Vehicle engine oil performance prediction method and system, electronic equipment and storage medium

Abstract

The invention provides a vehicle engine oil performance prediction method and system, electronic equipment and a storage medium, and relates to the technical field of vehicle engine oil informatization management. In order to solve the problem that in an engine oil performance prediction method in the prior art, a machine learning model takes signals collected by a sensor as features, and consequently feature extraction is not comprehensive, parameters of internal components of an engine lubricating system and parameters of an external system interacting with the engine lubricating system are obtained and associated; determining a characteristic value of an internal component failure mode according to the associated internal component parameters; obtaining label data of an internal component failure mode according to the engine oil sample; and training an engine oil performance prediction model according to the characteristic values and the label data. And based on the prediction model, the engine oil performance is predicted according to the operation data transmitted by the Internet of Vehicles in real time. And by associating internal component parameters and external system parameters of the system, the problems of limitation of direct measurement of the sensor and incomplete feature extraction are solved, so that the prediction result is more reliable.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle engine oil information management, and in particular to a vehicle engine oil performance prediction method and system, an electronic device and a storage medium. BACKGROUND

[0002] Vehicle engine oil is a key lubricant for the normal operation of the engine, and its main functions include reducing friction between internal parts of the engine and dissipating heat. Regularly checking the oil level and condition is crucial for protecting the engine's lifespan.

[0003] Fault modes of engine oil, such as high-temperature oxidation and increased wear and tear, can affect engine performance. By predicting engine oil performance degradation and remaining life, maintenance measures can be taken in advance to avoid engine failures caused by engine oil problems. Traditional maintenance relies on fixed mileage or time intervals, which may result in premature or late maintenance. Performance prediction methods can determine the optimal time for maintenance based on the actual state of the engine oil, reducing unnecessary maintenance.

[0004] Currently, vehicle engine oil performance prediction methods use machine learning models for prediction. Machine learning models generally use signals directly collected by sensors as feature values, which may not cover system-level interaction parameters, leading to incomplete feature extraction and affecting the prediction results. In addition, the machine learning models in the prior art are based on signal features and do not consider the interaction between system functions and the whole vehicle, making it difficult to identify the coupling effects across subsystems and unable to fully reflect the failure mechanism.

[0005] Therefore, it is of great significance to develop a vehicle engine oil performance prediction method to improve the reliability of engine oil performance prediction. SUMMARY

[0006] The present application proposes a vehicle engine oil performance prediction method to address the problem of incomplete feature extraction in existing machine learning models, which generally use signals collected by sensors as features. The method includes the following steps: Step S1: Obtain internal component parameters of the engine lubrication system and external system parameters interacting with the engine lubrication system; Step S2: Associate the internal component parameters with the external system parameters; Step S3: Determine the feature values of the internal component failure modes based on the associated internal component parameters, wherein the feature values of the internal component failure modes are parameters reflecting engine oil performance; Step S4: Collect engine oil samples from the engine lubrication system and obtain label data of the internal component failure modes based on the engine oil samples; Step S5, training a prediction model of engine oil performance by taking the characteristic value as input and the label data as output; Step S6, predicting the engine oil performance based on the prediction model according to the real-time transmission of the operation data of the Internet of Vehicles, and outputting a maintenance prompt information when the prediction result reaches a preset threshold.

[0007] Further, in the step S2, the internal component parameters are associated with the external system parameters by using a failure mode avoidance method, wherein the tools used by the failure mode avoidance method include a system boundary graph and an interface matrix and list; The corresponding relationship between the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system is determined by the system boundary graph; The interaction mode of the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system is analyzed by using the interface matrix and list.

[0008] Further, the characteristic value includes an oxidation damage amount and a wear amount, Wherein, the oxidation damage amount is calculated by an engine oil coking damage model, and the wear amount is calculated by a crankshaft bearing wear model.

