Multi-index fusion lubricating oil service life prediction method and system and medium
By determining the optimal combination of physicochemical indicators through principal component analysis and multiple linear fitting, a lubricating oil life prediction model was established, which solved the problems of sparse lubricating oil monitoring data and redundant indicators, and achieved accurate prediction of lubricating oil life.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUANGZHOU MECHANICAL ENGINEERING RESEARCH INSTITUTE CO LTD
- Filing Date
- 2026-02-06
- Publication Date
- 2026-05-19
AI Technical Summary
Existing lubricating oil monitoring methods result in sparse sample data, making it difficult to obtain dense trend information. Single physicochemical indicators cannot fully reflect the condition of lubricating oil, and there are complex coupling relationships and information redundancy between different indicators, leading to biased evaluation results and making it difficult to accurately reveal the deterioration pattern and service life of lubricating oil.
Principal component analysis was used to select principal components with a cumulative contribution rate greater than a set threshold as lubricant performance evaluation features. Combined with multiple linear fitting and scoring comparison, the optimal combination of physicochemical indicators was determined, and a multi-indicator fusion lubricant life prediction model was established based on weights to achieve accurate prediction of lubricant life.
It overcomes the information gaps under small sample data, quantitatively distinguishes the influence of combinations of physicochemical indicators, avoids redundant information interference, and achieves accurate evaluation of lubricating oil performance and life prediction, with high precision and objectivity.
Smart Images
Figure FT_1 
Figure FT_2 
Figure FT_3
Abstract
Description
Technical Field
[0001] This application relates to the field of lubricating oil life assessment technology, and in particular to a method, system and medium for predicting lubricating oil life by integrating multiple indicators. Background Technology
[0002] Lubricating oil is known as the "blood" of mechanical equipment. It plays a vital role in lubrication, wear resistance, and corrosion prevention, and is crucial for maintaining the safe operation of equipment and extending its service life.
[0003] During the use of lubricating oil, it will be affected by complex working conditions such as pressure, temperature, and environmental pollution, and will gradually undergo oxidative degradation, which will lead to significant changes in physicochemical properties (i.e., physical and chemical indicators) such as acid value, base value, and kinematic viscosity, ultimately affecting its lubrication effect.
[0004] In actual operating conditions, due to cost and operational convenience constraints, lubricating oil is typically monitored on a long-term, quarterly, or annual basis. This results in small data samples with large time spans, making it difficult to obtain dense trend information. Furthermore, the performance and lifespan of lubricating oil are often determined by multiple physicochemical indicators; a single indicator cannot comprehensively reflect the true state of the lubricating oil, leading to biased and inaccurate assessments. In addition, complex coupling relationships and information redundancy may exist between different physicochemical indicators. Current technologies typically rely on empirical judgment or mechanistic deduction to screen indicators. While this can cover key performance parameters to some extent, it still has shortcomings in practical applications: some indicators do not change significantly during lubricating oil use, making them unsuitable as effective bases for lifespan prediction; and some indicators may have highly overlapping characterization effects, leading to biased assessment results and making it difficult to accurately reveal the deterioration patterns and lifespan of lubricating oil. Summary of the Invention
[0005] The purpose of this application is to provide a method, system, and medium for predicting the life of lubricating oil by integrating multiple indicators. By deeply mining the coupling relationship of multiple indicators under limited data, the application achieves accurate prediction of the life of lubricating oil.
[0006] To achieve the above objectives, this application provides the following solution.
