WO2023274503A1

WO2023274503A1 - Locating an arc

Info

Publication number: WO2023274503A1
Application number: PCT/EP2021/067783
Authority: WO
Inventors: Ricardo Herrmann; Peter Schegner; Karsten Wenzlaff
Original assignee: Siemens Aktiengesellschaft; Technische Universität Dresden
Priority date: 2021-06-29
Filing date: 2021-06-29
Publication date: 2023-01-05

Abstract

The invention relates to a method for locating an arc (A) in low-voltage switchgear (20), comprising the following steps: - capturing arrival times (t1, t2, t3) at which an acoustic signal (10S) generated by the arc (A) arrives at each of at least two acoustic sensors (1, 2, 3) spatially spaced apart from one another; - determining at least one propagation time difference (Δt12, Δt13) of the acoustic signal (10S) by forming pairs of differences between two of the captured arrival times (t1, t2, t3) in each case; - determining the position (P_A) of the arc (A) on the basis of the at least one determined propagation time difference (Δt12, Δt13) according to the hyperbolic locating method.

Description

description

Locating an arc

The present invention relates to a method for locating an arc in a low-voltage switchgear, a device for carrying out the method, and the use of a sound-based hyperbolic locating method.

In low-voltage networks, short circuits are usually associated with parallel arcing faults. Particularly in high-performance distribution and switchgear systems, if they are not switched off quickly enough, they can lead to devastating damage to equipment, system components or entire switchgear systems. In order to avoid a long-lasting and large-scale failure of the power supply and to reduce personal injury, it is necessary to detect and extinguish such high-current, parallel accidental arcs in a few milliseconds.

Explicit arc fault detection systems are used to detect arc faults. The conventional arc fault detection systems available on the market consist of several components (multi-component systems) that have to be installed individually at the installation site. For example, fiber optic cables are installed in the areas of the system that are to be protected. The optical fibers capture the light emission generated by an arc and forward the optical signal to a centrally installed detection unit. Based on the evaluation of the optical signal and any other release conditions such as e.g. B. overcurrent, a trip signal for a short-circuiter be ready, which is usually used to extinguish the arc. The short circuiters can generate a short circuit when activated, e.g. B. by detonating an internally installed explosive charge. This means that arc voltage can no longer build up at the arc and the arc is extinguished. On Arc fault detection system is z. B. in WO 2017/050764 A1 (Siemens AG) 30.03.2017 described.

These systems indirectly combine arc fault location and arc fault detection when an evaluation of the sensor addressed is carried out. However, the accuracy of the location is only limited to the detection range of the sensor. An exact position of the arcing point cannot be determined with the existing systems.

The object of the present invention is therefore to provide improved arc location.

According to the invention, this object is achieved by a method according to claim 1, a device according to claim 7 and a use according to claim 9.

The inventive method for locating an arc in a low-voltage switchgear has the following steps: arrival times are recorded at which a sound signal generated by the arc arrives at at least two spatially spaced sound sensors. At least one transit time difference of the sound signal is determined by forming pairs of differences between two of the recorded arrival times. The position of the arc is determined on the basis of the at least one transit time difference determined using the hyperbolic locating method.

The device according to the invention for locating an arc in a low-voltage switchgear has at least two sound sensors that are spatially spaced apart from one another and are designed to detect the arrival times of a sound signal generated by the arc. The device also has a computing unit connected to the sound sensors. The computing unit is for determining at least one transit time difference of the sound signal by forming designed by pairwise differences of two of the recorded arrival times. The arithmetic unit is also designed to determine the position of the arc on the basis of the at least one determined travel time difference according to the hyperbolic locating method.

The use according to the invention of a sound-based hyperbolic locating method is for locating an arc in a low-voltage switchgear.

The locating method according to the invention requires at least two sensors that are synchronized in time in order to determine a transit time difference of a sound signal between two different sensor locations. In addition, the measured values of the two or more sensors must be transmitted to a computing unit, which can calculate a position of the arc on the basis of the difference in transit time measured. For these reasons, the locating method according to the invention can be referred to as a networked locating system.

Low voltage means voltages up to 1000 volts AC or 1500 volts DC. Low voltage means more specifically voltages that are greater than extra-low voltage with values of 50 volts AC or 120 volts DC.

