JP5578982B2 - Equipment failure evaluation system - Google Patents

Description

  Embodiments of the present invention relate to a device failure evaluation system that evaluates the state of a customer device such as an uninterruptible power supply system or an electric motor.

  Conventionally, as a method for evaluating the state of a consumer device that operates by receiving power supply from a power source, based on an electrical signal such as voltage or current measured by the consumer device, deterioration or failure of the consumer device is detected. Techniques for evaluating the possibilities have been proposed. In the apparatus failure evaluation system using this technology, a harmonic component which is one of power quality is analyzed based on the measured electric signal, and the possibility of the deterioration or failure of the customer apparatus is evaluated from the analysis result.

  However, there is a case where it is not possible to evaluate the possibility of the deterioration or failure of the customer device only by analyzing the harmonic component in the electric signal.

JP 2005-287238 A

  As described above, the apparatus failure evaluation system that evaluates the state of the consumer device by analyzing the harmonic component in the electrical signal has a problem that the state of the consumer device may not be evaluated.

  Therefore, the purpose is a device failure that can evaluate the state of the consumer device by combining the analysis of the harmonic component in the electrical signal measured from the consumer device and the analysis of the physical quantity other than the electrical signal. To provide an evaluation system.

  According to the embodiment, the apparatus failure evaluation system includes a harmonic abnormality evaluation unit, a frequency abnormality evaluation unit, and an apparatus failure estimation unit. A harmonic abnormality evaluation part produces | generates the 1st evaluation result which evaluated the state of the said consumer apparatus based on the electrical signal measured by the consumer apparatus which receives the supply of electric power and operate | moves. A frequency abnormality evaluation part produces | generates the 2nd evaluation result which evaluated the state of the said consumer apparatus based on the acoustic signal measured by the said consumer apparatus. The device failure estimation unit estimates a failure occurring in the consumer device or an abnormality that is a precursor of the failure based on the first and second evaluation results.

It is a figure which shows the structural example at the time of evaluating the state of a load apparatus by the apparatus failure evaluation system which concerns on 1st Embodiment. It is a figure which shows the example of the electrical signal measured by the electrical signal measurement part of FIG. It is a block diagram which shows an example of a function structure of the apparatus failure evaluation system of FIG. It is a block diagram which shows an example of a function structure of the harmonic abnormality evaluation part of FIG. It is a block diagram which shows an example of a function structure of the harmonic component calculation part of FIG. It is a figure which shows an example of the calculation result by the harmonic component calculation part of FIG. It is a figure which shows the example of the harmonic abnormal pattern previously recorded on the harmonic abnormal database of FIG. It is a block diagram which shows an example of a function structure of the harmonic component coincidence calculation part of FIG. It is a figure explaining an example of the normalization process in the normalization calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a block diagram which shows an example of a function structure of the frequency abnormality evaluation part of FIG. It is a block diagram which shows an example of a function structure of the frequency component calculation part of FIG. It is a figure which shows an example of a frequency normal pattern and a frequency abnormal pattern. It is a figure which shows the example of the frequency abnormality pattern previously recorded on the frequency abnormality database of FIG. It is a block diagram which shows an example of a function structure of the frequency component coincidence calculation part of FIG. It is a figure explaining an example of the normalization process in the normalization calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a figure explaining an example of the process in the coincidence degree calculation part of FIG. It is a block diagram which shows an example of a function structure of the apparatus failure estimation part of FIG. It is a figure which shows an example of the process table at the time of the comprehensive determination calculation part of FIG. 26 calculating a comprehensive determination. It is a figure explaining an example of the process in the abnormality extraction part of FIG. It is a figure which shows an example of the process table at the time of the comprehensive determination calculation part of FIG. 26 calculating a comprehensive determination. It is a figure which shows an example of the process table at the time of the comprehensive determination calculation part of FIG. 26 calculating a comprehensive determination. It is a schematic diagram which shows the structural example at the time of evaluating the state of a load apparatus by the apparatus failure evaluation system which concerns on 2nd Embodiment. It is a block diagram which shows an example of a function structure of the apparatus failure evaluation system of FIG. It is a block diagram which shows an example of a function structure of the temperature abnormality evaluation part of FIG. It is a figure which shows the example of the relationship between temperature and a temperature factor previously recorded on the temperature abnormality database of FIG. It is a block diagram which shows an example of a function structure of the apparatus failure estimation part of FIG. It is a figure which shows an example of the process table at the time of the comprehensive determination calculation part of FIG. 35 calculating a comprehensive determination. It is a figure which shows an example of the result of the comprehensive determination displayed on the display part of FIG.3 and FIG.32.

  Hereinafter, embodiments will be described with reference to the drawings.

(First embodiment)
Drawing 1 is a mimetic diagram showing the example of composition at the time of evaluating the state of load device 2 which is a consumer device by device failure evaluation system 1 concerning a 1st embodiment.

  The load device 2 is a device installed on the power consumer side such as an uninterruptible power supply system or an electric motor. The load device 2 is connected via a distribution line (which can also be a power transmission line) 8 to a bus 7 that receives predetermined power from the power transmission line 5 of the host power system 4 via the transformer 6. Further, the load device 2 includes equipment, devices, parts, and the like that receive power supply from the upper power system.

  A measuring device 3 is connected to the load device 2. The measurement device 3 includes an electrical signal measurement unit 31, an acoustic signal measurement unit 32, and an analog-digital (A / D) conversion unit 33.

  The electric signal measurement unit 31 measures an electric signal such as a voltage or current of the load device 2 and outputs the measured electric signal to the A / D conversion unit 33. The electric signal includes a harmonic component in addition to the fundamental wave. The harmonic component is represented as a collection of a plurality of sine waves each having a constant period with a different period. Here, when the fundamental wave is a primary signal, a signal whose period is ½ of the fundamental wave (frequency is twice) is referred to as a secondary harmonic, and the period is 1/3 of the fundamental wave (frequency is 3). Signal) is referred to as a third harmonic, and a 1 / n (frequency is n times) signal is referred to as an nth harmonic.

