Oil spectrum analysis method and oil spectrum analyzer thereof

Abstract

The present application relates to the technical field of oil spectral analysis, in particular to an oil spectral analysis method and an oil spectral analysis instrument thereof. The method comprises: obtaining spectral signals of different lubricating oil samples detected multiple times in a preset time period; screening overlapping peaks according to wavelength differences and peak widths corresponding to peak values between adjacent characteristic peaks of the spectral signals corresponding to each detection; determining characteristic peaks of additive elements according to peak value variation characteristics of the characteristic peaks in the overlapping peaks of the spectral signals corresponding to adjacent two detections in each sub time period; obtaining abnormal variation degrees of each characteristic peak of additive elements and screening abnormal peaks by combining difference conditions of abnormal variation factors of the same additive element characteristic peaks in adjacent sub time periods in the preset time period; and determining an enhancement coefficient of each additive element by comprehensively considering the abnormal variation degrees of each characteristic peak of additive elements and the number proportion of abnormal peaks, and generating a spectral analysis result. The present application improves the accuracy of the oil spectral analysis result.

Description

Technical Field

[0001] This invention relates to the field of oil spectral analysis technology, specifically to an oil spectral analysis method and an oil spectral analyzer. Background Technology

[0002] In engines used in critical equipment such as aircraft and ships, lubricating oil is a crucial component of the lubrication and cooling systems. Due to the high temperatures, pressures, and high-speed rotation generated during engine operation, lubricating oil is affected by various factors, leading to oxidation, additive failure, and other consequences. Using an oil spectrometer, abnormal conditions in the lubricating oil can be monitored in real time. By analyzing the chemical composition and structure of the oil, it is possible to understand and assess abnormal deterioration of the lubricating oil within the engine, which is of great significance for ensuring flight safety and engine maintenance.

[0003] An oil spectrometer is an instrument used to analyze the composition and content of lubricating oils. It offers numerous advantages, including fast measurement speed, the ability to simultaneously measure multiple elements, and ease of operation. Lubricating oils contain various additives (such as antioxidants and anti-wear agents), which are chemical substances added to improve their performance. These additives typically contain metallic elements such as zinc, phosphorus, calcium, and magnesium. During use, especially under high temperature and high pressure conditions, these additives gradually degrade, generating oxides and other corrosion products. The signals from these products change with usage; therefore, accurately detecting changes in the concentration of additive elements is crucial for assessing the condition of lubricating oils.

[0004] Existing methods for determining additive elements through principal component analysis extract multiple additive element indicators from oil spectral data into a multi-source feature index, which can accurately classify overlapping signals to some extent. However, due to the comprehensive processing of multiple indicators, the effect of abnormal deterioration of additives is weakened, resulting in a decrease in the accuracy of the results of lubricating oil spectral analysis. Summary of the Invention

[0005] To address the problem of low accuracy in existing methods for spectral analysis of lubricating oils, the present invention aims to provide an oil spectral analysis method and an oil spectral analyzer, the specific technical solution of which is as follows:

[0006] In a first aspect, the present invention provides a method for oil spectral analysis, the method comprising:

[0007] Acquire spectral signals from multiple tests of different lubricating oil samples within a preset time period;

[0008] Based on the wavelength difference and peak width of the peak values ​​of adjacent characteristic peaks in the spectral signals corresponding to each detection, overlapping peaks are screened; based on the peak value variation characteristics of the characteristic peaks in the overlapping peaks of the spectral signals corresponding to two adjacent detections within each sub-time period, the characteristic peaks of additive elements are determined, and the variation anomaly factor of the characteristic peaks of each additive element in the overlapping peaks within each sub-time period is obtained.

[0009] By combining the differences in the abnormal factors of the characteristic peaks of the same additive element in adjacent sub-time periods within a preset time period, the degree of abnormal change of the characteristic peaks of each additive element is obtained, and abnormal peaks are screened; by combining the degree of abnormal change of the characteristic peaks of each additive element and the proportion of the number of abnormal peaks, the enhancement coefficient of each additive element is determined.

[0010] By combining the spectral signal and the enhancement coefficient, the spectral analysis results of the oil are generated.

[0011] Preferably, the step of filtering overlapping peaks based on the wavelength difference and peak width between adjacent characteristic peaks of the spectral signal corresponding to each detection includes:

[0012] For any given detection:

[0013] The difference between the peak width and the corresponding wavelength difference between adjacent characteristic peaks of the spectral signal corresponding to any one detection is denoted as the first difference; the normalized result of the ratio of the first difference to the corresponding wavelength difference is used as the overlap parameter of adjacent characteristic peaks.

[0014] If the overlap parameter is greater than the preset overlap threshold, then the corresponding adjacent feature peaks are taken as overlapping peaks.

