Automobile reliability improving method, device and equipment based on FMEA data and medium

Abstract

The invention relates to an automobile reliability improving method, device and equipment based on FMEA data and a medium, and belongs to the technical field of automobile reliability design, the method is used for early warning risks in the automobile design stage and predicting faults in the automobile operation stage, and the method comprises the steps that automobile faults are layered based on the obtained FMEA data; a fault layering model is generated, and after semantic information extraction is carried out on the FMEA data, a fault database is established; blocks and attributes of the blocks are determined based on the fault database, and a block attribute library associated with the FMEA data is constructed based on the attributes of the blocks; in the design stage, risk early warning is performed on the block scheme in the design stage based on the block attribute library, and in the operation stage, the fault probability prediction is performed on the block scheme by utilizing the fault prediction model which is completely trained to obtain a fault prediction result, so that the reliability of the automobile is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of automobile fault prediction and reliability design, and in particular to a method, device, equipment and medium for improving automobile reliability based on FMEA data. Background Art

[0002] As a precision electromechanical product produced on a large scale, automobiles have increasingly complex internal structures, a growing number of components, and intricate inter-system coupling, posing significant challenges to ensuring vehicle reliability. This is particularly true given the rapid development of new technologies such as intelligent connectivity and autonomous driving, which are driving the continuous evolution of automotive electronic and electrical architectures and increasing hardware and software integration. These new functional safety requirements and usage scenarios are severely challenging traditional reliability design methods.

[0003] Currently, the automotive industry generally uses classic tools such as FMEA to conduct reliability analysis. Engineers rely on their experience to list the possible failure modes of each component or subsystem in the product, analyze the causes and impacts of the failures, and take corresponding preventive and detection measures.

[0004] Relying on manual experience is difficult to cope with the growing complexity and has limitations such as incomplete analysis, inaccurate risk assessment, and lack of quantitative basis. With the advancement of the digital and intelligent transformation of the automotive industry, a large amount of structured FMEA data and vehicle operation data are recorded and accumulated. How to fully tap the value contained in them to guide reliability design has become a problem that needs to be solved urgently. Summary of the Invention

[0005] In view of this, it is necessary to provide a method, device, electronic device and storage medium for improving automobile reliability based on FMEA data to solve the technical problems of incomplete analysis, inaccurate risk assessment and lack of quantitative basis when conducting reliability analysis on automobiles. To solve the above problems, the present invention provides a method for improving automobile reliability based on FMEA data, which is used to provide early warning of risks in the automobile design stage and predict failures in the automobile operation stage, including: Based on the acquired FMEA data, automobile faults are hierarchized to generate a fault hierarchical model, and after semantic information is extracted from the FMEA data, a fault database is established; Determining blocks associated with the fault hierarchical model and attributes of the blocks based on the fault database, and constructing a block attribute library associated with the FMEA data based on the attributes of the blocks; During the design phase, risk warning is performed on the block scheme of the design phase based on the block attribute library; During the operation phase, a well-trained fault prediction model is used to predict the fault probability of the block solution to obtain a fault prediction result.

[0006] In one possible implementation, stratifying automobile faults based on the acquired FMEA data to generate a fault stratification model includes: Determine the scope of the fault impact based on the FMEA data; The vehicle faults are layered according to the fault impact range to generate a fault layering model, wherein the fault layering model includes a component layer, a block layer, a functional system layer, a vehicle system layer, and a whole vehicle layer. The block is composed of multiple closely related components, the functional system is composed of multiple blocks, the vehicle system is composed of multiple functional systems, and the whole vehicle is composed of multiple vehicle systems. The relationship between the block layer and the component layer, the block layer, the functional system layer, the automobile system layer and the entire vehicle layer is determined based on the fault layer model.

[0007] In one possible implementation, the attributes of the block include structural parameters, material attributes, operating parameters, reliability indicators, functional parameters and maintenance parameters, wherein the structural parameters include at least size, weight, and shape; the material attributes include at least material type, strength, and hardness; the operating parameters include at least operating temperature, pressure, and speed; the reliability indicators include at least mean time between failures and failure rate; the functional parameters include at least braking force and reaction time; and the maintenance parameters include at least maintenance cycle and spare parts information.

