Lubricating oil identification method, device, equipment and medium

Abstract

The invention discloses a lubricating oil recognition method and device, equipment and a medium, and relates to the technical field of lubricating oil recognition. The spectrum data comprises chemical component information of the lubricating oil and is obtained by detecting target lubricating oil by adopting a high-resolution non-targeting technology; preprocessing the spectrum data of the target lubricating oil to obtain preprocessed spectrum data; performing feature extraction on the preprocessed spectrum data to obtain a spectrum feature corresponding to the target lubricating oil; inputting the spectral features corresponding to the target lubricating oil into a trained lubricating oil recognition model to obtain a prediction recognition result of the target lubricating oil; the trained lubricating oil identification model is a model obtained by taking the spectral features corresponding to the sample lubricating oil as input and taking the real identification result corresponding to the sample lubricating oil as a label for training. According to the invention, the efficiency and accuracy of lubricating oil identification are improved.

Description

Technical Field

[0001] The present application relates to the technical field of lubricating oil identification, and particularly to a lubricating oil identification method, device, equipment and medium. Background Art

[0002] In view of the current situation that the quality of current lubricating oil products is uneven and there are many counterfeit and shoddy products, it is difficult to effectively identify these counterfeit and shoddy products through technical means in market supervision, thus unable to effectively prevent and control the potential risks of product quality and safety brought about by this. At present, some manufacturers only identify based on information such as outer packaging and batch coding, but more and more illegal elements collect lubricating oil outer packaging barrels to avoid being identified as counterfeit products, which brings great difficulties to market supervision. Therefore, in order to avoid the situations of missed inspection, misjudgment and low efficiency when manually identifying lubricating oil products, there is an urgent need for a lubricating oil identification method with high identification accuracy. Summary of the Invention

[0003] The purpose of the present application is to provide a lubricating oil identification method, device, equipment and medium, which can improve the accuracy and efficiency of lubricating oil identification and reduce the situations of missed inspection and misjudgment when manually identifying lubricating oil products.

[0004] To achieve the above purpose, the present application provides the following solutions:

[0005] In a first aspect, the present application provides a lubricating oil identification method, including:

[0006] Obtain spectral data of the target lubricating oil; the spectral data includes chemical composition information of the lubricating oil and is obtained by detecting the target lubricating oil using a high-resolution non-targeted technology;

[0007] Preprocess the spectral data of the target lubricating oil to obtain preprocessed spectral data;

[0008] Extract features from the preprocessed spectral data to obtain spectral features corresponding to the target lubricating oil;

[0009] Input the spectral features corresponding to the target lubricating oil into a trained lubricating oil identification model to obtain a predicted identification result of the target lubricating oil; the trained lubricating oil identification model is a model trained with spectral features corresponding to sample lubricating oils as inputs and real identification results corresponding to the sample lubricating oils as labels.

[0010] Optionally, extracting features from the preprocessed spectral data to obtain spectral features corresponding to the target lubricating oil specifically includes:

[0011] Extract feature parameters in multiple dimensions from the preprocessed spectral data;

[0012] Using a feature selection method, the feature parameters of all dimensions are screened to obtain several key features; the key features are determined as the spectral features corresponding to the target lubricating oil.

[0013] Optionally, before inputting the spectral features corresponding to the target lubricating oil into the trained lubricating oil recognition model, the lubricating oil recognition method further includes:

[0014] Obtaining a sample data set; the sample data set includes the spectral features corresponding to several sample lubricating oils and the true recognition results corresponding to each sample lubricating oil;

[0015] Training the lubricating oil recognition model with the sample data set to obtain a trained lubricating oil recognition model.

[0016] Optionally, training the lubricating oil recognition model with the sample data set to obtain a trained lubricating oil recognition model specifically includes:

[0017] Training the lubricating oil recognition model with the sample data set to obtain an intermediate training model;

[0018] Evaluating the intermediate training model, and using the intermediate training model that meets the evaluation criteria as the trained lubricating oil recognition model.

[0019] Optionally, the lubricating oil recognition model is a fully connected neural network, a random forest model or a support vector machine.

[0020] Optionally, the spectral features include peak intensity, retention time and mass-to-charge ratio.