[0015] Preferably, the step of determining the characteristic peaks of additive elements based on the peak value variation characteristics of the characteristic peaks in the overlapping peaks of the spectral signals corresponding to two adjacent detections within each sub-time period, and obtaining the variation anomaly factor of the characteristic peak of each additive element in the overlapping peaks within each sub-time period, includes:

[0016] For any sub-time period:

[0017] Calculate the peak value difference between each characteristic peak in the overlapping peak of the spectral signal corresponding to each detection in any sub-time period and the corresponding characteristic peak in the spectral signal corresponding to the next adjacent detection. Record this difference as the second difference value corresponding to each characteristic peak in the overlapping peak of the spectral signal corresponding to each detection in any sub-time period. Use the sum of the second differences value corresponding to each characteristic peak in the overlapping peak of the spectral signal corresponding to all detections in any sub-time period as the peak variation factor corresponding to each characteristic peak in the overlapping peak.

[0018] Characteristic peaks with a peak variation factor greater than 0 are used as characteristic peaks of additive elements.

[0019] Calculate the first mean of the peak change factor corresponding to the characteristic peak of each additive element in each overlapping peak of all detections within any sub-time period; determine the negative correlation normalization result of the difference between the peak change factor corresponding to the characteristic peak of each additive element in each overlapping peak of each detection within any sub-time period and its corresponding first mean as the consistency index of the change of the characteristic peak of each additive element in each overlapping peak of each detection within any sub-time period.

[0020] Based on the change of consistency index of the change of characteristic peak of each additive element in each overlapping peak in two adjacent detections within any sub-time period, the change trend parameter of characteristic peak of each additive element in each detection of each overlapping peak in each sub-time period is determined.

[0021] By combining the differences in the variation trend parameters of each additive element characteristic peak in each overlapping peak and its adjacent next additive element in each detection within any sub-time period, and the peak variation factor, the variation anomaly factor of each additive element characteristic peak in each overlapping peak within any sub-time period is obtained.

[0022] Preferably, the step of determining the change trend parameter of each additive element characteristic peak in each overlapping peak detected in each of two adjacent detections within any sub-time period based on the change of the consistency index of the change in the characteristic peak of each additive element in each overlapping peak within any sub-time period includes:

[0023] The ratio between the consistency index of the change of each additive element characteristic peak in each overlapping peak in each detection within any sub-time period and the consistency index of the change of each additive element characteristic peak in each detection within each overlapping peak is used as the change trend parameter of each additive element characteristic peak in each detection within any sub-time period.

[0024] Preferably, the step of combining the difference in the variation trend parameters of each additive element characteristic peak in each overlapping peak and its adjacent next additive element in each detection within any sub-time period, and the peak variation factor, to obtain the variation anomaly factor of each additive element characteristic peak in each overlapping peak within any sub-time period includes:

[0025] Calculate the third difference between the change trend parameter of each additive element characteristic peak in each overlapping peak and its adjacent next additive element characteristic peak in each sub-time period;

[0026] Based on the peak change factor and the corresponding third difference for each characteristic peak in each overlapping peak detected in any sub-time period, the change anomaly factor of each additive element characteristic peak in each overlapping peak in any sub-time period is obtained. The peak change factor and the change anomaly factor are negatively correlated, and the third difference is positively correlated with the change anomaly factor.

[0027] Preferably, the step of combining the differences in abnormal factors of the characteristic peaks of the same additive element in adjacent sub-time periods within a preset time period to obtain the degree of abnormal change of the characteristic peaks of each additive element, and screening abnormal peaks, includes:

[0028] Based on the difference in the abnormality factor of the characteristic peak of the same additive element in adjacent sub-time periods in all sub-time periods, the degree of abnormal change of the characteristic peak of each additive element is obtained, and the difference in the abnormality factor is positively correlated with the degree of abnormal change.

[0029] Characteristic peaks whose abnormal changes exceed a preset abnormal change threshold are identified as abnormal peaks.

[0030] Preferably, determining the enhancement coefficient of each additive element by comprehensively considering the degree of abnormal change in the characteristic peaks of each additive element and the proportion of abnormal peaks includes:

[0031] For any additive element:

[0032] Calculate the first ratio between the degree of abnormal change of each characteristic peak of any abnormal additive element and the maximum degree of abnormal change of all characteristic peaks of any additive element;

[0033] The product of the average of the first ratios corresponding to all characteristic peaks of any additive element and the number of abnormal peaks of any additive element is taken as the attention level of any additive element.

[0034] The enhancement coefficient of any one of the additive elements is determined based on the level of attention received.

[0035] Preferably, determining the enhancement coefficient of any one of the additive elements based on the level of attention includes:

[0036] Calculate the sum of the attention values ​​for all additive elements;

[0037] The ratio of the attention level of any one of the additive elements to the sum value is determined as the enhancement coefficient of any one of the additive elements.

[0038] Preferably, the step of combining the spectral signal and the enhancement coefficient to generate the spectral analysis results of the oil includes:

[0039] Principal component analysis is used to process the spectral signal and extract a preset number of principal component directions;

[0040] Using any two principal component directions as the positive directions of the coordinate axes, the enhancement coefficient of each additive element is used as the projection weight, and the projection of each additive element in the principal component direction is used as the score of each additive element in the principal component direction.

[0041] The scores are used to generate spectral analysis results for the oil.