  FIG. 2 is a diagram illustrating an example of an electrical signal measured by the electrical signal measurement unit 31. In the present embodiment, voltages Vab and Vbc for two phases between lines are shown as an example. In FIG. 2, the horizontal axis is time, and the vertical axis is voltage. Assuming that three voltage phases are Va, Vb, and Vc, three line phases are Vab, Vbc, and Vca. In FIG. 2, Vab and Vbc are shown.

  The acoustic signal measuring unit 32 is installed inside the casing of the load device 2, the outer wall of the casing, or the vicinity of the apparatus casing. The acoustic signal measuring unit 32 includes, for example, a sound collecting microphone and an amplifier, and measures the acoustic signal of the load device 2. The acoustic signal measurement unit 32 outputs the measured acoustic signal to the A / D conversion unit 33. The acoustic signal includes signals in various frequency bands. The frequency is represented as a collection of a plurality of sine waves each having a constant period with a different period. Here, assuming that the fundamental wave is primary, a signal whose period is 1/2 of the fundamental wave (frequency is twice) is called secondary, and a signal whose period is 1/3 of the fundamental wave (frequency is 3 times). The third order signal is referred to, and a 1 / n (frequency is n times) signal is referred to as an nth order harmonic. For example, if the fundamental wave is 1 Hz, the second order is 2 Hz and the third order is 3 Hz. Generally, the frequency range of an acoustic signal that can be perceived by humans is 20 Hz to 20 kHz. Further, the frequency range of conversation between humans is 200 Hz to 8 kHz.

  The A / D converter 33 converts the electrical signal from the electrical signal measuring unit 31 into a digital signal. The A / D converter 33 converts the acoustic signal from the acoustic signal measuring unit 32 into a digital signal. When the electrical signal measuring unit 31 directly outputs an electrical signal as a digital signal and the acoustic signal measuring unit 32 directly outputs the acoustic signal as a digital signal, the A / D conversion unit 33 is unnecessary.

  FIG. 3 is a block diagram showing a functional configuration of the apparatus failure evaluation system 1 according to the first embodiment. The apparatus failure evaluation system 1 includes a harmonic abnormality evaluation unit 10, a frequency abnormality evaluation unit 20, an apparatus failure estimation unit 30, and a display unit 40.

  The harmonic abnormality evaluation unit 10 evaluates the state of the load device 2 based on the electrical signal from the measurement device 3, and outputs a first evaluation result for this evaluation to the device failure estimation unit 30. The frequency abnormality evaluation unit 20 evaluates the state of the load device 2 based on the acoustic signal from the measurement device 3 and outputs a second evaluation result regarding this evaluation to the device failure estimation unit 30. The device failure estimation unit 30 estimates a failure of the load device 2 and an abnormality that is a precursor thereof from the first evaluation result from the harmonic abnormality evaluation unit 10 and the second evaluation result from the frequency abnormality evaluation unit 20. To do. The device failure estimation unit 30 outputs this result to the display unit 40 as a device failure estimation result.

  FIG. 4 is a block diagram illustrating an example of a functional configuration of the harmonic abnormality evaluation unit 10 according to the first embodiment. The harmonic abnormality evaluation unit 10 in FIG. 4 includes a harmonic component calculation unit 11, a harmonic component storage unit 12, a harmonic component coincidence calculation unit 13, and a harmonic abnormality database 14.

  As shown in FIG. 5, the harmonic component calculation unit 11 includes an FFT (Fast Fourier Transform) processing unit 111.

  By the way, the observation data (measurement data) may be discrete sampling data due to the operating characteristics of the apparatus. When obtaining harmonic components from such sampling data, a technique called discrete Fourier transform is used. A method improved so that the discrete Fourier transform can be solved at high speed is called a fast Fourier transform (FFT), and is the most versatile method in harmonic analysis.

  Therefore, in the present embodiment, the harmonic component calculation unit 11 performs waveform analysis on the electric signal from the measurement device 3 by the FFT processing unit 111, and the ratio of the harmonic component included in the electric signal, that is, the harmonic component. The wave content is calculated for each order. The harmonic content is calculated by the equation (1) for each harmonic order.

Harmonic content (per order) = Harmonic magnitude (per order) ÷ Fundamental wave magnitude (1)
FIG. 6 is a diagram illustrating an example of a calculation result (hereinafter referred to as a harmonic component calculation result) obtained according to the equation (1). The harmonic component calculation unit 11 outputs the harmonic component calculation result to the harmonic component storage unit 12. In this embodiment, FFT is used to analyze the harmonic component for each order. However, the harmonic component may be analyzed by other methods.

  The harmonic component storage unit 12 stores the harmonic component calculation result from the harmonic component calculation unit 11.

  The harmonic abnormality database 14 records a plurality of harmonic abnormality patterns in advance. If there is an abnormality in the harmonic component in the electric signal, it is often a precursor to failure of the device and the device. The harmonic abnormality pattern is a pattern of harmonic content estimated to be a precursor of equipment and device failure. The harmonic abnormality pattern is obtained based on past statistics and the like, and is different for each device included in the load device 2 and for each type of failure. FIG. 7 is a diagram illustrating an example of a harmonic abnormality pattern recorded in advance in the harmonic abnormality database 14. In the harmonic abnormality database 14, for example, a harmonic abnormality pattern A, a harmonic abnormality pattern B, and a harmonic abnormality pattern C are recorded in advance. In addition, a vertical axis | shaft shows a harmonic content rate and a horizontal axis shows the order of a harmonic.

  As shown in FIG. 8, the harmonic component coincidence calculation unit 13 includes a normalization calculation unit 131 and a coincidence calculation unit 132.

  The normalization calculation unit 131 reads out the harmonic component calculation result stored in the harmonic component storage unit 12 and normalizes the total harmonic content of each order up to a preset order to be 1. To do. At this time, the normalized value of the harmonic content of the j-th order (j is a natural number of 2 or more) is calculated by Expression (2).

Normalized value of j-order harmonic content = j-order harmonic content / total value of harmonic content of each order up to a preset order (2)
The normalization calculation unit 131 outputs a normalization result of the harmonic component calculation result (hereinafter referred to as a harmonic component normalization result) to the coincidence calculation unit 132.

