JP2019095291A - Insulation deterioration diagnosis method and insulation deterioration diagnosis device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an insulation deterioration diagnosis method and an insulation deterioration diagnosis device capable of diagnosing a deterioration state of a resin insulator with high accuracy. An insulation deterioration diagnosis method for diagnosing deterioration of a resin insulating material is to irradiate the surface of the resin insulating material to be diagnosed with infrared light and disperse the reflected light or transmitted light of the resin insulating material to be diagnosed. A step of acquiring an infrared spectrum (step S10) and a step of extracting a plurality of numerical data from the shapes of peaks derived from a plurality of chemical species appearing in the acquired infrared spectrum and the shapes of straight portions of the infrared spectrum. (Step S22) and a step (step S30) of calculating the surface resistance of the diagnosis target based on the extracted plurality of numerical data. [Selection diagram] Fig. 2

Description

  The present disclosure relates to an insulation deterioration diagnosis method and an insulation deterioration diagnosis apparatus for diagnosing deterioration of a resin insulator.

  Resin insulators used in electrical equipment such as power distribution facilities deteriorate due to oxidation and hydrolysis of resin insulators, or reaction between fillers and acid gases in the atmosphere. When the resin insulator is deteriorated, the surface resistivity of the resin insulator is lowered to cause discharge or short circuit, which may result in failure of the electric device. Therefore, there is a need for a technique for accurately estimating the remaining life until failure or short circuit by grasping the deterioration state of the resin insulator.

  It is effective to measure the surface resistivity of the insulator for diagnosing the degradation of the insulator. However, since the electrical measurement of the surface resistivity is susceptible to external environmental noise such as humidity, a measurement error becomes a problem. Therefore, it has been practiced to diagnose the deterioration of the insulator by measuring chemical measurement items that are less affected by external environmental noise and have a strong correlation with the decrease in surface resistivity. .

  For example, JP-A-2010-71961 (Patent Document 1) is based on the fact that there is a correlation between the expression of hydroxyl groups (OH groups) on the surface of the polymer material and the deterioration of the surface resistance of the polymer material. Disclosed is a method of diagnosing the degree of deterioration of a polymeric material by quantifying OH groups on the surface of the polymeric material made of polyester resin by spectroscopic analysis.

Unexamined-Japanese-Patent No. 2010-71961

  In the diagnostic method described in Patent Document 1 above, since only the OH group on the surface of the polymer material is measured, the deterioration of the polymer material already progresses when a change appears in the detection amount of the OH group. , It may happen that the life is approaching. Therefore, it is feared that it is difficult to detect the deterioration of the insulator at an early stage before reaching the life.

  The present invention has been made to solve such problems, and an object thereof is to provide an insulation deterioration diagnosis method and an insulation deterioration diagnosis apparatus capable of diagnosing the deterioration state of a resin insulator with high accuracy. It is.

  The insulation degradation diagnosis method for diagnosing deterioration of a resin insulator according to the present disclosure irradiates the surface of a resin insulator to be diagnosed with infrared light, and disperses reflected light or transmitted light to obtain red as a diagnosis target. Extracting a plurality of numerical data from the step of acquiring an outer spectrum, the shape of peaks derived from a plurality of chemical species appearing in the acquired infrared spectrum, and the shape of a linear portion of the infrared spectrum; Calculating a surface resistivity of an object to be diagnosed based on a plurality of numerical data.

  An insulation degradation diagnostic device for diagnosing degradation of a resin insulator according to the present disclosure includes a spectroscopy device, an extraction unit, and a calculation unit. The spectroscope emits infrared light to the surface of the resin insulator to be diagnosed, and disperses the reflected light or transmitted light to obtain an infrared spectrum of the diagnosis target. The extraction unit extracts a plurality of numerical data from the shapes of peaks derived from a plurality of chemical species appearing in the infrared spectrum acquired by the spectroscopic device and the shape of a linear portion of the infrared spectrum. The calculation unit calculates the surface resistivity of the diagnosis target based on the plurality of numerical data extracted by the extraction unit.

  According to the present disclosure, it is possible to provide an insulation degradation diagnosis method and an insulation degradation diagnosis apparatus capable of diagnosing the degradation state of a resin insulator with high accuracy.

BRIEF DESCRIPTION OF THE DRAWINGS It is a figure which shows roughly the structure of the insulation degradation diagnostic apparatus according to Embodiment 1 of this invention. 7 is a flowchart showing an insulation deterioration diagnosis process performed in the insulation deterioration diagnosis device according to the first embodiment. It is a figure which shows the infrared spectrum of three unsaturated polyester resin in which a deterioration condition mutually differs. It is a figure explaining the method to extract the peak intensity and peak wave number of at least one chemical species from an infrared spectrum. It is a figure explaining the method to extract the peak intensity and peak wave number of at least one chemical species in a first derivative spectrum or a second derivative spectrum. It is a figure explaining the abundance method which extracts the number of intersections of the peak of at least 1 chemical species in an infrared spectrum, and at least 1 straight line parallel to the wave number axis of an infrared spectrum, and the wave number of an intersection. It is a figure explaining the method to extract the inclination of the linear part of an infrared spectrum. It is a figure which shows typically the relational expression of several numerical data and surface resistivity. 15 is a flowchart showing an insulation deterioration diagnosis process performed in the insulation deterioration diagnosis device according to the second embodiment. It is a figure which shows typically the calibration curve which shows the relationship between degradation degree and surface resistivity.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following drawings, the same or corresponding parts are denoted by the same reference characters and description thereof will not be repeated.

Embodiment 1
<Configuration of insulation deterioration diagnosis device>
FIG. 1 schematically shows a structure of an insulation deterioration diagnosis apparatus according to a first embodiment of the present invention. Insulation degradation diagnosis device 10 according to the first embodiment is mainly a device for diagnosing degradation of a resin insulator used for electrical equipment such as power distribution equipment. Insulation degradation diagnosis apparatus 10 according to the first embodiment performs diagnosis using the fact that a change due to deterioration of a resin insulator appears in infrared spectrum data obtained by infrared spectroscopy of a resin insulator to be diagnosed. It diagnoses the deterioration of the subject.