[0042] Secondly, the present invention provides an oil spectral analyzer, the spectral analyzer including a memory and a processor, the processor executing a computer program stored in the memory to implement the above-mentioned oil spectral analysis method.

[0043] The present invention has at least the following beneficial effects:

[0044] This invention first filters overlapping peaks in the spectral signal based on the wavelength difference and peak width between adjacent characteristic peaks in each detection. Then, it determines the characteristic peaks of additive elements based on the peak value variation characteristics of the characteristic peaks in the overlapping peaks of two adjacent detections. Weighting is then applied based on the sensitive time periods of the additive elements. Further analysis of the trend of additive element concentration over time allows for more accurate identification of abnormal elemental variations caused by changes in additive properties, determining the importance of each additive element. Subsequently, when differentiating overlapping peaks in the spectral signal, parameter enhancement is performed based on the importance of additive elements, improving the identification effect of abnormal additive deterioration, increasing the accuracy of judging the stability of lubricating oil properties, and providing a reliable technical means for lubricating oil quality monitoring. Attached Figure Description

[0045] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a flowchart of an oil spectral analysis method and the method performed by an oil spectral analyzer provided in an embodiment of the present invention. Detailed Implementation

[0047] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description, in conjunction with the accompanying drawings and preferred embodiments, describes an oil spectral analysis method and an oil spectral analyzer proposed according to the present invention.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0049] The following description, in conjunction with the accompanying drawings, details the specific scheme of the oil spectral analysis method and the oil spectral analyzer provided by the present invention.

[0050] An example of an oil spectral analysis method:

[0051] The specific scenario addressed in this embodiment is as follows: When using a spectrometer to detect the degradation of lubricating oil, the presence of worn metal from internal parts during engine operation causes the spectra produced by these metal elements under arc excitation to overlap with the characteristic peaks of some additive elements, leading to signal overlap and affecting the accuracy of the analysis of lubricating oil additive degradation, thus interfering with the analysis results. However, the spectral signals of worn metal elements are stable over time and have high concentrations, while the spectral signals of additive elements show a decreasing trend over time and are sensitive to changes in spectral signals. Therefore, this embodiment will analyze multiple detection results within a time period to enhance the fusion parameters of additive elements.

[0052] This embodiment proposes an oil spectral analysis method and an oil spectral analyzer. The analyzer includes a memory and a processor, and the processor executes a computer program stored in the memory, such as... Figure 1 As shown, the following steps are to be achieved:

[0053] Step S1: Obtain the spectral signals of different lubricating oil samples detected multiple times within a preset time period.

[0054] During prolonged engine operation, oil samples are periodically collected at designated sampling locations. The sampling interval is determined based on the equipment's operating status to ensure that the lubricating oil consumption is reflected. In this embodiment, samples are collected multiple times within a preset time period. This embodiment pre-sets that lubricating oil samples are collected every 12 hours for a preset time period of 30 days. In specific applications, the implementer can adjust the settings according to specific circumstances.

[0055] Each collected lubricating oil sample was placed in the sample chamber, and its spectrum was acquired using an oil spectrometer to obtain the raw spectral signal containing all component signals of the current sample. The raw spectral data was standardized to remove the influence of different signal intensities on the analysis results, and a smoothing filter was used to denoise the signal. The integrity of the spectral data was checked to remove outlier data points caused by instrument malfunction, sample contamination, etc. Then, the spectral data was normalized to ensure that the spectral intensity of each element was between 0 and 1, ensuring the comparability of data in subsequent analyses. It should be noted that all spectral signals mentioned below are normalized spectral signals. Each detection yields a corresponding spectral signal. The horizontal axis of the spectral signal represents the spectral wavelength; different elements emit different characteristic spectral lines under arc excitation, corresponding to different wavelengths. The vertical axis represents the spectral intensity.

[0056] Thus, this embodiment has collected the spectral signals corresponding to each detection within a preset time period.

[0057] Step S2: Based on the wavelength difference and peak width of the peak values ​​of adjacent characteristic peaks of the spectral signals corresponding to each detection, overlapping peaks are screened; based on the peak value change characteristics of the characteristic peaks in the overlapping peaks of the spectral signals corresponding to two adjacent detections within each sub-time period, the characteristic peaks of additive elements are determined, and the change abnormality factor of each characteristic peak of additive element in the overlapping peaks within each sub-time period is obtained.

[0058] When analyzing the deterioration of lubricating oils in use, existing methods using principal component analysis to distinguish spectral signals belonging to additive elements in overlapping signals weaken the indication of abnormal additive deterioration due to the comprehensive analysis of multiple indicators, thus reducing the accuracy of judging the stability of lubricating oil properties. Wear metal elements in overlapping signals typically have high concentrations and stable changes over time, while additive elements usually have low concentrations and gradually decompose and become ineffective over time, exhibiting a certain degree of sensitivity. Therefore, this step combines multiple detection results within a time period to analyze the changing trends of overlapping elements and enhances the signal of spectral data belonging to additive deterioration.

[0059] Since the metal elements and additive elements that wear out the engine may have similar spectral characteristics under electric arc excitation, resulting in signal overlap in the obtained spectrum, this step first analyzes the individual spectral signals to identify the overlapping signals in the spectral signals.