  In addition, the normalization calculation unit 131 normalizes a plurality of harmonic abnormality patterns recorded in the harmonic abnormality database 14 by the same method as that in Expression (2). The normalization calculation unit 131 outputs the normalization result of the harmonic abnormality pattern to the coincidence calculation unit 132.

  Here, the maximum order of the harmonic component calculation result may not match the maximum order of the harmonic abnormality pattern recorded in the harmonic abnormality database 14. In such a case, the normalization calculation unit 131 matches the order for normalizing the harmonic component calculation result with the order for normalizing the harmonic abnormality pattern recorded in the harmonic abnormality database 14. FIG. 9 is a diagram illustrating a normalization process in the normalization calculation unit 131 when the maximum order of the harmonic component calculation result is, for example, 25th and the maximum order of the harmonic abnormality pattern is 20th. . When the maximum order of the harmonic component calculation result is the 25th order and the maximum order of the harmonic abnormality pattern is the 20th order, the normalization calculation unit 131 unifies the order to be normalized by the 20th order, and the first to 20th order. Standardize the harmonic content up to the next. In this case, the denominator of Equation (2) is the total value of the harmonic content of each order up to the 20th order.

  The coincidence calculation unit 132 takes the absolute value of the difference between the harmonic component normalization result and the normalization result of the harmonic abnormality pattern for each order, and the total value (hereinafter referred to as harmonics) as shown in Equation (3). (Referred to as difference sum).

Harmonic difference total value = | Second-order component in harmonic component normalization result-Second-order component in normalization result of harmonic abnormality pattern | + | Third-order component in harmonic component normalization result-harmonic abnormality Third-order component of pattern normalization result | + ... + | n-order component in harmonic component normalization result-n-order component of normalization result of harmonic abnormal pattern | (3)
According to the equation (3), when the harmonic component normalization result and the harmonic abnormality pattern normalization result completely coincide with each other, the harmonic difference total value becomes zero. FIG. 10 is a diagram illustrating an example of a case where the harmonic component normalization result and the harmonic abnormality pattern normalization result completely match.

  Further, according to the equation (3), when the harmonic component normalization result and the harmonic abnormality pattern normalization result are completely different, the total value of the harmonic component normalization result is 1, and the harmonic abnormality pattern Since the total value of the standardization results of 1 is 1, the calculated harmonic difference total value is 2. FIG. 11 is a diagram illustrating an example in which the harmonic component normalization result and the harmonic abnormality pattern normalization result are completely different.

  Further, according to the equation (3), when the harmonic component normalization result and the harmonic abnormality pattern normalization result partially match, the harmonic difference total value is any value between 0 and 2 It becomes the value of. FIG. 12 is a diagram illustrating an example in which the harmonic component normalization result and the harmonic abnormality pattern normalization result partially match. According to the example shown in FIG. 12, the harmonic difference total value is a third-order value 0.6 of the harmonic component normalization result and a seventh-order value 0.6 of the normalization result of the harmonic abnormality pattern. The sum is 1.2.

  FIG. 13 is a diagram showing the relationship between the harmonic component normalization result and the harmonic abnormality pattern normalization result with respect to the harmonic difference total value. When the harmonic component normalization result and the harmonic abnormality pattern normalization result are completely different from each other, the total harmonic difference value is 2, and when the harmonic component is completely matched, it is 0. Moreover, when the harmonic component normalization result and the normalization result of the harmonic abnormality pattern partially match, the intermediate value between 0 and 2 is obtained.

  Here, the harmonic component coincidence is defined based on the harmonic difference total value. In the present embodiment, when the harmonic component normalization result and the harmonic abnormality pattern normalization result completely match, that is, when the harmonic difference total value is 0, the harmonic component coincidence is 100%. When the total difference is 2, that is, when the harmonic difference total value is 2, the harmonic component coincidence is set to −100%, and the middle between 100% and −100% is set to 0%. The harmonic component coincidence is obtained by Expression (4).

Harmonic component coincidence = (1−total harmonic difference total value) × 100 (%) (4)
FIG. 14 is a diagram showing the relationship of the harmonic component coincidence with the harmonic difference total value.

  The coincidence degree calculation unit 132 uses the equations (3) and (4) to calculate the harmonics for all the normalization results of all the harmonic abnormality patterns recorded in the harmonic abnormality database 14 with respect to the harmonic component normalization results. The wave component coincidence is calculated. The coincidence calculation unit 132 outputs the calculated plural harmonic component coincidences to the device failure estimation unit 30 as the first evaluation result.

  FIG. 15 is a block diagram illustrating an example of a functional configuration of the frequency abnormality evaluation unit 20 according to the first embodiment. The frequency abnormality evaluation unit 20 in FIG. 15 includes a frequency component calculation unit 21, a frequency component storage unit 22, a frequency component coincidence calculation unit 23, and a frequency abnormality database 24.

  As shown in FIG. 16, the frequency component calculation unit 21 includes an FFT processing unit 211. The frequency component calculation unit 21 calculates the frequency content of the acoustic signal by obtaining the frequency component of the acoustic signal from the measuring device 3 by the FFT processing unit 211. The frequency content at a predetermined frequency is calculated by the equation (5) based on the size of the frequency spectrum.

Frequency content = frequency spectrum size / total frequency spectrum size at sample frequency (5)
For example, if the sample frequency is 1 Hz to 1000 Hz, the denominator value is the sum of frequency spectra from 1 Hz to 1000 Hz.

  The frequency component calculation unit 21 outputs a calculation result (hereinafter referred to as a frequency component calculation result) obtained according to the equation (5) to the frequency component storage unit 22.

  In this embodiment, FFT, which is the most common method for analyzing the frequency component of an acoustic signal, is used. However, the frequency component may be analyzed by other methods.

  The frequency component storage unit 22 stores the frequency component calculation result from the frequency component calculation unit 21.