[0021] Optionally, the preprocessing includes denoising and baseline correction.

[0022] In a second aspect, the present application provides a lubricating oil recognition device, including:

[0023] A spectral data acquisition module for the target lubricating oil, configured to acquire spectral data of the target lubricating oil; the spectral data, including chemical composition information of the lubricating oil, is obtained by detecting the target lubricating oil using a high-resolution non-targeted technology;

[0024] A preprocessing module, configured to preprocess the spectral data of the target lubricating oil to obtain preprocessed spectral data;

[0025] A feature extraction module, configured to extract features from the preprocessed spectral data to obtain the spectral features corresponding to the target lubricating oil;

[0026] A lubricant identification module, which is used to input the spectral features corresponding to the target lubricant into a trained lubricant identification model to obtain the predicted identification result of the target lubricant; the trained lubricant identification model is a model trained with the spectral features corresponding to the sample lubricant as the input and the true identification result corresponding to the sample lubricant as the label.

[0027] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the above-mentioned lubricant identification method.

[0028] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned lubricant identification method is implemented.

[0029] According to the specific embodiments provided by the present application, the following technical effects are disclosed in the present application:

[0030] The present application provides a lubricant identification method, device, equipment and medium. By detecting the target lubricant through a high-resolution non-targeted technology, spectral data containing the chemical composition information of the target lubricant can be obtained; after preprocessing the spectral data containing the chemical composition information of the target lubricant, feature extraction is performed, and spectral features capable of characterizing the chemical composition information of the target lubricant can be obtained. Using the spectral features capable of characterizing the chemical composition information of the target lubricant as the input of a trained lubricant identification model for lubricant identification improves the accuracy of lubricant identification. Using the trained lubricant identification model to learn the law between the spectral features of the lubricant and the authenticity result of the lubricant, and using the trained lubricant identification model for lubricant identification can reduce the situation of missed inspection and misjudgment when manually identifying lubricant products, improve the accuracy of lubricant identification, and improve the efficiency of lubricant identification. Description of the Drawings

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0032] Figure 1 It is an application environment diagram of a lubricant identification method in an embodiment of the present application;

[0033] Figure 2 It is a flowchart of a lubricant identification method provided in an embodiment of the present application;

[0034] Figure 3Schematic diagram of the functional modules of a lubricating oil identification device provided by an embodiment of the present application;

[0035] Figure 4 Schematic diagram of the structure of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0036] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0037] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0038] The lubricating oil identification method provided by the embodiments of the present application can be applied to an application environment as Figure 1 shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set up separately, integrated on the server 104, placed in the cloud or on other servers. The terminal 102 can send the spectral data of the target lubricating oil to the server 104. After receiving the spectral data of the target lubricating oil, for the spectral data of the target lubricating oil, the server 104 preprocesses the spectral data of the target lubricating oil to obtain the preprocessed spectral data, extracts features from the preprocessed spectral data to obtain the spectral features corresponding to the target lubricating oil, and inputs the spectral features corresponding to the target lubricating oil into the trained lubricating oil identification model to obtain the predicted identification result of the target lubricating oil. The server 104 can feedback the obtained predicted identification result for the target lubricating oil to the terminal 102. In addition, in some embodiments, the lubricating oil identification method can also be implemented by the server 104 or the terminal 102 alone. For example, the terminal 102 can directly perform lubricating oil identification on the spectral data of the target lubricating oil, or the server 104 can obtain the spectral data of the target lubricating oil from the data storage system and perform lubricating oil identification on the spectral data of the target lubricating oil.

[0039] Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0040] In an exemplary embodiment, as Figure 2 shown, a lubricating oil identification method is provided. This method is executed by a computer device, and specifically can be executed alone by a computer device such as a terminal or a server, or can be jointly executed by a terminal and a server. In the embodiments of the present application, taking this method applied to Figure 1 the server 104 in it as an example for illustration, it includes the following steps 201 to step 204. Among them:

[0041] Step 201, obtain the spectral data of the target lubricating oil; the spectral data includes the chemical composition information of the lubricating oil and is obtained by detecting the target lubricating oil using a high-resolution non-targeted technology.