[0007] Firstly, this application provides a multi-index fusion method for predicting lubricating oil life. The method includes: acquiring a lubricating oil sample dataset; the lubricating oil sample dataset includes detection data of multiple physicochemical indicators of lubricating oil samples during use; based on the lubricating oil sample dataset, performing principal component analysis on the multiple physicochemical indicators, selecting principal components with a cumulative contribution rate greater than a set threshold as lubricating oil performance evaluation features, and obtaining a principal component score matrix; splitting and reorganizing the lubricating oil sample dataset according to different combinations of physicochemical indicators to obtain datasets with multiple combinations of physicochemical indicators; and based on the principal component score matrix... Using a dataset with multiple combinations of physicochemical indicators, a multivariate linear fitting and scoring comparison method is employed to determine the optimal combination of physicochemical indicators. Based on the principal component score matrix and the dataset corresponding to the optimal combination of physicochemical indicators, a linear fitting and scoring proportion method is used to assign weights to each physicochemical indicator in the optimal combination of physicochemical indicators. Based on the evolution law of each physicochemical indicator in the optimal combination of physicochemical indicators with mileage and the weights corresponding to each physicochemical indicator, a multi-indicator fusion lubricating oil life prediction model is established. According to the optimal combination of physicochemical indicators, the failure threshold of the corresponding physicochemical indicators in the lubricating oil to be tested is obtained, and the life of the lubricating oil to be tested is predicted using the lubricating oil life prediction model.
[0008] In a second aspect, this application also provides a computer system, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the multi-index fusion lubricating oil life prediction method described in the first aspect.
[0009] Thirdly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the multi-index fusion lubricating oil life prediction method described in the first aspect.
[0010] Based on the specific embodiments provided in this application, the following technical effects are disclosed.
[0011] This application fully considers the problem that lubricating oil monitoring methods in actual working conditions often result in lubricating oil sample datasets exhibiting small sample sizes and large ranges. Therefore, it proposes a method to effectively extract the evolution characteristics of multiple indicators under small sample data, overcoming the information gaps caused by data sparsity. Simultaneously, this application establishes a scoring system for combinations of multiple physicochemical indicators, which can quantitatively distinguish the degree of influence of different combinations of physicochemical indicators on lubricating oil performance, avoiding the problems of indicator duplication and information redundancy in traditional methods. Secondly, this application also proposes an objective weighting method based on scoring for individual physicochemical indicators in the selected optimal physicochemical indicator combinations. This allows for automatic calculation of the weights based on the differences in the contribution of each physicochemical indicator to lubricating oil performance, making the weight allocation more in line with objective laws. Finally, this application, through the fusion of multi-physicochemical indicator combination screening, objective weighting, and single-indicator life prediction, forms a systematic multi-indicator fusion lubricating oil life prediction model. This model overcomes the shortcomings of existing technologies that rely solely on a single indicator or subjective weighting, and can accurately reveal the deterioration patterns and service life of lubricating oil. Attached Figure Description
[0012] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0013] Figure 1 This is a flowchart of a lubricating oil life prediction method that integrates multiple indicators in one embodiment of this application.
[0014] Figure 2 This is a scree plot showing the cumulative contribution rate of principal components ranked in one embodiment of this application.
[0015] Figure 3 This is a data fitting graph showing the evolution of kinematic viscosity at 100°C with mileage in one embodiment of this application.
[0016] Figure 4 This is a data fitting graph showing the evolution of acid value with operating mileage in one embodiment of this application.
[0017] Figure 5 This is a data fitting graph showing the evolution of alkalinity with operating mileage in one embodiment of this application.
[0018] Figure 6 This is an internal structure diagram of a computer system according to another embodiment of this application. Detailed Implementation
[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0020] Currently, existing industrial lubricant monitoring methods generally suffer from long sampling periods (e.g., based on long mileage, quarterly, or annual data) and small sample sizes. This sparse data cannot meet the data density requirements of conventional trend analysis algorithms, making it difficult to accurately capture the deterioration trajectory of lubricant performance from gradual to abrupt changes. Therefore, there is an urgent need for a technical method based on small sample (limited sample size) monitoring data. This method involves in-depth mining of the coupling relationships between multiple indicators under limited data to screen out key indicators and their optimal combinations that have a significant impact on changes in lubricant performance, while also removing redundant information. This would enable accurate prediction of the remaining life of lubricants and solve the key technical challenge of effectively assessing lifespan under sparse data conditions.
[0021] The purpose of this application is to provide a method, system, and medium for predicting the life of lubricating oil by integrating multiple indicators. By deeply mining the coupling relationship of multiple indicators under limited data, the application achieves accurate prediction of the life of lubricating oil.