[0008] In a possible implementation, during the design phase, performing risk warning on the block solution in the design phase based on the block attribute library includes: Acquire a block scheme in a design phase, and determine block attributes of the block scheme and historical schemes similar to the block scheme based on the block attribute library; Obtaining the chunk attributes of the historical scheme, matching the chunk attributes of the chunk scheme with the chunk attributes of the historical scheme, and calculating the similarity of the chunks; A risk warning is performed on the block solution based on the similarity.

[0009] In a possible implementation, the similarity of the blocks is calculated as follows: , in, is the attribute vector of the chunks in the chunking scheme, is the attribute vector of the block of the historical scheme, is the inner product operation of vectors, is the attribute vector The module length, is the attribute vector The mold length.

[0010] In a possible implementation, during the operation phase, using a well-trained fault prediction model to perform fault probability prediction on the block solution to obtain a fault prediction result includes: Get historical fault data, Constructing a fault prediction model, training the fault prediction model using the block attributes in the block attribute library and historical fault data as training samples to obtain a fully trained fault prediction model; Based on the fully trained fault prediction model, a fault probability prediction is performed on the block scheme to obtain a fault prediction result, wherein the fault prediction model includes an input layer, a hidden layer, and an output layer, the input layer corresponds to the attributes of the block, the hidden layer includes multiple neurons, and the output layer corresponds to the fault probability.

[0011] In a possible implementation, during the operation phase, using a well-trained fault prediction model to perform fault probability prediction on the block solution to obtain a fault prediction result further includes: Acquire real fault data during the vehicle operation phase, compare the fault prediction result with the real fault data, and determine the prediction deviation of the fault prediction model; After analyzing the prediction deviation, the fault prediction model is improved by supplementing training samples, optimizing feature selection, and adjusting algorithm parameters, and the FMEA data is updated synchronously to improve the reliability of the vehicle.

[0012] On the other hand, the present invention also provides an automobile reliability improvement device based on FMEA data, which is used to provide early warning of risks in the automobile design stage and predict failures in the automobile operation stage, including: A fault stratification module is used to stratify automobile faults based on the acquired FMEA data, generate a fault stratification model, and establish a fault database after extracting semantic information from the FMEA data; a block attribute management module, configured to determine, based on the fault database, blocks associated with the fault hierarchical model and attributes of the blocks, and to construct, based on the attributes of the blocks, a block attribute library associated with the FMEA data; A risk warning module is used to issue a risk warning to the block scheme in the design phase based on the block attribute library during the design phase; The fault prediction module is used to predict the fault probability of the block solution using a well-trained fault prediction model during the operation phase to obtain a fault prediction result.

[0013] On the other hand, the present invention also provides an electronic device, comprising: a processor and a memory; The memory stores a computer-readable program executable by the processor; When the processor executes the computer-readable program, the steps of the above-mentioned method for improving automobile reliability based on FMEA data are implemented.

[0014] On the other hand, the present invention also provides a computer-readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in the automobile reliability improvement method based on FMEA data as described above.

[0015] The beneficial effects of the present invention are as follows: based on the acquired FMEA data, automobile faults are layered, a fault layering model is generated, and after semantic information is extracted from the FMEA data, a fault database is established. By establishing a structured fault database covering all levels, a systematic analysis of the automobile from the local to the whole, from the surface to the inside is achieved. A block attribute library associated with the FMEA data is constructed through the attributes of the blocks, and the representation of the blocks is standardized in an attributed form, so that the blocks are closely associated with the physical design parameters, thereby improving the comprehensiveness of the analysis. In the design stage, risk warnings are issued for the block solutions in the design stage based on the block attribute library, and risk warnings and experience reuse are achieved through attribute matching. In the operation stage, a well-trained fault prediction model is used to predict the failure probability of the block solution, thereby overcoming the subjectivity of manual experience judgment, improving the accuracy of risk assessment, making the analysis process more quantitative and objective, providing an objective basis for reliability target formulation and solution optimization, and comprehensively improving the reliability of automobile products. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flow chart of an embodiment of the method for improving automobile reliability based on FMEA data provided by the present invention; Figure 2 This is a structural schematic diagram of an embodiment of the vehicle reliability improvement device based on FMEA data provided by the present invention; Figure 3 This is a schematic structural diagram of an embodiment of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0017] The preferred embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, and are not used to limit the scope of the present invention.

[0018] The present invention discloses a method, apparatus, device, and medium for improving automobile reliability based on FMEA data, which can be used in a computer. The method, device, or computer-readable storage medium involved in the present invention can be integrated with the above-mentioned device or can be relatively independent.