[0060] Dramatic changes in spectral signals typically correspond to the emission characteristics of elements. Therefore, differentiating a single spectral data point can effectively detect peaks in the spectrum and accurately pinpoint their locations. The process of determining peaks is existing technology and will not be elaborated upon here.

[0061] The following embodiment will use any one detection within a preset time period as an example for explanation. Other detections can be processed using the method provided in this embodiment.

[0062] Specifically, for any given detection:

[0063] The difference between the peak width and the corresponding wavelength difference between adjacent characteristic peaks of the spectral signal corresponding to this detection is recorded as the first difference; the normalized result of the ratio of the first difference to the corresponding wavelength difference is used as the overlap parameter of adjacent characteristic peaks.

[0064] In this embodiment, a specific formula for calculating the overlap parameter is given. The overlap parameter between the j-th characteristic peak and the (j+1)-th characteristic peak of the spectral signal corresponding to the t-th detection can be expressed as:

[0065]

[0066] in, This represents the overlap parameter between the j-th and (j+1)-th characteristic peaks of the spectral signal corresponding to the t-th detection. This represents the peak width between the j-th and (j+1)-th characteristic peaks of the spectral signal corresponding to the t-th detection. This represents the wavelength corresponding to the j-th characteristic peak of the spectral signal in the t-th detection. This represents the wavelength corresponding to the (j+1)th characteristic peak of the spectral signal corresponding to the t-th detection. Indicates the absolute value sign. This represents the normalization function.

[0067] It should be noted that the peak width between the j-th characteristic peak and the (j+1)-th characteristic peak is specifically the wavelength between the starting trough of the j-th characteristic peak and the ending trough of the (j+1)-th characteristic peak.

[0068] This represents the wavelength difference between the j-th and (j+1)-th characteristic peaks of the spectral signal corresponding to the t-th detection. The larger this value, the greater the wavelength difference between these two characteristic peaks. This represents the first difference.

[0069] Using the above method, the overlap parameter of every two adjacent feature peaks in the spectral signal can be obtained. Feature peaks with overlap parameters greater than a preset overlap threshold are extracted. These extracted adjacent feature peaks constitute overlapping peaks, i.e., overlapping peaks are obtained in the spectral signal. An overlapping peak is composed of two or more adjacent feature peaks. Overlapping peaks are regions with small peak spacing or overlapping shapes. These regions may be caused by signal overlap from different types of elements. In this embodiment, the preset overlap threshold is 0.5. In specific applications, the implementer can set it according to the specific circumstances.

[0070] Abnormal denaturation and deterioration of additives can lead to significant enhancement or weakening of certain characteristic peaks in the spectrum. Therefore, fluctuations in these characteristics need to be closely monitored during principal component extraction (PCA). However, the presence of overlapping signals makes it difficult to distinguish the independent contribution of each element during PCA extraction, especially when the content of additive elements changes abnormally. Due to overlap, the abnormal signals may be masked by the signals of wear metal elements, thus weakening the signal of abnormal additive deterioration during PCA. Therefore, this step enhances the sensitivity of additive elements in the overlapping signal based on the different characteristics of wear metal elements and lubricating oil elements, ensuring that the effects of abnormal additive deterioration can be accurately identified during subsequent PCA.

[0071] Based on the lubricant's service life and the additive's expected lifespan, the duration of each sub-time period is set, and the preset time period is divided into multiple sub-time periods based on the duration of each sub-time period. For example, when the lubricant's service life and the additive's expected lifespan are approximately 90 days, the duration of each sub-time period is set to 10 days. This ensures that the duration of each sub-time period is reasonable while also guaranteeing sufficient spectral signals within each sub-time period to analyze any abnormal performance of the additive.

[0072] For any overlapping peak, theoretically, the concentration of additive elements in the lubricating oil should gradually decrease with increasing usage time, while the concentration of wear metal elements should gradually increase. Therefore, this embodiment will evaluate the peak changes of characteristic peaks in the overlapping peaks within each sub-time period.

[0073] For any sub-time period:

[0074] Calculate the difference between the peak value of each characteristic peak in the overlapping peak of the spectral signal corresponding to each detection in the sub-time period and the corresponding characteristic peak in the spectral signal of the next adjacent detection. Record this as the second difference value corresponding to each characteristic peak in the overlapping peak of the spectral signal corresponding to each detection in the sub-time period. The sum of the second differences value corresponding to each characteristic peak in the overlapping peak of the spectral signal corresponding to all detections in the sub-time period is used as the peak variation factor corresponding to each characteristic peak in the overlapping peak.

[0075] In this embodiment, a specific calculation formula for the peak variation factor is given, which is the peak variation factor within the sub-time period. The first overlapping peak The peak variation factor corresponding to each characteristic peak can be expressed as:

[0076]

[0077] in, Indicates the first time period within this sub-time period The first overlapping peak The peak variation factor corresponding to each characteristic peak This indicates the number of tests conducted within that sub-time period. This represents the spectral signal corresponding to the t-th detection within this sub-time period. The first overlapping peak The peak value of each characteristic peak This indicates that the spectral signal corresponding to the (t+1)th detection within this sub-time period is the one corresponding to the spectral signal corresponding to the tth detection. The first overlapping peak The peak value of the characteristic peak corresponding to each characteristic peak.

