency or the frequency of the on-board network, it is reliable and In a particularly effective way, the occurrence of Arc Tracking in the on-board network cable can be detected. By attracting a discrete current signal reduced predominantly by a factor of 10, which, compared with the signal characteristic, interpolated to estimate the frequency using only a very small number of read values, receives the information contained in the registered AC signal relatively completely, all data processing and signal estimation are simplified. To this end, the algorithm contains mainly two parallel program routes, and on the (first) program route for estimating the frequency, from the read speed, which is intentionally high for the desired high resolution of the AC signal, practically only three read values are processed. For the signal circuit itself or the (second) route of the program, it would be sufficient without noticeable loss of information of a relatively low read speed, so that the processed read speed, it is advisable, is reduced by reducing the number of read values of the discrete current signal.

The device operating in accordance with this method is preferably integrated into a safety switch of the on-board network of the aircraft, so that it serves in a simple and compact way to detect and disconnect emergency electric arcs that occur in the cable of the on-board network. For this purpose, the corresponding algorithm, it is advisable, is implemented in the form of the so-called ASIC (Application Specific Integrated Circuit) (specialized IC), which, due to its small design value, can be placed in a circuit breaker, which is usual for such an application, with especially small dimensions in accordance with the requirements .

The device is therefore suitable, in particular, for the detection of electric arcs caused by damage arising from voltage breakdowns due to insulation defects. A device that is firmly installed, it is advisable, in the on-board network of such an airborne aircraft, or a safety switch supplemented by this protective function, thus indicates arc faults and, if necessary, selectively trips them. Due to this, at least the burning time of the arcing arises is reduced so that subsequent damage is prevented. The safety switch supplemented with this protective function is therefore suitable, in particular, also for retrofitting already manufactured and / or intended airborne vehicles or aircraft in which traditional cable insulation is used. Such retrofitting is especially advisable, in particular, for economic reasons, since the usual service life of aircraft is currently 15-20 years.

Below examples of the invention are explained in more detail using the drawings, which depict:

- figure 1: flow chart of a method for detecting electric arcs in an onboard network cable;

- figure 2: a schematic diagram for explaining the principle of measurement and processing;

- figure 3: characteristic group of Arc signals and a single Arc signal;

- figure 4: a typical characteristic of the signal due to disconnection;

- figure 5: block diagram of a protective switch, which implements an algorithm for implementing the method;

- 6: in perspective, a General view of the protective switch.

Corresponding to each other details are indicated on all figures by the same reference position.

The voltage Ub used in the on-board network of an aircraft or aircraft is usually 115 V at a nominal frequency f 400 Hz. The corresponding time-constant or analog AC signal with the same frequency of 400 Hz can thus be described by the cosine theorem in a general and unbroken form, i.e. without interference, by the formula:

обозначает амплитуду, φ - фазу сигнала тока, а ω - круговую частоту, где f=ω/2π обозначает частоту сети или переменного тока.Wherein

denotes the amplitude, φ is the phase of the current signal, and ω is the circular frequency, where f = ω / 2π denotes the frequency of the network or AC.

Based on this theorem, in figure 1, the current I (t) in the supply line 2 of the on-board network or cable 1 of the on-board network is recorded by means of a sensor or a current transducer 3. The registered current signal I (t) contains, in addition to other disturbances, an Arc signal also arising due to the electric arc as an overlap. The registered current signal I (t) is read by means of an analog-to-digital converter 4 with a clock frequency f = 1 / T, for example, 50 kHz. The analog-to-digital Converter 4 has a resolution of at least 12 bits, mainly 16 bits. At a similar clock frequency f 50 kHz, current values (I (t) values) are read out every 0.02 ms, and

This signal, converted from a registered continuous current signal I (t) by means of an analog-to-digital converter 4 into a time-discrete signal I (k · T), is processed or evaluated using the algorithm shown in Fig. 1. For this, the signal I (k · T) in the first step 102 of the first route of the program by generating an absolute value according to the ratio

almost straighten, i.e. a negative half-wave appears to be a positive half-wave.

In the next step 103 of the program, the signal I '(k) is smoothed out using the digital filtering operation using the transfer function

and there is a reaction to impulse exposure

Moreover, the processing of the signal I '(k) in accordance with the operation of convolution or filtering leads to

Moreover, N = 9, if the corresponding average value is formed from 10 read values. N thus denotes the corresponding window width of the digital filter. In addition, k corresponds to the actual moment of time, while i denotes values backward in time. A digital filter or digital filtering operation thus works in the past, i.e. in a causal direction.

