CH684660A5 - A method for monitoring at least one electrical lead. - Google Patents

Description

1

CH 684 660 A5

2nd

description

The invention relates to a method for monitoring at least one electrical line for detecting a short circuit or overcurrent.

In traditional protection technology, the fault location or determination of the direction of energy flow during a short circuit or overcurrent in the power supply network is carried out by phase comparison of current and voltage (in phase or in phase), but with certain phase shifts - e.g. caused by network impedances, transient response of the filters for fundamental wave filtering of the voltage, etc. - difficulties can arise in the correct evaluation. In the known methods or arrangements of the type mentioned, the procedure in practice was such that measuring points were distributed over the line to be monitored, at which the phase relationship of current and voltage could be recorded. In the event of a short circuit, it was then possible to determine the two measuring points between which the short circuit occurred due to the 180 ° phase shift between current and voltage in a measuring center. Apart from the difficulties mentioned, these methods have the disadvantage that they usually require very expensive voltage converters, so that the overall costs are very high, especially in the case of complex networks with a corresponding number of measuring points.

The object of the present invention is to design a method of the type mentioned at the outset or a circuit arrangement for carrying out such a method in such a way that the disadvantages of the known methods and arrangements mentioned are avoided and that a short-circuit is particularly easy, safe and inexpensive. or overcurrent detection on one or more, possibly also networked, electrical lines is possible.

According to the present invention, this is achieved in a method of the type mentioned at the outset in that the time profile of at least one current is sampled, digitized and stored at at least one measuring point and that after the occurrence of the event to be recorded determined from the time profile, the stored measured values or the associated signal for the more detailed characterization of the event is or will be analyzed.

The switching arrangement according to the invention for carrying out such a method is characterized by the wording of claim 17.

According to the invention, the fault location is carried out with the current alone, that is to say without the aid of the voltage. This is possible in that the detection and evaluation of the current takes place taking into account its time profile, so that the expensive voltage converters mentioned are not necessary. This method can be adapted to any frequency including direct current, with limits regarding the feasibility at best in the possible processing speed of the electrical or electronic

Components exist. It can be used with individual spur lines as well as with meshed networks - for example for power supply via high-voltage lines or for the electrical supply network of the railway.

According to an embodiment of the method according to the invention, it is provided that the digitized measured values of the measuring points are continuously stored in a memory area, which is organized as a ring memory, and that the storage process is terminated a defined time after the occurrence of the event to be detected has been detected. This is a very simple and also hardware-saving further development, which nevertheless ensures that the data characteristic of the event to be recorded are available when required, without the need for extensive memory or the like.

To detect the occurrence of a short circuit, in a further embodiment of the invention, the digitized or stored measured values are continuously compared with a trigger threshold assigned to the respective current, and if the trigger threshold is exceeded by signal analysis, the time of the short circuit occurrence - or also / or the time of the short circuit end -or the time at which a circuit breaker opens for the first time - is determined. Setting or monitoring such a trigger threshold is very simple in connection with the sampled and digitized current signal. It is also very easy to carry out a signal analysis of the current curve stored in the memory at the time of the response, from which various characteristics of the short circuit, such as the time of entry mentioned or also the total current which has flowed until it was switched off via a circuit breaker or the like, are determined can be.

Since the end of the short circuit in a network node under consideration always represents the transition from a (basic) sine curve to a zero line, this transition is usually more pronounced than the entry of the short circuit; therefore that point in the curve can be searched using the same methods as the short circuit entry; this is not described separately here and below.

According to a particularly preferred embodiment of the method according to the invention, it is provided that the amplitude values of the crests of the individual half-waves of the respective current before and after exceeding the trigger threshold are related to one another and the amplitude value of the crests of the individual half-waves of the current after the trigger threshold has been exceeded to the trigger threshold and that with a ratio of the two amplitude values to one another of over 10 to 150%, in particular 10 to 80%, preferably 30 to 50%, and with a ratio of the amplitude values to the trigger threshold existing after the trigger threshold has been exceeded, of 110 to 150%, preferably 15 to 40%, the “short circuit” message to the trigger threshold, otherwise the “overload” message, er5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

2nd

3rd

CH 684 660 A5

4th

witnesses and is spent. It is therefore a level-triggered, interference-suppressed output with overcurrent test, which in most cases enables a clear distinction between short circuit and overload. The peak values of the individual current half-waves are determined before and after the trigger point, a decisive criterion being the rate of change of the current or these peak values.