The hyperbolic locating method, also known by the English term TDOA surveillance, is a well-known method for determining position (TDOA=Time Difference of Arrival). This exploits the fact that the propagation time of a signal is proportional to the distance between the transmitter and the receiver. The time difference between two different receivers defines a hyperboloid that indicates the potential location where the signal was sent. Are the recipients of the signal, e.g. B. sensors, and the transmitter of the signal in a common plane, z. B. on the earth's surface, the potential transmitter location is a hyper perbelast on this plane, which is characterized by a hyperbolic Position function (hyberbolal baseline) is defined with a pair of receivers at the foci of the hyperbola. All positions on this function have the same measured runtime difference to the transmitter. The position can then be calculated from the intersections of several of these functions for different pairs of receivers. For example, with three receivers, the 2D location where the signal was sent out is at the intersection of two hyperbolas. A description of the hyperbolic localization method can be found, for example, on pages 5 to 6 of the book Nixdorff, Kurt: Mathematical Methods of Sound Localization in the Atmosphere. - 1st edition - Braunschweig: Friedrich Vieweg & Sohn Verlagsgesellschaft mbH, 1977, ISBN-13 978-3-528-03067-4, https://doi.org/10.1007/97 8-3-322-85933-4_2 .

The time measurement signals at the receiving stations can be generated, for example, via a timer that is synchronized with one another via a synchronization signal that is received periodically. It is not necessary for the individual timers to display the same time at all times, but the time offset between the individual timers must be known.

The invention is based on the finding that the distinctive sound emission of the arc can be used to locate it in a low-voltage switchgear. Sound is better suited for this than the electromagnetic radiation emission of the arc, since sound has a relatively low propagation speed, i. H. significant transit time differences between the sensors arise, and sound propagates better in a switchgear than light, e.g. B. is less absorbed or shielded by installations and partitions of the low-voltage switchgear.

With the invention, it is possible to locate an arc in a switchgear with an accuracy of a few centimeters, and thus more precisely than with conventional optical detection systems. With the invention, parallel, high-current accidental arcing conditions in low-voltage switchgear and distribution systems can be located by a locating system that does not have to meet any time-related requirements for an accidental arcing protection system. The locating system consists of several sensors that evaluate non-electrical signals that are typical for an arc and determine the position of the arc by evaluating transit times.

The use of a positioning system results in the following advantages:

- With the locating system, a high degree of localization accuracy can be achieved: an arc can be located in a switchgear with an accuracy of a few centimetres.

- The locating system leads to lower installation costs: A complex installation of fiber optic cables in the systems under protection requirements, which is necessary for a conventional locating system, is not necessary.

- A central arc fault detection based on current and voltage measurement forms the protective function in the switchgear. The locating system locates an arc autonomously, i. H. independent of the detection system. The results of the arc location are only evaluated if the detection system has detected an arc. Since the detection of the arcs is not based on the locating system, the locating system cannot trigger an arc detection erroneously.

Advantageous refinements and developments of the inventions are specified in the dependent claims. The method according to the invention can also be developed in accordance with the dependent device claims, and vice versa. In an advantageous embodiment of the invention, the sound signal is an ultrasonic signal. This has the particular advantage that the signal emitted by the arc in the ultrasonic range is less affected by other noises, e.g. B.von egg nem circuit breaker, superimposed than in another acoustic range, z. B. in the listening area.

In an advantageous embodiment of the invention, the method has the following steps: At least one partial area of the low-voltage switchgear is formed. And at least two spatially spaced sound sensors are assigned to each sub-area. The term sub-area includes both a 2D and a 3D sub-area (=partial volume). It is therefore possible, please include various units of the low-voltage switchgear (panel, control cabinet, switchgear, withdrawable unit) to monitor with sensors. This has the particular advantage that an arc can be located in smaller areas, so that it can be located more easily and precisely.

In an advantageous embodiment of the invention, the sound sensors are arranged at the edge of the partial area. This has the particular advantage that the partial area delimited by the sensors can be monitored for the existence of an arc. The accuracy of the location depends on the area covered by the sensors. It is advantageous to place the sensors in the corners.

At least two sensors are preferably used to determine the position of an arc on a one-dimensional curve and at least one additional sensor spaced apart from the other sensors for each additional dimension. For example, if the position of an arc is only to be determined on a curve (1D measurement, one-dimensional measurement), it is sufficient to use two sensors and derive the position of the arc from the two arrival times of the signal at the two sensors. For each additional dimension of An additional sensor is required to determine the position. The position of the arc on a general spatial area can be determined by at least three sensors (2D measurement, two-dimensional measurement), while at least four sensors are required to determine the position in space, which must not lie in one plane.