[0009] Further, the engine oil coking damage model includes an engine oil delivery rate sub-model, a coking risk rate sub-model and an oxidation damage amount sub-model, and the calculation steps of the oxidation damage amount include: The blow-by rate in the internal component parameters of the engine lubrication system is obtained, and the blow-by rate is input into the engine oil delivery rate sub-model, and the engine oil delivery rate sub-model calculates the engine oil delivery rate according to the blow-by rate; The piston top land hot spot temperature in the internal component parameters of the engine lubrication system is obtained, and the piston top land hot spot temperature is input into the coking risk rate sub-model, and the coking risk rate sub-model calculates the coking risk rate according to the piston top land hot spot temperature; The engine speed in the internal component parameters of the engine lubrication system is obtained, and the engine speed, the engine oil delivery rate and the coking risk rate are input into the oxidation damage amount sub-model, and the oxidation damage amount sub-model calculates the oxidation damage amount according to the engine speed, the engine oil delivery rate and the coking risk rate.

[0010] Further, the wear amount is calculated by a crankshaft bearing wear model, specifically including: The peak combustion pressure and the total normal load of the crankshaft bearing in the internal component parameters of the engine lubrication system are obtained; The peak combustion pressure and the total normal load of the crankshaft bearing are input into the crankshaft bearing wear model, and the crankshaft bearing wear model calculates the wear amount based on the peak combustion pressure and the total normal load of the crankshaft bearing.

[0011] Furthermore, in step S4, the label data includes the oil's antioxidant index and iron wear content.

[0012] Furthermore, the prediction result includes remaining lifetime, and the calculation process for the remaining lifetime includes; The real-time operation data transmitted by the vehicle network is input into the prediction model. The antioxidant index in the tag data is used as the first evaluation dimension to obtain the first remaining lifetime. The iron wear content in the tag data is used as the second evaluation dimension to obtain the second remaining lifetime. Compare the first remaining lifetime with the second remaining lifetime, and take the smaller of the first remaining lifetime and the second remaining lifetime as the final remaining lifetime.

[0013] The present invention also provides a vehicle engine oil performance prediction system, wherein the prediction system employs the vehicle engine oil performance prediction method as described in any of the preceding claims, and specifically includes the following modules: The parameter acquisition module is used to acquire parameters of internal components of the engine lubrication system and parameters of external systems that interact with the engine lubrication system. The parameter association module is connected to the parameter acquisition module and is used to associate the internal component parameters with the external system parameters; The feature value calculation module is connected to the parameter association module and is used to determine the feature value of the internal component failure mode based on the associated internal component parameters. The feature value of the internal component failure mode refers to the parameter that reflects the performance of the engine oil. The tag data acquisition module is used to collect engine oil samples from the engine lubrication system and obtain tag data of internal component failure modes based on the engine oil samples. The prediction model training module is connected to the label data acquisition module and the feature value calculation module. It is used to train the prediction model of engine oil performance by taking the feature value as input and the label data as output. The prediction model is used to predict the engine oil performance based on the real-time operation data transmitted by the vehicle network and output the prediction result. The early warning module is connected to the prediction model training module and is used to judge the prediction results. When the prediction results reach a preset threshold, maintenance prompt information is output.

[0014] The application also provides an electronic device, comprising a processor and a memory; the processor is used to execute the steps of the vehicle engine oil performance prediction method according to any one of the above by calling the program or instruction stored in the memory.

[0015] The application also provides a computer readable storage medium comprising computer program instructions, which make a computer execute the steps of the vehicle engine oil performance prediction method according to any one of the above.

[0016] Compared with the prior art, the application has the following beneficial effects: Firstly, the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system are obtained and associated; the characteristic values of the internal component failure modes are determined according to the associated internal component parameters; the label data of the internal component failure modes are obtained according to the oil samples; and the prediction model of the engine oil performance is trained according to the characteristic values and the label data. Based on the prediction model, the engine oil performance is predicted according to the real-time transmission of the operation data of the Internet of Vehicles. By associating the internal component parameters and the external system parameters, the system level and the vehicle level are associated, and when the characteristic values of the internal component failure modes are determined according to the associated internal component parameters, all the interaction parameters of the internal component parameters are considered, which can make up for the limitations of direct measurement of sensors and the incompleteness of feature extraction, so that the prediction result is more reliable. Secondly, the corresponding relationship between the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system is determined by the system boundary graph; the interface matrix and the list are used to analyze the interaction mode of the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system, and finally the association relationship between the internal component parameters and the external system parameters is established, the characteristic values of the failure modes are determined based on the associated internal component parameters, and thus the prediction model is established. Compared with the machine learning model based on signal features in the prior art, the application considers the corresponding relationship and the interaction mode of the system level and the vehicle level, and can fully reflect the failure mechanism. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the specific embodiments of the application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the specific embodiments or the prior art description. Obviously, the drawings in the following description are some embodiments of the application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0018] Figure 1 is the flowchart of the vehicle engine oil performance prediction method provided by the embodiment of the application; Figure 2is a residual life prediction result graph of machine oil provided by an embodiment of the present application, taking an oxidation index as a label; Figure 3 is a residual life prediction result graph of machine oil provided by an embodiment of the present application, taking an iron content as a label; Figure 4 is a structural schematic diagram of a vehicle machine oil performance prediction system provided by an embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work belong to the scope of protection of the present application.