[0022] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0023] In one exemplary embodiment, a multi-index fusion method for predicting lubricating oil life is provided, such as... Figure 1 As shown, the specific method for predicting lubricating oil life by integrating multiple indicators is as follows.
[0024] Step S1: Obtain the lubricating oil sample dataset.
[0025] In this embodiment, oil monitoring technology is used to collect detection data of multiple physicochemical indicators of lubricating oil samples at different stages of use. During the data analysis phase, considering the excessive number of variables and similarities in the physicochemical indicators of lubricating oil, which could lead to decreased complexity and effectiveness in quantitative analysis, principal component analysis (PCA) is employed to address the issues of redundancy and similarity in the physicochemical indicators. This method divides the data into a calibration set and a test set. The number of principal components is determined by the cumulative contribution rate of the calibration set, and features are extracted during testing using a projection matrix. A higher cumulative contribution rate indicates more information contained in the principal components. Therefore, when extracting effective information from the physicochemical indicators of lubricating oil, PCA can reduce data dimensionality and complexity, and improve the accuracy and stability of quantitative analysis. Ultimately, the service life of the lubricating oil is predicted based on the selected physicochemical indicators.
[0026] The specific solution process for principal components is as follows.
[0027] set up( x 1, x 2, ..., x n () is a dataset of lubricating oil samples. n Each sample contains 10 samples, and each sample includes 10 samples. m If there are 1 / 2 dimensional variables, then the original matrix can be represented as , Indicates the nth element in the matrix i Line number j The elements of the column.
[0028] Calculate the covariance matrix of the population matrix, find its corresponding correlation coefficient matrix, and then find the eigenvalues of the correlation coefficient matrix, satisfying the following conditions: λ 1≥ λ 2≥… λ m ≥0, then the first i Contribution rate of each principal component The details are as follows.
[0029] .
[0030] Then before q Cumulative contribution rate of each principal component as follows.
[0031] .
[0032] Principal component scores The specific expression is as follows.
[0033] .
[0034] In the formula, for the original matrix Standardization is performed to obtain the standardized matrix. , Represents the first in the normalized matrix i Line number j Column elements, For the front k The load submatrix consists of adjusted eigenvectors.
[0035] Step S2: Based on the lubricating oil sample dataset, perform principal component analysis on multiple physicochemical indicators, select principal components with a cumulative contribution rate greater than a set threshold as lubricating oil performance evaluation features, and obtain the principal component score matrix.
[0036] In this embodiment, principal component analysis is performed on multiple physicochemical indicators to obtain a scree plot ranking the cumulative contribution rates of the principal components. Principal components with a cumulative contribution rate greater than 95% are selected as the main features for lubricant performance evaluation, used for subsequent screening and weighting of key physicochemical indicators. Based on the contribution rate, combined with... The test data of physical and chemical indicators are used to score each physical and chemical indicator, and the optimal combination of physical and chemical indicators for the final prediction of lubricating oil life is selected, while redundant indicators are removed.
[0037] Step S3: According to different combinations of physical and chemical indicators, the lubricating oil sample dataset is split and recombined to obtain datasets with multiple combinations of physical and chemical indicators.
[0038] In this embodiment, in order to facilitate the selection of physicochemical indicators that are effective in predicting the life of lubricating oil, it is necessary to combine different numbers of various physicochemical indicators.
[0039] A combination of two physicochemical indicators: .
[0040] Three physicochemical index combinations: .
[0041] ... K Individual combinations of physicochemical indicators: .
[0042] All combinations of physicochemical indicators: .
[0043] In the formula, C For all combinations of physicochemical indicators, It is a combination of two physicochemical indicators. It is a combination of three physicochemical indicators. for K A combination of physicochemical indicators , , and These are the numbers corresponding to the physicochemical indicators. K The quantity of physicochemical indicators.
[0044] Step S4: Based on the principal component score matrix and the dataset with multiple combinations of physicochemical indicators, the optimal combination of physicochemical indicators is determined by using multiple linear fitting and scoring comparison.
