A method for improving the repeatability of Fe element testing in oil analyzers

Abstract

The invention belongs to the field of oil analyzers, and provides a method for improving the repeatability of Fe element testing of oil analyzers: signals of standard oil with known Fe element content are collected under specific conditions, the collected signals are specifically processed by using chemometrics software using a partial least squares method, a correlation model between the Fe element content and a spectrum of an oil analyzer is established under specific parameters, the spectrum of a sample to be tested is collected under the same conditions, the spectrum of the sample to be tested is substituted into the correlation model, and the correlation model can quickly calculate the content of the Fe element in the sample to be tested. The method can significantly improve the repeatability of Fe element testing of the oil analyzer.

Description

Technical Field

[0001] The invention belongs to the field of oil analyzers and relates to a method for improving the repeatability of Fe element testing of an oil analyzer. Background Art

[0002] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention, and should not necessarily be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art.

[0003] Oil analyzer is a spectrometer that can directly measure the types and contents of multiple elements in engine lubricating oil. By testing the types and contents of multiple wear elements in engine lubricating oil, rapid and non-destructive monitoring of the engine can be achieved, and it has been widely used in many fields. Fe is the most common wear element in engine lubricating oil. The presence of Fe indicates that Fe-containing parts in the engine, such as gears, crankshafts, etc., have worn. The test repeatability is crucial for the evaluation of engine wear status monitoring.

[0004] The sampling method of the oil analyzer is to rotate the graphite disk electrode. However, since the graphite disk electrode is a disposable consumable, there are inevitably errors in porosity, density, size, etc. between the graphite disk electrodes, which leads to different actual injection volumes for each test of the oil analyzer. Different injection volumes naturally lead to different excited signal intensities, thus affecting the repeatability of the test results. To solve this problem, the existing correction method is to use hydrogen as an internal standard, thereby improving the test repeatability to a certain extent.

[0005] The existing technologies all establish working curves through the signal data of the characteristic spectrum line of the Fe element (usually 259.94nm); when the hydrogen element is used as the internal standard, the signal data corresponding to the characteristic spectrum line of the hydrogen element (usually 483.10nm) is also used. The signal data used is limited and is easily affected by factors such as signal-to-noise ratio, resolution, and optical path stability. Therefore, the existing methods have limited ability to improve test repeatability.

[0006] Therefore, finding a correction method to better improve the repeatability of Fe element testing in oil analyzers has always been one of the urgent problems to be solved in this field. Summary of the invention

[0007] In order to solve the above problems, the present invention provides a method for improving the repeatability of Fe element measurement of an oil analyzer based on a chemometric algorithm.

[0008] In order to achieve the above object, the present invention adopts the following technical solution:

[0009] The present invention provides a method for improving the repeatability of Fe element testing of an oil analyzer, comprising:

[0010] An oil analyzer is used to collect spectra of multiple standard oils of different concentrations with known Fe content ranging from 1 mg / kg to 100 mg / kg;

[0011] The collected spectra were processed in turn by first-order differentiation, 5-11-point Norris Derivativ smoothing, and standard canonical transformation;

[0012] The partial least square method was used in the wavelength range of 200nm to 500nm, and the number of principal factors was 6 to 8 to establish the correlation model between Fe element content and the spectrum of oil analyzer.

[0013] Use the oil analyzer to collect the spectrum of the sample to be tested under the same conditions;

[0014] Substituting the spectrum of the sample to be tested into the above correlation model, the correlation model can quickly calculate the content of Fe element in the sample to be tested.

[0015] Beneficial Effects of the Invention

[0016] The beneficial effects of the present invention include but are not limited to:

[0017] (1) The present invention involves all the data of the full band or the optimized post-band in the calculation, and extracts the effective information in the full band or the optimized post-band through a chemometric algorithm, thereby improving the adverse effects of factors such as signal-to-noise ratio, resolution, and optical path stability on the repeatability of Fe element testing, and improving the repeatability of the test results.

[0018] (2) The method of the present invention is simple, practical and easy to promote. DETAILED DESCRIPTION

[0019] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0020] A method for improving the repeatability of Fe element measurement in oil analyzers based on chemometric algorithms.