  In addition, as a resin insulator used as a diagnostic object, unsaturated polyester resin, an epoxy resin, a phenol resin, a silicone resin etc. can be used, for example.

  Referring to FIG. 1, the insulation degradation diagnosis device 10 according to the first embodiment includes a spectroscopy device 1, a surface resistivity calculation device 2, and an output device 3.

The spectroscope device 1 is constituted by an infrared spectroscope which irradiates infrared rays to the surface of a resin insulator to be diagnosed and splits reflected light or transmitted light. The spectroscope device 1 disperses infrared light having a wave number of 4000 to 650 cm ⁻¹ and detects light intensity for each of the dispersed wavelengths. Then, infrared spectrum data of a resin insulator to be diagnosed is output by converting the detected light intensity into an electric signal.

  In addition, by using an infrared light spectroscope as the spectroscope 1, analysis of a black diagnosis target which can not be measured by near infrared light becomes possible. As a method of acquiring an infrared spectrum, any of a transmission method, a reflection method, and an ATR (Attenuated Total Reflection) method can be used depending on a diagnostic object. However, when the ATR method is used, since it is not necessary to extract a sample for measurement from the resin insulator, an infrared spectrum can be obtained nondestructively. Further, in the ATR method, since the influence of the processing method of the sample does not appear in the measurement result, it is easy to grasp the temporal change of the infrared spectrum, and it is possible to obtain the infrared spectrum suitable for analysis.

  The infrared spectrum data acquired by the spectrometer 1 is input to the surface resistivity calculator 2. The surface resistivity calculating device 2 is configured of, for example, a personal computer. The surface resistivity calculation device 2 calculates the surface resistivity of the diagnosis target based on the infrared spectrum data given from the spectroscopy device 1. The “surface resistivity” is a value obtained by dividing the strength of the DC electric field in the surface layer of the resin insulator to be diagnosed by the current per unit length of the electrode. In practice, the surface resistivity corresponds to the surface resistance converted to the square surface on the surface to be diagnosed.

  The surface resistivity calculating device 2 includes a spectrum digitizing unit 4, a surface resistivity calculating unit 7, and an output unit 8. The spectrum digitizing unit 4 is a part for digitizing an infrared spectrum to be diagnosed, and includes a normalization unit 5 and a numerical data extraction unit 6.

The normalization unit 5 normalizes the infrared spectrum as pre-processing for digitizing the infrared spectrum. This normalization process is a process performed to eliminate differences in intensity due to measurement conditions and the like among a plurality of infrared spectra. Specifically, among the peaks derived from a plurality of chemical species appearing in the infrared spectrum, the peak intensities of the remaining chemical species on the basis of the intensity of the chemical species peak whose intensity hardly changes due to the deterioration of the resin insulator Standardize. For example, in the case where the diagnosis target is an unsaturated polyester resin, it is preferable to perform the normalization process on the basis of C—H stretching vibration (CH ₂ group, wave number 2919 cm ⁻¹ vicinity) hardly affected by deterioration.

  The numerical data extraction unit 6 extracts a plurality of numerical data from the normalized infrared spectrum. The infrared spectrum has a peak derived from a plurality of chemical species and a linear portion located between two adjacent peaks. In the specification of the present application, “a linear portion of an infrared spectrum” refers to a section of the infrared spectrum in which the spectral intensity is substantially constant with respect to a change in wave number, or the spectral intensity is substantially linear with respect to a change in wave number It means an interval that changes in That is, it means a section in which the spectrum intensity can be linearly approximated with respect to the wave number in the infrared spectrum data.

  The numerical data extraction unit 6 extracts a plurality of numerical data from the shapes of peaks derived from a plurality of chemical species appearing in the infrared spectrum and the shape of the linear portion of the infrared spectrum. Each of the plurality of extracted numerical data is data in which the numerical value changes, reflecting the change of the infrared spectrum due to the deterioration of the diagnosis target. In addition, according to the resin insulator used as a diagnostic object, the numerical data which can be made to reflect the change of an infrared spectrum suitably can be employ | adopted suitably as several numerical data. The method of extracting numerical data in the numerical data extraction unit 6 will be described later.

  The surface resistivity calculator 7 calculates the surface resistivity of the resin insulator to be diagnosed based on the plurality of numerical data extracted by the numerical data extractor 6. Specifically, the surface resistivity calculation unit 7 obtains an infrared spectrum and digitizes the infrared spectrum for each of a plurality of resin insulators whose surface resistivity is known in advance. Regarding the resin insulator to be diagnosed, a relational expression between a plurality of numerical data and a surface resistivity is derived in advance. Then, the surface resistivity calculation unit 7 uses the relational expression between the plurality of numerical data derived in advance and the surface resistivity, and the surface of the diagnosis target based on the plurality of numerical data extracted from the infrared spectrum of the diagnosis target. Calculate the resistivity. The surface resistivity calculation unit 7 outputs the calculated surface resistivity to the output unit 8.

  The output unit 8 is a portion for transmitting data indicating the surface resistivity of the diagnosis target to the output device 3 configured by a display, a printer or the like. When the output device 3 receives data indicating the surface resistivity from the output unit 8, the output device 3 displays the surface resistivity of the diagnosis target on the display screen of the display. Alternatively, print out the surface resistivity of the diagnosis target from the printer. The output device 3 may be configured to output the deterioration state of the diagnosis target in addition to the output of the surface resistivity of the diagnosis target.

<Method for diagnosing insulation deterioration>
Next, a method of diagnosing insulation deterioration according to the first embodiment of the present invention will be described.

  The insulation degradation diagnosis method according to the first embodiment mainly irradiates the surface of the resin insulator to be diagnosed with infrared light, and disperses the reflected light or transmitted light to obtain an infrared spectrum of the diagnosis target. And acquiring the plurality of numerical data from the shape of the acquired infrared spectrum, and calculating the surface resistivity of the diagnosis target based on the plurality of extracted numerical data.

  FIG. 2 is a flowchart showing an insulation deterioration diagnosis process performed in the insulation deterioration diagnosis apparatus 10 according to the first embodiment.

