JP2009300158A - Insulation monitoring device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To perform highly accurate insulation monitoring at a short time interval by using a monitoring signal having comparatively small power. <P>SOLUTION: A monitoring signal having a frequency different from a commercial frequency is injected into a class B grounding conductor 4 of a transformer 1, and a leakage current flowing back into the class B grounding conductor through a low voltage side electric path and the ground is detected, and an unnecessary component including the commercial frequency is removed from the leakage current, to thereby detect a measurement signal equivalent to a monitoring signal component of the leakage current. A reference signal equivalent to a monitoring signal injected from a reference input from the class B grounding conductor based on a class D grounding point of the low voltage side electric path is detected, and a ground insulation resistance component included in the measurement signal is extracted successively with high frequency resolution and time resolution, by DFT operation using a measurement signal of each reference signal n(>m) cycle portion shifted as much as each reference signal m cycle portion, which is synchronized with a prescribed phase of the reference signal. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an insulation monitoring apparatus, and more particularly to an insulation monitoring apparatus that monitors ground insulation resistance in a low-voltage side circuit of a power receiving transformer using a leakage current that circulates to a class B ground line through the circuit and the ground.

  Various loads, such as factory machinery or personal household computers, are connected to the low-voltage side circuit (100V, 200V, 600V) of the power receiving transformer. Therefore, it is necessary to prevent electrical leakage accidents.

  FIG. 9 is a diagram for explaining the prior art, and shows a schematic configuration of a conventional insulation monitoring device (Patent Documents 1 and 2). In the figure, a load 2 is connected to the low voltage side electric circuit of the power receiving transformer 1, and a first electric circuit 3 is grounded via a class B grounding wire 4, and a second electric circuit 5 is connected to the ground. Between the two, a ground impedance Z0 including a ground insulation resistance R0 and a ground electrostatic (floating) capacitance C0 is provided. On the low voltage side of such a transformer 1, a leakage current based on the ground impedance Z 0 and the ground resistance flows through the second electric circuit 5, the ground impedance Z 0, the ground, and the B-type ground line 4. The conventional insulation monitoring apparatus 200 is provided on the B type ground line 4 of such a low voltage side electric circuit.

  The insulation monitoring apparatus 200 generates a monitoring signal having a frequency different from the commercial frequency, and injects the monitoring signal into the B-type ground line 4 via the superimposing transformer 30, and the D-type grounding point ED of the low-voltage side electric circuit. And a reference signal detector 202 for detecting a reference signal corresponding to the monitoring signal detected from the B-type grounding wire 4 with respect to the B-type grounding wire 4, and the B-type grounding wire 4 via the low-voltage side electric circuit and the ground of the transformer 1 A measurement signal detection unit 203 that detects a leakage current that circulates, removes unnecessary components such as commercial frequencies and harmonics from the detection output, and detects a measurement signal corresponding to a monitoring signal component of the leakage current, and the reference And an arithmetic processing unit 204 for obtaining a ground insulation resistance component of a leakage current included in the measurement signal in phase synchronization with the signal, and a ground insulation resistance of the obtained measurement signal (that is, a monitoring signal component of the leakage current). It monitors the electric leakage by continuously monitoring component.

  In such a configuration, it is preferable that the ground impedance Z0 is relatively large and normal, but when insulation deterioration starts for some reason, the leakage current of the commercial component circulating to the B-type grounding wire 4 through the ground increases. Since a large commercial voltage (about several tens of volts) appears on the B-type grounding line 4 with the D-type grounding point ED as a reference, it is difficult to accurately detect the reference signal. As a result, the detection of the ground insulation resistance component included in the measurement signal also becomes unstable and inaccurate.

For this reason, conventionally, a low frequency signal (for example, 20 Hz, 0.5 V) with relatively large power is injected as a monitoring signal, and is included in the measurement signal by DFT calculation with a relatively coarse frequency resolution (for example, 2.5 Hz). The ground insulation resistance component was extracted.
JP 2005-181148 A Japanese Patent No. 3043278

  However, injecting a comparatively large power monitoring signal into the B-type ground line from the normal time is not preferable for not only the load device 2 but also peripheral devices. In addition, since the noise generated when the load device is turned ON / OFF includes not only a frequency component close to the frequency of the monitoring signal (for example, 20 Hz), such noise is completely eliminated by the filter means. If the frequency resolution of the DFT operation is as coarse as 2.5 Hz, for example, the frequency components obtained by the DFT operation near the center frequency (20 Hz) of the monitoring signal are 17.5 Hz, 20 Hz, Since the noise component close to the frequency of the monitoring signal such as 19 Hz or 21 Hz is included in the calculation result of 20 Hz, for example, sufficient S / N cannot be obtained.

  The present invention has been made in view of the above-described problems of the prior art, and an object of the present invention is to provide an insulation monitoring apparatus that can perform high-precision insulation monitoring in a short time interval using a relatively small power monitoring signal. There is to do.