On the second parallel route of the program, the time-discrete signal I (k) passes at the first step 201 of the program through a low-pass filter, and the number of read values after filtering is reduced, for example, by a coefficient R = 10 according to the relation

In the next program steps 202 and 203 of this program route, again similarly to program steps 102 and 103, the absolute value of the reduced signal I _R (k) is first formed with the result I ' _R (k), moreover

Then again smoothing of the signal I ' _R (k) according to the relations 4a-4c with the output signal:

This signal I _R '' (k) is supplied to the program module or step 301 for evaluating a single Arc signal or arc pulse I _SA (k), as well as to program step 302 for estimating the number or group I _GA (k) of Arc signals or arc signals.

On the first route of the program in step 104, the actual frequency of the signal I '' (k) is determined by the general ratio:

or in full ratio:

Where

where y (k) corresponds to the output signal I '' (k).

According to this relation (9), based on a given and, therefore, known reading time T or reading speed f = 1 / T, according to the values of y (k _l ), y (k _r ) and y (k _max ) or I '' ( k _l ), I '' (k _r ) and I '' (k _max ) in the respective places of reading or at the corresponding time steps k _r , k _l , k _max , the actual frequency ω (k) of the time-discrete signal I ( k · T) and, thus, the measured current characteristic I (t). For this, the cosine function is interpolated by three points or read values, as is explained in more detail below using FIG. 2.

With a reading time of 0.02 ms and an on-board network frequency of 400 Hz, 125 read values or pairs (k, y (k)) of values are obtained for the signal period. The actual signal I (k) can thus be described in terms of:

where ω denotes the actual frequency. The actual frequency ω is estimated using only three points:

If less than three points are used, the value required for a complete description of the function (1) of the cosine would have to be, i.e. amplitude i (t) or phase φ, deduce from another operation.

To determine the points P _l and P _{r, it is} preferable to set a threshold value A, above which and when shifted to a lower threshold level, the corresponding time value or k-value k _l , k _r, respectively, of the next read value is used, whereas as the corresponding amplitude value or the y-value y (k _l ), y (k _r ) use the corresponding threshold value A, so that is true:

In this case, the threshold value A can be matched with the maximum current I _max (t), when the threshold value A is automatically set, for example, at 10-70% of the maximum current I (t) and, thus, at the rated current, usually controlled by a safety switch depending on the cross-section of the wire. It is also advisable to recognize the frequency at the corresponding step of the program, in particular, when the frequency of the alternating current or the on-board network is variable. The recognition of the actual frequency ω can occur, for example, before the first step 102 of the program and before or after reading the measuring signal or current signal I (t) by means of PLL (Phase Locked Loop) (phase-locked loop).

At step 105 of the program, this actual frequency ω (k) obtained from the time-discrete current signal I (k) is compared with a threshold value Δω _crit representing the critical frequency ω '(k), for example, 50 Hz, and it is true:

The frequency ω (k) represents in this case half the frequency of the alternating current or the on-board network (f / 2 = 200 Hz). If Δω≥Δω _crit , then in step 105 of the program, an Arc flag is generated in the form of a state signal S _z , which is sent further to both step 301 of the program and step 302 of the program. The state signal S _z is high (S _z = 1) if the frequency ω (k) obtained from the actually recorded current characteristic I (k) exceeds the threshold value. Otherwise, the state signal S _{z is} low (S _z = 0).

While on the second route of the program with its steps or function modules 201-203 with a reduced discrete current signal I _R (k) almost all information about the processed signal I (k) is stored, on the first route of the program with its steps or function modules 102 -105 use only the information about the signal I (k), which is required to estimate the frequency. Due to this, the pulse repetition rate processed on the first route of the program is especially simple to control. It is also relatively simple to control the processing of the reduced pulse repetition rate on the second path of the program compared to the original discrete current signal I (k), and this program path or signal transmission circuit also provides a fairly accurate estimate of the information contained in the complete signal according to (11), in particular with respect to several amplitudes y (k _max ) of the signal.

In step 301, programs determine and analyze only single Arc-signals I _SA (k), while in step 302, a group I _GA (k) of Arc-signals is determined and analyzed. Figure 3 shows a typical characteristic of the Arc signal of group I _GA (k) (left) and a single Arc signal I _SA (k) (right).

