Method for Distinguishing Insulating Oil Manufacturers and Models Based on Infrared Spectroscopy

Abstract

The present invention provides a method for distinguishing manufacturers and models of insulating oil based on infrared spectroscopy, comprising: collecting an infrared transmittance map of the insulating oil in a liquid cell; selecting the infrared transmittance map with relatively obvious differential features in the infrared transmittance maps as a preset map; comparing the preset map with typical maps of transformer oils from different manufacturers and different models; and determining the manufacturer and model information corresponding to the insulating oil according to the comparison result. The beneficial effect of the present invention is that by comparing the insulating oil to be confirmed with transformer oils from different manufacturers and different models by the method of the present invention, it is possible to visually and quickly distinguish transformer oils from different manufacturers and different models, so as to determine the manufacturer and model information of the insulating oil to be confirmed, and the result of the method of the present invention is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of high-voltage insulation, and more particularly, to a method for distinguishing manufacturers and models of insulating oil based on infrared spectroscopy. Background Art

[0002] In recent years, due to the rapid growth of domestic demand for transformer oil, the demand for naphthenic crude oil transformer oil has far exceeded the production capacity of PetroChina Kunlun Karamay Oilfield. In this situation, non-naphthenic transformer oil or transformer oil blended with naphthenic transformer oil has flowed into the market in large quantities. In order to ensure the supply of high-quality oil sources for high-voltage transformers and the safe and reliable use of transformer oil, and to prevent adverse effects on the safe operation of power grid equipment, there is an urgent need for a method to quickly and effectively distinguish transformer oils from different manufacturers and models. Summary of the Invention

[0003] In view of this, the present invention proposes a method for distinguishing manufacturers and models of insulating oil based on infrared spectroscopy, aiming to solve the problem of quickly and effectively distinguishing transformer oils from different manufacturers and models.

[0004] On the one hand, the present invention proposes a method for distinguishing manufacturers and models of insulating oil based on infrared spectroscopy, including: collecting an infrared transmittance map of the insulating oil in a liquid cell; selecting the infrared transmittance map with relatively obvious differential features in the infrared transmittance maps as a preset map; comparing the preset map with the typical maps of transformer oils from different manufacturers and models; and determining the manufacturer and model information corresponding to the insulating oil according to the comparison results.

[0005] Further, by performing similarity discrimination between the preset map and the typical maps of transformer oils from different manufacturers and models, the manufacturer and model information corresponding to the insulating oil is determined.

[0006] Further, when performing the similarity discrimination, the curve of the preset map and the curve of the typical map are determined, and the curve of the preset map is compared with the curve of the typical map, and the manufacturer and model information corresponding to the insulating oil is determined according to the comparison results.

[0007] Further, the similarity score Score of the similarity discrimination XY , is calculated according to the following formula:

[0008]

[0009] where m is the number of given data curve categories, N is the number of coordinate points of each piece of data, x is the ordinate value of variable X, is the mean value of the ordinate of variable X, and y is the ordinate value of variable Y. is the mean value of the vertical coordinate of variable Y, and y is the vertical coordinate value of variable Y; represents the mean value of the vertical coordinate of variable Y, x i represents the i-th vertical coordinate value of variable X, y i represents the i-th vertical coordinate value of variable Y.

[0010] Furthermore, the insulating oil within the range of 4000 cm -1 -500 cm -1 in the liquid cell is scanned to determine the infrared transmittance map.

[0011] Furthermore, the spectrum within the range of 700 cm -1 -500 cm -1 in the infrared transmittance map is selected as the preset spectrum

[0012] Furthermore, the liquid cell is a liquid cell with a fixed optical path length or a variable optical path length.

[0013] Furthermore, the optical path length of the liquid cell is 0.1 mm - 1.5 mm.

[0014] Compared with the prior art, the beneficial effect of the present invention is that by comparing the insulating oil to be confirmed with transformer oils from different manufacturers and different models through the method of the present invention, it is possible to intuitively and quickly distinguish transformer oils from different manufacturers and different models, thereby determining the manufacturer and model information of the insulating oil to be confirmed, and the result of the method of the present invention is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0016] Figure 1 is the flowchart of the method for distinguishing the manufacturer and model of insulating oil based on infrared spectroscopy provided by the embodiment of the present invention;

[0017] Figure 2 is the spectrum of the mainstream transformer oil provided by the embodiment of the present invention;

[0018] Figure 3 is Figure 2 the infrared transmittance map of the spectrum of the mainstream transformer oil in -1 the range of 700 cm -1 -500 cm;

[0019] Figure 4The spectrogram of transformer oils of different models under the same production process of PetroChina provided by the embodiments of the present invention;

[0020] Figure 5 The spectrogram of transformer oils of different models under different production processes of CNOOC provided by the embodiments of the present invention. Detailed implementation manners

[0021] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be completely conveyed to those skilled in the art. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. Hereinafter, the present invention will be described in detail with reference to the drawings and in combination with the embodiments.