To determine the point in time at which the short-circuit occurs or at the end of a short-circuit, or the time at which a circuit breaker is opened, the stored time profile of the current is differentiated twice according to the time according to a further embodiment of the method, and then the value profile of this second derivative with regard to a sudden exceeding of one Short circuit of representative level is checked and finally such a sudden exceedance is detected as the short circuit occurrence or short circuit end or as the time at which a circuit breaker trips.

The start of the transition from the operating current to the short-circuit current can only be in the range below the trigger level. Since the range from the start of the transition to at least the trigger level is to be displayed as precisely as possible, the peak values can also be cut off in the case of large short-circuit currents, but this is meaningless for the evaluation. In principle, after the trigger threshold has been exceeded, either the point in time of the short-circuit occurrence or short-circuit end in a considered network node or the point in time at which a circuit breaker trips for the first time can be searched (in meshed networks, current can then flow at one measuring point afterwards, which is via another Circuit breaker with longer tripping time comes). Before the short-circuit occurs (short-circuit end), the recorded curve of the temporal current curve can either be zero (no operating current) or also in the size of the operating current with any harmonic content, whereby overloads are also possible. In the short-circuit current range, this curve is basically sinusoidal, although there may be deviations from the pure sinusoidal shape due to various circumstances, such as interference with a low-voltage network supplying the electronics, digitization errors, grounding problems, etc. The short-circuit current itself can naturally occur in any half-wave between 0 and 180 °, both with and without phase reversal or DC voltage component, which depends on various individual circumstances. The transition of the curve into the short-circuit area can also be either sharp-edged or noisy, or more or less smoothed. For all these reasons, it can be seen that various precautions are required for the evaluation of the stored signal curve in order to be able to do justice to all eventualities.

In order to refine such an evaluation or to eliminate disturbances, fluctuations in the current profile and the like, the stored time profile - as mentioned above - can be differentiated twice according to the time. This second derivative provides a corresponding step response when a short circuit occurs due to the signal change. The search for the y "jump word starts at the y trigger time and goes on the sinusoidal curve in the direction of the short-circuit entry. A reverse search direction (beginning after the short-circuit end) is usually not very useful, since the operating current range due to the harmonics continuously gives y" responses can deliver, the statement of which cannot be clearly interpreted.

If the signal is limited in the y direction (that is, in the direction of the current plotted over time), y "reactions also occur. These can be recognized as interference points on the basis of the y amplitude and can be ignored or eliminated during the evaluation.

In a further embodiment of the method, it is provided according to the invention that if the predetermined level is not abruptly checked for a short-circuit occurrence, the original curve is corrected again and again at those points by a polynomial correction or a regression line at which the third derivative after the time has suddenly exceeded a predetermined level. If necessary, the previously corrected original curve is used as the original curve.

If the (in principle) sinusoidal short-circuit curve has interference points, these can lead to incorrect statements by the y "step response mentioned above. A correction or check with an enlarged step size is then necessary. This correction can be carried out according to a polynomial of the second degree with as many points as possible (Usually, ten to thirty points are sufficiently precise as a maximum.) If the recorded signal runs into the y-limit, the number of points available for the correction calculation is reduced accordingly. If only four points are then available, then the amplitude is very large and the available signal train around the zero line no longer corresponds to the sinusoidal form, but rather to a straight line. For this reason, a regression line can be placed with more accuracy in practice through the remaining points (faster evaluation) History replaces the original sale, However, only “original values” may be used for the correction and not corrected values, since otherwise the curve can be greatly distorted. Such a correction using polynomials can eliminate interference points and smooth the curve in the sinusoidal short-circuit area. However, the actual short-circuit entry is made more angular in the course of the curve, as a result of which the y "step response is increased.