[0020] The specific embodiments of the present application will be described below.

[0021] In view of the problem that in the prior art, the machine learning model in the machine oil performance prediction method takes the signals collected by the sensor as the features, resulting in incomplete feature extraction, the present application is characterized in that: internal component parameters of an engine lubrication system and external system parameters interacting with the engine lubrication system are acquired and associated; feature values of internal component failure modes are determined according to the associated internal component parameters; label data of the internal component failure modes is acquired according to the machine oil samples; and a prediction model of machine oil performance is trained according to the feature values and the label data. Based on the prediction model, the machine oil performance is predicted according to the running data transmitted in real time by the Internet of Vehicles. By associating the internal component parameters and the external system parameters, the limitations of direct measurement by the sensor and the problem of incomplete feature extraction are made up, so that the prediction result is more reliable.

[0022] EMBODIMENT

[0023] An embodiment of the present application provides a vehicle machine oil performance prediction method, Figure 1 is a flowchart of the vehicle machine oil performance prediction method provided by an embodiment of the present application, as shown in Figure 1 The vehicle machine oil performance prediction method specifically includes the following steps: Step S1, acquiring internal component parameters of an engine lubrication system and external system parameters interacting with the engine lubrication system.

[0024] The engine lubrication system is a system that circulates engine oil in the system to realize lubrication, cooling and protection of key friction pairs of the engine, and maintains normal operation through energy and material exchange of internal components. The engine lubrication system is the carrier of the function of the engine oil, and its operating state directly affects the performance change of the engine oil. The working state of the core components of the engine lubrication system, such as the piston ring and the crankshaft bearing, is the main cause of the performance change of the engine oil.

[0025] The external system parameter interacting with the engine lubrication system refers to the external parameter affecting the operation of the lubrication system through energy, material or information exchange. For example, environmental parameters, working condition parameters of vehicle operation, etc. These parameters affect the engine load and the thermal load of the lubrication system, and indirectly change the performance of the engine oil. The internal component parameter is directly related to the performance change of the engine oil, and the external system parameter reflects the influence of the environment and the working condition on the lubrication system.

[0026] Step S2, associating the internal component parameter with the external system parameter.

[0027] By analyzing the internal influence relationship between the parameters of the internal components of the engine lubrication system and the external related parameters of the system, the analysis logic of the internal component parameters and the external system parameters is constructed to more comprehensively evaluate the performance of the engine oil and the state of the lubrication system.

[0028] For example, the internal component parameter is associated with the external system parameter, which includes using the failure mode avoidance method (FMA) to associate the internal component parameter with the external system parameter. The failure mode avoidance method is used to establish the correlation between the system level and the vehicle level to determine the failure mode and the failure mechanism corresponding to the failure mode.

[0029] The failure mode avoidance method refers to a method that takes market failures as the starting point, identifies interference factors between user use and adjacent systems, analyzes functional and durability failure modes, and outputs potential failure causes and physical mechanisms. The failure mode avoidance method is realized through a series of process steps and tools. The process steps include failure history review, identification of system failure modes and influences, analysis of potential causes of system failure, development of failure countermeasures and verification methods, etc. The tools used include function tree, system boundary diagram, interface matrix and list, parameter diagram (P diagram), causal performance robustness worksheet, design verification plan, etc. These tools are respectively used to understand system functions, identify internal and external interaction relationships, analyze interfaces, determine noise factors, and describe interference effects.

[0030] During the actual use of the semi-trailer, market failures can refer to various failure conditions of the engine oil lamp related system. The system level refers to the engine lubrication system, and the vehicle level refers to the entire semi-trailer, including the interaction between the engine lubrication system and other systems of the vehicle, the external environment and the user use scene.

