Arc detection device with multiple frequency band discrimination function

Abstract

The present invention proposes an arc detection device with a multiple frequency band discrimination function that divides an arc frequency band using multiple band pass filters and detects the occurrence of an arc for each divided frequency band, thereby removing switching noise and power supply noise that mainly occur only in a specific frequency band, thereby improving the accuracy of arc detection, while minimizing the use of expensive analog / digital converters, and increasing the accuracy of arc determination by detecting overlapping bands.To this end, the present invention provides an arc detection device with a multiple frequency band discrimination function that is connected to a line and applies a low pass filter, at least one band pass filter, and a high pass filter to a frequency band in which an arc is expected to occur, thereby dividing the frequency band into multiple detection bands and detecting the arc.

Description

[Technical field]

[0001] The present invention relates to an arc detection device with a multiple frequency band discrimination function, and more particularly to an arc detection device with a multiple frequency band discrimination function that divides the arc frequency band in which an arc occurs into multiple frequency bands and detects an arc for each divided frequency band, thereby minimizing erroneous detection caused by switching noise and inverter noise that are similar to an arc. [Background technology]

[0002] An arc is an intense light and heat that is generated when an electric current flows between two physically separated electrodes through a gas medium.

[0003] Arcs can be broadly classified into series arcs occurring within a single conductor, parallel arcs occurring between two conductors, and ground arcs occurring between ground and a single conductor.

[0004] Here, a series arc is an arc fault that occurs when the connection of a line or contact where current should flow normally becomes weak. Since such a series arc does not form a zero point of current, it is difficult to naturally extinguish the arc, which is highly dangerous, and it is also difficult to detect the fault.

[0005] Series arc faults like this can occur due to aging of the line, poor contact of connectors, etc. In the case of such series arc faults, even though an accident has occurred, it is difficult to detect them with a general overcurrent circuit breaker or earth leakage circuit breaker because the magnitude of the fault current is within the normal operating range. Summary of the Invention [Problem to be solved by the invention]

[0006] An object of the present invention is to provide an arc detection device with a multiple frequency band discrimination function that divides the arc frequency band in which an arc occurs into multiple bands and detects whether the same arc is occurring in each of the divided frequency bands, thereby minimizing erroneous detections caused by switching noise and noise generated during power conversion.

[0007] Another object of the present invention is to provide an arc detection device with a multiple frequency band discrimination function that detects an arc using a bandpass filter and a comparator in the arc frequency band, thereby minimizing the use of expensive ADCs (Analog to Digital Converters) with high processing speeds. [Means for solving the problem]

[0008] According to the present invention, the above-mentioned object is achieved by an arc detection device with a multiple frequency band discrimination function that is connected to a line, applies a low-pass filter, at least one band-pass filter, and a high-pass filter to a frequency band in which an arc is expected to occur, divides the frequency band into a plurality of detection bands, and detects the arc. Effect of the Invention

[0009] According to the present invention, the quality of electricity flowing through a line can be constantly monitored.

[0010] Furthermore, according to the present invention, by dividing the arc frequency band using a plurality of band-pass filters and detecting the presence or absence of an arc for each divided frequency band, switching noise and power supply noise that mainly occur only in specific frequency bands can be removed, thereby improving the accuracy of arc detection.

[0011] The present invention also minimizes the use of expensive analog-to-digital converters that operate at high speeds while applying bandpass for each of multiple frequencies, thereby reducing the manufacturing cost of the arc detection device with the overall multiple frequency band discrimination function, while further improving performance by dividing the arc detection process by frequency band. [Brief description of the drawings]

[0012] [Figure 1] 1 is a block diagram of an arc detection device with multiple frequency band discrimination capability according to an embodiment of the present invention; [Diagram 2] 1 is a reference diagram showing an example in which the frequency bands of a low-pass filter, a band-pass filter, and a high-pass filter are clearly distinguished. [Diagram 3] 1 is a reference diagram of an example in which a pass frequency band of a band-pass filter is configured to overlap with an adjacent frequency band. [Figure 4] FIG. 13 is a reference diagram showing an example in which arc detection is excluded for a frequency band in which switching noise or impulses frequently occur. [Diagram 5] 11 is a diagram illustrating an example of a reference voltage set in a comparator connected to a band-pass filter. FIG. [Figure 6] 11 is a diagram illustrating an example of a reference voltage set in a comparator connected to a band-pass filter. FIG. [Figure 7] FIG. 2 is a diagram showing an example of switching waveforms of a power supply device connected to a line. [Figure 8] FIG. 1 is a diagram showing an example of a signal in which an arc and inverter noise are mixed; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] The arc frequency band referred to in this specification may refer to all frequency bands. However, for the sake of understanding and convenience, this specification will mainly describe an embodiment of 0 Hz to 150 kHz. It should be noted in advance that the arc frequency band is not limited to the embodiments referred to in this specification.