In the case of a heavily grounded short circuit entry, the y "step response can also lie in the region of the noise of the curve. In this case, the

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

3rd

5

CH 684 660 A5

6

actual short-circuit occurrence with higher leads and interim corrections of the original (y) curve are worked out, although higher leads than the sixth can lead to misinterpretation due to noise and are better avoided.

In the latter context, advantages can also be achieved with a further embodiment of the method according to the invention, according to which the second derivative is determined at least twice in order to ascertain or verify an abrupt level violation, with different differences in the arguments and functional values.

According to a particularly preferred further embodiment of the invention, in a method of the type mentioned, in which a check is made at a number of locations that are spaced apart from one another for short-circuit occurrence or short-circuit end in a network node under consideration or for the point in time at which a circuit breaker trips from the stored measured values at each point at the time of the short-circuit occurrence or short-circuit end in a considered network node or at the time at which a circuit breaker trips, the slope of the tangent to the short-circuit entry or short-circuit end in a considered network node or the time at which a Circuit breaker switches off, corresponding points of the abrupt exceeding of the level are determined, where appropriate the slopes of the tangents are displayed with the correct sign.

Due to the direction of current flow - regardless of the polarity of the voltage currently prevailing at the point in time at which the short circuit occurs or at the end of a short circuit in a network node under consideration or at the time at which a circuit breaker opens - at each measuring point, of course, at the same time of the current short circuit occurrence or for all measuring points. At the end of a short circuit in a network node under consideration or at the point in time at which a circuit breaker trips, the fault location can be limited to the area between two measuring points. For the assessment of the direction of current flow, a criterion can only be found in the course of the current curve, which occurs with the same reliability in all measuring points. There are basically two options for this: the time of the short-circuit occurrence or the time of the short-circuit end (short-circuit current = 0) in a considered network node. The point in time of the short-circuit occurrence can always be used for the evaluation regardless of the switching states of the network. The recorded curve of the current signal can be subjected to a wide variety of procedures, which have already been mentioned above, in order to find the point in time at which the short-circuit occurs. Then, as mentioned, the direction of the slope of the signal after the time found is determined at all measuring points and output or reported as “current flow direction”. Instead of the at the time of the short-circuit occurrence

Measuring points prevailing current flow direction - which can be different depending on the time of the short-circuit occurrence after the respective half-wave of the current, but the energy flow direction always pointing the same direction to the fault location can also be used. Alternatively, the point in time at which a first circuit breaker trips can be used as a further possibility.

According to another preferred exemplary embodiment of the invention, it is provided that if the trigger threshold is exceeded from at least some of the data stored before and after the time the trigger threshold is exceeded, by forming the second derivative over time or by comparing the before, in particular at least one half-wave of the current before, the point in time at which the trigger threshold was exceeded, with the data stored after the point in time at which the trigger threshold was exceeded or by comparing the Fourier transforms of the before, in particular at least one half-wave of the current before, the point in time at which the trigger threshold was exceeded Data with the Fourjer transform of the data stored after the trigger threshold was exceeded, the time of the short-circuit occurrence or short-circuit end in a network node under consideration or the time at which a first power switch r switches off, is determined. The at least temporarily stored data of the digitized current curve thus enable rapid signal analysis and thus evaluation of the event to be monitored.

In order to determine a time range comprising the short-circuit occurrence or the short-circuit end or the time at which a circuit breaker opens for the first time, successive half-waves of the stored current profile can also be correlated with one another in such a way that the first with the second, then the second with the third, then the third with the fourth, etc. half-wave, and that the time of occurrence of those successive half-waves, the correlation values, is defined as the time range of the short-circuit occurrence or short-circuit end in a network node under consideration or as the time at which a first circuit breaker opens deviate the most from the other correlation values. So it starts again at the beginning of the curve (operating current), with each half-wave being compared with the next half-wave.