[0031] Based on the FMA method, the correlation between the system level and the vehicle level is established, the correlation between the system level component failure and the vehicle level working condition and environmental factors is realized, and finally the failure mode and its corresponding failure mechanism are determined. Through the establishment of the correlation between the system level and the vehicle level, the root cause of failure can be accurately located, and the systematicness and comprehensiveness of failure analysis are improved.

[0032] The tool used in the failure avoidance method includes a system boundary diagram and an interface matrix and list. The system boundary diagram is used to determine the corresponding relationship between the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system. The interface matrix and list are used to analyze the interaction mode of the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system.

[0033] The system boundary diagram is used to identify internal components and external systems interacting with them. When the system level is the engine lubrication system, the internal components of the engine lubrication system are determined through the system boundary diagram, including the crankcase ventilation system, the first piston ring, the piston pin, the connecting rod, the connecting rod bearing, the crankshaft, and the oil seal assembly. The above components are used to perform system level functions. The external systems interacting with the above internal components are determined through the system boundary diagram, including user factors, environment, exhaust system associated components, etc. These external systems belong to the vehicle level and affect the lubrication system function through various interactions.

[0034] The interface matrix and list are used for structured matrix analysis of all internal and external interfaces. The interaction type is defined to describe the correlation between the internal components and the external systems of the lubrication system, including physical contact, information exchange, and energy transfer. The interface matrix and list are used to analyze the interface interaction between internal components, for example, the interaction between the first piston ring and the main oil passage oil belongs to energy transfer, high temperature is transferred to the oil through the piston ring, causing oil degradation and coking. The mixed gas enters the crankcase oil through the piston ring, increasing the oil impurities, which belongs to material exchange. The interface matrix and list are used to analyze the interface interaction between internal components and external systems of the vehicle level, for example, the interaction between the crankshaft and oil seal assembly and the main oil passage oil includes energy transfer and material exchange.

[0035] In step S3, the characteristic value of the internal component failure mode is determined according to the correlated internal component parameters. The characteristic value of the internal component failure mode refers to the parameter reflecting the performance of the oil.

[0036] The characteristic value of the internal component failure mode is determined according to the correlated internal component parameters, which can reflect the influence of the external system parameters, for example, the influence of different vehicle working conditions or different use scenarios on the internal component parameters. The characteristic value refers to the parameter reflecting the performance of the oil, which can be used to quantitatively describe the severity of the failure mode.

[0037] The characteristic value includes an oxidation damage amount and a wear amount, wherein the oxidation damage amount is calculated by an engine oil coking damage model, and the wear amount is calculated by a crankshaft bearing wear model.

[0038] Specifically, the engine oil coking damage model includes an engine oil delivery rate sub-model, a coking risk rate sub-model, and an oxidation damage amount sub-model. The calculation steps of the oxidation damage amount include: obtaining a blow-by rate in the engine lubrication system internal component parameters, inputting the blow-by rate into the engine oil delivery rate sub-model, and calculating the engine oil delivery rate according to the blow-by rate by the engine oil delivery rate sub-model.

[0039] The engine oil delivery rate is calculated based on the engine oil delivery rate sub-model, and the calculation process is as follows: ; ; wherein BBY represents the blow-by rate, x is a value standardized by the interval range of the blow-by rate, and the engine oil delivery rate w oil is finally calculated by a cubic function.

[0040] The piston top land hot spot temperature in the engine lubrication system internal component parameters is obtained, and the piston top land hot spot temperature is input into the coking risk rate sub-model, and the coking risk rate is calculated according to the piston top land hot spot temperature by the coking risk rate sub-model.

[0041] The coking risk rate is calculated based on the coking risk rate sub-model, and the calculation process is as follows: ; ; wherein T HS represents the piston top land hot spot temperature, y is a value standardized by the interval range of the piston top land hot spot temperature, and wa is the coking risk rate.

[0042] The engine speed in the engine lubrication system internal component parameters is obtained, and the engine speed, the engine oil delivery rate, and the coking risk rate are input into the oxidation damage amount sub-model, and the oxidation damage amount is calculated according to the engine speed, the engine oil delivery rate, and the coking risk rate by the oxidation damage amount sub-model. The specific process of calculating the oxidation damage amount based on the oxidation damage amount sub-model is as follows: In a set time period, the relative damage increment of all set time periods is obtained according to the engine oil delivery rate, the coking risk rate, and the engine speed and the duration of the set time period. The specific formula is as follows: ; wherein ti represents the duration of the set time period i, Di,rel represents the relative damage amount of the set time period, w oil i represents the engine oil delivery rate in the set time period, n i represents the engine speed in the set time period, and w a represents the coking risk rate in the set time period.