[0045] In this embodiment, step S4 is as follows.
[0046] Step S41: Perform multiple linear fitting on the principal component score matrix with the dataset under each combination of physicochemical indicators, and solve for the corresponding coefficients of determination. The multiple linear fitting method is as follows.
[0047] .
[0048] In the formula, The intercept is... , , ... For regression coefficients, , , ... This refers to the test data for physicochemical indicators.
[0049] Step S42: Calculate the score corresponding to each combination of physicochemical indicators based on the coefficient of determination and the principal component contribution rate. The scoring method for each combination of physicochemical indicators is as follows.
[0050] .
[0051] In the formula, For the first c The scores corresponding to the combination of physicochemical indicators. For the first i The principal components and the first c The determining coefficient for the combination of physicochemical indicators For the first i The contribution rate of each principal component q For the front q One principal component.
[0052] Step S43: Compare the scores corresponding to all combinations of physicochemical indicators, and determine the combination of physicochemical indicators with the highest score as the optimal combination of physicochemical indicators.
[0053] Step S5: Based on the principal component score matrix and the dataset corresponding to the optimal combination of physicochemical indicators, weights are assigned to each physicochemical indicator in the optimal combination of physicochemical indicators using linear fitting and score proportion methods.
[0054] In this embodiment, step S5 is as follows.
[0055] Step S51: Perform linear fitting between the principal component score matrix and the dataset corresponding to the optimal combination of physicochemical indicators, and solve for the corresponding coefficients of determination. The linear fitting method is as follows.
[0056] .
[0057] In the formula, The test data representing physicochemical indicators, This is the intercept.
[0058] Step S52: Based on the coefficient of determination and the principal component contribution rate, calculate the score corresponding to each physicochemical index in the optimal physicochemical index combination. The scores corresponding to each physicochemical index in the optimal physicochemical index combination are as follows.
[0059] .
[0060] In the formula, The first in the optimal combination of physicochemical indices b The scores corresponding to each physicochemical indicator For the first i Among the principal components and the optimal combination of physicochemical indices, the first... b The coefficient of determination for each physicochemical index.
[0061] Step S53: Normalize the scores corresponding to each physicochemical index in the optimal physicochemical index combination to determine the weight of each physicochemical index in the optimal physicochemical index combination. The weights of each physicochemical index in the optimal physicochemical index combination are as follows.
[0062] .
[0063] In the formula, The first in the optimal combination of physicochemical indices b The weight of each physicochemical indicator The first in the optimal combination of physicochemical indices b The purpose of normalization is to control the score values for each physicochemical indicator within the range of [0,1]. B This represents the number of physicochemical indicators in the optimal combination of physicochemical indicators.
[0064] Step S6: Based on the evolution of each physicochemical index in the optimal combination of physicochemical indices with the operating mileage, and the weights of each physicochemical index, establish a multi-index fusion lubricating oil life prediction model.
[0065] Assuming that the evolution of physicochemical properties of lubricating oil over time conforms to a first-order reaction, a life prediction model based on a single physicochemical property can be established using the integral rate equation of a first-order reaction.
[0066] .
[0067] In the formula, The new oil value is the physicochemical index at room temperature; s The values of physicochemical properties at a certain test time; k This represents the oxidative degradation rate; t This refers to the lifespan of the lubricating oil.
[0068] Taking the logarithmic form of the above expression, we get... .
[0069] Based on this, the weights of each physicochemical index obtained from the solution are assigned to the life prediction model of the corresponding physicochemical index, and a multi-index fusion lubricating oil life prediction model is established as follows.
[0070] .
[0071] In the formula, To predict lubricating oil life based on multi-indicator fusion, and These are the first and second physicochemical indices in the optimal combination of physicochemical indices, respectively. b The weight of each physicochemical indicator and These are the first and second physicochemical indices in the optimal combination of physicochemical indices, respectively. b Time-sensitivity thresholds for individual physicochemical indicators and These are the first and second physicochemical indicators in the optimal combination of physicochemical indicators at room temperature, respectively. b New oil values for various physicochemical indicators and These are the first and second physicochemical indices in the optimal combination of physicochemical indices, respectively. b The oxidative degradation rate corresponding to each physicochemical index.