[0078] It should be noted that the spectral signal corresponding to the (t+1)th detection is the same as the spectral signal corresponding to the tth detection. The first overlapping peak The specific characteristic peak corresponding to the characteristic peak is: the characteristic peak in the spectral signal corresponding to the (t+1)th detection that corresponds to the spectral signal corresponding to the tth detection. The first overlapping peak Characteristic peaks with the same order value; for example: if the spectral signal corresponding to the t-th detection has the same characteristic peak order value; The first overlapping peak If the 5th characteristic peak is the spectral signal corresponding to the t-th detection, then the spectral signal corresponding to the (t+1)-th detection is the same as the spectral signal corresponding to the t-th detection. The first overlapping peak The characteristic peak corresponding to the t+1th detection is the 5th characteristic peak in the spectral signal.

[0079] This represents the spectral signal corresponding to the t-th detection within this sub-time period. The first overlapping peak The second difference corresponding to each characteristic peak. The peak variation factor reflects the changing trend of the characteristic peak over time. A positive result indicates that the concentration of the characteristic peak gradually decreases over time, possibly belonging to additive elements; a negative result indicates that the concentration of the characteristic peak gradually increases over time, possibly belonging to wear metal elements.

[0080] Using the above method, the peak variation factor corresponding to each characteristic peak in each overlapping peak within the sub-time period can be obtained, and the characteristic peak with a peak variation factor greater than 0 is taken as the characteristic peak of the additive element.

[0081] Next, the characteristic peaks belonging to additive elements in the overlapping peaks within the sub-time period are calculated for peak changes. When the characteristic peaks within the sub-time period exhibit fluctuations such as sudden drops, it indicates that the characteristic peaks are more likely to belong to abnormal additive elements.

[0082] Specifically, the mean value of the peak variation factor corresponding to each additive element characteristic peak in each overlapping peak of all detections within the sub-time period is calculated, and this mean value is recorded as the first mean value. The negative correlation normalization result of the difference between the peak variation factor corresponding to each additive element characteristic peak in each overlapping peak of each detection within the sub-time period and its corresponding first mean value is determined as the variation consistency index of each additive element characteristic peak in each overlapping peak of each detection within the sub-time period. In this embodiment, the negative correlation normalization method of the difference is specifically as follows: the function value of the negative exponential function of the difference with the natural constant as the base is used as the negative correlation normalization result of the difference. As another implementation method, the method provided in this embodiment can also be used for processing. The larger the variation consistency index, the more complex the peak variation of the j-th characteristic peak within the sub-time period, and the smaller the consistency within the sub-time period.

[0083] Furthermore, the ratio between the consistency index of the change of each additive element characteristic peak in each overlapping peak in each detection within the sub-time period and the consistency index of the change of each additive element characteristic peak in each overlapping peak in each detection within the next adjacent detection is used as the change trend parameter of each additive element characteristic peak in each overlapping peak in each detection within the sub-time period.

[0084] The above steps calculate the trend parameters of the characteristic peaks of all additive elements within the same sub-time period. By comparing the trends, abnormal changes in the characteristic peaks are identified. Ideally, the concentration of additive elements is stable, and their spectral intensity should be relatively stable within the time window. However, when the engine is under high temperature and high load for a long time during use, some additive elements may undergo abnormal denaturation and deterioration, leading to abnormal changes in the spectral signal. Therefore, when characteristic peaks with significantly different concentration change trends appear, it indicates that there may be abnormal changes in the overlapping signals.

[0085] Further, the absolute value of the difference between the characteristic peak of each additive element in each overlapping peak detected in each sub-time period and the change trend parameter of its next adjacent additive element is calculated, and this absolute value is recorded as the third difference. Based on the peak change factor corresponding to each characteristic peak in each overlapping peak detected in each sub-time period and the corresponding third difference, the change anomaly factor of each characteristic peak of each additive element in each overlapping peak in each sub-time period is obtained. The peak change factor and the change anomaly factor are negatively correlated, and the third difference is positively correlated with the change anomaly factor.

[0086] In this embodiment, a specific calculation formula for the abnormal change factor is given, which is the first factor within the sub-time period. The second test The first overlapping peak The abnormal factor of the variation of the characteristic peaks of each additive element can be expressed as:

[0087]

[0088] in, Indicates the first time period within this sub-time period The second test The first overlapping peak Abnormal factors in the variation of characteristic peaks of additive elements Indicates the first time period within this sub-time period The first overlapping peak Peak variation factor corresponding to the characteristic peaks of each additive element. Indicates the first time period within this sub-time period The second test The first overlapping peak The variation trend parameter of the characteristic peaks of each additive element. Indicates the first time period within this sub-time period The second test The first overlapping peak The parameter representing the changing trend of the next characteristic peak of an additive element from the characteristic peak of the previous additive element. Represents the zero constant. Indicates the absolute value sign. This represents an exponential function with the natural constant as its base.