[0021] The present invention achieves the above object through the following steps:

[0022] (1) Using an oil analyzer to collect spectra of multiple standard oils with different concentrations of known Fe content ranging from 1 mg / kg to 100 mg / kg;

[0023] The acquisition conditions are as follows: the spectrum acquisition range is 200nm to 500nm, pre-combustion is 3 to 6 seconds, and then the signal is continuously acquired 20 to 30 times at an acquisition frequency of 0.3 to 0.6 seconds and averaged, and each sample is acquired 5 times.

[0024] (2) The collected spectra were subjected to first-order differentiation, 5- to 11-point Norris Derivativ smoothing, and standard canonical transformation (SNV);

[0025] (3) using chemometric software using partial least squares (PLS) in the wavelength range of 200 nm to 500 nm, preferably in the wavelength range of 220 nm to 280 nm plus 320 nm to 460 nm, with the number of principal factors being 6 to 8, to establish a correlation model between the Fe element content and the oil analyzer spectrum;

[0026] (4) using an oil analyzer to collect a spectrum of the sample to be tested under the same conditions;

[0027] (5) Substitute the spectrum of the sample to be tested into the above correlation model, and the correlation model can quickly calculate the content of Fe element in the sample to be tested.

[0028] In some embodiments, the pre-combustion is 4 seconds.

[0029] In some embodiments, the acquisition frequency is 0.4 seconds.

[0030] In some embodiments, the signal is collected 25 times continuously and averaged.

[0031] In some embodiments, Norris Derivativ smoothing uses 7 points.

[0032] In some embodiments, the wavelength range is 220nm to 280nm plus 320nm to 460nm

[0033] In some embodiments, the number of prime factors is 7.

[0034] The present invention will be further described in detail below in conjunction with specific embodiments. It should be noted that the specific embodiments are intended to explain the present invention rather than to limit it.

[0035] Example 1

[0036] (1) Using an oil analyzer, collect spectra of standard oils with known Fe content of 1 mg / kg, 10 mg / kg, 30 mg / kg, 50 mg / kg, and 100 mg / kg;

[0037] The acquisition conditions were as follows: the spectrum acquisition range was 200 nm to 500 nm, pre-combustion was 4 seconds, and then the signal was continuously acquired 25 times at an acquisition frequency of 0.4 seconds and averaged, and each sample was acquired 5 times.

[0038] (2) The collected spectra were subjected to first-order differentiation, 7-point Norris Derivativ smoothing, and standard canonical transformation (SNV);

[0039] (3) The correlation model between Fe content and oil analyzer spectra was established using the partial least squares (PLS) method with chemometric software in the wavelength range of 220 nm to 280 nm plus 320 nm to 460 nm, with the number of principal factors being 7;

[0040] (4) using an oil analyzer to collect a spectrum of the sample to be tested under the same conditions;

[0041] (5) Substitute the spectrum of the sample to be tested into the above correlation model, and the correlation model can quickly calculate the content of Fe element in the sample to be tested.

[0042] Example 2

[0043] (1) Using an oil analyzer, collect spectra of standard oils with known Fe content of 1 mg / kg, 10 mg / kg, 30 mg / kg, 50 mg / kg, and 100 mg / kg;

[0044] The acquisition conditions were as follows: the spectrum acquisition range was 200 nm to 500 nm, pre-combustion was performed for 3 seconds, and then the signal was continuously acquired 20 times at an acquisition frequency of 0.5 seconds and averaged, and each sample was acquired 5 times.

[0045] (2) The collected spectra were subjected to first-order differentiation, 8-point Norris Derivativ smoothing, and standard canonical transformation (SNV);

[0046] (3) The correlation model between Fe content and oil analyzer spectra was established using the partial least squares (PLS) method in the wavelength range of 200 nm to 500 nm with 6 principal factors using chemometric software;

[0047] (4) using an oil analyzer to collect a spectrum of the sample to be tested under the same conditions;

[0048] (5) Substitute the spectrum of the sample to be tested into the above correlation model, and the correlation model can quickly calculate the content of Fe element in the sample to be tested.