[0072] Step S7: According to the optimal combination of physical and chemical indicators, obtain the failure threshold of the corresponding physical and chemical indicators in the lubricating oil to be tested, and use the lubricating oil life prediction model to predict the life of the lubricating oil to be tested.
[0073] By combining national standards for lubricating oil change, the oil change indicators for various physicochemical properties are obtained and used as failure thresholds for lubricating oil life assessment. Then, the failure thresholds of each indicator are substituted into a multi-indicator fusion lubricating oil life prediction model to achieve lubricating oil life prediction.
[0074] To verify the effectiveness of the multi-index fusion method for predicting lubricating oil life, this embodiment also assesses the life of a certain brand of diesel engine oil based on the test results of its physicochemical indicators at different operating mileages. The test data for each physicochemical indicator are shown in Table 1 below.
[0075] Table 1. Data on Physicochemical Indicators
[0076] Principal component analysis was performed on the above six physicochemical indicators. The crushing stone plot, ranking the cumulative contribution rates of the principal components for different physicochemical indicators, is shown below. Figure 2 ,from Figure 2 It can be seen that the cumulative contribution rate of the first principal component corresponding to the six physicochemical indicators is 96.22%, which is over 95%. Therefore, the first principal component is selected as the main feature for screening key physicochemical indicators of lubricating oil.
[0077] The physicochemical indicators are combined in different quantities according to the method in Table 2 below. Then, each combination of physicochemical indicators is scored according to the method in step S4 above.
[0078] Table 2. Combinations of Physicochemical Indicators and Corresponding Scoring Table
[0079] Based on Table 2, the optimal combination of physicochemical indicators for in-use diesel engine oils was selected, and the results are shown in Table 3. Table 3 shows that the scores for the optimal combination of three physicochemical indicators (96.16 points), four physicochemical indicators (96.19 points), five physicochemical indicators (96.21 points), and six physicochemical indicators (96.22 points) are similar. This indicates that the optimal combination of three physicochemical indicators can effectively characterize the four, five, and six physicochemical indicators, helping to maximize the characterization of oil performance with a smaller number of physicochemical indicator combinations. Therefore, some redundant physicochemical indicators can be eliminated during the life prediction process.
[0080] Table 3. Highest scores for each combination of physicochemical indicators.
[0081] Following step S5 above, each physicochemical indicator was scored individually, and the results are shown in Table 4. Table 4 shows that the scores for kinematic viscosity at 100℃, acid value, and alkali value were 93.74, 89.15, and 94.42, respectively. Then, based on these scores, each physicochemical indicator was weighted to obtain its corresponding weight.
[0082] Table 4. Scores and Weights of Physicochemical Indicators
[0083] according to By fitting the data on the evolution of kinematic viscosity, acid value, and alkalinity at 100℃ with mileage, the oxidative degradation rate of each physicochemical index was obtained. k See details Figures 3-5 .
[0084] The various physicochemical indicators k Substitution A lifetime prediction model based on single indicators of kinematic viscosity, acid value, and alkali value at 100℃ was established.
[0085] Lifetime prediction model for kinematic viscosity at 100℃: .
[0086] In the formula, The lifetime prediction results are based on kinematic viscosity at 100℃ as the evaluation index. The failure threshold is the kinematic viscosity at 100℃.
[0087] Acid value lifetime prediction model: .
[0088] In the formula, The lifetime prediction results are based on acid value as an evaluation index. The failure threshold is the acid value.
[0089] Lifetime prediction model for base number: .
[0090] In the formula, The lifetime prediction results are based on alkalinity as an evaluation index. This is the failure threshold for the alkalinity.
[0091] By assigning the weights corresponding to each physicochemical index to the single-index life prediction model, a multi-index fusion lubricating oil life prediction model is established.
[0092] .