  The frequency abnormality database 24 records a plurality of frequency abnormality patterns in advance. Here, the frequency abnormality pattern is a frequency pattern that is a precursor of failure of the device and the apparatus. The frequency abnormality pattern is different for each device included in the load device 2 and for each type of failure.

  Here, the frequency pattern will be described with reference to FIG. FIG. 17A is a diagram illustrating an example of a frequency pattern when the load device 2 is operating normally, that is, an example of a normal frequency pattern. According to FIG. 17A, the peak of the normal frequency pattern exists between 400 Hz and 500 Hz. FIG. 17B is a diagram illustrating an example of a frequency pattern when the load device 2 is abnormal, that is, a frequency abnormality pattern. According to FIG. 17B, in addition to the peak existing between 400 Hz and 500 Hz, there is another peak showing an abnormality in the vicinity of 1500 Hz.

  FIG. 18 is a diagram illustrating an example of a frequency abnormality pattern recorded in advance in the frequency abnormality database 24. In the frequency abnormality database 24, for example, a frequency abnormality pattern A, a frequency abnormality pattern B, and a frequency abnormality pattern C are recorded in advance. According to FIG. 15, the frequency abnormality pattern A has an abnormal frequency distribution in the vicinity of 1500 Hz, the frequency abnormality pattern B has an abnormal frequency distribution in the vicinity of 1200 Hz to 1300 Hz, and the frequency abnormality pattern C has a frequency near 1300 Hz. In addition, an abnormal frequency distribution exists in the vicinity of 1700 Hz to 1800 Hz. The vertical axis indicates the frequency magnitude (frequency content), and the horizontal axis indicates the frequency. Note that the frequency abnormality pattern is not limited to one in which the normal frequency distribution and the abnormal frequency distribution are completely separated in frequency. For example, the abnormal frequency distribution partially overlaps the normal frequency distribution and protrudes. There are cases where the distribution is uneven.

  As shown in FIG. 19, the frequency component coincidence calculation unit 23 includes a normalization calculation unit 231 and a coincidence calculation unit 232.

  The normalization calculation unit 231 reads out the frequency component calculation result stored in the frequency component storage unit 22, and the sum of the frequency content ratios in the preset sample frequency region is set to 1 for the read frequency component calculation result. Standardize in the same way. At this time, the normalized value of the frequency content of the j-th order (j is a natural number of 2 or more) is calculated by the equation (6).

Normalized value of j-th order frequency content rate = j-th order frequency content rate / total value of frequency content rates in a preset sample frequency region (6)
When the sample frequencies of Equation (5) and Equation (6) are the same, the j-th order component after normalization based on Equation (6) is the frequency content rate (j-th order) in Equation (5). Become.

  However, when the normal frequency distribution is larger than the abnormal frequency distribution as shown in FIG. 18, the difference between the normal frequency pattern and the abnormal frequency pattern is relatively small. Therefore, the frequency component calculation unit 21 and the normalization calculation unit 231 need to set the sample frequency in a region that includes a relatively large amount of abnormal frequency distribution.

  For example, the sample frequency is set to include a relatively large amount around 1500 Hz in the frequency abnormality pattern A, set to include a relatively large amount near 1200 Hz to 1300 Hz in the frequency abnormality pattern B, and 1300 Hz in the frequency abnormality pattern C. The setting is made so as to include a relatively large area around ˜1800 Hz.

  FIG. 20 is a diagram for explaining the sample frequency when the frequency abnormality pattern is normalized.

  FIG. 20A shows a case where the region of the sample frequency of the frequency abnormality pattern A shown in FIG. 18 is 1000 Hz to 2000 Hz. The normalization calculation unit 231 normalizes the frequency abnormality pattern A so that the total value of the frequency components at the sampling frequency of 1000 Hz to 2000 Hz is 1.

  FIG. 20B shows a case where the sample frequency region of the frequency abnormality pattern B shown in FIG. 18 is 1000 Hz to 1500 Hz. The normalization calculation unit 231 normalizes the frequency abnormality pattern B so that the total value of the frequency components at the sampling frequency of 1000 Hz to 1500 Hz is 1.

  FIG. 20C shows a case where the sample frequency region of the frequency abnormality pattern C shown in FIG. 18 is 1000 Hz to 2000 Hz. The normalization calculation unit 231 normalizes the frequency abnormality pattern C so that the total value of the frequency components at the sample frequencies of 1000 Hz to 2000 Hz is 1.

  The normalization calculation unit 231 outputs a normalization result of the frequency component calculation result (hereinafter referred to as a frequency component normalization result) and a normalization result of the frequency abnormality pattern to the coincidence calculation unit 232.

  The coincidence calculation unit 232 takes the absolute value of the difference between the frequency component normalization result and the frequency abnormality pattern normalization result for each frequency in the sample frequency domain, and calculates the total value (hereinafter referred to as Equation 7) , Referred to as the total frequency difference value).

Frequency difference total value = | frequency iHz component of frequency component normalization result−frequency iHz component of frequency abnormality pattern normalization result | + | frequency jHz component of frequency component normalization result−frequency abnormal pattern normalization result Component of frequency jHz of | +... + | Component of frequency nHz of frequency component standardization result-component of frequency nHz of standardization result of frequency abnormality pattern | (however, the sample frequency region is iHz to nHz). 7)
According to Expression (7), when the frequency component and magnitude of the frequency component normalization result and the frequency component and magnitude of the frequency abnormality pattern normalization result are completely coincident with each other, the total frequency difference value is 0. . FIG. 21 is a diagram illustrating an example of a case where the frequency component and size of the frequency component normalization result and the frequency component and size of the frequency abnormality pattern normalization result are completely the same. However, in FIG. 21, the sample frequency region is 1500 Hz to 1510 Hz.

  Further, according to Expression (7), when the frequency component normalization result and the frequency abnormality pattern normalization result are completely different, the total value of the frequency component normalization results is 1, and the frequency abnormality pattern normalization is performed. Since the total value of the results is 1, the calculated frequency difference total value is 2. FIG. 22 is a diagram illustrating an example in which the frequency component normalization result and the frequency abnormality pattern normalization result are completely different. However, in FIG. 22, the sample frequency region is 1500 Hz to 1510 Hz.