[0049] Example 3

[0050] (1) Using an oil analyzer, collect spectra of standard oils with known Fe content of 1 mg / kg, 10 mg / kg, 30 mg / kg, 50 mg / kg, and 100 mg / kg;

[0051] The acquisition conditions were as follows: the spectrum acquisition range was 200 nm to 500 nm, pre-combustion was performed for 5 seconds, and then the signal was continuously acquired 25 times at an acquisition frequency of 0.4 seconds and averaged, and each sample was acquired 5 times.

[0052] (2) The collected spectra were subjected to first-order differentiation, 5-point Norris Derivativ smoothing, and standard canonical transformation (SNV);

[0053] (3) The correlation model between Fe content and oil analyzer spectra was established using the partial least squares (PLS) method in the wavelength range of 200 nm to 500 nm with 7 principal factors using chemometric software;

[0054] (4) using an oil analyzer to collect a spectrum of the sample to be tested under the same conditions;

[0055] (5) Substitute the spectrum of the sample to be tested into the above correlation model, and the correlation model can quickly calculate the content of Fe element in the sample to be tested.

[0056] Example 4

[0057] (1) Using an oil analyzer, the spectra of standard oils with known Fe content of 1 mg / kg, 20 mg / kg, 40 mg / kg, 60 mg / kg, and 100 mg / kg were collected;

[0058] The acquisition conditions were as follows: the spectrum acquisition range was 200 nm to 500 nm, pre-combustion was performed for 6 seconds, and then the signal was continuously acquired 30 times at an acquisition frequency of 0.3 seconds and averaged, and each sample was acquired 5 times.

[0059] (2) The collected spectra were subjected to first-order differentiation, 11-point Norris Derivativ smoothing, and standard canonical transformation (SNV);

[0060] (3) The correlation model between Fe content and oil analyzer spectra was established using the partial least squares (PLS) method with chemometric software in the wavelength range of 220 nm to 280 nm plus 320 nm to 460 nm, with the number of principal factors being 7;

[0061] (4) using an oil analyzer to collect a spectrum of the sample to be tested under the same conditions;

[0062] (5) Substitute the spectrum of the sample to be tested into the above correlation model, and the correlation model can quickly calculate the content of Fe element in the sample to be tested.

[0063] Example 5

[0064] (1) Using an oil analyzer, the spectra of standard oils with known Fe content of 1 mg / kg, 30 mg / kg, 60 mg / kg, 90 mg / kg, and 100 mg / kg were collected;

[0065] The acquisition conditions were as follows: the spectrum acquisition range was 200 nm to 500 nm, pre-combustion was 4 seconds, and then the signal was continuously acquired 28 times at an acquisition frequency of 0.6 seconds and averaged, and each sample was acquired 5 times.

[0066] (2) The collected spectra were subjected to first-order differentiation, 9-point Norris Derivativ smoothing, and standard canonical transformation (SNV);

[0067] (3) The correlation model between Fe content and oil analyzer spectra was established using chemometric software using partial least squares (PLS) in the wavelength range of 200 nm to 500 nm with 8 principal factors;

[0068] (4) using an oil analyzer to collect a spectrum of the sample to be tested under the same conditions;

[0069] (5) Substitute the spectrum of the sample to be tested into the above correlation model, and the correlation model can quickly calculate the content of Fe element in the sample to be tested.

[0070] Comparative Example 1

[0071] (1) Using an oil analyzer to collect signal data of standard oils with known Fe content of 1 mg / kg, 10 mg / kg, 30 mg / kg, 50 mg / kg, and 100 mg / kg;

[0072] The acquisition conditions were as follows: the spectrum acquisition range was 200 nm to 500 nm, pre-combustion was 4 seconds, and then the signal was continuously acquired 25 times at an acquisition frequency of 0.4 seconds and averaged, and each sample was acquired 5 times.

[0073] (2) Establishing a working curve based on the signal data corresponding to the characteristic spectrum line 259.94 nm of the Fe element;

[0074] (3) using an oil analyzer to collect signal data of the sample to be tested under the same conditions;

[0075] (4) Substituting the signal data of the sample to be tested into the above working curve, the content of Fe element in the sample to be tested can be quickly calculated.