  In order to solve the above-described problem, an insulation monitoring apparatus according to the first aspect of the present invention includes a monitoring signal injection unit that injects a monitoring signal having a frequency different from the commercial frequency to the B-type ground line of the power receiving transformer, Current detecting means for detecting a leakage current circulating in the B-type ground line via the low-voltage side electric circuit and the ground of the transformer, and removing unnecessary components such as a commercial frequency and its harmonics from the detected leakage current The measurement signal detecting means for detecting the measurement signal corresponding to the monitoring signal component of the leakage current and the monitoring signal injected from the reference input from the B-type grounding line with the D-type grounding point of the low-voltage side circuit as a reference A DFT using a reference signal detecting means for detecting a reference signal corresponding to the reference signal, and a measurement signal for each reference signal n (> m) cycles synchronized with a predetermined phase of the reference signal and shifted by m cycles of the reference signal For calculation It includes a resistance component extraction means for sequentially extracting the ground insulation resistance component included in Ri該 measuring signal.

  In general, in the DFT operation, a frequency component that is an integral multiple of the fundamental frequency f0 is obtained. Therefore, the fundamental frequency f0 represents the frequency resolution of the DFT operation. Therefore, in the insulation monitoring using a monitoring signal of a specific frequency (for example, 20 Hz), the higher the frequency resolution, the more accurate the insulation monitoring can be performed. However, as the frequency resolution is higher, it is necessary to use a measurement signal for a longer time (that is, n is larger), and therefore the time interval (that is, the time resolution of the DFT operation) from which each operation result is obtained decreases. Therefore, in the present invention, the ground insulation resistance component included in the measurement signal is sequentially extracted by the DFT calculation using the measurement signal of each reference signal n cycles shifted by the reference signal m (<n) cycles. In this case, the larger n is, the better the extraction accuracy of the monitoring signal component is, and the smaller m is, the shorter the time interval can be obtained, the present invention can satisfy these simultaneously.

  In the second aspect of the present invention, the ground included in the measurement signal by DFT calculation using the measurement signals for n cycles of the reference signal synchronized with the predetermined phase of the reference signal and shifted by m cycles of the reference signal. Capacitance component extraction means for sequentially extracting capacitance components, and generating a suppression signal for canceling the ground capacitance component included in the leakage current monitoring signal component based on the extracted ground capacitance component signal Suppression signal generation means for supplying to the suppression unit of the current detection means, wherein the resistance component extraction means is based on a measurement signal in a state where the ground capacitance component included in the monitoring signal component of the leakage current is suppressed Thus, a ground insulation resistance component included in the measurement signal is extracted.

  In the present invention, the ground capacitance component of the monitoring signal component of the leakage current is suppressed so as to be substantially zero in the current detection means, so that the ground insulation resistance component can be reduced in a state with less noise than the remaining detection signals. Extract with high accuracy.

  In the third aspect of the present invention, the detected measurement signals are sequentially stored, and a free capacity for m reference signal cycles is provided for measurement signals for n reference signal cycles used in at least one DFT operation. A buffer memory having an added storage capacity and used by circulating the storage capacity is provided, and the resistance component extraction unit performs a DFT operation using a measurement signal of the buffer memory. Therefore, a finite storage capacity can be used efficiently.

  According to a fourth aspect of the present invention, there is provided display means for displaying each ground insulation resistance component signal generated at a timing shifted by the reference signal m cycles on the screen in real time. Therefore, the user can perform highly accurate insulation monitoring at short time intervals.

  In a fifth aspect of the present invention, the reference signal detection means includes a bandpass filter that extracts a commercial frequency component signal from a reference input of the type B ground line, and a bandpass filter that extracts the reference input of the type B ground line. And a filter means for extracting a reference signal corresponding to the frequency component of the monitoring signal from the reference input from which the commercial frequency component signal has been removed.

  In the present invention, the commercial component that is the main noise source is extracted from the reference input of the class B grounding wire, and the commercial component of the reference input is canceled (removed) by the extracted commercial component. Even if a large commercial voltage appears on the seed ground line, the commercial component can always be removed (cancelled) with high accuracy regardless of its size, and the reference signal corresponding to the low-power monitoring signal can be accurately detected from the remaining reference input. It can be detected. Further, by synchronizing the phase with such a reference signal, the ground insulation resistance component can be extracted with high accuracy and high reliability from the monitoring signal component of the leakage current circulating in the B-type ground line.

  In a sixth aspect of the present invention, cumulative addition means for cumulatively adding the measured signal of the detected leakage current over one cycle of the reference signal, and the cumulative addition result of the cumulative addition means has a predetermined threshold range. Determining means for determining whether or not there is a deviation, wherein the resistance component extraction means synchronizes with a predetermined phase of the reference signal and outputs a reference signal determined by the determination means as not to deviate from a predetermined threshold range. A ground insulation resistance component included in the measurement signal is extracted by the used DFT calculation.

  By the way, since the low frequency noise that is generated when the load device is turned on / off includes not only a frequency component close to the frequency of the monitoring signal (for example, 20 Hz), this type of noise component is a filter means. Then, it cannot be removed sufficiently. In this regard, in the present invention, the cumulative addition means cumulatively adds the leakage current measurement signal over one cycle of the reference signal, so that the cumulative addition result is the average of the low-frequency noise included in this measurement signal. It fluctuates greatly according to the level (DC level). Therefore, the determination unit determines whether or not the cumulative addition result deviates from a predetermined threshold range, and the resistance component extraction unit performs measurement for the reference signal n cycles determined not to deviate from the predetermined threshold range. A ground insulation resistance component included in the measurement signal is extracted by DFT calculation using the signal. Therefore, according to the present invention, insulation monitoring can be performed with high accuracy and high accuracy using an appropriate measurement signal while effectively avoiding the use of a measurement signal (monitoring signal) including low frequency noise that cannot be removed by the filter means. Can be done with reliability.