  Further, according to Equation (7), when the frequency component normalization result and the frequency abnormality pattern normalization result partially match, the frequency difference total value is any value between 0 and 2 It becomes. FIG. 23 is a diagram illustrating an example when the frequency component normalization result and the frequency abnormality pattern normalization result partially match. According to the example shown in FIG. 23, the frequency difference total value is the sum of the value 0.6 of the frequency 1503 Hz resulting from the frequency component normalization and the value 0.6 of the frequency 1507 Hz resulting from the normalization of the frequency abnormality pattern. 1.2. However, in FIG. 23, the sample frequency region is 1500 Hz to 1510 Hz.

  FIG. 24 is a diagram showing the relationship between the frequency component normalization result and the frequency abnormality pattern normalization result for the frequency difference total value. When the frequency component normalization result and the frequency abnormality pattern normalization result are completely different from each other, the total frequency difference value is 2, and when the frequency difference is completely the same, it is 0. Further, when the frequency component normalization result and the frequency abnormality pattern normalization result partially match, the value is an intermediate value between 0 and 2.

  Here, the frequency component coincidence is defined based on the total frequency difference value. In this embodiment, when the frequency component normalization result and the frequency abnormality pattern normalization result are completely coincident, that is, when the frequency difference total value is 0, the frequency component coincidence is 100%, If they are different, that is, if the total frequency difference value is 2, the frequency component coincidence is set to −100%, and the middle between 100% and −100% is set to 0%. The frequency component coincidence degree is obtained by Expression (8).

Frequency component coincidence = (1−total frequency difference value) × 100 (%) (8)
FIG. 25 is a diagram showing the relationship of the frequency component coincidence with the total frequency difference value.

  The coincidence calculation unit 232 uses the equations (7) and (8) to match the frequency component normalization result with the frequency component normalization result for each normalization result of all frequency abnormality patterns recorded in the frequency abnormality database 24. Calculate the degree. The coincidence degree calculation unit 232 outputs the calculated plural frequency component coincidence degrees to the device failure estimation unit 30 as the second evaluation result.

  FIG. 26 is a block diagram illustrating an example of a functional configuration of the device failure estimation unit 30 according to the first embodiment. The apparatus failure estimation unit 30 includes a comprehensive determination calculation unit 301, a comprehensive determination rearrangement unit 302, and an abnormality extraction unit 303.

  The comprehensive determination calculation unit 301 receives the first evaluation result from the harmonic abnormality evaluation unit 10 and the second evaluation result from the frequency abnormality evaluation unit 20. Comprehensive determination calculation unit 301 calculates comprehensive determination 1 and comprehensive determination 2 based on the first and second evaluation results.

  FIG. 27 is a diagram illustrating a processing table when the comprehensive determination calculation unit 301 according to the first embodiment calculates the comprehensive determination 1 and the comprehensive determination 2. In FIG. 27, for each failure, the harmonic component coincidence degree and the frequency component coincidence degree are added to calculate the overall determination 1.

  According to the example of FIG. 27, the harmonic component coincidence with the harmonic abnormal pattern of failure 1 in the load device 2 is 90%, and the frequency component coincidence with the frequency abnormal pattern of failure 1 is 20%. The overall judgment 1 is calculated as 110%. Similarly, the overall determination 1 of the failure 2 is calculated as 120%, the overall determination 1 of the failure 3 is calculated as 90%, the overall determination 1 of the failure 4 is calculated as -10%. Here, failure 1 to failure 4... Indicates various failures in a plurality of devices included in the load device 2.

  Comprehensive determination 2 is a result of dividing the result of comprehensive determination 1 by 2. Since the result of the comprehensive determination 1 is the sum of the harmonic component coincidence and the frequency component coincidence, the numerical value is 200% to −200%. On the other hand, since the result of comprehensive judgment 2 is ½ of the result, the numerical value is 100% to −100%. For this reason, the result of comprehensive judgment 2 is easier to understand as the degree of coincidence.

  The comprehensive determination rearrangement unit 302 receives the result of the comprehensive determination 2 from the comprehensive determination calculation unit 301. The comprehensive determination rearrangement unit 302 rearranges the faults 1 to 4 in descending order of the value of the comprehensive determination 2, and generates a rearrangement result of the comprehensive determination 2.

  The abnormality extraction unit 303 receives the sorting result of the comprehensive judgment 2 from the comprehensive judgment sorting unit 302. As shown in FIG. 28, the abnormality extraction unit 303 sets an abnormality extraction threshold value, and extracts a comprehensive determination 2 that exceeds the abnormality extraction threshold value. Then, the abnormality extraction unit 303 outputs the failure name related to the extracted comprehensive determination 2 to the display device 40. In addition, the abnormality extraction unit 303 outputs the sorting result of the comprehensive judgment 2 from the comprehensive judgment sorting unit 302 to the display unit 40.

  The display unit 40 displays the failure name output from the device failure estimation unit 30 and the sorting result of the comprehensive determination 2.

  As described above, in the first embodiment, the state of the load device 2 is evaluated based on the electrical signal from the load device 2 and also evaluated based on the acoustic signal of the load device 2. Then, the device failure estimation unit 30 is configured to estimate a failure occurring in the load device 2 or an abnormality that is a precursor of the failure based on these evaluation results. Accordingly, it is possible to estimate a failure occurring in the load device 2 or an abnormality that is a precursor of the failure from both analysis of harmonic components in the electrical signal and frequency analysis of the acoustic signal.

  Therefore, the apparatus failure evaluation system 1 according to the first embodiment combines the analysis of the harmonic component in the electrical signal measured from the consumer apparatus with the analysis of the physical quantity other than the electrical signal, so that the consumer apparatus Can be evaluated.