[0022] Refer to Figure 1 As shown, this embodiment provides a method for distinguishing insulating oil manufacturers and models based on infrared spectroscopy, including the following steps:

[0023] Step 1 S101: Collect the infrared transmittance diagram of the insulating oil in the liquid cell;

[0024] Step 2 S102: Select the infrared transmittance diagram with relatively obvious differential features in the infrared transmittance diagrams as the preset diagram;

[0025] Step 3 S103: Compare the preset diagram with the typical diagrams of transformer oils of different manufacturers and different models;

[0026] Step 4 S104: Determine the manufacturer and model information corresponding to the insulating oil according to the comparison result.

[0027] Specifically, the insulating oil in the range of 4000 cm -1 -500 cm -1 in the liquid cell is scanned to determine the infrared transmittance diagram. That is, the insulating oil in the liquid cell is the insulating oil for which the manufacturer and model information is to be confirmed. By scanning the infrared spectrum of the insulating oil to be confirmed, its infrared transmittance diagram is obtained.

[0028] Specifically, the diagram in the range of 700 cm -1 -500 cm -1 in the infrared transmittance diagram is selected as the preset diagram. The diagram data in the range of 700 cm -1 -500 cm -1 in the obtained infrared transmittance diagram is intercepted as the preset diagram for subsequent processing.

[0029] Specifically, an infrared spectrometer is used to scan the insulating oil in the liquid cell, and the scanning range is 4000 cm -1 -500 cm -1 , and its infrared transmittance map is selected. Specifically, the difference characteristics of the infrared transmittance map in the range of 700 cm -1 -500 cm -1 are relatively obvious. Therefore, it is intercepted as the final preset map for subsequent data comparison.

[0030] Specifically, after selecting the preset map, it is compared one by one with the infrared maps (i.e., typical maps) of insulating oils of existing known manufacturers and models, so as to confirm the manufacturer and model information of the insulating oil corresponding to the preset map. Specifically, for the infrared transmittance map in the range of 700 cm -1 -500 cm -1 with relatively obvious difference characteristics, direct observation, image recognition or other qualitative or quantitative methods can be used to compare the test data with the typical map, find the closest map, and then confirm the manufacturer and model information of the insulating oil to be confirmed.

[0031] It can be seen that by comparing the insulating oil to be confirmed with transformer oils of different manufacturers and models by the method of the present invention, it is possible to intuitively and quickly distinguish transformer oils of different manufacturers and models, so as to determine the manufacturer and model information of the insulating oil to be confirmed, and the result of the method of the present invention is more accurate.

[0032] Specifically, the above liquid cell is a liquid cell with a fixed path length or a variable path length. The path length of the liquid cell is 0.1 mm - 1.5 mm.

[0033] Specifically, when comparing the preset map with the typical map, the similarity between the preset map and the typical maps of transformer oils of different manufacturers and models is judged to determine the manufacturer and model information corresponding to the insulating oil.

[0034] Specifically, when performing similarity judgment, the curve of the preset map and the curve of the typical map are determined, and the curve of the preset map is compared with the curve of the typical map, and the manufacturer and model information corresponding to the insulating oil is determined according to the comparison result.

[0035] Specifically, the similarity score Score XY of the similarity judgment is calculated according to the following formula:

[0036]

[0037] Among them, m is the number of given data curve categories, N is the number of coordinate points of each piece of data, x is the vertical coordinate value of variable X, is the average value of the vertical coordinates of variable X, y is the vertical coordinate value of variable Y, is the average value of the vertical coordinates of variable Y, y——the vertical coordinate value of variable Y; represents the average value of the vertical coordinates of variable Y, x i represents the i-th vertical coordinate value of variable X, y i represents the i-th vertical coordinate value of variable Y.

[0038] It can be seen that by performing similarity discrimination on the preset spectrum and the typical spectrum, the data categories can be effectively distinguished, and the results can be expressed more intuitively. Furthermore, the manufacturer and model information of the insulating oil can be quickly and effectively confirmed, improving the work efficiency.