[0014] The band pass filter referred to in this specification may collectively refer to a low pass filter, a band pass filter, and a high pass filter based on the arc frequency band. Unless otherwise specified, it may refer to all of the low pass filter, the band pass filter, and the high pass filter, and when separately mentioned, it may be described separately as a low pass filter or a high pass filter instead of a band pass filter.

[0015] The present invention will now be described in detail with reference to the drawings.

[0016] FIG. 1 is a block diagram of an arc detection device with multiple frequency band discrimination capability according to one embodiment of the present invention.

[0017] Referring to FIG. 1 , an arc detection device with a multiple frequency band discrimination function according to an embodiment may include a current sensor 50, a band detection unit 100, a calculation unit 200, and a circuit breaker 300, and the calculation unit 200 may be composed of a signal processing unit 210 and an arc determination unit 220.

[0018] The current sensor 50 may be configured with a current measuring element such as a shunt resistor, a current transformer (CT) sensor, or a Hall sensor.

[0019] The current sensor 50 is directly connected to one of the pair of lines 10. The current sensor 50 calculates the value of the current flowing through the line 10 by referring to a voltage value generated by a current measuring element. The current sensor 50 can also provide the band detection unit 100 with a voltage value corresponding to the current generated by the current measuring element.

[0020] The band detection unit 100 detects an arc based on a voltage value for a current flowing through the line 10 via the current sensor 50. The band detection unit 100 divides the arc frequency band into a plurality of detection bands, and includes a low-pass filter 110a, band-pass filters 110b to 110n, or a high-pass filter 111) corresponding to each detection band.

[0021] The arc frequency band according to the embodiment can be divided into frequency bands according to the terms described below.

[0022] 1) Low frequency band: 0Hz~40kHz 2) Bandpass filter frequency band: 40khz~100khz 3) High frequency band: 100kHz or higher

[0023] In this embodiment, the low frequency band corresponds to a frequency band in which arcing is most intense, and also corresponds to a frequency band in which switching noise is most prevalent.

[0024] For the low frequency band, a configuration may be made in which a frequency conversion is separately performed using an analog / digital converter and a Fourier transform or a fast Fourier transform to distinguish between arcs and non-arc impulses. A low pass filter 110a is provided in the low frequency band, which will be described later.

[0025] The band of the bandpass filter corresponds to 40 kHz to 100 kHz, and the frequency band can be divided into one, two or more. In Fig. 1, the bandpass filters correspond to reference characters 110b to 110n.

[0026] The high frequency band corresponds to a frequency band of 100 kHz or more, and the high pass filter 111 is responsible for arc detection.

[0027] Arcs that occur on lines occur across the low frequency band, high frequency band and bandpass filter frequency band, but the difference is that unlike switching noise and impulses that only occur in specific frequency bands, arcs occur across the entire arc frequency band.

[0028] On the other hand, the frequency bands of low-pass filter 110a, band-pass filters 110b to 110n, and high-pass filter 111 may be independent of each other or may partially overlap each other. This will be described with reference to Figs. 2 to 4.

[0029] FIG. 2 shows an example in which the frequency bands of the low-pass filter 110a, the band-pass filters 110b to 110n, and the high-pass filter 111 are clearly separated from one another.

[0030] In FIG. 2, it can be seen that the frequency bands are divided so that low-pass filter 110a passes only the frequency band from 0 Hz to 40 kHz, band-pass filters 110b to 110n pass only the frequency band from 40 kHz to 100 kHz, and high-pass filter 111 passes only the frequency band above 100 kHz.

[0031] The success of the method of clearly dividing the arc frequency band by the low-pass filter 110a, the band-pass filters 110b to 110n, and the high-pass filter 111 can be determined depending on the degree of perfection of the filter gain of the band-pass filters. The gain characteristics of the filters will be described with reference to Table 1 below.