  As described above, according to the present invention, since highly accurate insulation monitoring can be performed at a short time interval using a relatively low power monitoring signal, the load device and peripheral devices are not affected significantly. Accurate insulation monitoring can be performed with high time resolution.

  DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, exemplary embodiments of the invention will be described in detail with reference to the accompanying drawings. FIG. 1 is a block diagram of an insulation monitoring apparatus according to an embodiment. In the figure, a load 2 is connected to a low-voltage side circuit of the power receiving transformer 1, and a first circuit 3 is grounded via a B-type grounding wire 4, and a second circuit 5 is connected to the ground. Between the two, a ground impedance Z0 including a ground insulation resistance R0 and a ground electrostatic (floating) capacitance C0 is provided. On the low voltage side of such a transformer 1, a leakage current based on the ground impedance Z 0 and the ground resistance flows through the second electric circuit 5, the ground impedance Z 0, the ground, and the B-type ground line 4. The insulation monitoring device 10 is provided on the B-type ground line 4 of such a low-voltage side electric circuit.

  The insulation monitoring device 10 includes monitoring signal injection means 20 that injects a monitoring signal W having a frequency different from the commercial frequency into the B-type ground line 4 of the transformer 1. The monitoring signal injection means 20 is a sinusoidal AC signal having a frequency (for example, 20 Hz) different from the commercial frequency by dividing the high-accuracy and high-stable clock signal CK provided from the arithmetic processing unit (MPU) 90. An oscillator (OSC) 21 that generates a monitoring signal and an amplifier 22 that amplifies the power so that the monitoring signal has a required voltage level (for example, about 0.07 V) on the B-type ground line 4 are obtained. The monitored signal W is injected into the B-type ground line 4 through a superposed transformer (CT) 30 as injection means.

  The insulation monitoring apparatus 10 further includes a reference signal detection means 40 for detecting a reference signal B corresponding to the injected monitoring signal W from a B-type ground line with a D-type grounding point ED of the low-voltage side electric circuit as a reference. The reference signal detection means 40 includes a commercial component removing unit 41 that removes a commercial frequency component contained in the reference input b from the reference input b of the B-type ground line 4, and a monitoring signal from the reference input from which the commercial component has been removed. Switched filter means for extracting a frequency component of W and adjusting the phase of the extracted signal to generate a reference signal B phase-synchronized with the ground capacitance component included in the leakage current monitoring signal component And a capacitor filter (SCF) 45.

  In the commercial component removal unit 41, the reference input b from the B-type ground line 4 is buffered (amplified by impedance matching and gain 1) by the buffer amplifier (BA) 42 on the one hand, and on the other hand from the B-type ground line 4. A commercial frequency component is extracted from the reference input b by a band pass filter (BPF) 43. Furthermore, the differential amplifier (AMP) 44 subtracts the commercial frequency component of the output of the BPF 43 from the reference input b of the output of the BA 42 to generate a reference input signal from which the commercial component is removed (cancelled).

  In the present embodiment, the commercial component that is the main noise source is extracted from the reference input b of the B-type grounding wire 4, and the commercial component of the reference input b is canceled (removed) by the extracted commercial component. Even if a large commercial voltage appears on the B-type ground line 4 due to deterioration or the like, the commercial component can always be removed (cancelled) with high accuracy regardless of its size, so that the remaining reference input is changed to the low-power monitoring signal W. The corresponding reference signal B can be detected with high accuracy. Further, by synchronizing the phase with the reference signal B, the ground insulation resistance component IgR can be extracted with high accuracy and high reliability from the monitoring signal component of the leakage current circulating in the B-type ground line 4.

  Further, the SCF 45 as the next-stage filter means has both a function as a band-pass filter and a function as a phase shifter that adjusts the phase of the reference signal B based on a clock frequency command designated by the MPU 90. First, by the function of the band-pass filter, unnecessary components such as the commercial frequency and its harmonics are further removed from the reference input b from which the commercial component is removed, and the reference signal B composed of the frequency component of the monitoring signal is extracted. By the function as the phase shifter, the phase of the reference signal B is matched with the phase of the ground capacitance component IgC included in the leakage current monitoring signal component. The reference signal B is sent to a later-described suppression signal generation unit 80 for suppressing the ground capacitance component included in the measurement signal M, and is also input to the next-stage synchronization signal generation unit 70. The synchronization signal S obtained in this way is sent to the arithmetic processing unit (MPU) 90.

  Further, the insulation monitoring device 10 includes a zero-phase current transformer (ZCT) 50 as current detection means for detecting a leakage current m that circulates from the low-voltage side electric circuit of the transformer 1 to the B-type grounding wire 4 through the ground. Measurement signal detection means 60 for detecting a measurement signal M corresponding to a monitoring signal component of the leakage current by removing unnecessary components such as a commercial frequency from the detected leakage current m.

  The measurement signal detection means 60 converts the leakage current m detected by the ZCT 50 into a voltage signal and amplifies it to a required level in the next stage, and outputs the commercial frequency and its harmonics from the output of the head amplifier 61. A low-pass filter (LPF) 62 that removes unnecessary components such as waves, and the output of the LPF 62 are sampled at a predetermined rate (for example, a rate of 32 samples per 50 ms of one cycle of the reference signal) and converted to a digital measurement signal M / D converter 63.