[0043] The relative damage amount in a plurality of set time periods is integrated to obtain an oxidation damage amount Drel. The specific formula is: ; The wear amount is calculated by a crankshaft bearing wear model, specifically including: obtaining the peak combustion pressure and the total normal load of the crankshaft bearing in the internal component parameters of the engine lubrication system. The peak combustion pressure and the total normal load of the crankshaft bearing are input into the crankshaft bearing wear model, and the crankshaft bearing wear model calculates the wear amount according to the peak combustion pressure and the total normal load of the crankshaft bearing. The specific calculation process includes: The wear amount is calculated according to the peak combustion pressure, the inertia force and the wear index.

[0044] The peak combustion pressure PFP is a key load factor causing the wear of the crankshaft bearing, and the formula for calculating the wear amount according to the peak combustion pressure PFP is: ;

[0045] Wherein, Drel' is the wear amount, k is the wear index, which is related to the lubrication state, and n is the engine speed. It reflects the influence of the peak combustion pressure and the wear index on the wear amount. The higher the speed and the greater the peak combustion pressure, the greater the relative wear amount.

[0046] The inertia force affects the wear through the total normal load W of the crankshaft bearing, and the calculation formula of the total volume Q of wear debris is: Q=K(WL) k / H; Wherein, k is the wear index, W is the total normal load of the crankshaft bearing, L is the sliding distance, H is the hardness of the softest contact surface, and K is a dimensionless constant.

[0047] The formula for calculating the wear amount according to the inertia force is: ; Wherein, n is the engine speed, k is the wear index, and n 2 is directly related to the inertia force, which reflects the influence of the inertia force and the wear index on the wear amount.

[0048] Step S4, collecting engine oil samples in the engine lubrication system, and obtaining label data of internal component failure modes according to the engine oil samples. The label data includes the antioxidant index and the iron wear content of the engine oil. The engine oil samples can be collected by manual sampling of engine oil on site and chemical detection to obtain the antioxidant index and the iron wear content.

[0049] The antioxidant index can be represented by total base number (TBN), which is a key indicator reflecting the oxidation capacity of the oil. As the oil is used, the TBN gradually decreases. The iron wear content is an indicator reflecting the degree of wear of the engine crankshaft bearing. The more severe the wear, the higher the iron content.

[0050] Step S5, the characteristic value is taken as the input quantity, and the label data is taken as the output quantity, and the prediction model of the oil performance is trained.

[0051] Through a long short-term memory (LSTM) neural network, the characteristic values in the internal components, such as the first ring of the piston coke damage, are taken as inputs, and the TBN is taken as label data for training, so that the model can learn the mapping relationship between the characteristics and the labels. The LSTM neural network includes four S function units, and the S function is the basic functional unit constituting the LSTM neural network structure. These units process information through input gates, forget gates, output gates, etc. Among them, the forget gate can decide whether to forget the information stored in the module. When the value it generates is close to zero, the value remembered in the module will be forgotten. The input gate controls whether the data is transmitted, and the output gate decides the memory information that can be output, which together realizes the effective processing of time series data by the LSTM.

[0052] For example, 4 sets of LSTM features and label data collected are divided into a training set and a test set, of which 3 sets of data are used as the training set and 1 set of data is used as the test set. The training set is used for the learning process of the LSTM neural network. The model learns the mapping relationship between the features and the label data in the training set, adjusts the internal parameters, and establishes a prediction model. The test set is used to evaluate the performance of the trained model and test the prediction ability of the model on the data not involved in the training, to verify the generalization and accuracy of the model. Finally, based on the prediction results of the test set, combined with the oil limit indicators determined by the national standard, the oil remaining life prediction results are obtained.

[0053] Table 1 shows the detection results of the user oil sample. Through the data of multiple oil performance indicators at different use mileages, the performance degradation of the oil with the increase of the use mileage is intuitively presented. When the use mileage of the oil increases from 0 km to 92478 km, the TBN decreases from 10.19 mgKOH / g to 6.89 mgKOH / g, and the iron content increases from 7 ppm to 22 ppm, clearly showing the degradation trend of the oil antioxidant performance and the increase of wear impurities, providing key label data for model training.