In step 301 of the program, using the signal I ' _{R R} (k), reduced in read speed, in step 203 and the state signal S _z , in step 105 of the program, it is examined whether at least L half-waves of signal I' _R (k) lie in the limits of the specified period of time. For this, the signal I '' _R (k) is formed, i.e. its increase is determined by the ratio

If S _z = 1, a grad (k)> grad _max , i.e. If grad (k) exceeds the threshold value, then the corresponding half-wave of the signal I '' _R (k) is considered as a potential Arc signal and the counter reading is increased. If the counter reading lies within the time interval above L, then at step 301 of the program, a notification signal representing the arc registration is generated, for example, high or 1. This signal is supplied to perform a logical OR operation (step 303) or to an OR logic element 5 supplying a signal S _ARC alerts to the indicator element 6, for example, in the form of a light emitting diode (LED).

In step 302 of the program, the group I _GA (k) of Arc signals is compared with a possibly (simultaneously) occurring characteristic of the interference signal due to a disconnection from the consumer side, a typical signal characteristic I _SE of which is shown in FIG. 4. True, a similar interference signal I _sa due to switching on or off a consumer or load connected via this on-board cable 1 has a characteristic similar to Arc signals I _GA (k). However, such a signal I _{SA of} interference is known to decay by an exponential function, which does not occur with a typical group of I _GA (k) Arc signals.

Using or based on this nature of the attenuation, it is therefore possible, due to disconnections, to distinguish a similar interference signal from a typical group I _GA (k) of Arc signals. To do this, first determine the maxima of all the half-waves of the Arc signals of group I _GA (k), the frequency ω of which lies above the critical frequency ω '. Then carry out the summation of the values of M following one after the other maxima, as well as determining the maximum values of M and the formation of the average value. Upon reaching the set fraction of the average value, for example, 90%, Arc-Tracking is detected, and at step 302 a warning signal representing the arc registration is generated, for example, high or 1. This signal is sent to perform a logical OR operation (step 303), as a result of which it is generated alert signal S _ac for indicator element 6 or for operation of the protective switch (figure 5).

The entire function of the program is predominantly integrated into the ASIC (Application Specific Integrated Circuit) (specialized IC), which, in turn, is placed as an integral part of the protective switch 7 (Fig.6), usually provided in an air or spacecraft. The dimensions of the safety switch 7 are coordinated in this case with the requirements usually imposed on air and spacecraft, and are approximately a = 50 mm, h = 40 mm and b = 20 mm.

Figure 5 shows the functional modules of such a protective switch 7 with a built-in analog-to-digital converter 4 and the ASIC 8 functional module, into which the algorithm depicted in figure 1 and operating according to the method described here is hard-coded. The network part 9 generates a supply current or a supply voltage for the function module 8 and the analog-to-digital converter 4 connected thereto, as well as for the built-in current sensor 3 and the remote sensing module 10 in the form of, for example, a pickup coil. The latter is connected with the function 11 of the protective switch, in particular with at least one trip unit, a lock and a switching contact 12. A switching contact 12 is connected between the LINE terminal A _Ln and the LOAD terminal A _{Ld of the} protective switch 7, which additionally contains a ground terminal A _G .

When the protective switch 7 is connected to the supply wire 2 of the on-board network or on-board cable 1, the current sensor 3 registers the current signal I (t), which goes when the switching contact 12 is closed via function 11 of the protective switch to the actuator, sensor and, thereby, to the consumer network or in general to the load 13 of the on-board network 1. Its processing for recording Arc signals (Arc Tracking) due to the arcing along the cable 1 of the on-board network of the arc occurs inside the protective switch 7 by means of an analog-to-digital converter 4 FPGA module 8 (gate array, programmable by the user). The latter generates, in the case of the described detection of Arc Tracking, an alarm signal S _acc sent to the indicator 6 and / or the remote sensing module 10. The latter provides a response signal S _A to the current, in particular, as protection line function 11 protective switch, which opens the switching contact 12 and thus separates the corresponding load 13 from the supply line 2 and, thus, from the on-board network 1.

The safety switch 7 thus provides, in addition to the traditional protection against overload current in the on-board network of such an aircraft, extended protection against emergency electric arcs, so that the existing on-board equipment is reliably protected.