Referring to FIG. 2, first, infrared spectrum is obtained by performing infrared spectroscopic analysis on the resin insulator to be diagnosed in spectroscopic device 1 (step S <b> 10)
Next, in the surface resistivity calculation device 2, a process of digitizing the infrared spectrum acquired by the spectroscopy device 1 is executed (step S20). In this digitizing process, first, as a pre-processing for digitizing an infrared spectrum, a process of normalizing the infrared spectrum is performed (step S21). This process is executed by the standardization unit 5 shown in FIG.

  Subsequently, a process of extracting a plurality of numerical data from the normalized infrared spectrum is performed (step S22). This process is executed by the numerical data extraction unit 6 shown in FIG.

  Next, the surface resistivity of the diagnosis target is calculated using the plurality of extracted numerical data (step S30). In step S30, the surface resistivity of the diagnosis target is calculated based on the plurality of numerical data extracted from the infrared spectrum of the diagnosis target using the plurality of numerical data derived in advance and the relational expression between the surface resistivity. Be done. The calculated surface resistivity of the diagnostic object is sent to the output device 3 and output from the output device 3 to the outside (step S40).

  In addition, the relational expression of several numerical data and surface resistivity which are used by the process of step S30 can be produced | generated by the process sequence shown to step S100-S120. In addition, the process shown to step S100-S120 is performed prior to the infrared spectroscopy analysis (step S10) of diagnostic object.

  Specifically, first, a plurality of resin insulators whose surface resistivity is known are prepared. The plurality of resin insulators include a new resin insulator having a high surface resistivity, and at least one degraded resin insulator having a decreased surface resistivity. Infrared spectroscopic analysis is performed on each of the plurality of resin insulators to obtain an infrared spectrum (step S100).

  Next, the infrared spectra of the plurality of acquired resin insulators are digitized (step S110). In step S110, a plurality of numerical data are extracted from the infrared spectrum for each resin insulator, and the plurality of extracted numerical data is associated with the surface resistivity of the resin insulator. As a result, a plurality of combinations of surface resistivity and a plurality of numerical data are generated.

  Based on the plurality of generated combinations, a relational expression between the plurality of numerical data and the surface resistivity is created (step S120). Specifically, the surface resistivity is used as an objective variable, and each numerical data is used as an explanatory variable, and multivariate analysis such as multiple regression analysis is performed to create a relational expression between the surface resistivity and a plurality of numerical data.

  Multivariate analysis is not limited to multiple regression analysis, and principal component regression, PLS (partial least squares method) regression analysis or the like may be used as long as correlation is recognized. Alternatively, these multivariate analyzes may be performed using commercially available multivariate analysis software.

<Step of extracting multiple numerical data from infrared spectrum>
Next, the processing procedure of the step (step S22) of extracting a plurality of numerical data from the infrared spectrum in the insulation degradation diagnosis process shown in FIG. 2 will be described in detail.

  As described above, the infrared spectrum has peaks and linear portions derived from a plurality of chemical species. In this step, a plurality of numerical data are extracted from the shapes of peaks of a plurality of chemical species and the shapes of straight portions. Below, the method to extract numerical data using an infrared spectrum when a diagnostic object is unsaturated polyester resin is explained.

FIG. 3 shows the infrared spectra of three unsaturated polyester resins which are different from each other in the state of deterioration. The horizontal axis of FIG. 3 shows a wave number (cm ^<-1> ), and a vertical axis shows a spectral intensity.

  In FIG. 3, spectrum k1 shows the infrared spectrum of a new unsaturated polyester resin, spectrum k2 shows the infrared spectrum of a degraded unsaturated polyester resin before reaching the life, and spectrum k3 shows the life The infrared spectrum of the unsaturated polyester resin of the end-of-life product reached.

In addition, the normalization process on the basis of the spectral intensity of CH group (wavenumber 2919 cm < ^-1 > vicinity) is performed to each of spectrum k1-k3. Further, in FIG. 3, in order to make each spectrum easy to see, each of the spectra k2 and k3 is shown to be shifted in the vertical axis direction with respect to the spectrum k1.

As shown in FIG. 3, peaks of a plurality of chemical species appear in the infrared spectrum of the unsaturated polyester resin. The types of chemical species and the size of each peak depend on the composition of the unsaturated polyester resin and the like. In the example of FIG. 3, a plurality of chemical species include a hydroxyl group (OH group, wave number 3600 to 3200 cm ⁻¹ ), a carbonyl group (C = O group, wave number 1730 cm ⁻¹ near), carbonate (wave number 1490 to 1410 cm ^{− 1),} nitrate (wavenumber 1360cm around ^-1), it contains at least two kinds of C-O group (wavenumber of around 1150 cm ^-1), and sulfate (wavenumber of around 1100 cm ^-1).

The type of chemical species appearing in the infrared spectrum and the shape of the chemical species peak (peak intensity, peak wave number, peak width, etc.) change depending on the deterioration state of the unsaturated polyester resin. For example, in the infrared spectrum shown in FIG. 3, in the new infrared spectrum (spectrum k1), the peak intensity of carbonate (wave number 1490 to 1410 cm ⁻¹ ) which is a composition component of unsaturated polyester resin is the largest, C = The O group (near wave number 1730 cm ⁻¹ ) and the C—O group (near wave number 1150 cm ⁻¹ ) have significantly lower peak intensities than carbonates. Further, OH groups (wave number 3600~3200cm ^-1), the peak of nitrate (wave number 1360 cm ^-1 vicinity) and sulfate (wavenumber of around 1100 cm ^-1) does not appear.

  On the other hand, in the infrared spectrum of the deteriorated product (spectrum k2), while the peak intensity of the carbonate decreases, peaks begin to appear in the nitrate and the OH group. Also, the peak intensity of the C = O group is also increasing. In the infrared spectrum (spectrum k3) of the end-of-life product, the peak intensities of the nitrate and OH groups are further increased.

Furthermore, the linear portion of the infrared spectrum is focused. The linear portion of the infrared spectrum corresponds to a section in which the spectral intensity can be linearly approximated with respect to the wave number, as described above. In the example of FIG. 3, the linear portion of the infrared spectrum is the infrared spectrum in the region of 2600 to 2000 cm ⁻¹ (the region RN4 in the drawing).