[0054] Table 1 Detection results of user oil sample

[0055] The current engine oil performance is determined by the current measured value of the label data, and the trend of the label data is the basis for model prediction. The remaining life is calculated based on the current engine oil state, combined with the predicted model of the label data change trend and the failure limit index, to calculate the difference between the mileage corresponding to the current label data and the mileage corresponding to the engine failure. Figure 2 is the engine oil remaining life prediction result graph provided by the embodiment of the present application with the antioxidant index as the label data, as shown in Figure 2 , including two prediction curves pre1 and pre2, which are the output results of the same prediction model under different training modes. Both prediction curves are consistent with the true value trend and have small deviations, which can further verify the reliability of the prediction model. From Figure 2 , the remaining mileage when the TBN decreases to the national standard limit value (<5.5 mgKOH / g) can be predicted, and the prediction result is 40,000 kilometers. Figure 3 is the engine oil remaining life prediction result graph provided by the embodiment of the present application with the iron content as the label, including two prediction curves pre3 and pre4, both of which are consistent with the true value trend and have small deviations. From Figure 3 , the remaining mileage when the iron content reaches the national standard limit value (<150 ppm) can be predicted, and the prediction result is 65,000 kilometers.

[0056] Step S6, based on the prediction model, the running data transmitted by the Internet of Vehicles in real time is used to predict the engine oil performance, and when the prediction result reaches the preset threshold, a maintenance prompt information is output.

[0057] The Internet of Vehicles (loV) refers to an information interaction system that realizes the information interaction between vehicles, infrastructure and cloud platforms through vehicle-mounted networks.

[0058] The prediction result includes the remaining life, and the calculation process of the remaining life includes: inputting the running data transmitted by the Internet of Vehicles in real time into the prediction model, taking the antioxidant index in the label data as the first evaluation dimension to obtain the first remaining life, and taking the iron wear content in the label data as the second evaluation dimension to obtain the second remaining life; comparing the sizes of the first remaining life and the second remaining life, and taking the smaller value of the first remaining life and the second remaining life as the final remaining life.

[0059] On the basis of the above-mentioned embodiments, taking the first ring coking damage of the piston as input, taking TBN as label data for training, the remaining mileage when TBN decreases to the limit value of the national standard (<5.5 mgKOH / g) can be predicted as 40000 kilometers, taking the relative wear of the crankshaft bearing as input, taking the iron content as the label for training, the remaining mileage when the iron content reaches the limit value of the national standard (<150 ppm) can be predicted as 65000 kilometers, and finally the remaining life of the current oil is determined as 40000 kilometers, so as to ensure that the oil is maintained before any index reaches the limit, and the engine is normally operated.

[0060] In the embodiment of the application, a failure mode avoidance method is adopted, that is, a passive predictive maintenance algorithm development is carried out by combining failure mechanism for correlation fusion analysis of system level and vehicle level, the key feature recognition efficiency and accuracy are improved, and predictive maintenance is realized without increasing the hardware cost of the vehicle end. By analyzing the failure history data, the typical working conditions of failure are identified, the physical and functional boundaries of the failure object are defined by boundary analysis, the demarcation between the system and the external environment is clear, and the relationship between the subsystems or components inside the system is clear, and the dependence relationship and interaction between different functions are analyzed from the functional point of view to identify the potential functional failure risk. At the same time, by using the failure mode avoidance method, the functional structure of the engine lubrication system and the interaction relationship between the subsystems are analyzed by using tools such as function tree, system boundary diagram, interface matrix and list, and the core failure mode and key influencing factors of the oil are identified. This failure mechanism analysis based on expert experience clearly defines the core research object for subsequent model construction, avoiding the invalid learning of irrelevant data by the data-driven method. The key feature values are calculated based on the physical mechanism, these physical models convert the abstract failure process into quantifiable feature data, and provide inputs for machine learning that conform to the actual failure mechanism instead of directly using raw sensor data. Professional knowledge and physical models ensure the effectiveness and relevance of the input features, and the data-driven method realizes dynamic prediction of the performance degradation of the oil by learning the mapping relationship between the features and the labels. The final model not only captures the data law, but also conforms to the physical failure logic, integrates expert knowledge and physical models into the machine learning model, and improves the robustness and prediction ability of the oil life prediction model.