In a further embodiment of the invention, it is provided that starting at the beginning of that part of the current curve stored before and after the trigger threshold is exceeded, from which point in time the short-circuit occurrence or short-circuit end occurs in a network node under consideration or the time at which a first circuit breaker trips is determined, a time window is subjected to a Fourier analysis that the time window is formed by the

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

4th

7

CH 684 660 A5

8th

Fourier analysis is gradually shifted over the entire range of the stored time function, and that as the time of the short circuit occurrence or short circuit end in a network node under consideration or as the time at which a circuit breaker trips, the first edge of that first time window was found in the direction of increasing time values where the spectrum shows a large proportion of additional frequencies. The stored curve is thus, starting on the side of the operating current, closing in small windows (for example 127 measuring points) closing on the right, subjected to a Fourier transformation with regard to the frequencies. If the kink of the short-circuit entry or end of a short-circuit in a network node under consideration or the time at which a circuit breaker opens appears in this window, the Fourier analysis recognizes those frequency parts which cause this kink, which defines the short-circuit entry on the time axis .

According to a further embodiment of the method according to the invention, provision can also be made for determining the zero crossings and the zero crossings to be expected from the operating current in order to determine the point in time of the short circuit occurrence or short circuit end in a network node under consideration or the point in time at which a circuit breaker trips in the range of the short-circuit current and that the half-wave during which a target-actual deviation of the zero crossings takes place due to the phase shift that occurs is determined as the time range of the short-circuit entry or short-circuit end in a considered network node or as the time at which a circuit breaker opens. The signal curve can also be filtered to eliminate the DC voltage component in the short-circuit area before the evaluation.

If the operating current is very small or zero before the short-circuit occurs, the strong bend in the curve can cause e.g. a clear result can be obtained using the derivation method. However, if the operating current is very large and the point in time of the short-circuit occurrence or short-circuit end in a network node under consideration or the point in time at which a circuit breaker opens is in the area of the zero-current crossing, the step response of the derivation method may not be entirely clear. In addition to the method of the mentioned section-wise approximation by means of second degree polynomials, the phase shift of the current can also contribute to the unambiguous decision. Due to the ohmic resistance, the short-circuit current has a different phase position compared to the operating current with its strongly inductive or capacitive resistance component.

According to a further particularly preferred embodiment of the invention, it is provided that for exact fault location in a parallel line with approximately the same ohmic resistance of the parallel lines at each end of each of the parallel lines

Lines a measuring point with a downstream measured value acquisition and evaluation circuit is arranged, whereby in the event of a short circuit, the total current at the ends of the parallel line and the current in an undisturbed parallel line are determined from the stored current curves of the measuring points, and from this the distance of the fault location based on the line length is calculated by the measuring point adjacent to the short circuit according to Kirchhoff's laws. Such parallel connections of two lines of the same cross-section and material are in the 110 kV network and also in a e.g. 15 kV rail network almost always the case on single and double-track lines with H-switching, so that the actual fault location within two measuring points can be determined relatively precisely due to the current distribution of the detected short-circuit currents. An operating current flowing in the area under consideration is reduced due to the very low voltage in this area at the time of the short circuit to values which only slightly distort the result of such a determination.

The circuit arrangement mentioned at the outset for carrying out the method according to the invention enables in a very simple manner the simultaneous scanning of all measuring inputs or, thus, the correctly timed monitoring of all measuring points, which makes short-circuit or overcurrent detection simple and reliable. The invention is explained in more detail below with reference to the schematic drawings. It shows:

1 shows an electrical line section to be monitored according to the invention,

2 shows an example of a circuit arrangement for carrying out the method according to the invention,

3 shows an example of a current-time curve in the area of a short circuit occurring,

4 shows a current-time curve when an overload occurs,

5 shows a current-time curve (y) in the region of a short-circuit occurrence (top) and the associated derivative (y ") according to the time (bottom),

6 shows a further current-time curve in the short-circuit area with an illustrated phase shift,

7 in the upper area a further current-time curve (y) in the short-circuit area and below an associated correlation curve,

8 another line section to be monitored with regard to short-circuit or overload entry with three measuring points,

9 shows the current-time curves associated with the measuring points according to FIG. 8 in the region of the short circuit entry,

10 shows the two upper curves from FIG. 9 with the gradients when the short circuit occurs,

11 shows a circuit arrangement according to the invention for fault location in a parallel line and

12 shows the current distribution in the event of a short circuit in a line arrangement according to FIG. 11.