  In addition, the insulation monitoring device 10 synchronizes with the predetermined phase of the reference signal B, and the measurement signal (that is, the leakage current monitoring signal) by DFT (Discrete Fourier Transform) calculation using the measurement signal M for n cycles of the reference signal. Component) Capacitance component extracting means 90b for extracting the ground capacitance component IgC included in M, and canceling the ground capacitance component included in the leakage current monitoring signal component based on the extracted ground capacitance component signal IgC And a suppression signal generator 80 for generating a suppression signal P for supplying the ground capacitance component in a direction to cancel the ground capacitance component from the suppression unit (tertiary winding) of the ZCT 50.

  The suppression signal generation unit 80 is a circuit for controlling the output level (amplitude) of the reference signal B generated by the SCF 45, and the ground capacitance obtained by the capacitance component extraction means 90b using the reference signal B from the SCF 45. By amplifying with the component IgC, a suppression signal P having such a magnitude that the ground capacitance component IgC of the monitoring signal detected by the ZCT 50 is canceled out magnetically is generated.

  The insulation monitoring device 10 further includes an arithmetic processing unit (MPU) 90 that performs various controls and processes in the insulation monitoring device 10 of the present embodiment. This arithmetic processing unit 90 is synchronized with a predetermined phase of the reference signal B (corresponding to a 90 ° delayed phase in this example), and is ground-insulated in the measurement signal M by DFT calculation using the measurement signal M for n cycles of the reference signal. Resistance component extraction means 90a for obtaining the resistance component IgR and the capacitance component extraction means 90b are provided. Details will be described later.

  In addition, the arithmetic processing unit 90 instructs the SCF 45 to specify a clock frequency that serves as a reference for adjusting the phase of the reference signal B. In addition to controlling each function of the display unit 101 and the operation unit 102, the alarm output unit 103 is controlled.

  Next, the basic operation of the insulation monitoring apparatus 10 configured as described above will be described. First, the monitoring signal injection means 20 generates a monitoring signal W having a frequency (for example, 20 Hz) different from the commercial frequency, and injects the monitoring signal W into the B-type ground line 4 of the power receiving transformer 1 through the superimposing transformer 30.

  On the other hand, in the measurement signal detection means 60, the leakage current flowing back to the B-type ground line 4 via the ground impedance Z0 and the ground is detected by the ZCT 50, and this current is converted into a voltage signal by the head amplifier 61 and the required level is reached. Amplify. Further, unnecessary components such as a commercial frequency and its harmonics are removed from the amplified output by the LPF 62, and the output is converted into a digital signal by the A / D converter 63.

  At the same time, in the reference signal detection means 40, the commercial component removal unit 41 removes the commercial component that is the main noise source from the reference input b from the B-type ground line 4. Further, the SCF 45 removes unnecessary components such as the commercial frequency and its harmonics from the output of the commercial component removing unit 41, and the phase of the reference signal B obtained in this way is the ground capacitance component IgC included in the leakage current measurement signal M. Adjust to the phase of. The reference signal B is sent to the suppression signal generation unit 80 for suppressing the ground capacitance component IgC included in the measurement signal M, and is also input to the next-stage synchronization signal generation unit 70, and is obtained here. The synchronized signal S is sent to the arithmetic processing unit 90.

  In the arithmetic processing unit 90, the resistance component extracting unit 90a performs the DFT operation using the measurement signal M for n cycles of the reference signal in synchronization with a predetermined phase of the reference signal B (corresponding to a 90 ° delayed phase in this example). The ground insulation resistance component IgR included in the measurement signal M is extracted, and the capacitance component extraction unit 90b performs the measurement by DFT calculation using the measurement signal M for n cycles of the reference signal in synchronization with a predetermined phase of the reference signal B. The ground capacitance component IgC included in the signal M is extracted.

  The suppression signal generation means 80 adjusts the amplitude of the reference signal B output from the SCF 45 based on the ground capacitance component signal IgC output from the capacitance component extraction means 90b every moment, and outputs the output suppression current signal P. Feedback control is performed so that the ground capacitance component IgC in the ZCT 50 is magnetically canceled (becomes 0) by being applied in the reverse direction to the suppression unit (third winding) of the ZCT 50. On the other hand, the resistance component extraction unit 90a can obtain the ground insulation resistance component IgR of the measurement signal M with high accuracy by using the measurement signal M in a state where the ground capacitance component IgC is sufficiently suppressed. Become.

  The basic configuration and operation of the insulation monitoring apparatus 10 have been described above. In the present embodiment, in order to enable highly accurate and reliable insulation monitoring even when the low-power monitoring signal W is used. In addition, some ideas are made in the arithmetic processing unit 90. Next, these will be described in detail.

  FIG. 2 is a functional block diagram of the arithmetic processing unit according to the embodiment. In the figure, 91 is a write control unit for writing an input measurement signal M to the buffer memory 92, and 92 is a reference signal m (<n for the measurement signal M for n cycles of the reference signal used in at least one DFT operation. ) A buffer memory 93 having a storage capacity including the free capacity for the cycle, 93 adds the input measurement signal M cumulatively over one cycle of the reference signal, and the cumulative addition result deviates from a predetermined threshold range. A cumulative addition determination unit 94 for determining whether or not there is a threshold update unit that optimally updates the predetermined threshold range by monitoring the cumulative addition result of the cumulative addition determination unit 93 in time series.