[0032] [Table 1]

[0033] As shown in Table 1, the gain of low-pass filter 110a decreases as the frequency increases, and the gain at point a1 is reduced to about half of the maximum value. As a result, when an arc occurs in the frequency band corresponding to point a1 of low-pass filter 110a, the maximum signal value of the generated arc can be received with the value reduced to about half.

[0034] As shown in Table 1, the gain of the band-pass filters 110b to 110n is significantly reduced at frequencies lower than the frequency corresponding to point a2 and higher than the frequency corresponding to point a2'. The band-pass filters 110b to 110n can ensure good reception gain in the frequency band between points a2 and a2', but if an arc occurs at a frequency lower than point a2 or higher than point a2', the signal value of the received arc will be significantly reduced and it may not be determined as an arc.

[0035] As shown in Table 1, the high-pass filter 111 exhibits a gain characteristic in which the gain at frequencies lower than point a3 is lower than the maximum value. If an arc were to occur at a frequency corresponding to point a3 or a point lower than a3, the maximum signal value of the arc that occurs would be determined to be too low, and there is a risk that it would not be identified as an arc.

[0036] The signal gains of the low-pass filter 110a, the band-pass filters 110b to 110n, and the high-pass filter 111 have been described above based on Table 1. To solve this problem, the applicant of the present invention proposes a method for adjusting the frequency bands assigned to the low-pass filter, the high-pass filter, and the band-pass filters 110a to 111 with reference to Figs. 3 and 4.

[0037] With additional reference to FIG. 3, the bandwidth of the filter for the arc frequency band may be implemented in such a way that it overlaps with adjacent frequency bands.

[0038] As shown in FIG. 3, the frequency band may be divided and allocated so that low-pass filter 110a filters the frequency band from 0 Hz to 45 kHz, band-pass filters 110b to 110n filter the frequency band from 35 kHz to 110 kHz, and high-pass filter 111 filters the frequency band of 90 kHz or higher.

[0039] The low-pass filter 110a filters a frequency band of 0 Hz to 45 kHz, which is wider than the frequency band shown in Fig. 2. This has the effect of expanding point a1 shown in Table 1 to the right on the frequency axis, which is the same as widening the frequency band of the low-pass filter.

[0040] High-pass filter 111 is extended to the lower frequency side relative to the frequency band shown in FIG. 2, and passes frequencies of 90 kHz and above, thereby ensuring signal gain for the 100 kHz frequency band that it was originally responsible for.

[0041] In Fig. 3, band 1 and band 2 have overlapping frequency regions. The frequency band of the low-pass filter 110a is responsible for band 1 and overlapping band b1, but the arc frequency band is divided so that the band-pass filters 110b to 110n also receive the overlapping band b1. Similarly, the band-pass filters 110b to 110n are configured to filter the frequency band corresponding to band 2 and the overlapping band b2 of the frequency band that was handled by the high-pass filter 111. As a result, the overlapping bands b1 and b2 are filtered so as to overlap with the filters that process the adjacent frequency bands, and no frequency band is lost in the arc detection, making it possible to determine the arc detection over the entire band.

[0042] On the other hand, the detection band may be configured in such a way that a specific frequency band is excluded, as shown in FIG.

[0043] FIG. 4 is a reference diagram showing an example in which arc detection is excluded for a frequency band in which switching noise and impulses frequently occur.

[0044] 4, the frequency band of 30 kHz to 50 kHz is excluded from the arc detection because, in an electric device to which an arc detection device having a multiple frequency band discrimination function according to the embodiment is applied, unnecessary noise that may be erroneously detected as an arc occurs in the frequency band of 30 kHz to 50 kHz. This is to prevent the arc detection device from making an erroneous detection by excluding the frequency band in which noise caused by the load connected to the line 10 exists from the arc detection.

[0045] The method of excluding the frequency bands shown in FIG. 4 from the detection band is determined according to the noise characteristics of the entire electrical system, and the exemplified frequency bands (30 kHz to 50 kHz) may differ depending on the electrical equipment to which the arc detection device having the multiple frequency band discrimination function according to this embodiment is connected.

[0046] This concludes the description of the characteristics of the low-pass filter 110a, the band-pass filters 110b to 110n, and the high-pass filter 111. The remaining components of the band detection section 100 will now be described.

[0047] First, the low-pass filter 110 a obtains a frequency band corresponding to the low frequency band from the sensed value of the current sensor 50 , and then transmits the obtained value to the arc determination unit 200 .