[0042] Step 202, preprocess the spectral data of the target lubricating oil to obtain the preprocessed spectral data.

[0043] Step 203, extract features from the preprocessed spectral data to obtain the spectral features corresponding to the target lubricating oil.

[0044] Step 204, input the spectral features corresponding to the target lubricating oil into the trained lubricating oil identification model to obtain the predicted identification result of the target lubricating oil; the trained lubricating oil identification model is a model trained with the spectral features corresponding to the sample lubricating oil as the input and the true identification result corresponding to the sample lubricating oil as the label.

[0045] Implement the above steps 201 to 204. By using high-resolution non-targeted technology to detect the target lubricating oil, spectral data containing the chemical composition information of the target lubricating oil can be obtained. After preprocessing the spectral data containing the chemical composition information of the target lubricating oil, feature extraction is performed, and spectral features capable of characterizing the chemical composition information of the target lubricating oil can be obtained. Using the spectral features characterizing the chemical composition information of the target lubricating oil as the input of the trained lubricating oil identification model for lubricating oil identification improves the accuracy of lubricating oil identification. By using the trained lubricating oil identification model to learn the law between the spectral features of lubricating oil and the authenticity results of lubricating oil, and using the trained lubricating oil identification model for lubricating oil identification, the situations of missed inspection and misjudgment during manual identification of lubricating oil products can be reduced, the accuracy of lubricating oil identification can be improved, and the efficiency of lubricating oil identification can be increased. Starting from the perspective of molecular big data, a digital fingerprint label at the molecular level is established for lubricating oil products, a characteristic spectral library of different brand lubricating oil products is established, and big data methods are used to characterize the characteristic differences of different brand products, and counterfeit and shoddy products are identified and appraised from the perspective of chemical composition, which can better support the market supervision of lubricating oil products.

[0046] Non-target screening technology for potential chemical substances in lubricating oil: Aiming at the material and use conditions of lubricating oil, a pretreatment method and a gas chromatography-high resolution mass spectrometry detection method are optimized and established to carry out non-target analysis on the volatile chemical substances migrating / releasing in lubricating oil samples. Two key aspects of work are carried out: First, a broad-spectrum screening of unknown chromatographic peaks with high responses in the samples (with a content ratio of more than 99.9%). For cases where there is a suitable match in the NIST spectral library, qualitative determination of unknown substances is achieved through a variety of qualitative means such as comprehensive scoring (spectral library retrieval, high-resolution filtering value HRF, etc.), retention index, PCI / NCI chemical ionization, and fine comparison of fragment ions. When there is no match with the spectral library, after determining the molecular formula, theoretical fragmentation, structure confirmation, and cleavage mechanism research of candidate molecular structures are carried out to obtain the best identification results. Second, chemometric methods are used to perform differential analysis on some specific samples, such as lubricating oils of different brands or different specifications and models, to find out the characteristic substances causing the differences and qualitatively determine the selected specific unknown substances.

[0047] In non-targeted screening, when broadly screening for highly responsive unknown peaks in samples, in cases where there is a suitable match in the NIST spectral library, accurate qualitative analysis is achieved through qualitative means such as comprehensive scoring, retention index, chemical ionization PCI / NCI to determine the molecular formula, fine comparison of fragment ions to distinguish isomers, and standard substance verification. If the spectral library cannot match, possible molecular structures are searched in the Chemspider and Pubchem databases, and the theoretical fragmentation of the candidate results is based on MassFrontier. Verification is carried out through accurate mass ions of primary / secondary mass spectrometry, fragmentation mechanism, standard substances, etc. Chemometric software such as Mass Profiler Professional is used to perform differential analysis such as principal component analysis and hierarchical clustering analysis on some samples to find characteristic compounds causing sample differences, and qualitative analysis of unknown substances is carried out on the selected specific substances.

[0048] Before step 204, the lubricating oil identification method further includes a training process of a lubricating oil identification model, including the following steps 301 to step 302.

[0049] Step 301: Obtain a sample data set; the sample data set includes spectral features corresponding to a number of sample lubricating oils and the true identification results corresponding to each sample lubricating oil;

[0050] Step 302: Use the sample data set to train the lubricating oil identification model to obtain a trained lubricating oil identification model.