  In the new infrared spectrum (spectrum k1), the straight part is almost parallel to the wave number axis (horizontal axis), while in the degraded product infrared spectrum (spectrum k2), the straight part is low from the high wave number side It is inclined so that the intensity decreases toward the wave number side. In the infrared spectrum (spectrum k3) of the end-of-life product, the slope of the linear portion is further increased.

Here, as the deterioration of the unsaturated polyester resin proceeds, the peak strength of the carbonate decreases and the peak strength of the nitrate increases because of the carbonate (for example, calcium carbonate) which is a constituent component of the unsaturated polyester resin. (CaCO ₃ )), polyester resin and glass resin react with nitrogen oxides (NO, NO ₂ ) in the atmosphere to form nitrate (eg, calcium nitrate (Ca (NO ₃ ) ₂ )) It is believed that.

Note that by the hydrogen sulfide constituents of the composition and the atmosphere of the unsaturated polyester resin (H 2 _S) or sulfur dioxide (SO ₂₎ are reacted in some cases sulfate is generated. Nitrates and sulfates are deliquescent and hygroscopic, and thus absorb atmospheric water and deliquesce.

  In addition, the reason why the peak strength of the OH group increases as the deterioration progresses is considered to be that the OH group is generated due to the hydrolysis of the surface of the unsaturated polyester resin and the like.

  Moreover, it is thought that the peak intensity of C = O group and C-O group changes because an ester group and a carboxyl group are produced | generated by oxidation or hydrolysis of the surface of unsaturated polyester resin. As for the CCO group and the C—O group, depending on the group to be generated, the peak intensity may increase or decrease, or the peak wave number may shift to the high wave number side or the low wave number side.

  In addition, as the deterioration progresses, the slope of the linear portion of the infrared spectrum increases because the surface state changes due to the contamination, oxidation or hydrolysis of the surface of the unsaturated polyester resin, and scattering of infrared light It is thought that the influence of In addition, since the influence of scattering of infrared light is larger on the high wave number side than on the low wave number side, as a result, the linear portion of the infrared spectrum is inclined.

  Thus, when the deterioration of the unsaturated polyester resin proceeds, the surface is covered with hygroscopic nitrates, sulfates, hydrophilic OH groups and the like. In addition, the surface of the resin insulator is oxidized due to corrosive gas in the air or aging deterioration. These factors reduce the surface resistivity of the resin insulator and eventually lead to dielectric breakdown.

  In the insulation degradation diagnosis method according to the first embodiment, a plurality of numerical data correlated with the decrease in surface resistivity is extracted from the infrared spectrum of the resin insulator to be diagnosed, and the plurality of extracted numerical values are extracted. The data is used to diagnose degradation of the subject being diagnosed. Specifically, in the step of extracting a plurality of numerical data from the infrared spectrum (step S20 in FIG. 2), at least two of the following (1) to (7) are extracted.

(1) Peak intensity of at least one chemical species (2) Peak wave number of at least one chemical species (3) Peak intensity of at least one chemical species in first derivative spectrum or second derivative spectrum (4) First derivative spectrum or Peak wave number of at least one chemical species in the second derivative spectrum (5) number of intersections of at least one chemical species peak and at least one straight line parallel to the wave number axis of the infrared spectrum (6) at least one Wave number at the intersection of the chemical species peak and at least one straight line parallel to the wave number axis of the infrared spectrum (7) Slope of at least one linear portion in the infrared spectrum Hereinafter, the unsaturated polyester resin shown in FIG. The method for extracting the numerical data (1) to (7) from the infrared spectrum will be described in detail.

(1) Peak intensity of at least one chemical species The peak intensity of at least one chemical species peak among a plurality of chemical species peaks appearing in the infrared spectrum is extracted. At least one chemical species is selected from among a plurality of chemical species whose intensity changes as the diagnostic target progresses.

FIG. 4 shows the wave number region of 3700 to 3100 cm ⁻¹ (region RN1 in FIG. 3) and the wave number region of 1750 to 1650 cm ⁻¹ (region RN2 in FIG. 3) in the infrared spectrum of FIG. 3. ing. A peak derived from an OH group appears in the wave number region RN1. A peak derived from a C = O group appears in the wave number region RN2. As shown in FIG. 4, the peak of the OH group and the peak of the C = O group both change in intensity or wave number as deterioration of the unsaturated polyester resin progresses. Therefore, as the numerical data, the intensity of the peak derived from the OH group or the intensity of the peak derived from the C = O group is extracted.

(2) Peak wave number of at least one chemical species The peak wave number of at least one chemical species peak among the plurality of chemical species peaks appearing in the infrared spectrum is extracted. In the example of FIG. 4, the wave number of the peak derived from the OH group (corresponding to the wave number p in the figure) or the wave number of the peak derived from the C = O group (corresponding to the wave number q in the figure) is extracted.

(3) Peak intensity of at least one chemical species in the first derivative spectrum or second derivative spectrum In FIG. 5 (A), the infrared spectrum in the region of 1600 to 1200 cm ⁻¹ wave number (region RN3 in FIG. 3) It is shown. In this wave number region, a peak derived from carbonate (wave number 1490 to 1410 cm ⁻¹ ) appears.

  The second derivative spectrum in this wave number domain is shown in FIG. 5 (B). In FIG. 5 (B), the spectrum k11 shows the second derivative spectrum of the new spectrum k1, the spectrum k12 shows the second derivative spectrum of the spectrum k2 of the deteriorated product, and the spectrum k13 shows the spectrum k3 of the life reached product Shows the second derivative spectrum of In FIG. 5B, in order to make each spectrum easy to see, each of the spectra k12 and k13 is shown to be shifted in the vertical axis direction with respect to the spectrum k11.