[0048] The output values ​​of band-pass filters 110b to 110n are individually provided to signal processing unit 210. In Fig. 1, the output value of band-pass filter 110b can be compared with a reference value set by the user in signal processing unit 210, and a comparison result value can be output. Also, the output values ​​of band-pass filters 110b to 110n can be compared with a plurality of reference voltages, rather than with one reference voltage, to generate respective comparison values.

[0049] For the purpose of arc detection, the output value of bandpass filter 110b may be configured to be compared with a reference voltage to determine whether an arc has occurred.

[0050] The signal processing unit 210 may be configured to compare the output value of the band-pass filter 110b with a reference voltage, and when a voltage greater than the reference voltage is applied from the band-pass filter 110b, determine that an arc has occurred and output a logical value of "1".

[0051] Similarly, the signal processing unit 210 can detect the occurrence of an arc by comparing the output value of the band-pass filter 110n with a reference voltage and outputting a logical value of “1” if the output value is greater than the reference voltage as a result of the comparison.

[0052] The same is true for the high-pass filter 111, and the signal processing unit 210 can detect the occurrence of an arc by comparing the output value of the high-pass filter 111 with a reference voltage set in the high-pass filter 111 and outputting a logical value of "0" or "1" depending on the comparison result.

[0053] Furthermore, the signal processing unit 210 can also generate data for evaluating the quality of power flowing through the line 10 in addition to detecting an arc by comparing the output values ​​of the filters 110b to 110n constituting the band detection unit 100 with different reference voltages.

[0054] For example, the signal processing unit 210 can compare the output value of the bandpass filter 110b with a reference voltage 1 to use for arc detection, and can compare the output value of the bandpass filter 110b with reference voltages 2, 3, and n, respectively, to use for measuring the quality of the power passing through the line 10.

[0055] Here, the reference voltage (reference voltage 1) required for arc detection and the reference voltages (reference voltage 2 to reference voltage n) used for power quality measurement may be different voltages, and are preferably set lower than the arc voltage. This will be described with reference to Figs. 5 and 6.

[0056] 5 and 6 show an example of a reference voltage to be compared with the output value of the bandpass filter 110b.

[0057] In FIG. 5, RV1 to RVn are shown as examples of reference voltages of the signal processing unit 210, and the dual reference voltage RV1 is assumed to be used for arc detection. The explanation will be given on the assumption that the voltage passing through the line 10 is direct current (dc).

[0058] The signal processor 210 associates a plurality of reference voltages RV1 to RVn with the direct current (dc) passing through the line 10, and determines the results of comparison with each of the reference voltages RV1 to RVn. If the direct current (dc) exhibits a waveform as shown in Fig. 5, the signal processor 210 outputs a logical value "1" for the reference voltage RV4 at t1, t2, and t3, but outputs a logical value "0" at t4 and t5, and outputs a logical value "1" at t6.

[0059] The fact that logical values ​​of 1, 1, 1, 1, 0, 0, 0, 1 are output as a result of comparison with reference voltage RV4 at regular time intervals on the time axis t means that the quality of the direct current (dc) is not uniform, and even if it is assumed that the current value of the direct current (dc) is constant, it means that the quality of the power flowing through line 10 decreases in proportion to the amount of change in the direct current (dc).

[0060] Next, FIG. 6 shows another example of the waveform of the DC current flowing through the line.

[0061] The waveform in FIG. 6 shows an example that includes power supply noise that occurs when AC is converted to DC.

[0062] Referring to FIG. 6, a DC w1 including power supply noise exceeds a reference voltage RV1 at points P1 and P2, and is detected as a logic value "1" when compared with the reference voltage RV1.

[0063] However, in the comparison with the other reference voltages RV2 to RVn, the DC w1 is not detected because it is small compared to the other reference voltages RV2 to RVn.

[0064] On the other hand, in FIG. 6, the direct current w2 is detected in the signal processing unit 210 as a result of comparison with the reference voltages RV1 to RV3.

[0065] In other words, this means that in DC w2, switching noise with a high maximum value is continuously detected with respect to the expected base voltage ref, indicating that the quality of DC w2 is lower than that of DC w1.

[0066] The quality of the direct current will be described with reference to FIGS.

[0067] Of the two graphs shown in FIG. 7, the upper graph shows the switching waveforms of a power supply device connected to line 10, and the lower graph shows the direct current (DC) output current waveform of the power supply device.