[0061] Embodiment 2 The embodiment of the application also provides a vehicle oil performance prediction system, Figure 4 is a structural schematic diagram of a vehicle oil performance prediction system provided by the embodiment of the application, as Figure 4 As shown in the figure, the vehicle oil performance prediction system adopts the vehicle oil performance prediction method in any one of the above-mentioned embodiments, and specifically includes the following modules: The parameter acquisition module 110 is used for acquiring the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system.

[0062] The parameter association module 120 is connected with the parameter acquisition module 110, and is configured to associate the internal component parameters with the external system parameters.

[0063] The characteristic value calculation module 130 is connected with the parameter association module 120, and is configured to determine the characteristic value of the internal component failure mode according to the associated internal component parameters, the characteristic value of the internal component failure mode being a parameter reflecting the oil performance.

[0064] The label data acquisition module 140 is configured to collect an oil sample in the engine lubricating system, and to acquire label data of the internal component failure mode according to the oil sample.

[0065] The prediction model training module 150 is connected with the label data acquisition module 140 and the characteristic value calculation module 130, and is configured to train a prediction model of the oil performance by taking the characteristic value as an input and the label data as an output, wherein the prediction model is configured to predict the oil performance according to real-time transmission of running data of the Internet of Vehicles, and to output a prediction result.

[0066] The early warning module 160 is connected with the prediction model training module 150, and is configured to judge the prediction result, and to output a maintenance prompt information when the prediction result reaches a preset threshold.

[0067] The embodiment establishes the association relationship between the system level and the vehicle level through the failure mode determination module 110, to determine the failure mode and the failure mechanism; the characteristic value calculation module 120 is configured to calculate the characteristic value of the failure mode according to the historical running data of the vehicle collected by the Internet of Vehicles; the label data acquisition module 130 is configured to acquire the label data according to the collected oil sample; the prediction model training module 140 is configured to train the prediction model of the oil performance according to the characteristic value and the label data; and the oil performance is predicted according to the real-time transmission of the running data of the Internet of Vehicles based on the prediction model. By associating the internal component parameters with the external system parameters, the limitations of direct measurement by the sensor and the incompleteness of feature extraction can be compensated, and the prediction result is more reliable. In addition, the low-cost component predictive maintenance model can be quickly constructed, the equivalent relationship between the user driving data and the service life can be quickly mined, the low-cost flexible maintenance model can be developed, and the demand of the user for reducing maintenance cost and improving profit margin can be met, which has important reference value.

[0068] Embodiment 3

[0069] An electronic device, comprising: a processor and a memory; The processor is configured to execute the steps of the vehicle oil performance prediction method according to any one of the embodiments by invoking the program or instruction stored in the memory.

[0070] Example 4

[0071] A computer-readable storage medium comprising computer program instructions to cause a computer to perform the steps of the vehicle oil performance prediction method of any of the preceding examples.

[0072] The computer-readable storage medium can take any combination of one or more of a variety of non-transitory computer-readable media. The computer-readable storage medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0073] Finally, it should be noted that the above-described embodiments are merely intended to illustrate the technical solutions of the present application, and are not intended to limit the present application; even though the present application has been described in detail with reference to the above-described embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the above-described embodiments, or make equivalent replacements to some or all of the technical features thereof; and such modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the technical solutions of the embodiments of the present application.

Claims

1. A method of vehicle oil performance prediction, characterized by, The method comprises the following steps: Step S1, obtaining engine lubrication system internal component parameters and external system parameters interacting with the engine lubrication system; Step S2, associating the internal component parameters with the external system parameters; Step S3, determining the characteristic value of the internal component failure mode according to the associated internal component parameters, wherein the characteristic value of the internal component failure mode refers to the parameter reflecting the oil performance; Step S4, collecting the oil sample in the engine lubrication system, and obtaining the label data of the internal component failure mode according to the oil sample; Step S5, training the prediction model of the oil performance by taking the characteristic value as the input and the label data as the output; Step S6, predicting the oil performance based on the prediction model according to the real-time transmission of the vehicle network operation data, and outputting the maintenance prompt information when the prediction result reaches the preset threshold.