  In addition, the calculation method of the comprehensive determination 1 and the comprehensive determination 2 in the comprehensive determination calculation unit 301 is not limited to the one shown in FIG. For example, the comprehensive determination calculation unit 301 may calculate the comprehensive determination 1 and the comprehensive determination 2 according to the processing table shown in FIG. Depending on the type of failure, there are cases where it is better to determine a failure based on an abnormality in an electrical signal, and cases where it is better to determine a failure based on an abnormality in an acoustic signal. Therefore, in the comprehensive determination calculation unit 301, it is set in advance for each failure whether the comprehensive determinations 1 and 2 are calculated based on which of the harmonic component coincidence and the frequency component coincidence.

  In the example of FIG. 29, when the comprehensive determinations 1 and 2 are calculated based on the harmonic component coincidence, bit 1 is added to the harmonic component coincidence column. Similarly, when calculating the overall determinations 1 and 2 based on the frequency component coincidence, bit 1 is added to the frequency component coincidence column.

  The overall determination calculation unit 301 sets the harmonic component coincidence values of failure 1, failure 2 and failure 3 as the value of comprehensive determination 1 in accordance with the processing table shown in FIG. The value of comprehensive judgment 1 is set as the value of comprehensive judgment 2.

  In addition, the comprehensive determination calculation unit 301 sets the frequency component coincidence value of the failure 5 and the failure 6 as the value of the comprehensive determination 1 in accordance with the processing table shown in FIG. The value of comprehensive judgment 1 is set as the value of comprehensive judgment 2.

  In addition, when it is better to comprehensively determine both the electrical signal and the acoustic signal according to the processing table shown in FIG. 29 according to the processing table shown in FIG. The frequency component coincidence degree of failure 4 is added to obtain the value of comprehensive determination 1. Then, a value obtained by dividing the value of comprehensive determination 1 by 2 is set as a value of comprehensive determination 2.

  The comprehensive determination calculation unit 301 outputs the calculated comprehensive determination 2 to the comprehensive determination rearrangement unit 302.

  The comprehensive determination calculation unit 301 may calculate the comprehensive determination 1 and the comprehensive determination 2 according to the processing table shown in FIG. Depending on the type of failure, it may be better to change the weight for failure evaluation based on electrical signals and the weight for failure evaluation based on acoustic signals. Therefore, in FIG. 30, a weight for failure evaluation based on an electrical signal and a weight for failure evaluation based on an acoustic signal are set in advance for each failure.

  In the example of FIG. 30, the weight for failure evaluation based on the electric signal is described in the weight column of the harmonic component matching degree. Similarly, the weight for failure evaluation based on the acoustic signal is described in the weight column of the frequency component coincidence degree. Here, the total value of the weight for the failure evaluation based on the electrical signal and the weight for the failure evaluation based on the acoustic signal is set to 1.

  The comprehensive determination calculation unit 301 sets the comprehensive determination 1 of failure 1, failure 2, failure 3, failure 4, failure 5, and failure 6 according to the processing table of FIG. , 0.7, 0.7, 0.1, 0 and weights 0, 0.2, 0.3, 0.3, 0.9, and 1 for failure evaluation based on acoustic signals, respectively. .

  The apparatus failure estimation unit 30 outputs the result of the comprehensive determination 1 to the display unit 40 and causes the display unit 40 to display the determination result.

(Second Embodiment)
FIG. 31 is a schematic diagram illustrating a configuration example when the state of the load device 2 that is a customer device is evaluated by the device failure evaluation system 1 according to the second embodiment. In FIG. 31, parts that are the same as those in FIG. 1 are given the same reference numerals, and only different parts will be described here.

  The measuring device 3 includes an electric signal measuring unit 31, an acoustic signal measuring unit 32, an analog-digital (A / D) conversion unit 33, and a temperature measuring unit 34.

  The temperature measuring unit 34 is installed inside the casing of the load device 2 or on the outer wall of the casing. The temperature measuring unit 34 measures the temperature of the load device 2.

  The A / D conversion unit 33 receives the temperature measurement result from the temperature measurement unit 34 and converts it into a digital signal. In addition, when the temperature measurement part 34 outputs temperature directly as a digital signal, A / D conversion is unnecessary.

  FIG. 32 is a block diagram illustrating a functional configuration of the apparatus failure evaluation system 1 according to the second embodiment. The apparatus failure evaluation system 1 includes a harmonic abnormality evaluation unit 10, a frequency abnormality evaluation unit 20, an apparatus failure estimation unit 30, a display unit 40, and a temperature abnormality evaluation unit 50.

  FIG. 33 is a block diagram illustrating an example of a functional configuration of the temperature abnormality evaluation unit 50 according to the second embodiment. The temperature abnormality evaluation unit 50 in FIG. 33 includes a temperature data storage unit 51, a temperature factor calculation unit 52, and a temperature abnormality database 53.

  The temperature data storage unit 51 receives temperature data from the measuring device 3 and stores it.

  The temperature abnormality database 53 records in advance the relationship between the temperature and the temperature factor indicating the risk of occurrence of an abnormality in the equipment in the load device 2 due to the temperature rise. FIG. 34 is a diagram showing an example of the relationship between temperature and temperature factor recorded in advance in the temperature abnormality database 53. The relationship between the temperature and the temperature factor differs for each device included in the load device 2 and for each type of failure. According to FIG. 34, it is considered that no abnormality occurs until the temperature reaches 30 ° C., and the temperature factor is zero. On the other hand, when the temperature is from 30 ° C. to 50 ° C., the temperature factor increases linearly, and when the temperature is 50 ° C. or higher, the temperature factor becomes 100.

  The temperature factor calculation unit 52 reads the temperature data stored in the temperature data storage unit 51 and the relationship between the temperature and the temperature factor recorded in the temperature abnormality database 53, and calculates the temperature factor based on the temperature data. . The temperature factor is calculated for each of a plurality of relationships recorded in the temperature abnormality database 53. The temperature factor calculation unit 52 outputs the calculated plurality of temperature factors to the device failure estimation unit 30 as a third evaluation result.

  FIG. 35 is a block diagram illustrating a functional configuration of the device failure estimation unit 30 according to the second embodiment. The apparatus failure estimation unit 30 includes a comprehensive determination calculation unit 301, a comprehensive determination rearrangement unit 302, and an abnormality extraction unit 303.