[0089] In this embodiment, the anti-zero constant is 0.01. In specific applications, the implementer can set it according to the specific situation. Indicates the first time period within this sub-time period The second test The first overlapping peak The third difference corresponding to the characteristic peak of each additive element. When the first... The second test The first overlapping peak The greater the difference in the third characteristic peak corresponding to each additive element, the better. The smaller the peak variation factor corresponding to the characteristic peak of each additive element, the more significant the difference between the peaks. The first overlapping peak The more abnormal the change in the characteristic peak of the additive element, that is, the more abnormal the change in the characteristic peak of the first additive element. The second test The first overlapping peak The greater the anomalous factor in the change of characteristic peaks of each additive element, the larger the change.

[0090] Using the above method, it is possible to obtain the variation anomaly factor of the characteristic peak of each additive element in the overlapping peaks of each sub-time period within a preset time period.

[0091] Step S3: Combine the differences in abnormal factors of the characteristic peaks of the same additive element in adjacent sub-time periods within the preset time period to obtain the degree of abnormal change of the characteristic peaks of each additive element, and screen out abnormal peaks; combine the degree of abnormal change of the characteristic peaks of each additive element and the proportion of abnormal peaks to determine the enhancement coefficient of each additive element.

[0092] The stability of user data varies across different sub-time periods. Analyzing the variation of the same characteristic peak data across different sub-time periods reveals that data with significant fluctuations indicates that the abnormal fluctuations of the corresponding additive elements during current use may be relatively high. Therefore, the corresponding characteristic peaks need to be given special attention during the fusion process.

[0093] Based on this, this embodiment obtains the degree of abnormal change of the characteristic peak of each additive element according to the difference of the abnormal factor of the change in the characteristic peak of the same additive element in adjacent sub-time periods in all sub-time periods of the preset time. The difference of the abnormal factor is positively correlated with the degree of abnormal change.

[0094] In this embodiment, a specific formula for calculating the degree of abnormal change is given, the first... The first overlapping peak The degree of abnormal change in the characteristic peaks of the additive elements can be expressed as:

[0095]

[0096] in, Indicates the first The first overlapping peak The degree of abnormal changes in the characteristic peaks of the additive elements. This indicates the number of sub-time periods within a preset time period. Indicates the first time in the preset time Within each sub-time period, the first Abnormal factors in the changes of characteristic peaks of additive elements. Indicates the first time in the preset time Within each sub-time period, the first Abnormal factors in the changes of characteristic peaks of additive elements. This represents the normalization function.

[0097] Indicates the first Sub-time period and the first Within each sub-time period, the first The difference in the anomalous factor of the characteristic peak changes of the same additive element. When the anomalous factor of the characteristic peak changes of the same additive element in adjacent sub-time periods within a preset time period is larger, the... The first overlapping peak The greater the degree of abnormal change in the characteristic peaks of the additive elements.

[0098] Characteristic peaks exhibiting abnormal changes exceeding a preset abnormal change threshold are identified as abnormal peaks. In this embodiment, the preset abnormal change threshold is 0.5; however, in specific applications, the implementer can set this threshold according to the specific circumstances.

[0099] The more similar the degree of abnormal changes of additive elements with similar properties in the spectrum, the greater the possibility of deterioration during the use of the corresponding type of additive. Therefore, the focus of attention on the characteristic peaks of additive elements in overlapping peaks is determined by analyzing the similarity of the degree of abnormal changes of all characteristic peaks of additive elements.

[0100] Specifically, for any additive element: obtain the maximum abnormal variation degree of all characteristic peaks of that additive element; calculate the ratio between the abnormal variation degree of each characteristic peak of that additive element and the maximum abnormal variation degree of all characteristic peaks of that additive element, and record this ratio as the first ratio. Multiply the average of the first ratios corresponding to all characteristic peaks of that additive element by the number of abnormal peaks of that additive element, and use this product as the attention level of that additive element. The more abnormal variation characteristic peaks there are, and the greater the degree of abnormality of these characteristic peaks, the greater the possibility of abnormal deterioration of the additive in the lubricating oil.

[0101] Using the above method, the attention level of each additive element can be obtained. When fusing multiple signal features, the feature peak with higher attention level indicates a greater possibility of additive deterioration, and the degree to which abnormal states need to be amplified during the fusion process is higher.

[0102] Therefore, the sum of the attention values ​​of all additive elements is calculated; for any additive element, the ratio between the attention value of that additive element and the sum value is determined as the enhancement coefficient of that additive element.

[0103] Thus, the enhancement coefficient of each additive element has been obtained using the method provided in this embodiment.

[0104] Step S4: Combine the spectral signal and the enhancement coefficient to generate the spectral analysis results of the oil.