[0019] A specific embodiment of the present invention discloses a method for improving automobile reliability based on FMEA data, which can be executed by a computer, specifically by one or more processors of the computer. Figure 1 This is a flowchart of the automobile reliability improvement method based on FMEA data provided by an embodiment of the present invention. Figure 1 , an automobile reliability improvement method based on FMEA data, used to provide early warning of risks in the automobile design phase and predict failures in the automobile operation phase, including: S101. Based on the acquired FMEA data, automobile faults are stratified to generate a fault stratification model. After semantic information is extracted from the FMEA data, a fault database is established. S102, determining blocks associated with the fault hierarchical model and attributes of the blocks based on the fault database, and constructing a block attribute library associated with the FMEA data based on the attributes of the blocks; S103. During the design phase, risk warning is performed on the block scheme in the design phase based on the block attribute library; S104. During the operation phase, the fault probability of the block solution is predicted using the well-trained fault prediction model to obtain a fault prediction result. Among them, FMEA (Failure Mode and Effect Analysis) is an important reliability design method. By evaluating and analyzing various possible risks of automobiles, the core is to identify potential failure modes and their potential causes, and evaluate their impact on automobile products or systems so as to take corresponding preventive measures. Through FMEA, automobile reliability analysis is carried out to obtain FMEA data; the fault layering model includes component layer, block layer, functional system layer, automobile system layer, and complete vehicle layer.

[0020] Compared to existing technologies, the FMEA data-based automobile reliability improvement method provided in this embodiment stratifies automobile faults based on acquired FMEA data, generates a fault stratification model, and establishes a fault database after extracting semantic information from the FMEA data. By establishing a structured fault database covering all levels, a systematic analysis of the vehicle from the local to the entire vehicle, from the surface to the inside, is achieved. Based on the fault database, the blocks associated with the fault stratification model and their attributes are determined. Based on the block attributes, a block attribute library associated with the FMEA data is constructed. Block representation is standardized using an attributed form, closely linking the blocks to physical design parameters and improving the comprehensiveness of the analysis. During the design phase, risk warnings are issued for block solutions based on the block attribute library. During the operation phase, failure probabilities are predicted for these block solutions using a fully trained fault prediction model to obtain fault prediction results. Attribute matching enables risk warnings and experience reuse, overcoming the subjectivity of manual experience judgment, improving the accuracy of risk assessment, and making the analysis process more quantitative and objective. This provides an objective basis for reliability target setting and solution optimization, thereby comprehensively improving the reliability of automobile products.

[0021] In some embodiments, in step S101, FMEA data is obtained, and after analyzing the FMEA data, the impact range of the automobile failure is determined, and the automobile failure is divided into five levels in order from small to large according to the failure impact range, and a fault hierarchical model is generated. The fault hierarchical model includes a component layer, a block layer, a functional system layer, a vehicle system layer and a whole vehicle layer. The component layer includes component failures, the block layer includes block failures, the functional system layer includes functional system failures, the vehicle system layer includes vehicle system failures, and the whole vehicle layer includes whole vehicle failures. The block is composed of multiple closely related components, the functional system is composed of multiple blocks, the vehicle system is composed of multiple functional systems, and the whole vehicle is composed of multiple vehicle systems. The component layer: located at the bottom layer, including basic components such as bolts, bearings, and sealing rings, the block layer: such as larger functional units such as cylinders, pistons, and clutches, the functional system layer: such as engine systems, gearbox systems, suspension systems, etc., the vehicle system layer: including power systems, transmission systems The vehicle layer is located at the top layer and represents the entire vehicle system. Block failure refers to the failure of a module with independent sub-functions composed of multiple closely related components. Functional system failure refers to the failure of a system composed of multiple blocks that work together to achieve a major function. Blocks are the key abstraction between components and systems, representing modules with independent sub-functions composed of several closely related parts. The relationship between the block layer and the component layer, block layer, functional system layer, vehicle system layer and vehicle layer is determined based on the fault hierarchy model. That is, each layer of the fault hierarchy model has a vertical relationship of inclusion and being included, support and dependence. Components are the most basic physical units, and their failure rate directly determines the reliability of the upper-level blocks and systems. Blocks are the main carriers of function realization, supporting the needs of functional systems upward and forcing the selection of key components downward. Each functional system may involve multiple blocks, and the realization of the macro functions of the whole vehicle requires the coordination of multiple systems.