[0051] Among them, step 301 specifically includes: obtaining a number of genuine lubricating oils and counterfeit lubricating oils as sample lubricating oils. Using high-resolution non-targeted technology to collect spectral, mass spectral or chromatographic data of the sample lubricating oils, these data contain a large amount of chemical composition information and can reflect the compositional differences of the lubricating oils. The collected spectral, mass spectral or chromatographic data of the sample lubricating oils are used as the spectral data of the sample lubricating oils. The high-resolution non-targeted technology can be high-resolution mass spectrometry or nuclear magnetic resonance, etc.

[0052] Preprocess the collected data, including denoising and baseline correction, etc., to improve the data quality and facilitate subsequent analysis. Then preprocess the spectral data of each sample lubricating oil to obtain the preprocessed spectral data of each sample lubricating oil.

[0053] Extract multiple characteristic parameters from the preprocessed spectral data, such as peak intensity, retention time, mass-to-charge ratio, etc. These characteristics will be used as the input of the machine learning model. Peak intensity: To reflect the compound concentration, it is an important indicator for identifying authenticity. Retention time: In chromatographic analysis, it represents the residence time of the compound in the chromatographic column and is used to assist in compound identification. Mass-to-charge ratio (m / z): In mass spectrometry analysis, it represents the mass-to-charge ratio of the compound and is used to determine the molecular weight of the compound. These characteristics jointly describe the chemical composition of the lubricating oil sample and provide comprehensive information for the model.

[0054] In high-resolution non-targeted techniques (such as high-resolution mass spectrometry or chromatography), "peak intensity" refers to the signal intensity, which is usually related to the concentration of specific compounds in the sample. Taking the spectral data of the pre-treated sample lubricating oil as an example of mass spectrometry, the specific steps are as follows:

[0055] 1. "Obtaining the mass spectrometry graph": Obtain the mass spectrometry graph of the sample lubricating oil, and each peak corresponds to a mass-to-charge ratio.

[0056] 2. "Baseline correction": Perform denoising and baseline correction to remove the baseline noise in the mass spectrometry graph and ensure the accuracy of the peak intensity.

[0057] 3. "Peak detection": Determine the retention time, identify the peaks in the mass spectrometry graph, and determine the mass-to-charge ratio and intensity of each peak.

[0058] 4. "Peak integration": Calculate the peak area of each peak as the peak intensity.

[0059] 5. "Normalization": Perform normalization on the peak intensity to obtain the normalized peak intensity to eliminate systematic errors.

[0060] 6. "Feature extraction": Use the retention time, mass-to-charge ratio, and normalized peak intensity as features and input them into a machine learning model or a deep learning model.

[0061] It should be noted that: peak overlap: overlapping peaks in the spectral data need to be separated by methods such as deconvolution; noise interference: the influence of noise needs to be reduced by methods such as filtering; instrument calibration: ensure the stable state of the instrument and avoid response differences.

[0062] After obtaining the characteristic parameters in multiple dimensions, not all peaks have discriminative significance. In this application, key features are screened through a feature selection method, and the selected key features are used as the spectral features corresponding to the sample lubricating oil. In this embodiment, the key features screened through feature selection include peak intensity, retention time, and mass-to-charge ratio.

[0063] After obtaining the spectral features corresponding to each sample lubricating oil, label the sample lubricating oil to distinguish genuine and fake products, and use the labeling result as the label corresponding to the sample lubricating oil to form a labeled sample data set for supervised learning. After obtaining the sample data set, perform model selection and training: select a suitable machine learning algorithm (such as support vector machine (SVM), random forest, neural network, etc.), and use the labeled data to train the model. It should be noted that ensure the balance of the number of genuine and fake samples in the training set to avoid model bias. The goal of the model is to learn the characteristic differences between genuine and fake lubricating oils. Then the lubricating oil identification model can be a fully connected neural network (FCN), a random forest model, or a support vector machine.