  Reference numeral 95 denotes a ground insulation resistance component IgR included in the measurement signal M by DFT calculation using the measurement signal M for n cycles of the reference signal in the buffer memory 92 in synchronization with the synchronization signal S from the synchronization signal generator 70. And a ground separation component that separates the electrostatic capacitance component IgC from the ground. The CR separation unit 95 includes a resistance component extraction unit 90a for obtaining the ground insulation resistance component IgR and a capacitance component extraction unit 90b for obtaining the ground capacitance component IgC.

  Further, 96 monitors each ground insulation resistance component signal IgR separated by the CR separation unit 95 in time series, and the display output of the ground insulation resistance component IgR is unstable due to the measurement signal M affected by noise. A weighting processing unit 97 that suppresses the occurrence of electric leakage, and 97 is a leakage determination unit that determines the presence / absence of leakage based on the grounded insulation resistance component IgR subjected to the weighting process. These functional blocks are realized by program execution of the MPU 90.

  Next, the operation of each unit will be described in detail. First, the write control unit 91 sequentially writes the input measurement signal M in the buffer memory 92 in synchronization with a predetermined timing of the synchronization signal S (for example, the zero phase of the ground insulation resistance component IgR). In parallel with this, the cumulative addition determination unit 93 starts cumulative addition of the measurement signal M, cumulatively adds the measurement signal M over one cycle of the reference signal (for example, 32 samples per 50 ms), and the cumulative addition result Is within a range between predetermined thresholds TH1 and TH2.

  By the way, this measurement signal M is obtained by sufficiently suppressing the ground capacitance component IgC contained in the leakage current monitoring signal component by the ZCT 50 as described above and removing unnecessary components exceeding 20 Hz by the LPF 62. Originally, only the ground insulation resistance component IgR of the monitoring signal should remain in the measurement signal M.

  However, the actual measurement signal M often includes a noise signal having a frequency close to the monitoring signal (20 Hz). For example, when the noise generated when the load 2 is turned on / off is viewed in a section of 20 Hz (50 ms), the average level may vibrate greatly at a frequency close to 20 Hz. The LPF 62 cannot be completely removed. Therefore, the cumulative addition determining unit 93 effectively detects the presence / absence of the influence of the noise signal by cumulatively adding the measurement signal M over one cycle of the reference signal. This will be described in detail below.

  FIG. 3 is a flowchart of the cumulative addition determination unit according to the embodiment. In step S11, the generation of the synchronization signal S is waited, and when it occurs, a predetermined initialization process is performed in step S12. Specifically, the cumulative addition register = 0, the addition counter of the measurement signal M = 1, and the invalid flag = 0 are initialized. Further, in step S13, the generation of the measurement signal M is awaited. When the measurement signal M is generated, the measurement signal M is acquired in step S14, and the measurement signal M is added to the cumulative addition register in step S15. In step S16, it is determined whether or not the addition number counter ≧ K (for example, 32). If NO, the addition number counter is incremented by 1 in step S17, and the process returns to step S13.

  Thus, when the measurement signal M is cumulatively added for K (32) times in the end, the determination in step S16 is YES, and the flow proceeds to step S18. In step S18, it is determined whether or not the cumulative addition value is within a range between the predetermined thresholds TH1 and TH2 (TH1 <cumulative addition value <TH2). If NO (out of range), the reference signal 1 is determined in step S19. It is assumed that an invalid flag = 1 (invalid) indicating that the measurement signal M for the cycle is invalid. If YES (within range), the process of step S19 is skipped. In step S20, the flag state (valid / invalid) is notified to the write control unit 91 and the threshold update unit 94. Then, the process returns to step S11.

  FIG. 4 is an operation explanatory diagram of the cumulative addition determination unit 93 according to the embodiment, and shows the cumulative addition progress and the cumulative addition result when several types of typical noise are superimposed on the monitoring signal W. FIG. 4A shows a case where noise is not included. Since the measurement signal M (= monitoring signal W) in this case is a sinusoidal AC signal, the cumulative addition process over one cycle of the reference signal increases uniformly in the first half as shown in the figure, but in the second half, In the same manner, it finally decreases to “0”. This cumulative addition result SUM is effective because it is within the range of the thresholds TH1 and TH2.

  FIG. 4B shows a case where a noise signal N having a frequency higher than 20 Hz is superimposed on the monitoring signal W. The cumulative addition progress SUM in this case also increases in the first half and decreases in the second half as in FIG. 4A, but the average level of the noise signal N is somewhat offset to the + side. The cumulative addition result SUM is located slightly below the positive threshold TH2. This measurement signal M is also effective because the influence of noise is not so great. However, since such high frequency noise is actually removed by the LPF 62, there is no problem.

  FIG. 4C shows a case where a spike-like noise signal N whose average level greatly fluctuates toward the + side is superimposed on the monitoring signal W. Such a noise signal is generated when the load 2 is turned on. It appears typically in conjunction. In this case, the cumulative addition progress SUM also greatly increases in the first half and starts to decrease in the second half, but since the average level of the noise signal is greatly offset in this period, the cumulative addition result SUM is also + The threshold value TH2 on the side is greatly deviated to the + side. Such a low-frequency noise component N is problematic because it cannot be removed by the LPF 62, but the measurement signal M affected by the noise component N is appropriately “invalid” by the cumulative addition determination unit 93.