[0022] After the fault hierarchical model is generated, the semantic information in the FMEA data is extracted to establish a structured fault database covering each level in the fault hierarchical model. The fault database adopts a graph database structure, with nodes representing entities at each level and edges representing the relationships between entities. The fault database is formally represented as follows: , , , in, It is an entity, representing parts, blocks, functional systems, automotive systems, and complete vehicles. The fault database stores information such as the name, code, and function description of five levels, namely, the whole vehicle, vehicle system, functional system, module, and component, in the form of a node tree, as well as the subordinate relationships between adjacent levels, providing a unified framework for subsequent analysis.

[0023] In some embodiments, in step S102, based on the fault database, the blocks and the attributes of the blocks associated with the fault hierarchical model are determined, and based on the attributes of the blocks, a block attribute library associated with the FMEA data is constructed, a group of common basic blocks are predefined in the fault database, and the inherent attributes of the blocks such as key materials, structures, and specification parameters are summarized. After associating the attributes of the blocks with the FMEA data, a block attribute library is constructed. By associating the attributes of the blocks with the FMEA data, the most common fault models under different designs can be summarized, and the attributes of the blocks include structural parameters, material attributes, operating parameters, reliability indicators, functional parameters, and maintenance parameters. The structural parameters include at least size, weight, and shape, and the material attributes include at least Material type, strength, and hardness; operating parameters include at least operating temperature, pressure, and speed; reliability indicators include at least mean time between failures and failure rate; functional parameters include at least braking force and reaction time; maintenance parameters include at least maintenance cycle and spare parts information. The block attribute library is a relational database that stores the block or component attribute parameters accumulated from previous FMEAs, including four categories: structural parameters, material properties, operating parameters, and reliability indicators. When designing the data table of the block attribute library, full consideration is given to consistency with the company's existing PLM system to reduce duplicate entry; the block attribute library introduces data version management, recording the timestamp, person, and reason for each parameter change to ensure traceability; the block attribute library is represented by a multidimensional vector space: ,in, is the attribute vector of the block, For the The value of each attribute of a block is a continuous value (such as size, weight), a discrete value (such as material type), or a Boolean value (such as whether it has a certain function).

[0024] In some embodiments, in step S103, during the design phase, a risk warning is performed on the block scheme of the design phase based on the block attribute library. The block scheme of the design phase is obtained, and the attributes of the blocks in the block scheme are identified through the block attribute library. Historical schemes similar to the block scheme and the attributes of the blocks in the historical schemes are determined. The block attributes of the block scheme and the block attributes of the historical schemes are matched, and the similarity of the blocks is calculated. The calculation formula for the similarity of the blocks is: , in, is the attribute vector of the chunks in the chunking scheme, is the attribute vector of the block of the historical scheme, is the inner product operation of vectors, is the attribute vector The module length, is the attribute vector Length of the module; Risk warning is issued for block solutions based on similarity. When a historical solution with a high similarity is a high-risk solution, the designer is prompted that "there may be risks in citing this solution."

[0025] The process of using the block attribute library to provide risk warning for block schemes in the design phase is as follows: obtain the attributes of the newly designed blocks in the design phase, perform attribute vector preprocessing on the attributes of the newly designed blocks, match them with the block attributes of historical schemes in the block attribute library, calculate the similarity of the blocks, query the failure probability of the newly designed blocks in the block attribute library based on the similarity, and perform risk assessment on the newly designed blocks. When the newly designed block scheme is a high-risk scheme, a risk warning is issued.