[0064] The lubricant identification model includes an input layer, a hidden layer, and an output layer. The input layer accepts three spectral features (including peak intensity, retention time, and mass-to-charge ratio) as inputs, and the peak intensity input into the lubricant identification model is the peak intensity after normalization processing. When the spectral features of the sample lubricant are input into the lubricant identification model, they can be represented in the form of feature vectors. For example: Sample 1: [peak intensity 1, retention time 1, mass-to-charge ratio 1], Sample 2: [peak intensity 2, retention time 2, mass-to-charge ratio 2]... Sample N: [peak intensity N, retention time N, mass-to-charge ratio N]. These feature vectors will serve as the inputs of the model. Suppose there is a lubricant sample with the following mass spectrometry and chromatography data: Peak intensity: 1500 (in arbitrary units), Retention time: 5.3 minutes, Mass-to-charge ratio: 456.3. Represent it as the feature vector [1500, 5.3, 456.3]. After inputting it into the model, the model may output "1" (true) or "0" (false). That is, when the lubricant identification model outputs "1", it indicates that the lubricant sample is genuine, and when the lubricant identification model outputs "0", it indicates that the lubricant sample is counterfeit. Hidden layer: Learn the relationships between features through algorithms such as neural networks and decision trees. Output layer: Output the predicted identification result (true or false).

[0065] Model training: 1) Feature standardization: Perform standardization processing on spectral features such as peak intensity, retention time, and mass-to-charge ratio to eliminate the dimensional differences. 2) Label assignment: Assign labels (true or false) to each sample lubricant. 3) Train the model: Use the labeled data, that is, the sample data set to train the model and learn the relationship between the spectral features and labels of the lubricant. By taking peak intensity, retention time, and mass-to-charge ratio as input features, the machine learning model can effectively learn the chemical feature differences between genuine and counterfeit lubricants, thus achieving efficient and accurate identification.

[0066] Use cross-validation or an independent test set to evaluate the model performance, and optimize the model by adjusting hyperparameters or selecting different algorithms to ensure its generalization ability.

[0067] Then step 302 specifically includes: Training the lubricant identification model with the sample data set to obtain an intermediate training model; Evaluating the intermediate training model, and taking the intermediate training model that meets the evaluation criteria as the trained lubricant identification model. Among them, indicators such as accuracy, recall rate, and F1 score can be used to evaluate the model performance, and the thresholds of each indicator are set as the evaluation criteria. The intermediate training model that meets the evaluation criteria, that is, the intermediate training model not less than the thresholds of each indicator, is taken as the trained lubricant identification model.

[0068] After obtaining the trained lubricant identification model, perform model prediction. That is, for a new lubricant sample, extract its spectral features, including peak intensity, retention time, and mass-to-charge ratio, and input them into the trained lubricant identification model. The model outputs the predicted identification result of the new lubricant sample, and the predicted identification result is true or false.

[0069] In step 201 above, spectral, mass spectral or chromatographic data of the target lubricating oil are collected using a high-resolution non-targeted technique, and the spectral, mass spectral or chromatographic data are used as the spectral data of the target lubricating oil. These data contain a large amount of chemical composition information and can reflect the compositional differences of the lubricating oil. The high-resolution non-targeted technique can be high-resolution mass spectrometry or nuclear magnetic resonance, etc.

[0070] The spectral data of the target lubricating oil obtained in step 201 are preprocessed to obtain preprocessed spectral data. The preprocessing may include denoising and baseline correction. Then, feature extraction is performed on the preprocessed spectral data. Step 203 may include the following steps 401 to 402.

[0071] Step 401: Feature extraction is performed on the preprocessed spectral data to obtain feature parameters in multiple dimensions.

[0072] Step 402: Using a feature selection method, the feature parameters in all dimensions are screened to obtain several key features; the key features are determined as the spectral features corresponding to the target lubricating oil. In this embodiment, the key features screened by feature selection include peak intensity, retention time, and mass-to-charge ratio.

[0073] The peak intensity, retention time, and mass-to-charge ratio obtained above are used as the spectral features corresponding to the target lubricating oil, and the spectral features corresponding to the target lubricating oil are input into the trained lubricating oil identification model to obtain the predicted identification result of the target lubricating oil. The peak intensity input into the trained lubricating oil identification model is the peak intensity after normalization processing.