  FIG. 4D shows a case where a spike-like noise signal N whose average level greatly fluctuates on the minus side is superimposed on the monitoring signal W. Such a noise signal is when the power supply 2 of the load 2 is turned off. It typically appears in conjunction with In this case, the cumulative addition progress SUM is greatly waved in the middle because the monitoring signal W and the noise signal N are substantially opposite in phase, but the average level of the noise signal N is largely offset to the minus side. The cumulative addition result SUM also deviates greatly from the minus threshold TH1 to the minus side. Such a measurement signal M is also “invalid” by the cumulative addition determination unit 93.

  In practice, the monitoring signal W and the noise signal N are superimposed with various amplitudes and phases. However, the cumulative addition process and the cumulative addition result SUM are described above. Can easily be guessed from examples.

  As described above, if the measurement signal M includes a noise signal having a relatively low frequency, the ground insulation resistance component IgR cannot be accurately monitored. Therefore, the measurement signal M for one cycle is subjected to the DFT calculation. It is better not to use it. Therefore, in the present embodiment, when the cumulative addition result SUM for one cycle of the reference signal deviates from a predetermined threshold range, a signal indicating that the measurement signal M in the section is “invalid” is sent to the write control unit 91 and The threshold update unit 94 is notified.

  Returning to FIG. 2, when the write control unit 91 receives a notification of “invalid” from the cumulative addition determination unit 93, the write control unit 91 outputs the measurement signal M for one cycle of the current reference signal written in the buffer memory 92 to the previous time. It is rewritten (that is, overwritten) by the measurement signal M for one cycle of the valid reference signal. In this way, the measurement signal M having a large noise component is efficiently excluded from the DFT calculation, and is replaced by the latest measurement signal M that is “valid”.

  On the other hand, the threshold update unit 94 provides preset thresholds TH1 and TH2 to the cumulative addition determination unit 93 and monitors “valid” / “invalid” notifications sent from the cumulative addition determination unit 93. The threshold ranges TH1 and TH2 are updated as appropriate. This will be described in detail below.

  FIG. 5 is a flowchart of the threshold update unit according to the embodiment. In step S31, a predetermined initialization process is performed. Specifically, an invalid number counter for counting the number of invalid notification receptions = 0 and an effective number counter for counting the number of valid notification receptions = 0 are initialized. In step S32, the generation of the synchronization signal S (that is, corresponding to the zero phase of the ground insulation resistance component IgR) is waited. Wait for the end phase of the resistance component IgR). Eventually, if there is this flag notification, it is determined in step S34 whether the notification is an invalid flag = 1, and if YES (invalid), the invalid number counter is incremented by one in step S35. If NO (valid), the effective number counter is incremented by 1 in step S36. In step S37, it is determined whether or not the sum of both counters ≧ L (for example, 8). If NO, the process returns to step S32 to repeat the above processing.

  Thus, when the sum of both counters (that is, the total number of occurrences of flag notification) becomes L (= 8) in the determination in step S37, the flow proceeds to step S38, where it is determined whether or not the invalid number counter = L. To do. In the case of YES, all of the cumulative addition results SUM for 8 consecutive reference signal cycles (0.4 sec) have deviated from the ranges of the thresholds TH1 and TH2, so that the current threshold range is too narrow. In step S39, this threshold range is widened at a predetermined rate or by a predetermined amount. For example, the threshold value range is widened by multiplying the current threshold value TH1 and / or TH2 by a predetermined ratio (for example, 1.1 to 1.5) or adding a constant value to the current threshold value TH1 and / or TH2. If NO, the process of step S39 is skipped.

  In step S40, it is determined whether or not the effective number counter = L. If YES, all the cumulative addition results SUM for eight consecutive reference signals are within the range of the thresholds TH1 and TH2. Assuming that the current threshold range is too wide, in step S41, the threshold range is narrowed at a predetermined rate or by a predetermined amount. For example, the threshold value range is narrowed by multiplying the current threshold value TH1 and / or TH2 by a predetermined ratio (for example, 0.6 to 0.9) or subtracting a certain value from the current threshold value TH1 or TH2. In the case of NO, since valid and invalid are mixed in this section, it is assumed that this threshold range is appropriate, and the process returns to step S31 without updating. In this way, the threshold range is updated so that the threshold range automatically adapts to the measurement environment reflecting the current noise magnitude.

  Next, an example of such a threshold update process will be specifically described. FIG. 6 is an explanatory diagram of the operation of the threshold update unit according to the embodiment. It is assumed that thresholds TH2 and TH1 are initially set up and down across the level “0”. In the section of the first reference signal 8 cycles (0.4 sec) starting at time t1, threshold determination is performed using the thresholds TH1 and TH2, and in this example, all of the eight consecutive cumulative addition results SUM are within this threshold range. Therefore, the threshold range is determined to be wide, and the threshold range is narrowed in the next section. In the next section, since only the fifth and seventh cumulative addition results SUM deviate from the threshold range at this time, this threshold range is determined to be appropriate and is maintained as it is.

  On the other hand, in the first 0.4 sec section starting at time t2, since all of the eight cumulative addition results SUM deviate from the current threshold range, this threshold range is determined to be narrow, and at the next time point, the threshold range Has been widely used. The same applies hereinafter. Thus, only useful measurement signals M suitable for the current measurement environment are sequentially stored in the buffer memory 92. It should be noted that both the threshold range is widened and the threshold range is narrowed, a predetermined limit is set, and the threshold range cannot be updated beyond these limits.