[0026] In some embodiments, in step S104, during the operation phase, a well-trained fault prediction model is used to predict the failure probability of the block solution, obtain a fault prediction result, give a quantitative fault probability estimate, and recommend alternative solutions. Designers can refer to the early warning information and expert advice to optimize the solution or take targeted preventive measures. The specific process of fault prediction for the block solution is: obtain historical fault data, build a fault prediction model, and the fault prediction model mainly includes a classification model and a regression model. The classification model takes the block attribute as input and the failure mode as output to predict the most likely failure type. The regression model takes the block attribute as an independent variable and the failure rate as a dependent variable to quantify the failure risk. The block attributes and historical fault data in the block attribute library are taken as training samples to train the fault prediction model. A global optimization mechanism is introduced to improve the training efficiency. The cross-validation method is used to evaluate the performance of the fault prediction model to obtain a fully trained fault prediction model. Based on the fully trained fault prediction model, the fault probability of the block solution is predicted to obtain the fault prediction result. The fault prediction model includes an input layer, a hidden layer, and an output layer. The input layer corresponds to the various attributes of the block, and each node represents an attribute feature. The hidden layer contains multiple neurons for feature extraction and nonlinear transformation. There may be multiple hidden layers, representing a deep learning structure. The output layer outputs the fault probability. Taking a neural network as an example, its structure can be expressed as: , in, is the input block attribute vector, is the output failure probability, 、 are weights, 、 are the bias vectors, is the activation function.

[0027] In some embodiments, real fault data from the automobile operation stage is obtained, that is, during the mass production and market operation stage of automobile products, the Internet of Vehicles platform, after-sales system, quality monitoring system, etc. are connected to collect real fault data and working condition data through multiple channels. For heterogeneous data with different degrees of structure, cleaning and fusion processes are designed to improve data quality. The fault prediction results are compared with the real fault data to determine the prediction deviation of the fault prediction model. By comparing the prediction results of each stage in the automobile product development process with the final mass production and operation conditions, the deviation items are found, the root causes of the deviations are traced back, a targeted reflection and improvement list is formed, and the prediction deviations are analyzed. That is, big data from the entire life cycle provides a new perspective for in-depth understanding of the mechanism of failure and tracing design defects. Traditional FMEA is limited to the design stage and it is difficult to accurately predict failures under mass production conditions. Big data truly records the status of the product from production to use. The root cause of the failure can be traced from the bottom up through data correlation analysis. For example, for a certain failure reported by a user, it can be associated with the after-sales maintenance record to understand the fault location and severity. The system further associates manufacturing data to discover potential process defects, ultimately tracing them back to component design and identifying weak links that lead to failures. For inaccurate predictions, the fault prediction model is improved by supplementing training samples, optimizing feature selection, and adjusting algorithm parameters, and the FMEA data is updated simultaneously. Regular FMEA data checks are conducted to eliminate outdated knowledge, update ineffective countermeasures, and extract new knowledge points to keep the FMEA data advanced and timely. By updating the FMEA data, the block attribute library is updated, and the reliability experts' experience judgment rules are used to assist engineers in making key decisions such as scheme comparison, risk assessment, and corrective measures. By matching the best solution in the block attribute library, effective fault response solutions are intelligently recommended, shortening the decision-making cycle, and forming a data-driven double closed loop of fault prediction and experience extraction to ensure that the analysis conclusions are highly consistent with the front-line reality. Through data-driven closed-loop learning, the prediction effect is continuously improved, and the reliability of the vehicle is improved. Online learning algorithms, such as stochastic gradient descent, are used to improve the fault prediction model: , in, for The model parameters at time t, is the learning rate, is the loss function.

[0028] In order to use intelligent reliability analysis in daily engineering practice, a software platform is designed to support the entire process of reliability design. The introduction of the software platform also makes it possible to standardize the collection and full utilization of reliability big data. The software structure includes a data management module, which is responsible for data import, storage and management; a model training module, including data preprocessing, feature engineering and model training functions; a fault prediction module, which realizes the fault prediction function of real-time data; a risk assessment module, which performs risk assessment and early warning based on the prediction results; a visualization module, which provides a visual display of various data and results; a user interface module, which provides a user interaction interface; a system management module, which is responsible for management functions such as user permissions and system configuration. The use of the software platform allows digital and structured FMEA results to play a leading role in every link. The support of the software platform frees engineers from complicated information search and data collection, and focuses their main energy on defect analysis and optimization design, promoting the emergence of more innovative results.

[0029] In order to better implement the automobile reliability improvement method based on FMEA data in the embodiment of the present invention, based on the automobile reliability improvement method based on FMEA data, correspondingly, Figure 2 As shown, an embodiment of the present invention further provides an automobile reliability improvement device based on FMEA data, which is used to provide early warning of risks in the automobile design stage and predict failures in the automobile operation stage. The automobile reliability improvement system 200 based on FMEA data includes: The fault stratification module 201 is used to stratify automobile faults based on the acquired FMEA data, generate a fault stratification model, and establish a fault database after extracting semantic information from the FMEA data; A block attribute management module 202 is configured to determine blocks associated with the fault hierarchical model and attributes of the blocks based on the fault database, and to construct a block attribute library associated with the FMEA data based on the attributes of the blocks; The risk warning module 203 is used to provide risk warning for the block scheme in the design phase based on the block attribute library during the design phase; The fault prediction module 204 is used to perform fault probability prediction on the block solution using a well-trained fault prediction model during the operation phase to obtain a fault prediction result.