[0068] As shown in Fig. 7, a power supply device that supplies DC to a line 10 has the characteristic of repeatedly generating pulses with a certain magnitude and frequency in the process of switching AC to convert it to DC in a normal state where no arc is generated. Every time the AC is sliced, a triangular current waveform flows into the line 10, and this type of current waveform is characterized by being found in a specific frequency band (mainly occurring in the 100 kHz band). Such a waveform that occurs regularly with a certain magnitude in a specific frequency band does not need to be judged as an arc.

[0069] FIG. 8 shows an example of a signal in which arcs and inverter noise coexist.

[0070] Referring to FIG. 8, it can be seen that arcs occur across the entire arc frequency band, with the amplitude being particularly large in the low frequency band.

[0071] On the other hand, the switching noise generated by the inverter was only observed in the bands of 15 kHz, 33 kHz, and 48 kJZ, and its amplitude was similar to that of the arc.

[0072] The switching noise generated by the inverter differs from that shown in Fig. 6 in that it occurs in three places because the inverter changes the switching period to maintain the output voltage.

[0073] When the quality of the voltage flowing through the line 10, i.e., the quality of the power, changes drastically, the inverter changes the switching noise to pass a constant voltage through the line 10. However, when looking at the graph of Fig. 7, which is defined in terms of the relationship between frequency and amplitude, only three peak amplitudes IV1, IV2, and IV3 are observed, and when defined in terms of the time axis and amplitude, only one of the inverter switching noises IV1, IV2, and IV3 occurs at any given time, and the other two noises do not appear.

[0074] That is, when the band detection unit 100 detects an arc over time for the arc frequency band, only one switching noise is detected, i.e., the inverter switching noise is regarded as one impulse and does not occur in the entire frequency band of the arc frequency band.

[0075] The arc waveform and non-arc waveform flowing through the line 10 have been described above with reference to Figures 7 and 8. The arc determination unit 220 will now be described.

[0076] The arc determination unit 200 compares the filtering result value of the band detection unit 100 with a plurality of reference voltages provided in the signal processing unit 210, receives the comparison result, and performs counting for each of the filters 110a to 111 provided in the band detection unit 100. For example, the output value of the band pass filter 110b may be compared with one reference voltage, but for power quality measurement, it can be compared with two or more reference voltages to perform arc detection and power quality measurement simultaneously.

[0077] The arc determination unit 200 may count the number of times that each of the filters 110a to 111 exceeds a reference voltage applied to each filter, and then determine that an arc has occurred when the counted value exceeds a preset reference value.

[0078] When determining only whether or not an arc has occurred, the arc determination unit 200 may compare the output values ​​output from the filters 110a-111 with a single reference voltage for arc determination, and count the number of times that the logical value "1" has occurred as a result of the comparison, thereby determining whether or not an arc has occurred.

[0079] On the other hand, unlike the band-pass filters 110b to 110n and the high-pass filter 111, the low-pass filter 110a can be subjected to analog / digital conversion by the signal processing unit 210.

[0080] The low-band signal filtered by the low-pass filter 110a may be supplied to the frequency determination unit 241 after being digitally converted through analog / digital conversion.

[0081] The signal processing unit 210 performs an FFT analysis on the current value sensed by the current sensor 50 to first detect an estimated signal predicted as an arc signal based on the magnitude and frequency of the pulse in an arc state, and can determine whether an arc has occurred based on the magnitude and frequency of the pulse generated in a device (e.g., a power supply or an inverter) connected to the line 10 and the estimated signal.

[0082] The reason for performing a separate FFT analysis only on low-pass filter 110a is that 4) there is a possibility that signal processing unit 210, which has a comparison function, may mistakenly determine that inverter noise or noise generated from the power supply is an arc, and 5) there is a concern that noise generated from a load other than the power supply or inverter may be confused with an arc.

[0083] 6) In addition, depending on the load connected to the line 10, noise with large amplitude may be generated periodically. In a structure that simply compares the output value of the low-pass filter 110a with a reference voltage, the noise may be erroneously determined to be an arc.

[0084] The arc determination unit 220 may determine whether an arc has occurred by performing an FFT analysis on the signal that has been FFT-processed by the signal processing unit 210, or may determine whether an arc has occurred by comparing and counting the output values ​​of each of the filters 110a-111 with a reference voltage, or may create power quality data of the DC flowing through the line 10 by applying a plurality of reference voltages to each of the filters 110a-111. Furthermore, when the arc determination unit 220 determines that an arc has occurred on the line 10, it can interrupt the DC flowing through the line 10 by switching on the circuit breaker 300.