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

5

9

CH 684 660 A5

10th

1 shows a line network in which, according to the invention, the time function of the current is to be recorded and evaluated. The sub-network in question consists in particular of a line 1 with an independent feed point 2, 3 at each end. In the course of the line, n individual measuring points 4 for detecting the instantaneous current values are arranged with downstream evaluation devices 5 and output units 6. In the measuring points 4, the current signal is obtained, for example, from a current transformer or a magnetic field sensor built into an isolator and processed with a measuring transducer for further processing. The measurement signal obtained in this way is sampled, digitized, stored and evaluated in the evaluation device 5. The results of the evaluation are transferred to the output or display device 6, which either displays them directly on site and / or transmits them via a telecontrol system or the like to a control center, where they are either further processed, linked, or output.

FIG. 2 shows a block diagram of an example of an evaluation device 5 from FIG. 1. The measurement signals Sig 1 to Sig n supplied by the measuring points (4 in FIG. 1) are sent to each of the sampling / hold circuits 7 controlled by a sampling clock AT Analog inputs of an analog multiplexer 8, which, depending on the state of the control lines MUX-ADR, connects a measurement signal to the analog / digital converter 9. This converter 9 converts the measurement signal into a sequence of digital words (e.g. with a word length of 12 bits), which are present at the data input of a memory 10. This memory 10, like the upstream units 7, 8, 9, is controlled by a control and computation block 11 via control and address lines (AT, MUX-ADR, A / D-STRG, MEM-STRG, MEM-ADR). The data words which have just been written into or read from the memory are present at further inputs (DAT) of the control and arithmetic block 11. The control itself takes place in such a way that all of the measurement signals Sig 1 to Sig n are sampled simultaneously by the sample and hold circuits 7, switched through to the A / D converter 9 via the analog multiplexer 8 and stored in the memory 10 as digitized measurement values will. A separate memory area organized as a ring buffer is provided for each measurement signal and is addressed by the control and arithmetic block 11 via the address lines MEM-ADR. As soon as a memory area is fully occupied, the subsequent data is stored again at the beginning of the area, whereby the oldest data is overwritten.

The control and arithmetic block 11 has a further input (CLOCK) at which a clock signal generated by the system clock generator 12 is present. With further outputs, output units such as Output relay 13, controlled. The control and arithmetic block 11 is usually formed by a specialized microcomputer that executes a program that controls the required work steps.

3 and 4, based on current signals actually obtained or recorded, for example, with an arrangement according to FIGS. 1 and 2, the possible distinction according to the invention between the occurrence of a short circuit (FIG. 3) or an overload (FIG. 4) clarifies. As soon as the recorded current signal exceeds the trigger level 15, the amplitude ratio of the current is evaluated before and after the trigger time. A rough assessment can be made beforehand using the ratio of amplitude A to amplitude B. For a more precise distinction, the steadily increasing current rise is observed. If the current rises evenly, i.e. without a sudden change in amplitude, (indicated by line 16 in FIG. 4), it is clearly an overload. If, on the other hand, the current increases with a sudden change in amplitude, an overload can still be assumed if the level is exceeded below 20% of the trigger value and the amplitude ratio A to B is less than 40%. Only those cases that do not meet this criterion with the numerical values given here as an example are to be regarded as a short circuit. As it turns out, around 90% of all circuit breaker trips can be clearly and correctly assigned to either a short circuit or an overload.

5 to 7, some possibilities according to the invention for limiting or determining the point in time of a short-circuit occurrence by evaluating the digitized or stored current curve will now be described. 5 shows the current signal denoted by y in the region of the short-circuit entry, the current rise here being very smooth due to various circumstances, and hence the or from the signal form y alone does not allow a clear time indication for the short-circuit entry. After twice differentiating the signal curve y over time, a curve shape is shown symbolically in FIG. 5 below and labeled y ", which shows a corresponding step response when a short circuit occurs. This y" step response can also lie in the area of the noise of the curve, where then the short-circuit occurrence can be worked out with higher derivatives and correction calculations.