In the second derivative spectrum, a peak (minimum value) corresponding to the peak included in the infrared spectrum appears. In other words, the second derivative spectrum can cause a peak buried in the infrared spectrum to appear as a sharpened peak (minimum value). As shown in FIG. 5 (B), a brand new second derivative spectrum (spectrum k11) has a local minimum corresponding to the peak of carbonate. On the other hand, in the second derivative spectrum (spectrum k12) of the deteriorated product, in addition to the minimum value corresponding to the carbonate peak, the minimum value derived from the nitrate (wave number 1360 cm ⁻¹ ) appears. In the second derivative spectrum (spectrum k13) of the end-of-life product, the minimum value derived from nitrate is further sharpened. As for the change in such second derivative spectrum, in the infrared spectrum shown in FIG. 5A, as deterioration of the unsaturated polyester resin progresses, a nitrate peak appears and its peak intensity gradually increases. It corresponds to the thing.

  Therefore, the intensity of the peak of at least one chemical species is extracted from the second derivative spectrum. In FIG. 5B, the intensity of the peak derived from nitrate is extracted from the second derivative spectrum. Specifically, intensities α, β and γ of peaks derived from nitrate are respectively extracted from the spectra k11, k12 and k13. By comparing the extracted intensities α, β and γ with each other, it is possible to catch the appearance of the nitrate peak in the infrared spectrum.

  Although not shown, in the first derivative spectrum, the maximum value or the minimum value appears at the wave number corresponding to the inflection point of the corresponding infrared spectrum, and the maximum value or the minimum value of the corresponding infrared spectrum is supported. A zero point appears in the wave number to be Therefore, peak intensities of at least one chemical species may be extracted from the first derivative spectrum.

(4) Peak wave number of at least one chemical species in the first derivative spectrum or second derivative spectrum As shown in FIG. 5 (B), the second derivative spectrum is derived from the peak embedded in the infrared spectrum Minimum value appears. Therefore, by extracting the wave number of the peak of at least one chemical species from the second derivative spectrum, it is possible to capture the chemical species that appears as the deterioration of the diagnostic object progresses.

  In addition, in the first derivative spectrum, the maximum value and the minimum value appear at the wave number corresponding to the inflection point of the corresponding infrared spectrum, and the zero point appears at the wave number corresponding to the maximum value or the minimum value of the infrared spectrum. Thus, by extracting the wave number of the peak of at least one chemical species from the first derivative spectrum, it is possible to capture the change in the shape of the corresponding infrared spectrum.

(5) Number of intersection points of at least one chemical species peak and at least one straight line parallel to the wave number axis of the infrared spectrum In FIG. 6, the wave number is 3700 to 3100 cm ⁻¹ (a region in FIG. The infrared spectrum at RN1) is shown. In this wave number region, a peak derived from an OH group appears as the deterioration progresses.

  At least one straight line L1 parallel to the wave number axis (horizontal axis) of the infrared spectrum is drawn with respect to the OH group peak. In the example of FIG. 6, a plurality of straight lines L1 parallel to the wave number axis are drawn at equal intervals. The number of intersections of the plurality of straight lines L1 and the OH group peak is extracted.

  In the example of FIG. 6, since the straight line portion and the straight line L1 in the spectrum k1 of the new product are substantially parallel, the number of intersection points is zero. On the other hand, in the spectrum k2 of the deteriorated product, since the peak of the OH group appears, the number of intersections with the straight line L1 is two. Furthermore, in the spectrum k3 of the end-of-life product, the number of points of intersection with the straight line L1 is increased to 14 due to the increase in the peak intensity of the OH group.

  The number of intersections of a peak of a certain chemical species and at least one straight line L1 can be an indicator indicating whether the chemical species peak appears or disappears in the infrared spectrum. Specifically, as shown in FIG. 6, when the number of intersection points changes from 0 to 2, it can be seen that a peak of a specific chemical species appears in the infrared spectrum. In addition, when the number of intersections increases from 2 further, it turns out that the peak of this chemical species is large. Conversely, when the number of intersection points decreases, it can be seen that the peak of this chemical species decreases. It can be seen that the peak of this species disappeared when the number of crossing points decreased to zero.

  Therefore, as numerical data, for at least one chemical species peak among the plurality of chemical species peaks appearing in the infrared spectrum, an intersection point between the peak and at least one straight line parallel to the wave number axis of the infrared spectrum Extract the number of

(6) Wave number at the intersection of the peak of at least one chemical species and at least one straight line parallel to the wave number axis of the infrared spectrum In FIG. 6, additionally, the peak of one chemical species and at least one straight line L1 The wave number at the intersection with can be an index indicating whether the position of the peak of a particular chemical species is shifted to the low wave number side or the high wave number side. Specifically, when the wave number at the intersection of one straight line L1 and the peak of a specific chemical species decreases as the deterioration progresses, the position of this peak is shifted to the low wave number side I understand that. Conversely, when the wave number at the intersection of one straight line L1 and the peak of a specific chemical species increases as the deterioration progresses, it is understood that the position of this peak is shifted to the high wave number side.

  In addition, at two intersections existing on one straight line L1 intersecting a specific chemical species, the difference between the wavenumber at one intersection and the wavenumber at the other intersection can be an index indicating the width of the peak.

  Thus, as numerical data, the wave number at the intersection of the peak of at least one chemical species and at least one straight line parallel to the wave number axis of the infrared spectrum is extracted.

(7) Inclination of at least one linear portion in infrared spectrum FIG. 7 shows an infrared spectrum in the region of 2600 to 2000 cm ⁻¹ (the region RN4 in FIG. 3) of the wavenumber. In this wave number region, a linear portion of the infrared spectrum appears.

  The linear portion of this infrared spectrum is linearly approximated, and the inclination is extracted. For example, an arbitrary orthogonal coordinate system is set with the wave number axis (horizontal axis) of the infrared spectrum as the x axis and the spectral intensity axis (vertical axis) as the y axis, and the inclination of the linear portion in this orthogonal coordinate system is the numerical value Extract. In the example of FIG. 7, in the new spectrum k1, the linear portion can be approximated by y = a1 (ie, slope = 0). In the spectrum k2 of the degraded product, the straight line portion can be approximated by y = a2x, and in the spectrum k3 of the end-of-life product, the straight line portion can be approximated by y = a3x.

  In the infrared spectrum of the unsaturated polyester resin, the influence of the scattering of infrared light is greater on the high wave number side than on the low wave number side, so that the deterioration progresses and the slope of the linear portion changes. Therefore, by quantifying the inclination of the linear portion of the infrared spectrum, it is possible to catch the indication of the deterioration of the diagnosis object from the change of the numerical data.