  The comprehensive determination calculation unit 301 obtains the first evaluation result from the harmonic abnormality evaluation unit 10, the second evaluation result from the frequency abnormality evaluation unit 20, and the third evaluation result from the temperature abnormality evaluation unit 50. receive. The comprehensive determination calculation unit 301 calculates the comprehensive determination 1 based on the first to third evaluation results.

  FIG. 36 is a diagram illustrating a processing table when the comprehensive determination calculation unit 301 according to the second embodiment calculates the comprehensive determination 1. In FIG. 36, a weight for failure evaluation based on an electrical signal, a weight for failure evaluation based on an acoustic signal, and a weight for failure evaluation based on temperature are set in advance for each failure.

  In the example of FIG. 36, the weight for failure evaluation based on the electrical signal is described in the weight column of the harmonic component matching degree. Further, the weight for failure evaluation based on the acoustic signal is written in the weight column of the frequency component coincidence degree. In addition, a weight for failure evaluation based on temperature is described in a weight column of a temperature factor. Here, the total value of the weight for the failure evaluation based on the electrical signal, the weight for the failure evaluation based on the electrical signal, and the weight for the failure evaluation based on the temperature is set to 1.

  The overall determination calculation unit 301 performs the overall determination 1 of failure 1, failure 2, failure 3, failure 4, failure 5 and failure 6 according to the processing table of FIG. 0.7, 0.4, 0.4, 0, 0, weights 0, 0.1, 0.3, 0.3, 0, 0.5 for failure evaluation based on acoustic signals, and failure evaluation based on temperature Are calculated based on the weights of 0.8, 0.2, 0.3, 0.3, 1, 0.5, respectively.

  The apparatus failure estimation unit 30 outputs the result of the comprehensive determination 1 to the display unit 40 and causes the display unit 40 to display the determination result.

  As mentioned above, in the said 2nd Embodiment, while evaluating the state of the load apparatus 2 based on the electric signal and acoustic signal from the load apparatus 2, it evaluates based on the temperature of the load apparatus 2. FIG. Then, the device failure estimation unit 30 is configured to estimate a failure occurring in the load device 2 or an abnormality that is a precursor of the failure based on these evaluation results. Thereby, it is possible to estimate a failure occurring in the load device 2 or an abnormality that is a precursor of the failure from a plurality of viewpoints of analysis of harmonic components in the electrical signal, frequency analysis of the acoustic signal, and temperature analysis. It becomes.

  Therefore, the apparatus failure evaluation system 1 according to the second embodiment combines the analysis of the harmonic component in the electrical signal measured from the consumer apparatus with the analysis of the physical quantity other than the electrical signal, so that the consumer apparatus Can be evaluated.

(Other embodiments)
Although not described in each of the above embodiments, the apparatus failure evaluation system 1 receives the electrical signal, the acoustic signal, and the temperature measured by the measuring apparatus 3 as needed, and repeats the evaluation of the state of the load apparatus 2 at regular intervals. It doesn't matter if you do. That is, the device failure identification unit 30 includes a first evaluation result from the harmonic abnormality evaluation unit 10, a second evaluation result from the frequency abnormality evaluation unit 20, and a third evaluation result from the temperature abnormality evaluation unit 30. And receive from time to time. And the apparatus failure specific | specification part 30 displays the result of comprehensive determination on the display part 40 for every fixed time based on these results. FIG. 37 is a diagram illustrating an example of the comprehensive determination result displayed on the display unit 40. In FIG. 37, the results of comprehensive determination for each failure are plotted for 24 hours a day.

  Thereby, the apparatus failure evaluation system 1 can display the evaluation results of the load apparatus 2 in time series. For this reason, the user of the device failure evaluation system 1 can determine whether the abnormality occurring in the load device 2 is abrupt or a sign of failure.

  Note that the present invention is not limited to the above embodiments as they are, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, constituent elements over different embodiments may be appropriately combined.

DESCRIPTION OF SYMBOLS 1 ... Device failure evaluation system 10 ... Harmonic abnormality evaluation part 11 ... Harmonic component calculation part 111, 211 ... FFT processing part 12 ... Harmonic component preservation | save part 13 ... Harmonic component coincidence calculation part 131, 231 ... Normalization calculation Units 132, 232 ... coincidence calculation unit 14 ... harmonic abnormality database 20 ... frequency abnormality evaluation unit 21 ... frequency component calculation unit 22 ... frequency component storage unit 23 ... frequency component coincidence calculation unit 24 ... frequency abnormality database 30 ... device failure Estimating unit 301 ... comprehensive determination calculating unit 302 ... comprehensive determination rearranging unit 303 ... abnormality extracting unit 40 ... display unit 50 ... temperature abnormality evaluating unit 51 ... temperature data storage unit 52 ... temperature factor calculating unit 53 ... temperature abnormality database 2 ... load Device 3 ... Measuring device 31 ... Electric signal measuring unit 32 ... Acoustic signal measuring unit 33 ... A / D converter 34 ... Temperature measuring unit 4 ... Upper power system 5 ... Power transmission 6 ... transformer 7 ... bus 8 ... distribution lines

Claims (8)