[0074] The lubricating oil identification method provided in this application further includes a continuous update process of the lubricating oil identification model, specifically: as new sample lubricating oils are added, the lubricating oil identification model is updated regularly, that is, the spectral features of the new sample lubricating oils are added to the original sample data set to obtain a new sample data set, and the trained lubricating oil identification model is retrained using the new sample data set to obtain a new trained lubricating oil identification model, so as to maintain the accuracy and robustness of the lubricating oil identification model.

[0075] Apply the trained lubricating oil identification model to new samples. By inputting high-resolution non-targeted data, the model outputs the authenticity identification result. Combine chemical knowledge to explain the identification result of the model, identify the key differential components of genuine and fake lubricating oils, and provide a basis for quality control. Integrate the entire process into an automated system to achieve rapid and automatic identification of the authenticity of lubricating oils. Use high-resolution non-targeted technology to provide comprehensive chemical composition information of lubricating oils, and use machine learning models to automatically identify complex patterns, improving the efficiency and accuracy of lubricating oil identification. By combining these two technologies, efficient and accurate identification of the authenticity of lubricating oils can be achieved, reducing the occurrence of harmful events caused by counterfeit and shoddy products, creating a healthy consumption environment, providing technical support for protecting the legitimate rights and interests of consumers, enhancing the early warning and response capabilities for sudden quality and safety incidents. The relevant research results can also guide lubricating oil production enterprises to recognize the hazards and differential information of counterfeit and shoddy lubricating oils, provide technical support for cracking down on counterfeit and shoddy products, recover losses for regular production enterprises, provide a good market order for consumers, and will produce good social benefits.

[0076] The application scenarios of the lubricating oil identification method provided by this application include: (1) Quality control: Monitor the quality of lubricating oil in real time during the production process. (2) Market supervision: Rapidly detect counterfeit and shoddy lubricating oil products in the market. (3) R & D support: Analyze the formulations of different lubricating oils and optimize product performance.

[0077] This application also provides an application scenario that applies the above-mentioned lubricating oil identification method. Specifically: The lubricating oil identification method provided in this embodiment can be applied in the lubricating oil market supervision scenario. The lubricating oil market supervision scenario includes a data collection link, a lubricating oil identification link, and a supervision link; the spectral data of the target lubricating oil enters the lubricating oil identification link from the data collection link, and the corresponding prediction and identification result is obtained through a human-machine collaborative method and enters the downstream supervision link. The lubricating oil identification method provided in this embodiment belongs to the machine marking link in the lubricating oil identification link. Specifically, in the process of the lubricating oil identification link for the target lubricating oil, the target lubricating oil can be marked based on the collaborative method of machine marking and manual marking, that is, adding the corresponding authenticity identification label to the spectral data of the target lubricating oil.

[0078] Based on the same inventive concept, the embodiment of this application also provides a lubricating oil identification device for implementing the above-mentioned lubricating oil identification method. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more of the following embodiments of the lubricating oil identification device can refer to the limitations on the lubricating oil identification method in the above text, and will not be repeated here.

[0079] In an exemplary embodiment, as Figure 3As shown, a lubricating oil identification device is provided, which includes the following modules:

[0080] A spectral data acquisition module T1 for the target lubricating oil, which is used to acquire the spectral data of the target lubricating oil; the spectral data, including the chemical composition information of the lubricating oil, is obtained by detecting the target lubricating oil using a high-resolution non-targeted technology.

[0081] A preprocessing module T2, which is used to preprocess the spectral data of the target lubricating oil to obtain the preprocessed spectral data.

[0082] A feature extraction module T3, which is used to extract features from the preprocessed spectral data to obtain the spectral features corresponding to the target lubricating oil.

[0083] A lubricating oil identification module T4, which is used to input the spectral features corresponding to the target lubricating oil into a trained lubricating oil identification model to obtain the predicted identification result of the target lubricating oil; the trained lubricating oil identification model is a model trained with the spectral features corresponding to the sample lubricating oil as the input and the true identification result corresponding to the sample lubricating oil as the label.