[0105] The spectral data of oil samples are complex, with numerous elements to be analyzed and some correlation between them. Principal component analysis (PCA) is used to reduce the dimensionality of the spectral data, transforming high-dimensional data into low-dimensional data. This helps to extract the principal components of additive elements and wear metal elements from the complex signals, simplifying the spectral data and facilitating subsequent analysis of additive failure. The enhancement coefficients of different signal components in overlapping peaks are obtained using the above method, allowing PCA to focus more on abnormal changes in additive elements when extracting principal components.

[0106] Based on known lubricant formulations and the sources of wear-related metal elements, the wear-related metal elements and additive elements in the original spectral data were preliminarily identified. Common additive elements include calcium, phosphorus, zinc, and magnesium; wear-related metal elements include iron, aluminum, chromium, copper, lead, and tin, originating from internal engine metal components.

[0107] A three-dimensional coordinate system is established with spectral wavelength as the X-axis, elemental concentration as the Y-axis, and time series as the Z-axis. Principal component analysis (PCA) is used to process the spectral signal, extracting a preset number of principal component directions. The preset number is set by the implementer according to specific circumstances and will not be elaborated further here. Using any two principal component directions as the positive axes, the enhancement coefficient of each additive element is used as the projection weight to calculate the projection of each additive element onto the principal component directions. The calculated projection is then used as the score for each additive element on the principal component directions. This embodiment, by introducing elemental enhancement coefficients during PCA, enhances the performance characteristics of abnormal additive element signals in overlapping peaks. This amplifies the abnormal change trends of additive elements during signal separation of overlapping peaks, effectively enhancing the identification effect of abnormal additive degradation during PCA.

[0108] Subsequently, based on the scores of each additive element in the principal component direction calculated using the above method, the main characteristics of the oil reflected by the principal component were preliminarily determined. The spectral signals were initially classified into wear metal element signals and additive element signals. The variation trend of additive elements was obtained, thus obtaining the spectral analysis results of the oil.

[0109] The acquisition motherboard acquires spectral signals through a CCD acquisition board and analyzes the raw spectral data of the oil through the above steps to generate a spectral analysis report of the oil. The spectral signals, spectral analysis report, and possible abnormal changes in additives and predicted problems are packaged into a data package. Abnormal wear conditions are marked and the analysis data is transmitted to a touch screen for feedback to relevant personnel.

[0110] This embodiment first filters overlapping peaks in the spectral signal based on the wavelength difference and peak width between adjacent characteristic peaks in each detection. Then, it determines the characteristic peaks of additive elements based on the peak value variation characteristics of the characteristic peaks in the overlapping peaks of two adjacent detections. Weighting is performed based on the time period in which the additive elements are sensitive. By further analyzing the trend of additive element concentration changes over time, it is possible to more accurately identify abnormal elemental variations caused by changes in additive properties, determine the importance of each additive element, and then enhance parameters based on the importance of additive elements when distinguishing overlapping peaks in the subsequent spectral signal. This enhances the identification effect of abnormal deterioration of additives, improves the accuracy of judging the stability of lubricating oil properties, and provides a reliable technical means for lubricating oil quality monitoring.

[0111] An example of an oil spectral analyzer:

[0112] This embodiment provides an oil spectral analyzer including a memory and a processor. The processor executes the computer program stored in the memory to implement the above-described oil spectral analysis method. An oil spectral analysis method has been described in detail in an embodiment of an oil spectral analysis method, so it will not be described in detail here.

[0113] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method of spectral analysis of oil, characterized in that, The method includes: Acquire spectral signals from multiple tests of different lubricating oil samples within a preset time period; Based on the wavelength difference and peak width of the peak values ​​of adjacent characteristic peaks in the spectral signals corresponding to each detection, overlapping peaks are screened; based on the peak value variation characteristics of the characteristic peaks in the overlapping peaks of the spectral signals corresponding to two adjacent detections within each sub-time period, the characteristic peaks of additive elements are determined, and the variation anomaly factor of the characteristic peaks of each additive element in the overlapping peaks within each sub-time period is obtained. By combining the differences in the abnormal factors of the characteristic peaks of the same additive element in adjacent sub-time periods within a preset time period, the degree of abnormal change of the characteristic peaks of each additive element is obtained, and abnormal peaks are screened; by combining the degree of abnormal change of the characteristic peaks of each additive element and the proportion of the number of abnormal peaks, the enhancement coefficient of each additive element is determined. By combining the spectral signal and the enhancement coefficient, spectral analysis results of the oil are generated; The enhancement coefficient of each additive element is determined by comprehensively considering the degree of abnormal changes in the characteristic peaks of each additive element and the proportion of abnormal peaks, including: For any additive element: Calculate the first ratio between the degree of abnormal change of each characteristic peak of any abnormal additive element and the maximum degree of abnormal change of all characteristic peaks of any additive element; The product of the average of the first ratios corresponding to all characteristic peaks of any additive element and the number of abnormal peaks of any additive element is taken as the attention level of any additive element. Calculate the sum of the attention values ​​for all additive elements; The ratio of the attention level of any one of the additive elements to the sum value is determined as the enhancement coefficient of any one of the additive elements.