<Step of calculating surface resistivity of diagnostic object based on plural numerical data>
Next, the process procedure of the process (step S30) of calculating the surface resistivity of a diagnostic object based on several numerical value data among the insulation degradation diagnostic processes shown in FIG. 2 is demonstrated in detail.

  In this step, the surface resistivity of the diagnosis target is calculated by substituting a plurality of numerical data extracted from the infrared spectrum into a plurality of numerical data derived in advance and a relational expression of surface resistivity. In FIG. 8, a plurality of (n numbers) of numerical data obtained from an infrared spectrum are variables X 1, X 2,. X1, X2, ... Xn) are schematically shown. This relational expression is generated by executing the processing of steps S100 to S120 in FIG.

  As shown in FIG. 8, by substituting the numerical data X to be diagnosed into this relational expression Y = F (X1, X2,... Xn), the surface resistivity Y to be diagnosed is calculated. Can.

  As explained above, in the insulation degradation diagnosis method according to the first embodiment, a plurality of numerical data are extracted from the shape of the infrared spectrum to be diagnosed, and the surface resistance of the diagnosis target is based on the plurality of extracted numerical data. By calculating the rate, the deterioration of the diagnosis object is diagnosed. That is, according to the insulation degradation diagnosis method according to the first embodiment, the degradation of the diagnostic object is diagnosed based on the plurality of numerical data extracted from the shape of the infrared spectrum.

  As shown in Patent Document 1, also in the conventional degradation diagnosis method, a method of diagnosing degradation of a resin insulator by quantifying an infrared spectrum has been adopted. However, in the conventional degradation diagnosis method, it is feared that it takes a long time to catch signs of degradation of the resin insulator because it only measures the peak intensity of one kind of chemical species. As a result, it may happen that deterioration of the resin insulation has already progressed when a change appears in the peak intensity of the specific chemical species.

  On the other hand, in the first embodiment, in order to extract a plurality of numerical data from the shape of the infrared spectrum, not only the change of the peak intensity of a specific chemical species but also the minute change of the infrared spectrum Can be In particular, in the first embodiment, since a plurality of numerical data are extracted from the shapes of peaks derived from a plurality of chemical species appearing in an infrared spectrum and the shape of a linear portion of the infrared spectrum, infrared rays associated with deterioration It is possible to accurately capture minute changes in the spectrum. Then, the surface resistivity can be estimated with high accuracy by calculating the surface resistivity based on the plurality of extracted numerical data. As a result, since the degradation state of the diagnosis object can be diagnosed with high accuracy, it is possible to detect the degradation of the resin insulator at an early stage up to the life.

  In the insulation degradation diagnosis method according to the first embodiment, it is preferable that the plurality of numerical data extracted from the infrared spectrum be a digitization of a portion of the infrared spectrum in which a change due to age is significant. . However, the mode of change of the infrared spectrum due to deterioration varies depending on the type of the resin insulator. Therefore, by freely setting the numerical data to be extracted according to the resin insulator to be diagnosed, it is possible to realize highly accurate deterioration diagnosis regardless of the type of the diagnosis object.

Second Embodiment
In the first embodiment described above, as a method of generating a relational expression between a plurality of numerical data and surface resistivity, a configuration is described in which multivariate analysis is performed using surface resistivity as an objective variable and multiple numerical data as an explanatory variable. However, in the second embodiment, a configuration will be described in which a plurality of numerical data are converted into a single index using multivariate analysis, and a calibration curve indicating the relationship between the index and the surface resistivity is generated.

  The insulation degradation diagnosis device 10 according to the second embodiment is the same as the configuration of the insulation degradation diagnosis device 10 shown in FIG. 1, and thus the detailed description will not be repeated.

  FIG. 9 is a flowchart showing an insulation deterioration diagnosis process performed in the insulation deterioration diagnosis apparatus 10 according to the second embodiment.

  The insulation diagnosis process shown in FIG. 9 includes steps S25 and S35 instead of the steps S20 and S30 as compared with the insulation diagnosis process shown in FIG. Further, the insulation diagnosis process shown in FIG. 9 includes steps S140 and S150 instead of step S120. The other steps are the same as those in the insulation diagnosis process of FIG. 2, and thus the detailed description will not be repeated.

  Referring to FIG. 9, when the plurality of numerical data is extracted from the infrared spectrum acquired by spectroscopic device 1 in surface resistivity calculation device 2 (step S22), the plurality of numerical data is subjected to multivariate analysis To replace the single indicator (step S25). For example, using the Mahalanobis-Taguchi system method (hereinafter, referred to as MT method), a plurality of numerical data are represented as a single index (the Mahalanobis distance). In the second embodiment, the distance between the Mahalanobises is defined as "deterioration degree", taking the MT method as an example.

  Next, the surface resistivity of the diagnosis target is calculated using the calculated degree of deterioration (step S35). In step S35, the surface resistivity of the diagnosis target is calculated based on the calculated degree of degradation using a calibration curve indicating the relationship between the degree of degradation previously derived and the surface resistivity.

  A calibration curve indicating the relationship between the degree of deterioration and the surface resistivity used in the process of step S35 can be generated by the process procedure shown in steps S100 to S150. The processes of steps S100 to S150 are performed prior to infrared spectroscopic analysis of a diagnosis target.

  Specifically, first, a plurality of resin insulators whose surface resistivity is known are prepared. The plurality of resin insulators include a new resin insulator and at least one degraded resin insulator. Infrared spectroscopic analysis is performed on each of the plurality of resin insulators to obtain an infrared spectrum (step S100).

  Next, the infrared spectra of the plurality of acquired resin insulators are digitized (step S110). In step S110, a plurality of numerical data are extracted from the infrared spectrum for each resin insulator.

  Next, the plurality of extracted numerical data is replaced with the degree of deterioration using the MT method (step S140). Specifically, a plurality of numerical data extracted from a new infrared spectrum is used as a unit space, and a plurality of numerical data extracted from the infrared spectrum of a deteriorated product is used as a diagnostic target to determine the Mahalanobis distance. Then, with the calculated Mahalanobis distance as the degree of deterioration, the surface resistivity of the resin insulator is linked. As a result, a plurality of combinations of surface resistivity and degradation degree are generated.