2. The vehicle oil performance prediction method according to claim 1, characterized by, In the step S2, the internal component parameters and the external system parameters are associated by using a failure mode avoidance method, wherein the tools used by the failure avoidance method include system boundary graph and interface matrix and list; The corresponding relationship between the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system is determined by the system boundary graph; The interaction mode of the internal component parameters of the engine lubrication system and the external system parameters interacting with the engine lubrication system is analyzed by using the interface matrix and list.

3. The vehicle oil performance prediction method according to claim 1, characterized by, The characteristic value includes oxidation damage and wear amount; The oxidation damage is calculated by an oil coking damage model, and the wear amount is calculated by a crankshaft bearing wear model.

4. The vehicle oil performance prediction method according to claim 3, characterized by, The oil coking damage model includes an oil delivery rate sub-model, a coking risk rate sub-model, and an oxidation damage amount sub-model, and the calculation steps of the oxidation damage amount include: Obtaining the blow-by rate in the internal component parameters of the engine lubrication system, inputting the blow-by rate into the oil delivery rate sub-model, and calculating the oil delivery rate by the oil delivery rate sub-model according to the blow-by rate; Obtaining the piston top land hot spot temperature in the internal component parameters of the engine lubrication system, inputting the piston top land hot spot temperature into the coking risk rate sub-model, and calculating the coking risk rate by the coking risk rate sub-model according to the piston top land hot spot temperature; Obtaining the engine speed in the internal component parameters of the engine lubrication system, inputting the engine speed, the oil delivery rate, and the coking risk rate into the oxidation damage amount sub-model, and calculating the oxidation damage amount by the oxidation damage amount sub-model according to the engine speed, the oil delivery rate, and the coking risk rate.

5. The vehicle oil performance prediction method according to claim 3, characterized by, The wear amount is calculated by the crankshaft bearing wear model, specifically including: Obtaining the peak combustion pressure and the total normal load of the crankshaft bearing in the internal component parameters of the engine lubrication system; The peak combustion pressure and the total normal load of the crankshaft bearing are input into the crankshaft bearing wear model, and the crankshaft bearing wear model calculates a wear amount according to the peak combustion pressure and the total normal load of the crankshaft bearing.

6. The vehicle oil performance prediction method of claim 1, wherein In the step S4, the label data includes an oxidation resistance index and an iron wear content of the engine oil.

7. The vehicle oil performance prediction method according to claim 6, characterized by, The prediction result includes a remaining life, and a calculation process of the remaining life includes: The running data transmitted in real time by the Internet of Vehicles is input into the prediction model, the oxidation resistance index in the label data is taken as a first evaluation dimension to obtain a first remaining life, and the iron wear content in the label data is taken as a second evaluation dimension to obtain a second remaining life; The first remaining life and the second remaining life are compared, and a smaller one of the first remaining life and the second remaining life is taken as a final remaining life.

8. A vehicle oil performance prediction system characterized by comprising: The prediction system adopts the vehicle engine oil performance prediction method in claims 1-7, and specifically includes the following modules: A parameter acquisition module is configured to acquire internal component parameters of an engine lubrication system and external system parameters interacting with the engine lubrication system; A parameter association module is connected with the parameter acquisition module and configured to associate the internal component parameters with the external system parameters; A characteristic value calculation module is connected with the parameter association module and configured to determine characteristic values of internal component failure modes according to the associated internal component parameters, the characteristic values of the internal component failure modes being parameters reflecting engine oil performance; A label data acquisition module is configured to collect engine oil samples in the engine lubrication system and acquire label data of the internal component failure modes according to the engine oil samples; A prediction model training module is connected with the label data acquisition module and the characteristic value calculation module and configured to take the characteristic values as input quantities and the label data as output quantities to train a prediction model of engine oil performance, wherein the prediction model is configured to predict the engine oil performance according to running data transmitted in real time by the Internet of Vehicles and output a prediction result; An early warning module is connected with the prediction model training module and configured to judge the prediction result and output a maintenance prompt information when the prediction result reaches a preset threshold.

9. An electronic device, comprising: A processor and a memory; The processor is configured to execute steps of the vehicle engine oil performance prediction method in any one of claims 1-7 by calling programs or instructions stored in the memory. 10.A computer readable storage medium, comprising computer program instructions, the computer program instructions causing a computer to execute steps of the vehicle engine oil performance prediction method in any one of claims 1-7.

Images