2. The method for oil spectral analysis according to claim 1, characterized in that, The step of filtering overlapping peaks based on the wavelength difference and peak width between adjacent characteristic peaks of the spectral signal corresponding to each detection includes: For any given detection: The difference between the peak width and the corresponding wavelength difference between adjacent characteristic peaks of the spectral signal corresponding to any one detection is denoted as the first difference; the normalized result of the ratio of the first difference to the corresponding wavelength difference is used as the overlap parameter of adjacent characteristic peaks. If the overlap parameter is greater than the preset overlap threshold, then the corresponding adjacent feature peaks are taken as overlapping peaks.

3. The method for oil spectral analysis according to claim 1, characterized in that, The step involves determining the characteristic peaks of additive elements based on the peak value variation characteristics of the characteristic peaks in the overlapping peaks of the spectral signals corresponding to two adjacent detections within each sub-time period, and obtaining the variation anomaly factor of each additive element characteristic peak in the overlapping peaks within each sub-time period, including: For any sub-time period: Calculate the peak value difference between each characteristic peak in the overlapping peak of the spectral signal corresponding to each detection in any sub-time period and the corresponding characteristic peak in the spectral signal corresponding to the next adjacent detection. Record this difference as the second difference value corresponding to each characteristic peak in the overlapping peak of the spectral signal corresponding to each detection in any sub-time period. Use the sum of the second differences value corresponding to each characteristic peak in the overlapping peak of the spectral signal corresponding to all detections in any sub-time period as the peak variation factor corresponding to each characteristic peak in the overlapping peak. Characteristic peaks with a peak variation factor greater than 0 are used as characteristic peaks of additive elements. Calculate the first mean of the peak change factor corresponding to the characteristic peak of each additive element in each overlapping peak of all detections within any sub-time period; determine the negative correlation normalization result of the difference between the peak change factor corresponding to the characteristic peak of each additive element in each overlapping peak of each detection within any sub-time period and its corresponding first mean as the consistency index of the change of the characteristic peak of each additive element in each overlapping peak of each detection within any sub-time period. Based on the change of consistency index of the change of characteristic peak of each additive element in each overlapping peak in two adjacent detections within any sub-time period, the change trend parameter of characteristic peak of each additive element in each detection of each overlapping peak in each sub-time period is determined. By combining the differences in the variation trend parameters of each additive element characteristic peak in each overlapping peak and its adjacent next additive element in each detection within any sub-time period, and the peak variation factor, the variation anomaly factor of each additive element characteristic peak in each overlapping peak within any sub-time period is obtained.

4. The oil spectral analysis method according to claim 3, characterized in that, The method for determining the trend parameters of each additive element characteristic peak in each overlapping peak detected in each of two adjacent detections within any sub-time period, based on the change of consistency index of the change in the characteristic peak of each additive element in each overlapping peak detected in each sub-time period, includes: The ratio between the consistency index of the change of each additive element characteristic peak in each overlapping peak in each detection within any sub-time period and the consistency index of the change of each additive element characteristic peak in each detection within each overlapping peak is used as the change trend parameter of each additive element characteristic peak in each detection within any sub-time period.

5. The method for oil spectral analysis according to claim 3, characterized in that, The method combines the differences in the variation trend parameters of each additive element characteristic peak in each overlapping peak and its adjacent next additive element within any sub-time period, along with the peak variation factor, to obtain the variation anomaly factor of each additive element characteristic peak in each overlapping peak within any sub-time period, including: Calculate the third difference between the change trend parameter of each additive element characteristic peak in each overlapping peak and its adjacent next additive element characteristic peak in each sub-time period; Based on the peak change factor and the corresponding third difference for each characteristic peak in each overlapping peak detected in any sub-time period, the change anomaly factor of each additive element characteristic peak in each overlapping peak in any sub-time period is obtained. The peak change factor and the change anomaly factor are negatively correlated, and the third difference is positively correlated with the change anomaly factor.

6. The method for oil spectral analysis according to claim 1, characterized in that, The method involves combining the differences in abnormal factors of characteristic peak changes of the same additive element within adjacent sub-time periods within a preset time period to obtain the degree of abnormal change of characteristic peaks of each additive element, and screening abnormal peaks, including: Based on the difference in the abnormality factor of the characteristic peak of the same additive element in adjacent sub-time periods in all sub-time periods, the degree of abnormal change of the characteristic peak of each additive element is obtained, and the difference in the abnormality factor is positively correlated with the degree of abnormal change. Characteristic peaks whose abnormal changes exceed a preset abnormal change threshold are identified as abnormal peaks.

7. The method for oil spectral analysis according to claim 1, characterized in that, The step of combining the spectral signal and the enhancement coefficient to generate the spectral analysis results of the oil includes: Principal component analysis is used to process the spectral signal and extract a preset number of principal component directions; Using any two principal component directions as the positive directions of the coordinate axes, the enhancement coefficient of each additive element is used as the projection weight, and the projection of each additive element in the principal component direction is used as the score of each additive element in the principal component direction. The scores are used to generate spectral analysis results for the oil.

8. An oilseed spectral analyzer, the spectral analyzer comprising a memory and a processor, characterized in that, The processor executes a computer program stored in memory to implement the method of claim 1.

Images