  Based on the plurality of combinations generated, a calibration curve indicating the relationship between the degree of deterioration and the surface resistivity is created (step S150).

  FIG. 10 schematically shows a calibration curve showing the relationship between the degree of deterioration obtained from the infrared spectrum and the surface resistivity. This calibration curve is generated by executing the processing of steps S100 to S150 of FIG. As shown in FIG. 10, the surface resistivity R of the diagnostic object can be calculated by substituting the degradation degree D1 of the diagnostic object into the calibration curve.

  As described above, in the insulation degradation diagnosis method according to the second embodiment, a plurality of numerical data are extracted from the shape of the infrared spectrum to be diagnosed, and the plurality of extracted numerical data is a single index. The deterioration of the object to be diagnosed is diagnosed by calculating the surface resistivity of the object to be diagnosed using a calibration curve showing the relationship between the degree of deterioration and the surface resistivity. That is, also in the insulation degradation diagnosis method according to the second embodiment, as in the insulation degradation diagnosis method according to the first embodiment described above, a diagnosis target is obtained based on a plurality of numerical data extracted from the shape of the infrared spectrum. Deterioration is diagnosed. Therefore, also in the insulation degradation diagnosis method according to the second embodiment, the surface resistivity can be accurately estimated based on the plurality of extracted numerical data. As a result, since the deterioration state of the diagnosis object can be diagnosed with high accuracy, it is possible to detect the deterioration of the resin insulator at an early stage up to the life.

  In the following third to sixth embodiments, a deterioration diagnosis example of a diagnostic object using the insulation deterioration diagnosis method according to the first and second embodiments described above will be described.

Third Embodiment
In the insulation deterioration diagnosis according to the third embodiment, a plurality of numerical data is extracted from the shape of a peak derived from a single chemical species appearing in an infrared spectrum, and the surface to be diagnosed using the plurality of extracted numerical data The resistivity can be calculated.

  As an example, the peak intensity or peak wave number of a single chemical species in the infrared spectrum and the peak intensity or peak wave number of the chemical species in the first derivative spectrum or the second derivative spectrum are extracted. These numerical data are effective when diagnosing the deterioration state of the resin insulation in which the peak of another chemical species appears buried in the peak of one chemical species as it degrades. Preferably, the single species is nitrate or sulfate. In the infrared spectrum of the unsaturated polyester resin shown in FIG. 3, when quantifying the shape of the nitrate peak, the nitrate peak intensity and the nitrate peak intensity in the second derivative spectrum are extracted by: The signs of deterioration can be caught at an early stage.

  As another example, the number or point of intersection of the peak intensity or peak wave number of a single chemical species in the infrared spectrum with at least one straight line parallel to the peak of the chemical species and the wave number axis of the infrared spectrum Extract the wave number of These numerical data are effective when diagnosing the deterioration state of the resin insulation in which the peak of the chemical species appears or the peak position shifts with deterioration. In the infrared spectrum of the unsaturated polyester resin shown in FIG. 3, when quantifying the shape of the OH group or C = O group peak, the peak intensity of the chemical species, the peak of the chemical species and the infrared spectrum By extracting the number of intersections with at least one straight line parallel to the wave number axis of and the wave number of the intersections, it is possible to catch signs of deterioration at an early stage.

Fourth Embodiment
In the insulation degradation diagnosis according to the fourth embodiment, a plurality of numerical data are extracted from the shapes of peaks of a plurality of chemical species appearing in an infrared spectrum, and the surface resistivity of the diagnosis target is extracted using the plurality of extracted numerical data. The configuration can be calculated.

  In the insulation degradation diagnosis according to the fourth embodiment, for example, while the peak strength of two or more chemical species increases with deterioration, the peak strength of one chemical species increases while the peak of another chemical species increases. It is effective when diagnosing the deterioration state of the resin insulator etc. which intensity | strength reduces.

Embodiment 5
In the insulation degradation diagnosis according to the fifth embodiment, a plurality of numerical data are extracted from the shapes of the peak of at least one chemical species appearing in the infrared spectrum and the shape of the linear portion of the infrared spectrum and extracted. The surface resistivity of the diagnosis target can be calculated using numerical data.

  In the insulation degradation diagnosis according to the fifth embodiment, for example, when a degradation state of a resin insulator in which the peak intensity of at least one chemical species changes and the slope of the linear portion of the infrared spectrum changes with degradation is diagnosed. It is effective for

Sixth Embodiment
When the numerical value data of any one of the numerical data (1) to (7) extracted in the first embodiment has a correlation with the surface resistivity to be diagnosed, a resin insulator having a known surface resistivity A relational expression between the numerical data and the surface resistivity may be created, and the surface resistivity of the diagnosis target may be estimated using the created relational expression.

  As described above, according to the insulation degradation diagnosis method according to the first to sixth embodiments, the numerical data to be extracted can be set according to the state of degradation appearing in the resin insulator to be diagnosed. Therefore, regardless of the type of the diagnosis target, the deterioration of the diagnosis target can be diagnosed with high accuracy.

  In the first to sixth embodiments, the method for calculating the surface resistivity when the diagnosis target is unsaturated polyester resin is exemplified, but when the diagnosis target is an epoxy resin, a phenol resin, a silicone resin, etc. The surface resistivity can be calculated using the same method. In addition, since the change of the peak by insulation degradation is large compared with the infrared spectrum of other resin insulators, the infrared spectrum of unsaturated polyester resin can be quantified with higher precision.

  It should be understood that the embodiments disclosed herein are illustrative and non-restrictive in every respect. The scope of the present invention is indicated not by the embodiments described above but by the claims, and is intended to include all modifications within the meaning and scope equivalent to the claims.

  DESCRIPTION OF SYMBOLS 1 spectroscopy apparatus, 2 surface resistivity calculation apparatus, 3 output apparatus, 4 infrared spectrum digitization part, 5 standardization part, 6 numerical data extraction part, 7 surface resistivity calculation part, 8 output part, 10 insulation degradation diagnosis device .