  Next, the storage mode of the buffer memory 92 will be described with reference to FIG. In this embodiment, the measurement signal M is A / D converted at a rate of 32 samples per one cycle (50 ms) of the reference signal, and such measurement signals M are collected for 16 cycles (0.8 sec, 512 samples). This is called block data B1, and such block data is collected for 8 blocks (4096 samples) and called group data G1. The buffer memory 92 uses a reference signal m (for example, 16 samples) for a measurement signal M (4096 samples) for a reference signal n (for example, 16 cycles × 8 blocks B1 to B8 = 128) cycles used in at least one DFT operation. ) It has a storage capacity including the free capacity B1 ′ for the cycle, and uses these limited storage capacity cyclically.

  Further, the operation of the CR separation unit 95 will be described with reference to FIG. The CR separation unit 95 basically synchronizes with a predetermined timing of the synchronization signal S (that is, corresponding to a 90 ° delayed phase of the reference signal B), and outputs the measurement signal M for the reference signal n (= 128) cycles. The ground insulation resistance component IgR and the ground capacitance component IgC included in the measurement signal M are separated by the used DFT calculation. The ground insulation resistance component IgR can be separated using a sin function, and the ground capacitance component IgC can be separated simultaneously using a cos function. By the way, such a DFT calculation may be performed every 6.4 sec when the measurement signal M for 4096 samples can be used. However, since the ground insulation resistance component IgR is calculated only every 6.4 sec, the time The resolution is bad.

  Therefore, preferably, the CR separation unit 95 synchronizes with the predetermined timing of the synchronization signal S and outputs the measurement signal M for each reference signal n (= 128) cycles shifted by the reference signal m (= 16) cycles. The ground insulation resistance component IgR and the ground capacitance component IgC included in the measurement signal are separated by the used DFT calculation.

  Specifically, according to FIG. 7, first, the ground insulation resistance component IgR1 and the ground capacitance component IgC1 included in the measurement signal are separated by DFT calculation using the measurement signal M of the group G1, and then , IgR2 and IgC2 included in the measurement signal are separated by DFT calculation using the measurement signal M of group G2. At this time, since the memory for the first block (B1) of the buffer memory 92 is empty, the newly input measurement signal M is sequentially stored in this empty area (indicated by B2 ′ in the figure). Yes. In the same manner, the process proceeds in the same manner as described above so as to effectively use the limited memory.

  In this embodiment, since the DFT calculation is performed using the measurement signal M for 6.4 sec, the fundamental frequency f0 of the DFT calculation is approximately 0.156 Hz. Since frequency components that are integral multiples of the fundamental frequency f0 can be obtained in the DFT operation, this fundamental frequency f0 = 0.156 Hz is the frequency resolution of the DFT operation. Based on this, when each frequency component is viewed around the frequency 20 Hz of the measurement signal M, the frequency component of 127 × f0 is 19.844 Hz, the frequency component of 128 × f0 is just the frequency of 20.00 Hz, and the frequency of 129 × f0. The component is 20.156 Hz, and it can be seen that the ground insulation resistance component IgR of the measurement signal M (20.00 Hz) can be detected with very high frequency resolution with high accuracy even when the monitoring signal W with low power is used.

  In the present embodiment, since the ground insulation resistance component IgR is obtained every 0.8 sec, the user can monitor the insulation state with high time resolution. Further, since the ground capacitance component IgC is also obtained every 0.8 sec, the feedback response of the ground capacitance component suppression control is fast.

   Further, the ground insulation resistance component IgR of the measurement signal M obtained every 0.8 sec is input to the weighting processing unit 96 of FIG. Then, when the current ground insulation resistance component signal IgR increases at a predetermined rate or more than the previous ground insulation resistance component signal IgR, the weighting processing unit 96 sets the current ground insulation resistance component signal IgR to be less than 1. The current ground insulation resistance component signal IgR is weighted with a small weight, and in other cases, the current ground insulation resistance component signal IgR is directly used as the current ground insulation resistance component signal IgR. This will be specifically described below.

  FIG. 8 is a diagram for explaining the operation of the weighting processing unit according to the embodiment. In the figure, the first insulation resistance component IgR1 is obtained when 6.4 seconds have elapsed from time t1, and the value is displayed on the display unit 101. Furthermore, after 0.8 sec, the second insulation resistance component IgR2 is obtained. Since this is smaller than the previous resistance component IgR1, the value is displayed as it is. Further, the next insulation resistance component IgR3 is larger than the previous insulation resistance component IgR2, but since it does not exceed the predetermined increment limit θ, the value is displayed as it is.

  Thereafter, the process proceeds in the same manner. In this example, since the fifth insulation resistance component IgR5 exceeds a predetermined increment limit θ with respect to the fourth resistance component IgR4, the insulation resistance component IgR5 has a predetermined magnification (for example, 0.6 to 0). .8) is multiplied to be converted into a new insulation resistance component IgR5 ′, and this value is displayed on the display unit 101. Similarly, since the eighth insulation resistance component IgR8 exceeds the predetermined increment limit θ, it is multiplied by a predetermined magnification to become a new insulation resistance component IgR8 ', and this value is displayed on the display unit 101. Thus, the sudden increase in the insulation resistance component IgR accompanying the instantaneous increase in the noise component is effectively mitigated, and the monitor is not unstable.