[0039] In specific implementation, taking the spectra of I-40 of Sinopec, KI45X of CNPC, S4 of Shell, KI50X of CNPC, and HI50X of CNOOC Taizhou as examples for illustration, the following method can be adopted. When finding the closest spectrum, first obtain the infrared spectrum of the insulating oil to be confirmed, intercept the preset spectrum, and obtain the curve of the preset spectrum. At the same time, obtain the curves of the spectra of I-40 of Sinopec, KI45X of CNPC, S4 of Shell, KI50X of CNPC, and HI50X of CNOOC Taizhou respectively. Then, these curves are the spectral curves of the current mainstream transformer oils: I-40 of Sinopec, KI45X of CNPC, S4 of Shell, KI50X of CNPC, and HI50X of CNOOC Taizhou in sequence.

[0040] Preferably, when obtaining the spectral curves of I-40 of Sinopec, KI45X of CNPC, S4 of Shell, KI50X of CNPC, and HI50X of CNOOC Taizhou, it is preferably to intercept the infrared transmittance map with relatively obvious difference features of each spectrum, that is, the infrared transmittance map in the range of 700cm -1 -500cm -1 is compared to improve the acquisition efficiency of the comparison result, that is, the work efficiency can be effectively improved.

[0041] Specifically, obtain the curve of a preset spectrum and calculate the similarity between this curve and the curves of five typical spectra.

[0042] Specifically, when tested by traditional similarity calculation methods, the three correlation coefficients have poor discrimination effects on K145X and K150X. Among them, the Manhattan distance algorithm can effectively distinguish data categories, but the distance values are not intuitive enough. In view of this, a similarity discrimination method that improves the correlation coefficient through the Manhattan distance is given, so that it can effectively distinguish data categories and more intuitively express the results.

[0043] Specifically, the similarity score Score X，Y is equal to the Pearson correlation coefficient ρ X，Y multiplied by a correction coefficient λ X，Y , where λ is a coefficient with a value range in (0, 1) obtained from the Manhattan distance, and is used to correct the correlation of data with large similarities but belonging to different categories.

[0044] Specifically, the reason for using the exponential function for the correction coefficient is as follows: The value range of the exponential function on the negative half-axis is between (0, 1), which conforms to the probability range; the exponential function is monotonically increasing on the negative half-axis. The greater the Manhattan distance, the smaller the value after taking the negative number, the smaller its exponential value, and the smaller the finally calculated similarity score, which meets the requirements of the correction coefficient.

[0045] Specifically, through the above improvement, the calculation formula of the final similarity discrimination algorithm is as follows:

[0046]

[0047] Among them, Score XY is the similarity score of variables X and Y, with a value range in (0, 1), indicating the similarity degree between the two; λ X，Y is the correction coefficient of the correlation coefficient, calculated through the Manhattan distance, with a value range in (0, 1); ρ X，Y is the Pearson correlation coefficient, with a value range in (0, 1); d XY is the Manhattan distance between variables X and Y; m is the number of given data curve categories; N is the number of coordinate points of each data. Taking the infrared data intercepted from 700cm -1 -500cm -1 as an example, then N = 200; x is the ordinate value of variable X; is the average ordinate value of variable X; y is the ordinate value of variable Y; is the average ordinate value of variable Y, x i represents the i-th ordinate value of variable X, y i represents the i-th ordinate value of variable Y.

[0048] Specifically, data verification:

[0049] 1) Intercept 700cm-1 -500 cm -1 Infrared data;

[0050] 2) Given five fixed curves: I-40, K150X-1, K145X-1, S4, H150X-1;

[0051] 3) Test the correlation between the new data and the given five data, and the one with the highest similarity score is the determined category.

[0052] Table 1 shows the correlation between Test Data 1: K145X and the given five data curves

[0053]

[0054] Table 1

[0055] It can be seen from Table 1 that the correlation coefficients of K145X with K145X-1 and K150X-1 are both very high. At this time, improving the similarity score can widen the score gap between the two, which can better distinguish the two types of data, and thus can effectively classify K145X.

[0056] Table 2 shows the correlation between Test Data 2: K150X-2 and the given five data curves

[0057]

[0058] It can be seen from Table 2 that the correlation coefficients of K150X-2 with K150X-1 and K145X-1 are both very high. At this time, improving the similarity score can widen the score gap between the two, which can better distinguish the two types of data, and thus can effectively classify K150X-2.