In an advantageous embodiment of the invention, three sound sensors are used for locating. Only two coordinates of the arc in the low-voltage switchgear are determined using the hyperbolic locating method as the intersection curve of two hyperboloids, but the third coordinate is based on additional information about the low-voltage switchgear. Additional information can be the physical dimensions (height, width, depth) of the low-voltage switchgear and the routing of cables in the low-voltage switchgear. This has the particular advantage that the required number of sensors can be reduced.

In an advantageous embodiment of the invention, two sound sensors are used for locating. In this case, only one coordinate of the arc in the low-voltage switchgear is determined by the hyperbolic locating method, but the second and third coordinates are determined on the basis of further information about the low-voltage switchgear. Additional information can be the physical dimensions (height, width, depth) of the low-voltage switchgear, the position and dimensions of parts of the low-voltage switchgear and the routing of lines in the low-voltage switchgear. If a low-voltage switchgear has partial areas that are sealed off from one another, two sensors can be arranged in each of the partial areas, e.g. B. each only above and below. This has the special advantage that the required number of sensors can be reduced. In an advantageous embodiment of the invention, the low-voltage switchgear has at least one sub-area to which at least two spatially spaced sound sensors are assigned, with the sound sensors being arranged at the edge of the sub-area. This has the particular advantage that an arc is located in smaller areas, so that it can be located more easily and more precisely.

In an advantageous embodiment of the invention, two spatially spaced acoustic sensors are assigned to a partial area, the acoustic sensors being arranged at the edge of the partial area. This has the particular advantage that the partial area delimited by the sensors can be monitored for the existence of an arc.

In an advantageous embodiment of the invention, a method for locating an arc in a low-voltage switchgear according to one of claims 1 to 6 is combined with a method for detecting the arc on the basis of voltage values and current values in the low-voltage switchgear. Similarly, in an advantageous embodiment of the invention, a device for locating an arc in a low-voltage switchgear is combined with a device for detecting the arc in the low-voltage switchgear on the basis of voltage values and current values. These combinations have the special advantage that no special protection-relevant requirements have to be observed for the locating system because the locating system is not part of the arc protection system. The locating system is both a good supplement for rapid, central arc fault detection and can also be used as an alternative to classic arc fault detection systems that have not previously had explicit locating. The location system can be evaluated centrally in combination with an arc detection system or work as an independent system. A connection of the locating system to an arc detection system can be done wirelessly or via a bus system. With regard to the transmission times, there are no special time requirements, since the signal is not required for detection and is therefore required as part of the protective function for arc fault protection.

The combinations mentioned above therefore have two aspects: On the one hand, such a combination has a protective and a location function for low-voltage switchgear, because part of the low-voltage switchgear can be monitored by an arc detection system and an arc location system, e.g. B. in each sub-area a photodiode and an additional positioning system. The advantage of this is that the protection system can be made very compact, e.g. B. by measuring current and / or voltage, and the locating system is not part of the protection system. On the other hand, an existing arc detection system in a low-voltage switchgear can be supplemented with an arc detection system.

A computer program product is also proposed which can be loaded directly into the internal memory of a digital computing device and comprises software code sections which are used to carry out the steps of the method described herein when the product is run on the computing device. The computer program product can be stored on a data carrier such as a USB memory stick, a DVD or a CD-ROM, a flash memory, EEPROM or an SD card. The computer program product may also be in the form of a signal loadable over a wired or wireless network.

The method is preferably implemented in the form of a computer program for automatic execution. The invention is thus on the one hand also a computer program with program code instructions that can be executed by a computer and on the other hand a storage medium with such a computer program, i.e. a computer program product with program code stuffs, and finally also a low-voltage switchgear, in whose memory such a computer program is loaded or loadable as a means for carrying out the procedural procedure and its refinements.

If method steps or method steps are described below, this refers to actions that take place due to the computer program or under the control of the computer program, unless it is expressly stated that individual actions are initiated by a user of the computer program. At a minimum, any use of the term "automatic" means that the action in question is due to or under the control of the computer program.

Instead of a computer program with individual program code instructions, the method described here and below can also be implemented in the form of firmware. It is clear to the person skilled in the art that instead of an implementation of a method in software, an implementation in firmware or in Firm - and software or in firmware and hardware is possible. It should therefore apply to the description presented here that the term software or the term computer program also includes other implementation options, namely in particular an implementation in firmware or in firmware and software or in firmware and hardware.