List of Reference Items

1 - on-board network

2 - supply line

3 - current sensor

4 - analog-to-digital converter

5 - logical element OR / logical operation OR

6 - indicator element

7 - safety switch

8 - ASIC function module

9 - network part

10 - remote operation module

11 - circuit breaker function

12 - switching contact

13 - load

102-303 - program steps

I (t) - AC signal

I (k) - time discrete current signal

I _SA - (single) Arc signal

I _GA - Arc Signal Group

I _SE - interference signal characteristic

S _arc - alert

S _z - status signal

Claims (16)

1. A method for detecting emergency electric arcs in a cable (1), in particular, on-board network cable of an airborne aircraft, in which a registered AC signal (I (t)) is read discretely in time and using a discrete signal generated in this case (I ( k)) of the current, by interpolating a certain number of read values (y (k)), determine the angular function (I (k)), which models the characteristic of the alternating current and from which the actual frequency (ω) of the alternating current is derived, and from the comparison result At the natural frequency (ω) of an alternating current with a given frequency (ω ′), a conclusion is drawn about the occurrence of an emergency electric arc. 2. The method according to claim 1, in which to derive the actual frequency (ω) of the alternating current from the number of read values (y (k)), only the read value (P _max ) is used, representing the maximum value, as well as the time before the maximum value left read value (P _l ) and lying in time behind the maximum value (right) read value (P _r ). 3. The method according to claim 2, wherein the left (P _l ) and right (P _r ) reference values are given the same y-value. 4. The method according to claim 2 or 3, in which a threshold value (A) is set, which is given as a y-value exceeding its read value (P _l ) and the read value (P _r ) below it. 5. The method according to one of claims 1 to 3, in which the actual frequency (ω) of the alternating current is determined using the algorithm according to the ratio:

Where

moreover, 1 / T denotes the clock frequency, and k _l , k _max , k _r denote the time cycles (k · T) of the read values (y) given to the corresponding amplitude values y (k _l ), y (k _max ), y (k _r ) (k)). 6. The method according to claim 1, wherein in the first processing step (102) an absolute value of the discrete current signal (I (k)) is formed. 7. The method according to claim 6, in which at the step (201) of processing the number of read values (I (k)) of the discrete current signals, they are reduced by the coefficient R, and in the first step (202) of the processing, the absolute value of the reduced discrete signal (I _R (k)) current. 8. The method according to one of claims 6 or 7, in which, at the second processing step (103, 203), a discrete current signal (I ′ (k)) pre-processed by forming an absolute value and / or a discrete signal (I _R ′ (K)) of the current, reduced by the coefficient R and pre-processed due to the formation of an absolute value, continue to be processed by digital filtering according to the convolution operation

moreover, N denotes the width of the digital filtering window, and h (k) is the resulting response to the pulsed action, with h (k) = 1 / N for k = 0 to N-1, and the rest h (k) = 0 . 9. The method according to claim 8, in which using the signal (S _r ), the state derived from the result of the comparison of frequencies, and the output signal (I _R '' (k)) derived from the reduced discrete signal (I _R (k) ) current, single interference signals (ISA) and / or group (IGA) of interference signals are evaluated in relation to their formation due to an electric arc. 10. The method according to claim 1, wherein the alert signal (S _arc ) generated when an emergency electric arc is detected is used to trip the protective switch (7), which, if triggered, separates the load (13) from the cable (1). 11. The method according to claim 1, in which the detected group (I _GA (k)) of Arc signals is distinguished from the noise resulting from the disconnection of the signal (I _SE ) by the nature of its attenuation. 12. Device for detecting emergency electric arcs (Arc Tracking) in cable (1), in particular, on-board network cable of an airborne aircraft, comprising a current sensor (3) for detecting an alternating signal (I (t)) passing through cable (1) current, an analog-to-digital converter (4), which from a signal (I (t)) of an alternating current generates a certain number of time-discrete readout values (y (k)), and a functional module (8), the algorithm of which, by evaluating the selected readout values (at _1r , Y _max ) determines the real frequency (ω) current deviation, and when it deviates from the set frequency (ω '), it generates a signal (S _arc ) alerts. 13. The device according to item 12, in which ASIC (Application Specific Integrated Circuit) (specialized IP) is used as the functional module (8). 14. The device according to item 12 or 13, in which the current sensor (3) and the analog-to-digital converter (4), as well as the functional module (8) are integrated into the protective switch (7), the switching contact (12) of which, upon receipt of the signal (S _arc ) alarm is open and disconnects the connection between the input terminal (A _Ln ) for the supply wire (2) of the cable (1) and the output terminal (A _Ld ) for the load (13). 15. The device according to 14, in which the supply network part (9) connected from the input side to the input terminal (A _Ln ), from the output side is connected to a functional module (8) and a current sensor (3). 16. A safety switch, in particular for the on-board network of an airborne aircraft, comprising a device according to claim 12 or 13 for detecting and tripping emergency electric arcs arising in the cable (1).

Images