[0093] According to the national standard GB / T 7607 "Diesel Engine Oil Change Indicators", the failure thresholds for kinematic viscosity, acid value and base value at 100℃ were obtained, as detailed in Table 5.
[0094] Table 5 Physicochemical Indicators: New Oil Value, Oil Change Indicators, and Failure Threshold
[0095] The failure thresholds of each physicochemical index were substituted into the corresponding single-index life prediction model and the multi-index fusion lubricating oil life prediction model to achieve lubricating oil life prediction for both single and multi-index indicators. The life prediction results for single and multi-index indicators are shown in Table 6.
[0096] Table 6. Results of Single-Indicator and Multi-Indicator Life Prediction
[0097] In summary, this application has the following advantages.
[0098] (1) The correlation coefficient R is obtained by performing multiple linear fitting between the principal component score matrix and the detection data of multiple physicochemical indicators. 2 By combining the contribution rates of each principal component, a scoring system for multiple combinations of physicochemical indicators was constructed. This system can quantitatively reflect the degree of influence of different combinations of physicochemical indicators on lubricating oil performance and generate corresponding scores. This method effectively identifies and distinguishes the differences between various physicochemical indicators, avoids interference from redundant information, and thus achieves a more accurate assessment of lubricating oil performance, providing objective and reliable data support for life prediction.
[0099] (2) After completing the screening of the optimal combination of physicochemical indicators, the principal component score matrix is linearly fitted to the detection data of each physicochemical indicator to obtain the correlation coefficient R. 2 Furthermore, by combining the contribution rates of each principal component, a scoring system for individual physicochemical indicators is established, and each indicator is quantitatively scored. Subsequently, based on the scoring results of each physicochemical indicator, an objective weighting method for the physicochemical indicators is established, and the weights of each physicochemical indicator are then calculated. This weighting method avoids the uncertainty and arbitrariness of traditional subjective experience-based weighting, and achieves an objective quantification of the role of physicochemical indicators in lubricating oil life assessment.
[0100] (3) The overall process can more comprehensively and scientifically reflect the deterioration law of lubricating oil performance and realize high-precision prediction of lubricating oil life, which has significant practical value and promotion significance.
[0101] In another exemplary embodiment, a computer system is provided, which may be a server or a terminal, and its internal structure diagram may be as follows: Figure 6 As shown, the computer system includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores lubricating oil sample datasets and datasets with various combinations of physicochemical indicators. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the aforementioned multi-indicator fusion lubricating oil life prediction method.
[0102] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer system to which the present application is applied. A specific computer system may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.
[0103] In another exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.
[0104] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).
[0105] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.
[0106] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.
[0107] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.
Claims
1. A method for predicting lubricating oil life through multi-index fusion, characterized in that, The multi-index fusion method for predicting lubricating oil life includes: Obtain a lubricating oil sample dataset; the lubricating oil sample dataset includes detection data of multiple physicochemical indicators of lubricating oil samples during use; Based on the lubricating oil sample dataset, principal component analysis was performed on multiple physicochemical indicators. The principal components with a cumulative contribution rate greater than a set threshold were selected as lubricating oil performance evaluation features, and the principal component score matrix was obtained. The lubricating oil sample dataset was split and reorganized according to different combinations of physical and chemical indicators to obtain datasets with multiple combinations of physical and chemical indicators. Based on the principal component score matrix and the dataset under various combinations of physicochemical indicators, the optimal combination of physicochemical indicators is determined by using multiple linear fitting and scoring comparison. Based on the principal component score matrix and the dataset corresponding to the optimal combination of physicochemical indicators, a linear fitting and scoring ratio method is used to assign weights to each physicochemical indicator in the optimal combination of physicochemical indicators. Based on the evolution of each physicochemical index in the optimal combination of physicochemical indices with the operating mileage, and the weights of each physicochemical index, a multi-index fusion lubricating oil life prediction model is established. According to the optimal combination of physicochemical indicators, the failure threshold of the corresponding physicochemical indicators in the lubricating oil to be tested is obtained, and the life of the lubricating oil to be tested is predicted using the lubricating oil life prediction model.