The properties, features and advantages of this invention described above, and the manner in which they are achieved, will become clearer and more clearly understood from the following description of the exemplary embodiments, which are illustrated in more detail with reference to the drawings. It shows in each case schematically and not true to scale

1 shows the time profile of measured variables that are recorded during the ignition of an arc;

2 shows a first switchgear in an oblique view; 3 shows the first switchgear of FIG. 2 in a front view; 4 shows a second switchgear in an oblique view; FIG. 5 shows the second switchgear of FIG. 4 in a front view; 6 shows a hyperbola locating method with three sensors in a switchgear;

7 shows a switch cabinet; 8 shows a method for locating hyperbola with two sensors in a switchgear;

9 shows a lateration method; FIG. 10 shows the time profile of ultrasonic and UV measured variables that are recorded during the ignition of an arc;

11 shows a lateration method with two detector units in a switchgear;

12 shows a lateration method with a detector unit in a switchgear;

13 shows a cross correlation; 14 shows an envelope curve enveloping a sound signal; and FIG. 15 shows a threshold method.

An arc quickly reaches temperatures in the range of a few 10,000 K. An arc therefore has intensive electromagnetic radiation. In addition, due to the high temperature, the air surrounding the arc expands quickly, which can be perceived as noise emitted by the arc.

FIG. 1 shows the time profile of various measured variables that were recorded during and up to approximately 29 ms after the ignition of an arc; the arc ignites at time t=0.

Graphic a shows the voltage U _LB across an arc and the current i _LB through the arc.

Graph b shows relative voltage values of a sound sensor S _s and an ultrasonic sensor Sus _/ detecting sound generated by the arc; the measured voltage Voltage values u _s of the sensors normalized to the amount of the maximum voltage value |u _{s | max} .

Graph c shows relative voltage values of an IR sensor SIR, a VIS sensor S _ViS and a UV sensor Suv detecting radiation generated by the arc; the measured voltage values u _s of the sensors are normalized to the magnitude of the maximum voltage value |us _{| max} .

This property of an arc, the emission of electromagnetic radiation and sound waves that begins after it is ignited, can be used to locate the arc.

2 shows a first cuboid switchgear 20 in an oblique view. The switchgear 20 has a floor 24, two side walls 21, 22, a ceiling and a rear wall 26.

The front wall 25 has been temporarily removed to allow access to the interior. The bottom 24 and the top 23 are parallel to the x-z plane, the two side walls 21, 22 are parallel to the y-z plane and the rear wall 26 and the front wall 25 are parallel to the x-y plane.

At the two ends of the front edge of the floor 24 and at the two ends of the left side edge of the ceiling 23, a sound sensor 1 to 4 was positioned, which is designed to ultrasonic signals, which are generated by an arc gene in the interior of the switchgear 20, to recieve. The sound sensors 1 to 4 have mutually synchronized clocks. The sound sensors 1 to 4 are each connected to a computing unit 35 via signal lines 34; thus, the sound sensors 1 to 4 can transmit the arrival times t1 to t4 of an ultrasonic signal, which was generated by an arc in the interior of the switchgear 20, to the computing unit 35. The arithmetic unit can calculate three independent runtime differences based on the four arrival times tl to t4 and from this according to the Hyperbolic location method, the 3D coordinates P _A (x _A , y _A , z _A ) of the arc in the interior of the switchgear 20 determine.

FIG. 3 shows the first switching system 20 shown in FIG. 2 in a front view. In this switchgear 20, the z-coordinate z _A of the position P _A of the arc is not essential for locating the arc, because the dimension of the switchgear 20 along the z-axis is relatively small and the possible positions of an arc in z Direction be limited, z. B. because there is only one current-carrying conductor in the z-dimension. It is therefore sufficient for locating the arc to merely determine the 2D coordinates P _A (x _A , y _A ) of the arc in the interior of the switchgear 20 .

At the two ends of the front edge of the floor 24 and at the front end of the left side edge of the ceiling 23, a sound sensor 1 to 3 was positioned, which is designed to ultrasonic signals, which are generated by an arc gene in the interior of the switchgear 20, to recieve. The sound sensors 1 to 3 have mutually synchronized clocks. The sound sensors 1 to 3 are each connected to a computing unit 35 via signal lines 34; thus, the sound sensors 1 to 3 can transmit the arrival times t1 to t3 of an ultrasonic signal, which was generated by an arc in the interior of the switchgear 20, to the computing unit 35. Based on the three arrival times t1 to t3, the computing unit can calculate two independent transit time differences and use this to determine the 2D coordinates P _A (x _A , y _A ) of the arc in the interior of the switchgear 20 using the hyperbola locating method.

4 shows a second cuboid switchgear 20 in an oblique view, which is designed similarly to the switchgear 20 shown in FIG the Subdivide switchgear 20 into three switch cabinets 28.1, 28.2 and 28.3.