Claims (14)

An insulation deterioration diagnosis method for diagnosing deterioration of a resin insulator, comprising:
Irradiating a surface of a resin insulator to be diagnosed with infrared light and dispersing the reflected light or transmitted light to obtain an infrared spectrum of the diagnostic target;
Extracting a plurality of numerical data from the shapes of peaks derived from a plurality of chemical species appearing in the infrared spectrum and the shape of a linear portion of the infrared spectrum;
Calculating a surface resistivity of the diagnosis target based on the plurality of numerical data extracted. The plurality of numerical data are
Peak intensity of at least one of the plurality of chemical species,
Peak wave number of at least one of the plurality of chemical species,
Peak intensity of at least one chemical species in the first derivative spectrum of the infrared spectrum,
Peak wave number of at least one chemical species in the first derivative spectrum,
Peak intensity of at least one chemical species in the second derivative spectrum of the infrared spectrum,
Peak wave number of at least one chemical species in the second derivative spectrum,
The number of intersections of at least one chemical species peak with at least one straight line parallel to the wave number axis of the infrared spectrum,
At least two of the wave number of the intersection of the peak of at least one chemical species and at least one straight line parallel to the wave number axis of the infrared spectrum, and the slope of at least one linear portion in the infrared spectrum The insulation degradation diagnostic method according to claim 1.   The insulation degradation diagnosis method according to claim 2, wherein the at least one chemical species includes at least one of an OH group, a COO group, a C—O group, a nitrate and a sulfate. The plurality of numerical data are
The peak intensity of the first chemical species among the plurality of chemical species,
Peak wave number of the first chemical species,
Peak intensity of the first chemical species in the first derivative spectrum,
Peak wave number of the first chemical species in the first derivative spectrum,
Peak intensity of the first chemical species in the second derivative spectrum,
Peak wave number of the first chemical species in the second derivative spectrum,
The number of intersections of the peak of the first chemical species and at least one straight line parallel to the wave number axis of the infrared spectrum, and the peak of the first chemical species and parallel to the wave number axis of the infrared spectrum The insulation degradation diagnosis method according to claim 1, wherein the method includes at least two of wave numbers of intersections with at least one straight line.   The insulation degradation diagnosis method according to claim 4, wherein the first chemical species is a nitrate or a sulfate.   In the step of calculating the surface resistivity of the diagnosis target, the plurality of numerical data obtained in advance and the relationship between the surface resistivity of the resin insulator and the plurality of numerical data extracted are used. The insulation degradation diagnostic method according to any one of claims 1 to 5, wherein the surface resistivity of the diagnostic target is calculated.   In the step of calculating the surface resistivity of the diagnosis target, the plurality of numerical data extracted is converted into a single index by multivariate analysis, and the single index and the resin insulation obtained in advance The insulation degradation diagnostic method according to claim 6, wherein the surface resistivity of the diagnosis target is calculated based on the converted single index using a calibration curve indicating a relationship with a surface resistivity of an object.   Using the plurality of numerical data extracted from the shapes of the peaks derived from the plurality of chemical species and the shapes of the linear portion appearing in the infrared spectrum of each of the plurality of resin insulators whose surface resistivity is known The insulation degradation diagnostic method according to claim 6, further comprising the step of generating a relational expression between a plurality of numerical data and the surface resistivity of the resin insulator.   The insulation degradation diagnosis method according to any one of claims 1 to 8, wherein the resin insulator is any of unsaturated polyester resin, epoxy resin, phenol resin, and silicone resin. An insulation deterioration diagnosis apparatus for diagnosing deterioration of a resin insulator, comprising:
A spectroscope for acquiring an infrared spectrum of the diagnostic object by irradiating the surface of the resin insulator to be diagnosed with infrared light and dispersing the reflected light or the transmitted light;
An extraction unit for extracting a plurality of numerical data from the shapes of peaks derived from a plurality of chemical species appearing in the infrared spectrum acquired by the spectroscopic device and the shape of a linear portion of the infrared spectrum;
And a calculation unit that calculates a surface resistivity of the diagnosis target based on the plurality of numerical data extracted by the extraction unit. The plurality of numerical data are
Peak intensity of at least one of the plurality of chemical species,
Peak wave number of at least one of the plurality of chemical species,
Peak intensity of at least one chemical species in the first derivative spectrum of the infrared spectrum,
Peak wave number of at least one chemical species in the first derivative spectrum,
Peak intensity of at least one chemical species in the second derivative spectrum of the infrared spectrum,
Peak wave number of at least one chemical species in the second derivative spectrum,
The number of intersections of at least one chemical species peak with at least one straight line parallel to the wave number axis of the infrared spectrum,
At least two of the wave number of the intersection of the peak of at least one chemical species and at least one straight line parallel to the wave number axis of the infrared spectrum, and the slope of at least one linear portion in the infrared spectrum The insulation degradation diagnostic device according to claim 10. The plurality of numerical data are
The peak intensity of the first chemical species among the plurality of chemical species,
Peak wave number of the first chemical species,
Peak intensity of the first chemical species in the first derivative spectrum,
Peak wave number of the first chemical species in the first derivative spectrum,
Peak intensity of the first chemical species in the second derivative spectrum,
Peak wave number of the first chemical species in the second derivative spectrum,
The number of intersections of the peak of the first chemical species and at least one straight line parallel to the wave number axis of the infrared spectrum, and the peak of the first chemical species and parallel to the wave number axis of the infrared spectrum The insulation deterioration diagnosis device according to claim 10, wherein the insulation deterioration diagnosis device includes at least two of wave numbers of intersections with at least one straight line.   The calculation unit is configured to calculate a surface resistivity of the diagnosis target based on the plurality of numerical data extracted by using a relational expression between the plurality of numerical data acquired in advance and the surface resistivity of the resin insulator. The insulation degradation diagnostic device according to any one of claims 10 to 12, wherein The calculation unit converts the plurality of extracted numerical data into a single index by performing multivariate analysis, and a relationship between the single index acquired in advance and the surface resistivity of the resin insulator. The insulation degradation diagnosis device according to any one of claims 10 to 12, wherein the surface resistivity of the diagnosis target is calculated based on the single index converted using a calibration curve indicating.

Images