  Furthermore, in the case from the time t2, after the generation of the 16th insulation resistance component IgR16, the insulation resistance components IgR17 to IgR27 of each measurement signal M that follows are comparatively caused by normal insulation deterioration except for IgR22 on the way. It is rising moderately. In this case, since neither of them exceeds the predetermined increment limit θ, the resistance value is displayed as it is without being subjected to weighting processing. In this example, after the predetermined threshold value TH3 is exceeded, for example, at the fourth time (IgR27), the leakage determination unit 97 detects the insulation failure ALM, and an “alarm” to that effect is output to the alarm output unit 103. . In addition, in the case where each obtained insulation resistance component IgR rises uniformly and steeply due to a sudden short circuit accident in the load 2 etc., although there is a slight time delay, it is practically promptly and reliably insulated. A defective ALM is detected.

  In the above embodiment, the threshold range is determined every 8 consecutive reference signal cycles (0.4 sec). However, the present invention is not limited to this. The determination range of the threshold range can be arbitrarily set. In the above embodiment, the operation of each unit has been specifically described with an example of numerical values. However, it goes without saying that the present invention is not limited to these numerical examples.

It is a block diagram of the insulation monitoring apparatus by embodiment. It is a block diagram of the arithmetic processing part by embodiment. It is a flowchart of the accumulation addition determination part by embodiment. It is operation | movement explanatory drawing of the accumulation addition determination part by embodiment. It is a flowchart of the threshold value update part by embodiment. It is operation | movement explanatory drawing of the threshold value update part by embodiment. It is operation | movement explanatory drawing of CR isolation | separation part by embodiment. It is operation | movement explanatory drawing of the weighting process part by embodiment. It is a figure explaining a prior art.

Explanation of symbols

1 Transformer (Receiving transformer)
2 Load 3 1st electric circuit 4 B class grounding wire 5 2nd electric circuit 10 Insulation monitoring device 20 Monitoring signal generation part 21 Oscillator (OSC)
22 Amplifier 30 Injection means (superimposed transformer: CT)
40 Reference signal detection means 41 Commercial component removal unit 42 Buffer amplifier (BA)
43 Band pass filter (BPF)
44 Differential Amplifier (AMP)
45 Switched Capacitor Filter (SCF)
50 Detection means (Zero phase current transformer: ZCT)
60 Measurement signal detection means 61 Head amplifier 62 Low pass filter (LPF)
63 A / D converter 70 Synchronization signal generator 80 Suppression signal generator 90 Arithmetic processor (MPU)

Claims (6)

Monitoring signal injection means for injecting a monitoring signal having a frequency different from the commercial frequency into the B-type ground line of the power receiving transformer;
Current detecting means for detecting a leakage current circulating in the B-type ground line via the low-voltage side electric circuit and the ground of the transformer;
Measurement signal detection means for detecting a measurement signal corresponding to a monitoring signal component of the leakage current by removing unnecessary components such as a commercial frequency and its harmonics from the detected leakage current;
A reference signal detecting means for detecting a reference signal corresponding to the injected monitoring signal from a reference input from the B-type grounding line with respect to a D-type grounding point of the low-voltage side electric circuit;
A ground insulation resistance component included in the measurement signal is sequentially obtained by DFT calculation using the measurement signal for each reference signal n (> m) cycles that are shifted by m cycles of the reference signal in synchronization with a predetermined phase of the reference signal. An insulation monitoring device comprising: a resistance component extracting means for extracting. Capacitance for sequentially extracting ground capacitance components included in the measurement signal by DFT calculation using the measurement signals for n cycles of the reference signal that are shifted by m cycles of the reference signal in synchronization with the predetermined phase of the reference signal A component extraction unit and a suppression signal for canceling out the ground capacitance component included in the monitoring signal component of the leakage current based on the extracted ground capacitance component signal and supplying the suppression signal to the suppression unit of the current detection unit Suppression signal generation means for
The resistance component extraction unit extracts a ground insulation resistance component included in the measurement signal based on a measurement signal in a state where the ground capacitance component included in the monitoring signal component of the leakage current is suppressed. The insulation monitoring apparatus according to claim 1. The detected measurement signals are sequentially stored, and have a storage capacity obtained by adding a free capacity for m reference signal cycles to measurement signals for n reference signal cycles used in at least one DFT operation, It has a buffer memory that is used to cycle through the storage capacity,
The insulation monitoring apparatus according to claim 1, wherein the resistance component extraction unit performs DFT calculation using a measurement signal of the buffer memory.   3. The insulation monitoring apparatus according to claim 1, further comprising display means for displaying each ground insulation resistance component signal generated at a timing shifted by the reference signal m cycles on a screen in real time.   The reference signal detection means removes a commercial frequency component signal extracted from the reference input of the type B ground line by a band pass filter and a commercial frequency component signal extracted by the band pass filter from the reference input of the type B ground line. 2. The insulation according to claim 1, further comprising: a commercial component removing unit configured to extract a reference signal corresponding to a frequency component of the monitoring signal from a reference input from which the commercial frequency component signal has been removed. Monitoring device. A cumulative addition unit that cumulatively adds the measurement signal of the detected leakage current over one cycle of the reference signal, and determines whether or not the cumulative addition result of the cumulative addition unit deviates from a predetermined threshold range. Determination means,
The resistance component extraction unit is synchronized with a predetermined phase of the reference signal, and the ground insulation resistance component included in the measurement signal by DFT calculation using the reference signal determined not to deviate from the predetermined threshold range by the determination unit The insulation monitoring apparatus according to claim 1, wherein:

Images