A sound sensor 1 to 4 was positioned at each of the two ends of the front edge of the floor 24.1 of the left control cabinet 28.1 and at the two ends of the left side edge of the ceiling 23 of the left control cabinet 28.1 gene generated in the interior of the switchgear 20 to receive. The sound sensors 1 to 4 have mutually synchronized clocks. The sound sensors 1 to 4 are each connected via signal lines to a computing unit (not shown); thus the sound sensors 1 to 4 can transmit the arrival times t1 to t4 of an ultrasonic signal, which was generated by an arc in the interior of the left switch cabinet 28.1 of the switchgear 20, to the computing unit 35. Based on the four arrival times tl to t4, the computing unit can calculate three independent transit time differences and, based on the hyperbolic location method, the 3D coordinates P _A (x _A , y _A , z _A ) of the arc in the interior of the left control cabinet 28.1 of the switchgear 20.

FIG. 5 shows the second switching system 20 shown in FIG. 3 in a front view. In this switchgear 20, for the same reasons as in the _switchgear _shown in _FIG .

FIG. 6 illustrates a 2D hyperbola location method in a switchgear 20 based on three sound sensors 1, 2 and 3, which are positioned on the components of the switchgear 20 delimiting the interior. A Cartesian coordinate system (x, y) is introduced to determine the location, the position of which can be defined arbitrarily. The length li between a sensor i and the arc position P _A results from equation (1) from the transit time Ati of the sound propagating with the sound speed v _s , ie the difference between the time tO at which the arc ignites and the arrival time ti at which the sound emitted by the arc at the arc position P _A arrives at the sensor i (i =

1, 2, 3). li = Ati vs = (ti - _tO ) _vs (1)

Since the time ^tO is not known, it is eliminated by considering the length differences between the sensors:

Ali _j = li - l _j = [(ti - tO) v _s ] - [(tj - tO) v _s ] =

= (ti - tj) v _s = Ati _j v _s

The time difference Ati _j between a pair of sensors, known in English as TDOA (= Time Difference of Arrival), is measured. The lengths li can be replaced with the arc coordinates (x _A , y _A ) and the known sensor positions (xi, yi) using the Pythagoras theorem. The resulting equation (2), from the sensor combination i and j, corresponds to a hyperbolic equation. With the three sensors 1, 2 and 3, two independent hyperbola equations can have the following form:

be set up. Since Ati _j , v _s and the sensor coordinates (xi, yi) and (X _j , y _j ) are known, the desired arc coordinates (x _A , y _A ) can be calculated from them. In FIG. 5, this corresponds geometrically to the determination of the 2D position P _A (x _A , y _A ) of the arc A as the intersection of the two hyperbolic branches h 2 and ^hi3 , which are described mathematically by hyperbolic equations as in equation (2). .

For a three-dimensional localization of an arc A, Equation (2) must be extended by a z-coordinate. Based on four arrival times tl to t4 of a sound signal on four sensors, three independent hyperbolic equations can have the following form:

set up and the required arc coordinates (x _A , y _A , z _{A )} are calculated. In geometric terms, this corresponds to the determination of the 3D position P _{A (} x _A , y _A , z _{A )} of arc A as the intersection of three hyperboloids, which are described by three hyperbolic equations according to equation (3).

The following table provides an overview of the boundary conditions of the hyperbola localization method:

Where d denotes the number of dimensions of the space in which the location is located, e.g. For example, if d = 2 for a 2D room (= plane) and d = 3 for a 3D room, and m the number of sound sensors, the following must be true: m > d + 1 In other words: In order to determine d coordinates of the position P _A of the arc, at least d+1 sensors must each measure the arrival time of a sound signal from the arc.

FIG. 7 shows a switch cabinet 28.1, which, as outlined in FIGS. 4 and 5, can be part of a larger switchgear 20. FIG. Fastening elements 32, 33 are arranged in the interior of the control cabinet 28.1, which divide the interior into smaller sub-spaces: horizontally running fastening elements 32 separate inserts 29 arranged one above the other in the right part of the interior, while a vertical fastening element 33 one in the left Part of the interior of the vertical channel 31 power distribution channel from the racks 29 arranged to the right of it separates. The current distribution channel 31 can have any width; he can e.g. B. have a width of 60 to 80 mm. A busbar 36 runs in the vertical direction in the power distribution channel 31 , via which electrical energy is brought to different electrical equipment 30 which are arranged in the racks 29 . "Operating resources" is a collective term for all devices that can be installed in the switch cabinet. To detect arcs on the busbar 36, three sound sensors 1, 2, 3 are arranged on the floor and on the ceiling of the power distribution channel 31.