[0085] The determination method of the arc determination unit 220 is as follows: 7) a method of determining that an arc has occurred on the line 10 when all of the comparison results of the signal processing unit 210 are identified as an arc detection; 8) a method of determining that an arc has occurred on the line 10 when, as a result of the comparison by the signal processing unit 210, the result value determined to be an arc is greater than the result value determined not to be an arc; and 9) a method of determining that an arc has occurred when, as a result of the comparison by the signal processing unit 210, the result value determined to be an arc is less than the result value determined not to be an arc, but the output value of the low-pass filter 110a is determined to be the occurrence of an arc. 10) a method of determining that an arc has occurred on the line 10, giving priority to the FFT analysis result when the FFT analysis result of the signal processing unit 210 indicates an arc and at least one of the output values ​​of the filters 110a to 111 is determined to be an arc; and 11) a method of determining that an arc has occurred on the line 10, giving priority to the arc determination result of the high-pass filter 111, when it is determined that at least one arc has occurred in the filters 110a to 110n other than the high-pass filter 111.

[0086] Here, item 11) is a method applicable when the arc occurring on the line 10 is mainly in the high band, and item 9) is a method applicable when the arc occurring on the line 10 is mainly in the low band.

[0087] Since the frequency band of the arc may vary depending on the line 10 to which the arc detection device of the embodiment is connected and the load connected to the line 10, it is desirable to detect the arc using one of the above items 7) to 11) that is suitable for the entire electrical system.

[0088] The method according to the embodiment of the present invention described above can be created as a program to be executed by a computer, and the program can be stored in a computer-readable recording medium, examples of which include a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, and an optical data storage device.

[0089] The computer readable recording medium can be distributed among computer systems connected via a network, and the computer readable code can be stored and executed in a distributed manner.Functional programs, codes, and code segments for implementing the above method can be easily construed by a programmer skilled in the art to which the present invention pertains.

[0090] Furthermore, while the preferred embodiments of the present invention have been illustrated and described above, the present invention is in no way limited to the specific embodiments described above, and it goes without saying that various modifications can be made by a person having ordinary knowledge in the technical field to which the invention pertains without departing from the gist of the present invention as claimed in the claims, and such modifications should not be understood individually from the technical ideas and prospects of the present invention.

Claims

1. An arc detection device with a multiple frequency band discrimination function that is connected to a line and applies a low-pass filter, one or more band-pass filters, and a high-pass filter to a frequency band in which an arc is expected to occur, thereby dividing the frequency band into a plurality of detection bands and detecting the arc.

2. a band detection unit that compares the magnitude of the waveform detected through a band pass filter with a preset arc reference value for each of the detection bands; a calculation unit that counts the detection results for each detection band and determines that an arc has occurred when the count value deviates from a preset reference count value; 2. The arc detection device with multiple frequency band discrimination function according to claim 1, comprising:

3. In the band pass filter, 3. The arc detection device with multiple frequency band discrimination function according to claim 2, wherein the filtering frequency band is any one of bands obtained by dividing the arc frequency band, and has at least one region overlapping with an adjacent frequency band.

4. In the band pass filter, 3. The arc detection device with multiple frequency band discrimination function according to claim 2, wherein the filtering frequency band is one of the bands obtained by dividing the arc frequency band and is independent of adjacent frequency bands.

5. The band detection unit 3. The arc detection device with multiple frequency band discrimination capability of claim 2, comprising a combination of two or more of a low-pass filter, a band-pass filter, and a high-pass filter.

6. The calculation unit is 6. The arc detection device with multiple frequency band discrimination function according to claim 5, wherein an output value of the low-pass filter is subjected to analog / digital conversion and then a frequency analysis is performed to compare a pulse size of a predefined arc state with a frequency condition to determine whether an arc has occurred.

7. The calculation unit is comparing each signal output from the band detection unit with a preset voltage value, and outputting the comparison result as a detection count; 3. The arc detection device with multiple frequency band discrimination function according to claim 2, further comprising a signal processing unit for counting the number of detections.

8. The calculation unit is a signal processor that counts arc detection results that exceed a reference value for each detection band; 8. The arc detection device with multiple frequency band discrimination function according to claim 7, further comprising an arc determination unit that determines that the arc has occurred when the number of times the arc has been detected counted by the signal processing unit exceeds a preset reference number of times of detection.

Images