[0084] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 4 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store lubricating oil identification data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it implements a lubricating oil identification method.

[0085] Those skilled in the art can understand that Figure 4 the structure shown in

[0086] In an exemplary embodiment, a computer device is provided, which includes a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0087] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0088] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0089] Those of ordinary skill in the art can understand that all or part of the processes of implementing the above method embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium, and when the computer program is executed, it can include the processes of the above method embodiments. Among them, any reference to a memory, database, or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memories. 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), magnetoresistive 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 be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0090] In each of the embodiments provided in this application, the database involved may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., and is not limited thereto. In each of the embodiments provided in this application, the processor may be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., and is not limited thereto.

[0091] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0092] Specific examples are used in this article to elaborate on the principles and implementation manners of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application; at the same time, for those of ordinary skill in the art, according to the idea of this application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to this application.

Claims

1. A lubricating oil identification method, characterized in that, The lubricating oil identification method includes: Obtaining spectral data of the target lubricating oil; the spectral data includes chemical composition information of the lubricating oil and is obtained by detecting the target lubricating oil using a high-resolution non-targeted technique; Preprocessing the spectral data of the target lubricating oil to obtain preprocessed spectral data; Performing feature extraction on the preprocessed spectral data to obtain spectral features corresponding to the target lubricating oil; Inputting the spectral features corresponding to the target lubricating oil into a trained lubricating oil identification model to obtain a predicted identification result of the target lubricating oil; the trained lubricating oil identification model is a model trained with spectral features corresponding to sample lubricating oils as inputs and true identification results corresponding to the sample lubricating oils as labels.

2. The lubricating oil identification method according to claim 1, wherein Performing feature extraction on the preprocessed spectral data to obtain spectral features corresponding to the target lubricating oil, specifically including: Performing feature extraction on the preprocessed spectral data to obtain feature parameters in multiple dimensions; Using a feature selection method to screen the feature parameters in all dimensions to obtain several key features; determining the key features as the spectral features corresponding to the target lubricating oil.

3. The lubricating oil identification method according to claim 1, characterized in that, Before inputting the spectral features corresponding to the target lubricating oil into the trained lubricating oil identification model, the lubricating oil identification method further includes: Obtaining a sample data set; the sample data set includes spectral features corresponding to several sample lubricating oils and true identification results corresponding to each sample lubricating oil; Training the lubricating oil identification model using the sample data set to obtain a trained lubricating oil identification model.

4. The lubricating oil identification method according to claim 3, characterized in that, Training the lubricating oil identification model using the sample data set to obtain a trained lubricating oil identification model, specifically including: Training the lubricating oil identification model using the sample data set to obtain an intermediate trained model; Evaluating the intermediate trained model and taking the intermediate trained model that meets the evaluation criteria as the trained lubricating oil identification model.

5. The lubricating oil identification method according to claim 1, characterized in that The lubricating oil identification model is a fully connected neural network, a random forest model, or a support vector machine.

6. The lubricating oil identification method according to claim 1, characterized in that, The spectral features include peak intensity, retention time, and mass-to-charge ratio.

7. The lubricating oil identification method according to claim 1, characterized in that The preprocessing includes denoising and baseline correction.

8. A lubricating oil identification device, characterized in that, The lubricating oil identification device includes: A spectral data acquisition module for the target lubricating oil, configured to obtain spectral data of the target lubricating oil; the spectral data includes chemical composition information of the lubricating oil and is obtained by detecting the target lubricating oil using a high-resolution non-targeted technique; A preprocessing module, configured to preprocess the spectral data of the target lubricating oil to obtain preprocessed spectral data; A feature extraction module, configured to perform feature extraction on the preprocessed spectral data to obtain spectral features corresponding to the target lubricating oil; A lubricating oil identification module, configured to input the spectral features corresponding to the target lubricating oil into a trained lubricating oil identification model to obtain a predicted identification result of the target lubricating oil; the trained lubricating oil identification model is a model trained with spectral features corresponding to sample lubricating oils as inputs and true identification results corresponding to the sample lubricating oils as labels.

9. A computer device, comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the lubricating oil identification method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the lubricating oil identification method described in any one of claims 1-7.

Images