[0059] In summary, it can be seen that improving the similarity discrimination algorithm has a good effect on distinguishing data categories. At the same time, by comparing the preset spectrum with the typical spectrum through the above method, the insulating oil manufacturers and models corresponding to the preset spectrum can be effectively and quickly distinguished.

[0060] The above liquid cell is a liquid cell with a fixed path length or a variable path length, and the recommended path length is 0.1 mm - 1.5 mm.

[0061] Refer to Figure 2 as shown, specifically, Figure 2 The several curves in are the spectra of the current mainstream transformer oils on the market: I-40 of Sinopec, KI45X of PetroChina, S4 of Shell, KI50X of PetroChina, and HI50X of CNOOC Taizhou.

[0062] Combined with Figure 3As shown, specifically, in the spectrograms of I-40 of Sinopec, KI45X of PetroChina, S4 of Shell, KI50X of PetroChina, and HI50X of CNOOC Taizhou, the infrared transmittance in the range of 700 cm -1 -500 cm -1 is as shown in Figure 3 This figure includes the transformer oil of CNOOC Qingdao and the paraffin-based and naphthenic-based transformer oils of Taizhou.

[0063] Specifically, when using the above method to distinguish transformer oils of different models from the same manufacturer, there are the following two situations:

[0064] Refer to Figure 4 As shown, one is taking PetroChina as an example, with different models of transformer oils under the same production process. As Figure 4 shown, its spectrogram contains KI50X, KI25X, KI45X, and duplicate samples of each sample. Although the differences in the spectrograms of different models of transformer oils are not as significant as in other cases in this situation, the differences are much greater than the repeatability of each model, which is sufficient to distinguish different models. In addition, it can be confirmed by combining the kinematic viscosity of the oil sample at -30 °C.

[0065] Refer to Figure 5 As shown, the other is taking CNOOC as an example, with different models of transformer oils under different production processes. As Figure 5 shown, its spectrogram contains HI50X, naphthenic-based transformer oil, paraffin-based transformer oil, and duplicate samples of each sample. In this case, the differences in the spectrograms of different models of transformer oils are similar to those between different manufacturers, and it is relatively easier to distinguish compared to the first situation.

[0066] It can be seen that by using the method in this embodiment, it is possible to effectively distinguish the situation where there is repeatability in transformer oils of the same model from the same manufacturer. From Figure 4 and Figure 5 shown in the two spectrograms, it can be seen that the test repeatability of transformer oils of the same model from the same manufacturer is very stable, and the test error is much lower than the differences between different manufacturers or different models of transformer oils.

[0067] When the above method is specifically implemented, if it is uncertain whether a certain transformer oil is Karamay transformer oil, using this method for testing can effectively determine whether it is a commercially available Karamay transformer oil. At the same time, if it is uncertain whether a certain transformer oil is KI45X or KI50X transformer oil, using this method for testing can effectively determine whether it is a commercially available KI45X or KI50X transformer oil.

[0068] It can be understood that the present invention guards against scanning the insulating oil in the liquid cell by using an infrared spectrometer, and the scanning range is 4000 cm -1-500 cm -1 Select its infrared transmittance spectrum. Using the transmittance spectrum instead of the absorption spectrum, the differences between different transformer oils are more significant and can be identified by the observation method. Meanwhile, select the infrared transmittance spectrum from 700 cm -1 -500 cm -1 For the infrared transmittance spectrum, use direct observation, image recognition or other qualitative or quantitative methods to compare the test data with the typical spectrum and find the closest spectrum. Specifically, compare the insulating oil to be confirmed with transformer oils from different manufacturers and different models, which can intuitively and quickly distinguish transformer oils from different manufacturers and different models, thereby determining the manufacturer and model information of the insulating oil to be confirmed, and the results of the method of the present invention are more accurate.

[0069] Specifically, when the above embodiments are specifically implemented, they can be carried out according to the following process:

[0070] 1. Sampling and storage of samples

[0071] 1.1 Sampling is carried out in accordance with the provisions of GB7597.

[0072] 1.2 The samples should be covered to avoid heat and direct light, and filtered with dry filter paper before use.

[0073] 2. Preparation of reagents and tools

[0074] 2.1 A 10 mL glass syringe, absorbent paper for laboratory use, a gas mask, and a rubber bulb.

[0075] 2.2 Petroleum ether, alcohol, and carbon tetrachloride of analytical grade.

[0076] 3. Selection and installation of the liquid cell

[0077] 3.1 A liquid cell with a path length of 0.1 mm - 1.5 mm and a sodium chloride or zinc selenide window. The sodium chloride cell should be stored in a dry and sealed manner to prevent moisture absorption from the air and operated under an infrared drying lamp.