8 illustrates an ID hyperbolic localization method in a control cabinet 28.1, which is constructed similarly to the control cabinet shown in FIG Height H of the cabinet extending power rail 36 are not shown equipment electrically powered, which are arranged in six horizontally one above the other on parent racks 29.1 to 29.6. Here are each at the lower and the upper end of the left outer wall 21, each at a horizontal distance B from the to the middle of the cabinet arranged busbar 36, sound sensors 2, 3 arranged, which can measure the arrival times of a burning arc at the position P _A emitted sound signal. Since an arc in the current distribution duct 31 must lie on the busbar 36, two sound sensors 2, 3 are sufficient for precise location of the arc h ₂ 3(l); the position P _A( 1) of the arc is therefore defined as the intersection of the first hyperload h _{23 (} 1) with the busbar 36; the arc may have caused damage, especially in the second slot 29.2, which covers an area of the busbar 36 in which the arc is located at position P _A . Similarly, a position P _{A (} 2) of a second arc can be determined.

FIG. 9 visualizes the property of an arc A positioned at a position P _A in a 2D representation (xy coordinate system), both electromagnetic radiation and electromagnetic waves UV, e.g. B. ultraviolet light, as well as sound waves US, z. B. ultrasound send the. In this case, an arcing event, e.g. B. an ignition of the arc, the arc signal 10 generated is noticeable to an observer both in the form of a sound signal 10S and in the form of an electromagnetic radiation signal 10EM, with the two different types of signals 10S, 10EM being assigned to the same arc signal 10 and therefore being emitted at the same time tO but arrive at an observer or a sensor S at significantly different arrival times ti or t _EM,i . The cause of these different arrival times is the well-known fact that sound waves in air and electromagnetic waves have a propagation speed that differs by a factor of about 10 ⁶ : the speed of sound v _s is approx. 343 m/s, the speed of light c is approx. 3 x 10 ⁸ m/s. This different propagation speed is shown in FIG. 9 by the arrows of different lengths for the electromagnetic wave UV and the sound wave. le US indicated. Both types of waves 10EM and 10S reach a detector unit 38, which has both a sound sensor Su _s/ suitable for detecting the sound signal 10S, e.g. B. Ult Schnellall, as well as a light sensor Suv, suitable for Detek tion of the electromagnetic radiation signal 10EM, z. B. UV light has. A microphone or an ultrasonic sensor can be used as the sound sensor Sus. A UV, light or IR sensor can also be used to detect the radiation. In addition, the detector unit 38 has a computing unit 35, for example a microcontroller, for the electronic processing of the sensor signals received from the two sensors Sus and Suv.

To locate an arc A, use can be made of the above-mentioned fact that the electromagnetic waves 10EM and sound waves 10S emitted by an arc A propagate at very different speeds. A dynamic threshold value can be used for the evaluation the maximum modulation during the event. FIG. 10 shows an example of an evaluation of the transit time difference between a UV signal UV and an ultrasonic signal US by a detector unit, as shown in FIG. After the arc is ignited, a strong UV signal UV reaches the UV sensor after approx. 0.2 ms, whereas the sound signal at the US sensor only takes approx.

3 ms delayed compared to the UV signal. The modulation in % is shown on the y-axis in the upper window. Due to the high propagation speed of the electromagnetic signal UV, it is justified to equate the arrival time of the electromagnetic signal UV with the ignition time tO of the arc, i. H. the point in time at which the arc emits both the electromagnetic signal and the acoustic signal US.

FIG. 11 illustrates a 2D latation method in a switchgear 20 based on two detector units 38, 39, each comprising a sound sensor and an electromagnetic radiation sensor, which are attached to the interior space delimiting Components of the switchgear 20 are positioned. The distinctive progression of sound and radiation is used to determine the distance to the arc; this uses the fact that the radiation propagates around six powers of ten faster than the sound. Exceeding a limit value on the radiation sensor is defined as the point in time tO at which the arc ignites. The transit time Ati of the sound signal is determined from the time ti at which a limit value on the sound sensor is exceeded and the time tO. With the known speed of sound v _s , the propagation time Ati to a distance lium can be calculated according to equation (1). The distance li determined with a sensor i can already severely restrict the location where the arc burned. If the position of the arc is to be determined even more precisely, the distances of several sensors can be combined. For this purpose, a Cartesian coordinate system (x, y) is introduced, the position of which can be freely defined.