6, a current curve is shown in the area of the short circuit, in which the time of the short circuit occurs directly in the area of the current zero crossing. If in this case the operating current becomes rather large, the step response of the derivation method may not be clear. Here the occurring phase shift of the current can be used to determine the time range of the short-circuit occurrence.

Due to the ohmic resistance, the short circuit current has a different phase position compared to the operating current with its strongly inductive resistance component. The zero crossings, which have the time interval x1, are determined from the operating current. The expected further zero crossings in the range of

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

6

11

CH 684 660 A5

12th

Short-circuit current set. In the area of the half-wave, in which the time of the short-circuit occurs, the soli-actual deviation of the distance between the zero crossings (here x2) leads to the expected statement, since the phase shift now takes place here. The curve in the short-circuit area can also be filtered to eliminate the DC voltage component. The zero crossing at the end of the half-wave containing the short circuit is detected and the sign of the tangent in this zero crossing is inverted and output as a current direction criterion.

According to FIG. 7, the curve (current versus time) shown above and designated y is examined for the time of the short-circuit occurrence by starting at the beginning of the curve (operating current) to compare one half-wave with the next half-wave. The half-wave that deviates greatly from the previous half-wave contains the time of the short-circuit occurrence. The correlation values are plotted in Fig. 7 below.

8 to 10, the fault location in a control center (e.g. transformer station or load distributor) can be located in the area between two measuring points 4 (in the ÖBB contact line network, e.g. train stations) due to the direction of the current flow - regardless of the polarity of the voltage at the time of the short-circuit occurrence , in general, e.g. in the 110 kV network, substations). The load resistance is designated RL here and is formed, for example, by a moving electric locomotive. The associated operating currents are entered in Fig. 8 with IB1 and IB2. In the event of a short circuit (indicated by arrow 17), opposite flow directions of the short-circuit currents IK1 and IK2 are now determined in the adjacent measuring points MP1 and MP2 after the trigger threshold has been exceeded - the error therefore lies between the corresponding measuring points 4.

For the assessment of the current flow directions, a criterion must now be specified in the course of the current curve, which occurs with the same reliability in all measuring points. There are in principle two options entered in FIG. 9: either the time of the short-circuit occurrence (indicated by the dashed line 18) or the time of the short-circuit disconnection (marked by the arrows 19 in FIG. 9).

The time of the short-circuit occurrence can always be used for the evaluation, regardless of the switching states of the network. The current flow curve is initially subjected to various procedures as described above, for example, in order to find the point in time at which the short-circuit occurs. The direction of the slope of the signal after the point in time found, or the inverted slope of the zero point tangent, is then determined (identified in each case by 20 in FIG. 10) and further processed or reported as “current flow direction”. Since it is only a matter of determining the difference occurring in the direction of flow of the short-circuit current, it is also irrelevant whether the short-circuit current rise occurs during a positive or a negative voltage half-wave or whether the short circuit causes a phase reversal or not.

As a further possibility, the evaluation of the short-circuit end is also entered or indicated in FIG. 10. The time at which the first circuit breaker is switched off after the short circuit (e.g. in the line labeled 21) can be determined in an analogous manner as described above and the direction of the slope (20) of the respective curves before this point can be used.

After the fault location has been limited to the area between two measuring points, for example, as described with reference to FIGS. 8 to 10, lines 1 connected in parallel in accordance with FIGS. 11 and 12 can also be used based on the current distribution according to the Kirchhoff principle, assuming the same line cross sections and Line lengths (this is almost always the case with high-voltage lines), the fault location can be calculated exactly. An operating current measuring in the area under consideration decreases due to the very low voltage in this area at the time of the short circuit to values which only slightly falsify the result of this determination.

The location of the fault is again marked by arrow 17. From the cable lengths or assigned resistances R1, R2 and the currents 11, 12 and 13 (see Fig. 12) the distance of the fault location from the measuring point adjacent to the short circuit (assigned distance resp Resistance R1) can be calculated according to Kirchhoff with R1 / R3 = (I2 + 2 * I3) / (I1 + I2).