[0076] Comparative Example 2

[0077] (1) Using an oil analyzer to collect signal data of standard oils with known Fe content of 1 mg / kg, 10 mg / kg, 30 mg / kg, 50 mg / kg, and 100 mg / kg;

[0078] The acquisition conditions were as follows: the spectrum acquisition range was 200 nm to 500 nm, pre-combustion was 4 seconds, and then the signal was continuously acquired 25 times at an acquisition frequency of 0.4 seconds and averaged, and each sample was acquired 5 times.

[0079] (2) Using the signal data corresponding to the characteristic spectrum line of Fe element at 259.94 nm and the signal data corresponding to the characteristic spectrum line of hydrogen element at 483.10 nm as internal standards, a working curve is established;

[0080] (3) using an oil analyzer to collect signal data of the sample to be tested under the same conditions;

[0081] (4) Substituting the signal data of the sample to be tested into the above working curve, the content of Fe element in the sample to be tested can be quickly calculated.

[0082] In Examples 1 to 5 and Comparative Examples 1 to 2, standard oils having Fe contents of 10 mg / kg, 30 mg / kg, 50 mg / kg and 100 mg / kg, and used engine lubricants having Fe contents of 17 mg / kg and 62 mg / kg, respectively, were measured 7 times, and the relative standard deviations were calculated as shown in the table below.

[0083]

[0084] It can be seen from the table that the relative standard deviations measured in Examples 1 to 5 are all between 2.5% and 3.3% for standard oils with different concentrations of Fe element and used engine lubricating oils, which are better than 5.8% to 6.4% in Comparative Example 1 and 4.3% to 4.9% in Comparative Example 2.

[0085] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for improving the repeatability of Fe element testing of an oil analyzer, characterized in that: include: An oil analyzer is used to collect spectra of multiple standard oils with different concentrations of known Fe content ranging from 1 mg / kg to 100 mg / kg; The collected spectra were processed in turn by first-order differentiation, 5-11-point Norris Derivativ smoothing, and standard canonical transformation; The partial least square method was used in the wavelength range of 200nm to 500nm, and the number of principal factors was 6 to 8 to establish the correlation model between Fe element content and the spectrum of oil analyzer. Use the oil analyzer to collect the spectrum of the sample to be tested under the same conditions; Substitute the spectrum of the sample to be tested into the above correlation model, and the correlation model can quickly calculate the content of Fe element in the sample to be tested; The oil analyzer is a rotating disk electrode emission spectrometer.

2. The method for improving the repeatability of Fe element testing of an oil analyzer according to claim 1, characterized in that: The acquisition conditions are as follows: the spectrum acquisition range is 200nm to 500nm, pre-combustion is 3 to 6 seconds, and then the signal is continuously acquired 20 to 30 times at an acquisition frequency of 0.3 to 0.6 seconds and averaged, and each sample is acquired at least 5 times.

3. The method for improving the repeatability of Fe element testing of an oil analyzer according to claim 1, characterized in that: Pre-burn for 4 seconds.

4. The method for improving the repeatability of Fe element testing of an oil analyzer according to claim 1, characterized in that: The acquisition frequency is 0.4 seconds.

5. The method for improving the repeatability of Fe element testing of an oil analyzer according to claim 1, characterized in that: The signal was collected 25 times continuously and averaged.

6. The method for improving the repeatability of Fe element testing of an oil analyzer according to claim 1, characterized in that: The spectra of standard oils with known Fe content of 1 mg / kg, 10 mg / kg, 30 mg / kg, 50 mg / kg and 100 mg / kg were collected.

7. The method for improving the repeatability of Fe element testing of an oil analyzer according to claim 1, characterized in that: NorrisDerivativ smoothing uses 7 points.

8. The method for improving the repeatability of Fe element testing of an oil analyzer according to claim 1, characterized in that: The wavelength range is 220nm~280nm and 320nm~460nm.

9. The method for improving the repeatability of Fe element testing of an oil analyzer according to claim 1, characterized in that: The number of principal factors used is 7.

Images