The length li between a sensor i and the arc can be replaced with the arc coordinates (x _A , y _A ) and the known sensor positions (xi, yi) using the Pythagorean theorem. The resulting equation (4):

(ti - to)

Ati

corresponds to a circle equation. With the two sensors 1 and 2, two such independent circuit equations can be set up. Since Ati _j , v _s and the sensor coordinates are known, the desired arc coordinates (x _A , y _A ) can be calculated from them; there are generally two equivalent solutions, since the arc position can be symmetrical to the line connecting the sensors. In FIG. 11, this corresponds geometrically to the determination of the 2D positions P _A (x _A , y _A ) and P' _A (X' _A , y' _A ) of the arc A as the intersection points of the two circles ki and k2, which mathematically described by circle equations as in equation (4) be practiced To determine at which of the two mathematically equivalent 2D positions, intersection P _{A (} x _A , y _A) and intersection P ' _A (X' _{A /} y' _A) , the arc A burns, must sen either more information about the switchgear 20 such. B. the wiring, or a runtime measurement must be carried out to a third sensor. If the sensors were placed at the edge of the surveillance area, one of the two intersection points is outside the surveillance area.

For a three-dimensional localization of an arc A, Equation (4) must be extended by a z-coordinate. Based on three sound propagation times Ati of a sound signal at three sensors, three independent spherical equations of the following form can be calculated:

are set up and the arc coordinates sought (x _A , y _A , z _{A )} are calculated therefrom; here, too, there are several mathematically equivalent solutions. Geometrically this corresponds to the determination of the 3D position P _{A (} x _A , y _A , z _{A )} of the arc A as the intersection of three spheres, which are described by three sphere equations according to Equation (5).

The following table provides an overview of the boundary conditions of the lateration method:

12 illustrates an ID latation method in a switch cabinet 28.1, which is constructed similarly to the switch cabinet shown in FIG Switch cabinet extending busbar 36 are not shown resources electrically ver provides that are arranged in six horizontally stacked A drawers 29.1 to 29.6 are arranged. A detector unit 38 with a sound and an electromagnetic radiation sensor is arranged at the lower end of the left-hand outer wall 21, at a horizontal distance B from the current rail 36 arranged towards the middle of the cabinet Can measure position P _A burning arc emitted sound signal and an electromagnetic radiation signal. Since an arc in the current distribution channel 31 must lie on the conductor rail 36, a detector unit 38 is sufficient for precise localization of the arc. the position P _A of the arc is therefore defined as the intersection of the circle k with the busbar 36; the arc has therefore eventual damage especially in the fifth drawer 29.5 caused, which covers an area of the busbar 36, in which the position P _A of the arc is gens.

FIGS. 13 to 15 relate to the determination of transit time differences in the hyperbola location method. For the hyperbe location method, the transit time differences between two or more sensors must be determined; a cross-correlation, in particular a cross-correlation of an envelope curve of signal values or a threshold value method can be used for this purpose. Thresholding is less computationally intensive than cross-correlation and will thus often be preferred. In the case of low-current arcs, e.g. B. Longitudinal arcs, there is no significant increase in noise; Instead, a sound intensity profile specific to the event is created, the difference in transit time between two or more sensors can be analyzed more precisely by means of cross-correlation of the envelope curve than with a threshold value method. This means that high-current arcing faults can be evaluated using the simpler threshold value method, while low-current or heavily shadowed arcing faults can be located more precisely using the cross-correlation of the envelope curve.

FIG. 13 illustrates a cross correlation. Graphic a shows the original measurement curves ui and U2 of a sound signal recorded by two sensors 1, 2. For a cross-correlation, the measurement curves ui and U2 are first smoothed by forming their magnitude and including a median filter with a window width of z. B. 1000 values is applied. The one with the equation

As can be seen in graphic c, the cross-correlation R12 determined has a clear maximum at t _diff = -3.165 ms, which corresponds to the runtime difference between the two sensors 1 and 2. Graph b shows the th measurement curves ui and U2 of sensors 1 and 2, measurement curve U2 recorded by sensor 2 being shifted by Δt=−3.165 ms in relation to measurement curve ui recorded by sensor 1. It can be seen that the courses of the two measurement curves Ui and U2 are very similar when they are shifted relative to one another by the transit time difference Δt = -3.165 ms.