[0030] like Figure 3 As shown, the present invention also provides an electronic device 300 based on the FMEA data to improve automobile reliability. The electronic device 300 can be a computing device such as a mobile terminal, a desktop computer, a notebook, a PDA, or a server. The electronic device 300 includes a processor 301, a memory 302, and a display 303. Figure 3Only some of the components of the electronic device 300 are shown, but it should be understood that implementation of all of the shown components is not required, and more or fewer components may be implemented instead.

[0031] In some embodiments, the memory 302 may be an internal storage unit of the electronic device 300, such as the hard drive or memory of the electronic device 300. In other embodiments, the memory 302 may also be an external storage device of the electronic device 300, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. equipped on the electronic device 300. Furthermore, the memory 302 may include both an internal storage unit of the electronic device 300 and an external storage device. The memory 302 is used to store application software installed on the electronic device 300 and various data, such as program code installed on the electronic device 300. The memory 302 may also be used to temporarily store data that has been output or is about to be output. In one embodiment, the memory 302 stores an automobile reliability improvement program based on FMEA data. This automobile reliability improvement program based on FMEA data can be executed by the processor 301, thereby implementing the automobile reliability improvement method based on FMEA data according to various embodiments of the present invention.

[0032] In some embodiments, the processor 301 may be a central processing unit (CPU), a microprocessor, or other data processing chip, configured to execute program codes or process data stored in the memory 302 , such as a method for improving automobile reliability based on FMEA data.

[0033] In some embodiments, display 303 can be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. Display 303 is used to display identification information for the vehicle reliability improvement program based on FMEA data and to display a visual user interface. Components 301-303 of electronic device 300 communicate with each other via a system bus.

[0034] In some embodiments, when the processor 301 executes the automobile reliability improvement program based on FMEA data in the memory 302, the various steps of the automobile reliability improvement method based on FMEA data as described in the above embodiments are implemented. Since the automobile reliability improvement method based on FMEA data has been described in detail above, it will not be repeated here.

[0035] Accordingly, an embodiment of the present application also provides a computer-readable storage medium, which is used to store computer-readable programs or instructions. When the program or instructions are executed by a processor, the steps or functions of the automobile reliability improvement method based on FMEA data provided in the above-mentioned method embodiments can be implemented.

[0036] In summary, the automobile reliability improvement method, device, equipment and medium based on FMEA data provided by the present invention are used to warn of risks in the automobile design stage and predict faults in the automobile operation stage. Automobile faults are stratified based on the acquired FMEA data to generate a fault stratification model, and a fault database is established after semantic information is extracted from the FMEA data. Blocks and their attributes are determined based on the fault database, and a block attribute library associated with the FMEA data is constructed based on the attributes of the blocks. In the design stage, risk warnings are issued for the block schemes in the design stage based on the block attribute library. In the operation stage, the failure probability of the block schemes is predicted using a well-trained fault prediction model to obtain fault prediction results, thereby improving the reliability of the automobile.

[0037] Those skilled in the art will appreciate that all or part of the process steps of the above-described embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, such as a magnetic disk, an optical disk, a read-only memory, or a random access memory.

[0038] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed by the present invention should be covered by the scope of protection of the present invention.

Claims

1. A method for improving automobile reliability based on FMEA data, which is used to warn of risks in the automobile design stage and predict failures in the automobile operation stage, characterized in that: include: Based on the acquired FMEA data, automobile faults are hierarchized to generate a fault hierarchical model, and after semantic information is extracted from the FMEA data, a fault database is established; Determine the blocks associated with the fault hierarchical model and the attributes of the blocks based on the fault database, and construct a block attribute library associated with the FMEA data based on the attributes of the blocks; In the design phase, risk warning is performed on the block scheme in the design phase based on the block attribute library; During the operation phase, a well-trained fault prediction model is used to predict the fault probability of the block solution to obtain a fault prediction result.