With the described method according to the invention, the described possibilities for short-circuit or overcurrent monitoring also result in a number of further advantageous possibilities, which are only briefly mentioned in examples below.

For example, it can be easily checked in every busbar system whether there is a busbar earth fault. Such an earth fault occurs when some of the currents flowing in do not flow to the earth in an uncontrolled manner, but via the busbar. If all line outlets (substation or station switchgear) are now detected by the measuring system, then, in line with Kirchhoff's law (sum of the currents equal to zero), a grounding of the busbar can be carried out easily due to the simultaneous acquisition of measured values (sample / hold with downstream analog multiplexer) be recognized.

If the protection-off commands are recorded in a substation, it can be checked in real time based on the signal detection of the current whether the short-circuit or overload current was switched off in the specified time, whereby a higher-level circuit breaker can also be switched off if necessary ( Response time below 1 ms).

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

7

13

CH 684 660 A5

14

Claims (17)

Claims 1. A method for monitoring at least one electrical line for detecting a short circuit or overcurrent, characterized in that the time profile of at least one current is sampled, digitized and stored at at least one measuring point and that after the occurrence of the event to be recorded, which is determined from the time profile, the stored measured values or the associated signal for a more detailed characterization of the event are or will be analyzed. 2. The method according to claim 1, characterized in that the digitized measured values of the measuring points are continuously stored in a memory area, which is organized as a ring memory, and that the storage process is terminated for a defined time after the occurrence of the event to be detected has been detected. 3. The method according to claim 1 or 2, characterized in that to detect the occurrence of a short circuit, the digitized or stored measured values are continuously compared with a trigger threshold assigned to the respective current, and that when the trigger threshold is exceeded by signal analysis, the time of the short circuit occurrence or the end of the short circuit in a network node under consideration, or the time at which a circuit breaker trips is determined. 4. The method according to claim 3, characterized in that the amplitude values of the crests of the individual half-waves of the respective current before and after exceeding the trigger threshold are related to one another and the amplitude value of the crest of the individual half-waves of the current after the trigger threshold is exceeded to the trigger threshold and that at a ratio of the two amplitude values to one another of 10 to 150%, in particular 10 to 80%, preferably 30 to 50%, and / or in a ratio of the amplitude values to the trigger threshold present after the trigger threshold has been exceeded of more than 110 to 150% , preferably 115 to 140%, the “short circuit” message of the trigger threshold, otherwise the “overload” message is generated and output. 5. The method according to claim 3, characterized in that to determine the time of the short circuit onset or short circuit end in a network node under consideration or the time at which a circuit breaker trips, the stored time profile of the current is differentiated twice after the time, then the value profile of this second derivative is checked with regard to a sudden exceeding of a level representative of the short circuit and such a sudden exceeding is detected as short-circuit occurrence or as short-circuit end or as a point in time at which a circuit breaker trips. 6. The method according to claim 5, characterized in that, if the predetermined level is not abrupt, it is checked for a short-circuit occurrence or for a short-circuit end in a network node under consideration or for a point in time at which a circuit breaker trips. that the original curve is corrected at the time points by a polynomial correction or a regression line at which the third derivative after the time has suddenly exceeded a predetermined level. 7. The method according to claim 6, characterized in that the previously corrected original curve is used as the original curve. 8. The method according to any one of claims 5 to 7, characterized in that the derivation is determined at least twice to determine or verify a sudden level exceedance, with different differences in the arguments and function values. 