FIG. 14 illustrates an envelope curve HK (dashed line) for a sound signal 10S (solid line). An envelope curve of a signal follows the progression of the amplitude of the sound signal 10S which changes over time (t-axis) and is plotted along the y-axis. So the envelope is like a filter that filters out changes in the sound signal on small time scales and only keeps changes in the sound signal on larger time scales. The envelope curve, as the envelope of the fine structure of a sound signal, essentially reflects the course of the sound level.

FIG. 15 illustrates a threshold method. Graphic a shows the measurement curves ui and U2 of a sound signal recorded by two sensors 1, 2. To determine the transit time difference Δt, a threshold value comparison is applied to the measurement curves ui and U2 of both sensors 1 and 2. The point in time at which the absolute value of the measurement signal u _s( t) exceeds a fixed threshold value +u* or -u* for the first time is used as At, see graphic a:

or exceeds a percentage threshold value p* of the maximum modulation Iu _s I _max for the first time:

Graphic b shows the function F(u _s ,t) for the two measurement curves Ui and U2. The time difference between the two jump points at which the functions F(ui,t) and F(u2,t) jump from 0 to 1 , corresponds to the transit time difference At.

Claims

patent claims

1. A method for locating an arc (A) in low voltage switchgear (20), comprising the following steps:

- detecting arrival times (t1, t2, t3) at which a sound signal (10S) generated by the arc (A) arrives at at least two spatially spaced sound sensors (1, 2, 3);

- Determining at least one transit time difference (Atl2, Atl3) of the sound signal (10S) by forming pairs of differences of two of the detected arrival times (tl, t2, t3);

- Determining the position (P_A) of the arc (A) on the basis of the at least one determined transit time difference (Atl2, Atl3) according to the hyperbolic locating method.

2. The method according to claim 1, wherein the acoustic signal (10S) is an ultrasonic signal.

3. The method according to any one of claims 1 and 2, with the following steps:

forming at least one sub-area (28.1, 28.2, 28.3) of the low-voltage switchgear (20);

Assigning at least two spatially spaced sound sensors (1, 2, 3) to each sub-area (28.1,

28.2, 28.3).

4. The method according to claim 3, wherein the sound sensors (1, 2, 3) are arranged at the edge of the partial area (28.1, 28.2, 28.3).

5. The method according to any one of the preceding claims, wherein three sound sensors (1, 2, 3) are used for locating, with only two coordinates (xA, yA) of the arc (A) in the low-voltage switchgear (20) by the hyperbolic locating method as Intersection curve of two hyperboloids can be determined, but the third coordinate (zA) on basis further information about the low-voltage switchgear (20) is determined.

6. The method according to any one of claims 1 to 4, wherein two sound sensors (1, 2, 3) are used for locating, only one coordinate (xA) of the arc (A) in the low-voltage switchgear (20) being determined by the hyperbolic locating method is, but the second and the third coordinate (yA, zA) are determined on the basis of further information about the low-voltage switchgear (20).

7. Device for locating an arc (A) in a low-voltage switchgear (20), having:

- At least two spatially spaced sound sensors (1, 2, 3) for detecting arrival times (t1, t2, t3) of a sound signal (10S) generated by the arc (A);

- A computing unit (35) connected to the sound sensors (1, 2, 3), which is designed to determine at least one transit time difference (Atl2, Atl3) of the sound signal (10S) by forming pairs of differences between two of the recorded arrival times (Tl, t2, t3) and for determining the position (P_A) of the arc (A) on the basis of the at least one determined travel time difference (Atl2, Atl3) according to the hyperbolic locating method.

8. The device according to claim 7, wherein the low-voltage switchgear (20) has at least one partial area (28.1, 28.2, 28.3) to which at least two spatially spaced sound sensors (1, 2, 3) are assigned, the sound sensors (1, 2, 3) are arranged at the edge of the partial area (28.1, 28.2, 28.3).

9. The device according to claim 8, wherein two spatially spaced sound sensors (1, 2, 3) are assigned to a partial area (28.1, 28.2, 28.3), the sound sensors (1, 2, 3) being located at the edge of the partial area (28.1, 28.2,

28.3) are arranged. 10. Using a sound-based hyperbolic locating method to locate an arc (A) in a low-voltage switchgear (20).

11. Combination of a method for locating an arc (A) in a low-voltage switchgear (20) according to one of Claims 1 to 6 with a method for detecting the arc (A) on the basis of voltage values and current values in the low-voltage switchgear (20 ).

12. Combination of a device for locating an arc (A) in a low-voltage switchgear (20) according to one of claims 7 to 9 with a device for detecting the arc (A) in the low-voltage switchgear (20) on the basis of voltage values and current values.