2. The lubricating oil life prediction method based on multi-index fusion according to claim 1, characterized in that, The combination method of the physicochemical indicators is as follows: ； In the formula, C For all combinations of physicochemical indicators, It is a combination of two physicochemical indicators. It is a combination of three physicochemical indicators. It is a combination of four physicochemical indicators. for K A combination of physicochemical indicators K The quantity of physicochemical indicators.
3. The lubricating oil life prediction method based on multi-index fusion according to claim 1, characterized in that, Based on the principal component score matrix and the dataset with multiple combinations of physicochemical indicators, the optimal combination of physicochemical indicators is determined using multiple linear fitting and score comparison methods, specifically including: The principal component score matrix is fitted to the dataset under each combination of physicochemical indicators using multiple linear regression, and the corresponding coefficient of determination is calculated. Based on the determination coefficient and principal component contribution rate, calculate the score corresponding to each combination of physicochemical indicators; Compare the scores corresponding to all combinations of physicochemical indicators, and determine the combination of physicochemical indicators with the highest score as the optimal combination of physicochemical indicators.
4. The lubricating oil life prediction method based on multi-index fusion according to claim 3, characterized in that, The score corresponding to each combination of physicochemical indicators is as follows: ； In the formula, For the first c The scores corresponding to the combination of physicochemical indicators. For the first i The principal components and the first c The determining coefficient for the combination of physicochemical indicators For the first i The contribution rate of each principal component q For the front q One principal component.
5. The method for predicting lubricating oil life by fusing multiple indicators according to claim 1, characterized in that, Based on the principal component score matrix and the dataset corresponding to the optimal combination of physicochemical indicators, a linear fitting and scoring ratio method is used to assign weights to each physicochemical indicator in the optimal combination of physicochemical indicators, specifically including: The principal component score matrix is linearly fitted to the dataset corresponding to the optimal combination of physicochemical indicators, and the corresponding coefficient of determination is calculated. Based on the determination coefficient and principal component contribution rate, calculate the score corresponding to each physicochemical index in the optimal physicochemical index combination; The scores corresponding to each physicochemical index in the optimal physicochemical index combination are normalized to determine the weight of each physicochemical index in the optimal physicochemical index combination.
6. The lubricating oil life prediction method based on multi-index fusion according to claim 5, characterized in that, The scores corresponding to each physicochemical index in the optimal combination of physicochemical indices are as follows: ； In the formula, The first in the optimal combination of physicochemical indices b The scores corresponding to each physicochemical indicator For the first i Among the principal components and the optimal combination of physicochemical indices, the first... b The coefficient of determination for each physicochemical index For the first i The contribution rate of each principal component q For the front q One principal component.
7. The method for predicting lubricating oil life by fusing multiple indicators according to claim 5, characterized in that, The weights of each physicochemical index in the optimal combination of physicochemical indices are as follows: ； In the formula, The first in the optimal combination of physicochemical indices b The weight of each physicochemical indicator The first in the optimal combination of physicochemical indices b The scores corresponding to each physicochemical indicator B This represents the number of physicochemical indicators in the optimal combination of physicochemical indicators.
8. The method for predicting lubricating oil life by fusing multiple indicators according to claim 1, characterized in that, The lubricating oil life prediction model is as follows: ； In the formula, To predict lubricating oil life based on multi-indicator fusion, and These are the first and second physicochemical indices in the optimal combination of physicochemical indices, respectively. b The weight of each physicochemical indicator and These are the first and second physicochemical indices in the optimal combination of physicochemical indices, respectively. b Time-sensitivity thresholds for individual physicochemical indicators and These are the first and second physicochemical indices in the optimal combination of physicochemical indices, respectively. b New oil values for various physicochemical indicators and These are the first and second physicochemical indices in the optimal combination of physicochemical indices, respectively. b The oxidative degradation rate corresponding to each physicochemical index.
9. A computer system, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the lubricating oil life prediction method with multi-index fusion as described in any one of claims 1-8.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the lubricating oil life prediction method with multi-index fusion as described in any one of claims 1-8.