9. The method according to any one of claims 3 to 8, characterized in that at a plurality of locations spaced apart from one another for a short-circuit occurrence or for a short-circuit end in a network node under consideration or for a point in time at which a circuit breaker trips, it is checked that from the stored measured values at each point at the time of the short-circuit occurrence or short-circuit end in a network node under consideration or at the time at which a circuit breaker trips, the slope of the tangent to the times corresponding to the short-circuit entry or the short-circuit end or the time at which a circuit breaker trips the abrupt level exceedance is determined, where appropriate the slopes of the tangents are displayed with the correct sign. 10. The method according to any one of claims 3, 4 or 9, characterized in that when the trigger threshold is exceeded from at least part of the data stored before and after the time of the trigger threshold being exceeded by forming the second derivative over time or by comparing the before, in particular at least one half wave of the current before the time the trigger threshold is exceeded with the data stored after the time the trigger threshold is exceeded or by comparing the Fourier transform of the before, in particular at least one half wave of the current before the time of the Exceeding the trigger threshold stored data with the Fourier transform of the data stored after the time the trigger threshold was exceeded, the time of the short-circuit occurrence or short-circuit end in a considered network node or the time at which a circuit breaker trips, b is determined. 11. The method according to claim 3 or 4, characterized in that to determine a time range comprising the short-circuit occurrence or the short-circuit end in a network node under consideration or the time at which a circuit breaker trips, successive half-waves of the stored current profile 10th 15 20th 25th 30th 35 40 45 50 55 60 65 8th 15 CH 684 660 A5 16 which are correlated in such a way that the first is correlated with the second, then the second with the third, then the third with the fourth etc. half-wave, and that the time range of the short-circuit occurrence is defined as the occurrence time of those successive half-waves whose correlation values are most strongly from deviate from the other correlation values. 12. The method according to claim 3 or 4, characterized in that, beginning at the beginning of each part of the current curve stored before and after the point in time at which the trigger threshold is exceeded, from which the point in time of the short-circuit occurrence or short-circuit end in a considered IM network node or the point in time which a circuit breaker trips, is determined, a time window is subjected to a Fourier analysis, that the time window is gradually shifted over the entire range of the stored time function to form the Fourier analysis, and that the time of the short-circuit occurrence or The end of a short circuit in a network node under consideration or as the time at which a circuit breaker trips, which is defined in the direction of increasing time values, the first edge of that first time window in which the spectrum shows a large proportion of additional frequencies. 13. The method according to claim 3 or 4, characterized in that to determine the point in time of the short-circuit occurrence or short-circuit end in a network node under consideration or the point in time at which a circuit breaker trips, determines the zero crossings from the operating current and the zero crossings to be expected in the range of the short-circuit current and that the half-wave during which a target-actual deviation of the zero crossings takes place as a result of the phase shift that occurs is determined as the time range of the short-circuit entry or short-circuit end in a network node under consideration or the time at which a first circuit breaker opens. 14. The method according to claim 13, characterized in that the signal curve for eliminating the DC voltage component in the short-circuit area is filtered before the evaluation. 15. The method according to any one of claims 9 to 14, characterized in that the half-crossing at the end of the half-wave which contains the short-circuit entry or the short-circuit end in a network node under consideration or the time at which a circuit breaker opens, and the sign of the tangent is inverted in this zero-crossing and used as a current direction criterion. 16. The method according to any one of claims 3 to 15, characterized in that for exact fault location in a parallel line with approximately the same ohmic resistance of the parallel lines at each end of each of the parallel lines a measuring point with a downstream measured value acquisition and evaluation circuit is arranged, in the event of a short circuit from the stored current profiles of the measuring points Total current 11 resp. 12 at the ends of the parallel line and the current 13 in an undisturbed parallel line is determined, and from this the distance R1 of the fault location, based on the line length R3, from the measuring point R1 adjacent to the short circuit according to Kirchhoff's laws with R1 / R3 = (I2 + 2 * I3) / (11 +12) is calculated. 17. Circuit arrangement for carrying out a method according to one of claims 1 to 16, characterized by at least two measuring inputs (4) which via a sample and hold circuit (7) and a common analog multiplexer (8) to the analog input of an A. / D converter (9), the output of which is connected to a memory (10), the sample and hold circuits (7), the analog multiplexer (8) and the A / D converter (9 ) and the memory (10) are connected to a control and computing block (11) via address and control lines and the measurement inputs (4) are scanned simultaneously. 5 10th 15 20th 25th 30th 35 40 45 50 55 60 65 9