2. The automobile reliability improvement method based on FMEA data according to claim 1 is characterized in that: The method of stratifying automobile faults based on the acquired FMEA data and generating a fault stratification model includes: Determine the impact scope of the failure based on the FMEA data; The automobile faults are layered according to the fault impact range to generate a fault layering model, wherein the fault layering model includes a component layer, a block layer, a functional system layer, an automobile system layer and a whole vehicle layer, wherein the block is composed of a plurality of closely related components, the functional system is composed of a plurality of blocks, the automobile system is composed of a plurality of functional systems, and the whole vehicle is composed of a plurality of automobile systems; The relationship between the block layer and the component layer, the block layer, the functional system layer, the automobile system layer and the whole vehicle layer is determined based on the fault layer model.

3. The automobile reliability improvement method based on FMEA data according to claim 1 is characterized in that: The attributes of the block include structural parameters, material attributes, operating parameters, reliability indicators, functional parameters and maintenance parameters, wherein the structural parameters include at least size, weight and shape; the material attributes include at least material type, strength and hardness; the operating parameters include at least operating temperature, pressure and speed; the reliability indicators include at least mean time between failures and failure rate; the functional parameters include at least braking force and reaction time; and the maintenance parameters include at least maintenance cycle and spare parts information.

4. The automobile reliability improvement method based on FMEA data according to claim 3 is characterized in that: In the design phase, risk warning is performed on the block scheme in the design phase based on the block attribute library, including: Acquire a block scheme in a design phase, and determine block attributes of the block scheme and historical schemes similar to the block scheme based on the block attribute library; Acquire the block attributes of the historical scheme, match the block attributes of the block scheme with the block attributes of the historical scheme, and calculate the similarity of the blocks; Based on the similarity, a risk warning is performed on the block solution.

5. The automobile reliability improvement method based on FMEA data according to claim 4 is characterized in that: The calculation formula of the similarity of the blocks is: ， in, is the attribute vector of the chunks in the chunking scheme, is the attribute vector of the block of the historical scheme, is the inner product operation of vectors, is the attribute vector The module length, is the attribute vector The mold length.

6. The automobile reliability improvement method based on FMEA data according to claim 5 is characterized in that: In the operation phase, the fault probability prediction of the block solution is performed using a well-trained fault prediction model to obtain a fault prediction result, including: Get historical fault data, Constructing a fault prediction model, training the fault prediction model using the block attributes in the block attribute library and historical fault data as training samples to obtain a fully trained fault prediction model; Based on the fully trained fault prediction model, the fault probability of the block scheme is predicted to obtain a fault prediction result, wherein the fault prediction model includes an input layer, a hidden layer, and an output layer, the input layer corresponds to the attribute of the block, the hidden layer includes multiple neurons, and the output layer corresponds to the fault probability.

7. The automobile reliability improvement method based on FMEA data according to claim 6 is characterized in that: In the operation phase, the fault probability prediction of the block solution is performed using a well-trained fault prediction model to obtain a fault prediction result, and further includes: Acquire real fault data during the operation phase of the vehicle, compare the fault prediction result with the real fault data, and determine the prediction deviation of the fault prediction model; After analyzing the prediction deviation, the fault prediction model is improved by supplementing training samples, optimizing feature selection, and adjusting algorithm parameters, and the FMEA data is updated synchronously to improve the reliability of the vehicle.

8. An automobile reliability improvement device based on FMEA data, used to warn of risks in the automobile design stage and predict failures in the automobile operation stage, characterized in that: include: A fault stratification module is used to stratify automobile faults based on the acquired FMEA data, generate a fault stratification model, and establish a fault database after extracting semantic information from the FMEA data; A block attribute management module, used to determine the blocks associated with the fault hierarchical model and the attributes of the blocks based on the fault database, and to construct a block attribute library associated with the FMEA data based on the attributes of the blocks; A risk warning module is used to provide risk warning for the block scheme in the design phase based on the block attribute library during the design phase; The fault prediction module is used to predict the fault probability of the block scheme using a well-trained fault prediction model during the operation phase to obtain a fault prediction result.

9. An electronic device, characterized in that: including memory and processor; The memory stores a computer-readable program executable by the processor; When the processor executes the computer-readable program, the steps of the method for improving automobile reliability based on FMEA data as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in the automobile reliability improvement method based on FMEA data as described in any one of claims 1-7.

Images