A harmonic abnormality evaluation unit that generates a first evaluation result that evaluates the state of the consumer device based on an electrical signal measured by the consumer device that operates by receiving power supply;
A frequency abnormality evaluation unit that generates a second evaluation result that evaluates the state of the consumer device based on an acoustic signal measured by the consumer device;
Based on the first and second evaluation results, comprising a failure occurring in the customer device, or a device failure estimation unit for estimating an abnormality that is a precursor of failure ,
The harmonic abnormality evaluation unit is
A waveform analysis of the electrical signal, and a harmonic component calculation unit for calculating a harmonic component calculation result for the harmonic component in the electrical signal;
Harmonic abnormality database that pre-records the harmonic component pattern of the harmonic when a failure occurs as a harmonic abnormal pattern;
The harmonic component coincidence between the harmonic component calculation result and the harmonic abnormality pattern is calculated, and the harmonic component coincidence is output to the device failure estimation unit as the first evaluation result. With a degree calculator
The frequency abnormality evaluation unit is
Analyzing the waveform of the acoustic signal and calculating a frequency component calculation result for the acoustic signal; and
A frequency abnormality database for pre-recording a frequency component pattern of a frequency when a failure occurs as a frequency abnormality pattern;
A frequency component coincidence calculation unit that calculates a frequency component coincidence between the frequency component calculation result and the frequency abnormality pattern, and outputs the frequency component coincidence as the second evaluation result to an apparatus failure estimation unit; ,
The apparatus failure estimation unit is
A comprehensive determination calculation unit that performs a comprehensive determination by adding the harmonic component coincidence and the frequency component coincidence for the same failure ,
An abnormality extraction unit that extracts a failure having a matching degree in the comprehensive determination higher than a preset threshold as a failure occurring in the customer device or an abnormality that is a precursor of the failure. An equipment failure evaluation system. The harmonic abnormality database records in advance the harmonic abnormality pattern for each of a plurality of possible failures,
The harmonic component coincidence calculation unit calculates a plurality of harmonic component coincidences from the harmonic component calculation result and the harmonic abnormality pattern for each of the plurality of failures,
The frequency abnormality database records in advance the frequency abnormality pattern for each of the plurality of failures,
The frequency component matching degree calculating unit, and the frequency component calculation result from said plurality of failure for each of the frequency abnormality pattern, device failure evaluation according to claim 1, wherein calculating a plurality of frequency components match degree system. The comprehensive determination calculation unit weights the harmonic component coincidence and the frequency component coincidence for the same failure, and adds the weighted harmonic component coincidence and the frequency component coincidence to make a comprehensive determination. And
The abnormality extraction unit extracts a failure having a degree of coincidence in the comprehensive determination higher than a preset threshold as a failure occurring in the consumer device or an abnormality that is a precursor of the failure. The apparatus failure evaluation system according to claim 1 or 2 . A harmonic abnormality evaluation unit that generates a first evaluation result that evaluates the state of the consumer device based on an electrical signal measured by the consumer device that operates by receiving power supply;
A frequency abnormality evaluation unit that generates a second evaluation result that evaluates the state of the consumer device based on an acoustic signal measured by the consumer device;
A temperature abnormality evaluation unit that generates a third evaluation result that evaluates the state of the consumer device based on a temperature measured by the consumer device that operates by receiving power supply ; and
A device failure estimation unit that estimates a failure occurring in the customer device or an abnormality that is a precursor of the failure based on the first to third evaluation results;
Was immediately Bei,
The harmonic abnormality evaluation unit is
A waveform analysis of the electrical signal, and a harmonic component calculation unit for calculating a harmonic component calculation result for the harmonic component in the electrical signal;
Harmonic abnormality database that pre-records the harmonic component pattern of the harmonic when a failure occurs as a harmonic abnormal pattern;
The harmonic component coincidence between the harmonic component calculation result and the harmonic abnormality pattern is calculated, and the harmonic component coincidence is output to the device failure estimation unit as the first evaluation result. With a degree calculator
The frequency abnormality evaluation unit is
Analyzing the waveform of the acoustic signal and calculating a frequency component calculation result for the acoustic signal; and
A frequency abnormality database for pre-recording a frequency component pattern of a frequency when a failure occurs as a frequency abnormality pattern;
A frequency component coincidence calculation unit that calculates a frequency component coincidence between the frequency component calculation result and the frequency abnormality pattern, and outputs the frequency component coincidence as the second evaluation result to an apparatus failure estimation unit; ,
The temperature abnormality evaluation unit is
A temperature abnormality database that records in advance a temperature factor pattern indicating a relationship between temperature and a temperature factor indicating a risk of occurrence of a failure due to temperature; and
A temperature factor calculation unit that calculates the temperature factor from the temperature measured by the consumer device and the temperature factor pattern, and outputs the temperature factor to the device failure estimation unit as the third evaluation result ; Prepared,
The apparatus failure estimation unit is
Comprehensive judgment that weights the harmonic component coincidence, frequency component coincidence, and temperature factor for the same failure, and adds up the weighted harmonic component coincidence, frequency component coincidence, and temperature factor, respectively, for comprehensive judgment A judgment calculation unit;
An abnormality extraction unit that extracts a failure having a matching degree in the comprehensive determination higher than a preset threshold as a failure occurring in the customer device or an abnormality that is a precursor of the failure. equipment failure evaluation system shall be the. The harmonic abnormality database records in advance the harmonic abnormality pattern for each of a plurality of possible failures,
The harmonic component coincidence calculation unit calculates a plurality of harmonic component coincidences from the harmonic component calculation result and the harmonic abnormality pattern for each of the plurality of failures,
The frequency abnormality database records in advance the frequency abnormality pattern for each of the plurality of failures,
The frequency component coincidence calculation unit calculates a plurality of frequency component coincidences from the frequency component calculation result and the frequency abnormality pattern for each of the plurality of failures,
The temperature abnormality database records in advance the temperature factor pattern for each of a plurality of possible failures,
The temperature factor calculation unit, the temperature measured by the consumer device, and a plurality of failures for each of the temperature factor pattern, device failure evaluation according to claim 4, wherein calculating the plurality of temperature factors system. Equipment failure evaluation system according to claim 1 or 4, further comprising a display unit for displaying the estimation result from the device failure estimator. The harmonic abnormality evaluation unit generates the first evaluation result for each preset period,
The frequency abnormality evaluation unit generates the second evaluation result for each period,
The device failure estimation unit estimates a failure occurring in the consumer device or an abnormality that is a precursor of the failure based on the first and second evaluation results for each period. The apparatus failure evaluation system according to claim 1. The harmonic abnormality evaluation unit generates the first evaluation result for each preset period,
The frequency abnormality evaluation unit generates the second evaluation result for each period,
The temperature abnormality evaluation unit generates the third evaluation result for each period,
The device failure estimation unit estimates a failure occurring in the consumer device or an abnormality that is a precursor of the failure based on the first, second, and third evaluation results for each period. The apparatus failure evaluation system according to claim 4 .

Images