[0078] 3.2 Clean the windows and gaskets of the liquid cell for later use (clean with a 1:1 mixture of petroleum ether and anhydrous ethanol of analytical grade) and blow dry with a rubber bulb.

[0079] 3.3 When installing the liquid cell, pay attention to the correspondence between the oil inlet / outlet and the window, and tighten the hand-tightening screws diagonally.

[0080] 3.4 Clean the installed liquid cell, place it in the instrument test position for background collection, and measure the O-H stretching vibration absorption value (corrected peak height) from 700 cm-1 to 500 cm-1 to be <0.005. Otherwise, clean it again with anhydrous ethanol (and carbon tetrachloride if necessary, wearing a gas mask) until the requirements are met.

[0081] 3.5. If the requirements in 3.4 cannot be met, the window pane shall be cleaned, polished or discarded.

[0082] 4. Preparation for the test

[0083] 4.1. Prepare a liquid cell with an optical path length of 0.1 mm - 1.5 mm that meets the test requirements and is equipped with a sodium chloride or zinc selenide window pane.

[0084] 4.2. Slowly inject the sample into the slightly inclined liquid cell (with the outlet at a higher position) using a syringe that has been cleaned with the sample. Cover the other outlet of the liquid cell with absorbent paper to prevent oil from overflowing and contaminating the window; there should be no air bubbles in the liquid cell.

[0085] 4.3. Place the liquid cell filled with the test sample into the test position of the infrared spectrometer. After collecting the sample, compare the shapes and heights of the curves in the spectral region. Different types of transformer oils can be distinguished by the naked eye. In addition, further qualitative and quantitative analysis of data discretization can be carried out, and the types of oils can be judged according to the similarity ranking with known curves.

[0086] It can be seen that by comparing the insulating oil to be confirmed with transformer oils from different manufacturers and different models using the method of the present invention, different transformer oils from different manufacturers and different models can be intuitively and quickly distinguished, thereby determining the manufacturer and model information of the insulating oil to be confirmed, and the results of the method of the present invention are more accurate.

[0087] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A method for distinguishing insulating oil manufacturers and models based on infrared spectroscopy, characterized in that, Including: Collecting the infrared transmittance map of the insulating oil in the liquid cell; Selecting the infrared transmittance map with relatively obvious differential features in the infrared transmittance maps as the preset map; Take the spectrum within the range of 700 cm -1 - 500 cm -1 in the infrared transmittance graph as the preset graph; Comparing the preset map with the typical maps of transformer oils from different manufacturers and of different models; By performing similarity discrimination between the preset map and the typical maps of transformer oils from different manufacturers and of different models, To determine the manufacturer and model information corresponding to the insulating oil; When performing the similarity discrimination, determining the curve of the preset map and the curve of the typical map, comparing the curve of the preset map with the curve of the typical map, and determining the manufacturer and model information corresponding to the insulating oil according to the comparison result; the calculation formula of the similarity discrimination algorithm is as follows: Wherein, ScoreXY is the similarity score of variables X and Y, and the value ranges between (0, 1), indicating the similarity degree between the two; λ X，Y is the correction coefficient of the correlation coefficient, calculated by the Manhattan distance, and its value ranges between (0, 1); ρ X，Y is Pearson The correlation coefficient ranges from (0, 1); d XY is the Manhattan distance between variables X and Y; m is the number of given data curve categories; N is the number of coordinate points for each data, taking the intercepted 700 cm -1 -500 cm -1 infrared data as an example, then N = 200; x is the ordinate value of variable X; is the mean value of the ordinate of variable X; y is the ordinate value of variable Y; is the mean value of the ordinate of variable Y, x i represents the i-th ordinate value of variable X, y i represents the i-th ordinate value of variable Y; Determining the manufacturer and model information corresponding to the insulating oil according to the comparison result.

2. The method for distinguishing insulating oil manufacturers and models based on infrared spectroscopy according to claim 1, characterized in that, Scan the insulating oil within the range of 4000 cm -1 - 500 cm -1 in the liquid pool to determine the infrared transmittance map.

3. The method for distinguishing insulating oil manufacturers and models based on infrared spectroscopy according to claim 1, wherein, The liquid cell is a liquid cell with a fixed optical path length or a variable optical path length.

4. The method for distinguishing insulating oil manufacturers and models based on infrared spectroscopy according to claim 1, wherein The optical path length of the liquid cell is 0.1 mm - 1.5 mm.

Images