A method for analyzing real-time data on a vehicle
By using streaming computing and memory processing technologies, efficient analysis of real-time vehicle data is achieved, solving the problems of vehicle data analysis latency and insufficient fault identification capabilities, and enabling immediate early warning and high-precision diagnosis.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING AEROSPACE MEASUREMENT & CONTROL TECH
- Filing Date
- 2025-12-12
- Publication Date
- 2026-06-26
AI Technical Summary
Existing vehicle data analysis methods suffer from high latency, insufficient fault identification capabilities, and the inability of complex algorithms to run in real time, resulting in the inability to achieve real-time early warning and the failure to fully explore the value of in-depth data.
Employing streaming computing and in-memory processing technologies, this system achieves millisecond-level real-time analysis and online diagnosis through multi-source heterogeneous data streaming access and standardization, real-time feature extraction, lightweight parallel fault diagnosis model inference, multi-dimensional correlation analysis and root cause inference, real-time decision-making and hierarchical early warning.
It enables real-time early warning and handling of vehicle faults, improves fault diagnosis accuracy, identifies complex faults and locates the root cause, reduces false alarms and false negatives, and solves the problem of limited on-board computing resources.
Smart Images
Figure CN122286277A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data analysis technology, and in particular to a method for analyzing real-time vehicle data. Background Technology
[0002] Current vehicle systems are equipped with numerous sensors, control units, and intelligent devices. These devices generate massive amounts of real-time data during operation, directly reflecting the vehicle's health and operational safety. However, current data analysis still relies on offline methods, resulting in high data processing latency and hindering real-time fault warnings and rapid responses. Diagnostic methods also largely depend on threshold judgments, lacking the ability to identify complex or correlated faults, leading to false alarms and missed alarms. Furthermore, the deeper value of this massive amount of data has not been fully explored, and the lack of analysis on data correlations and trends leaves predictive maintenance and operational optimization without effective data support. Simultaneously, limitations in onboard computing resources and insufficient storage and computing power prevent complex analytical algorithms from running in real-time on vehicles.
[0003] Therefore, there is an urgent need for an analysis method that can adapt to the vehicle environment and has high-speed, real-time, and high-precision diagnostic capabilities. Summary of the Invention
[0004] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide a method for analyzing real-time vehicle data. This invention achieves millisecond-level real-time analysis and online diagnosis through streaming computing and memory processing technology, enabling immediate early warning and handling of faults during operation.
[0005] To achieve the above and other related objectives, the present invention provides a method for analyzing real-time vehicle data, comprising: S1. Multi-source heterogeneous data streaming access and standardization: Through the vehicle data acquisition unit, the real-time data streams of various vehicle subsystems are accessed, and the different protocols and formats of the real-time data streams are parsed and standardized, and the data is uniformly encapsulated into standard data objects. S2. Real-time feature extraction: Set a time memory buffer for each key monitoring parameter in the vehicle, and calculate the real-time feature vector within the buffer; S3. Lightweight Parallel Fault Diagnosis Model Inference: The real-time feature vector is input into the pre-trained lightweight diagnostic model for parallel inference and outputs diagnostic information.
[0006] S4. Multi-dimensional correlation analysis and root cause inference: Based on the diagnostic information, start the correlation analysis engine; S5. Real-time decision-making and tiered early warning: Based on the diagnostic information, generate early warning information of different levels; S6. Diagnostic Results and Data Feedback: The diagnostic information, early warning information, key feature data in the feature vector, and raw data in the real-time data stream are packaged and transmitted to the ground data center in real time through the communication network.
[0007] In one embodiment of the present invention, the standard data object includes a timestamp, a data source ID, and a data value.
[0008] In one embodiment of the present invention, step S2, which involves setting a time memory buffer for each key monitoring parameter in the vehicle, and calculating the real-time feature vector within the buffer, includes: S21. Determine monitoring parameters: Select the key vehicle parameters to be monitored and set a buffer in memory; S22. Calculate based on the set trigger each time a new data point arrives; S23. After receiving the real-time data stream, the computational processing unit performs unit computation according to the stream processing framework. S24. After receiving the real-time data stream according to the aforementioned stream processing framework, perform calculations using the mean method, variance method, standard deviation method, peak value method, linear fitting slope method, and frequency domain feature method. S25. After all the computing processing units have completed their calculations, they assemble all the features into the real-time feature vector.
[0009] In one embodiment of the present invention, in step S24: The formula for the mean method is: , The formula for the variance method is: , The formula for the standard deviation method is: , The peak method formula is: , The formula for the linear fitting slope method is: , Where n is the number of data points. It is the sum of the products of x and y. and It is the sum of x and y. It is the sum of x squared; The formula for the frequency domain feature method is: , Where N is the total length of the original signal, and k is the frequency index, with a value range of 0, 1, ..., N / 2. 10,1,...,N / 2 1. X(k) is the k-th frequency component of the N-point Fourier transform result, G(k) is the N / 2-point Fourier transform result calculated from all even-numbered points in the original signal, and H(k) is the N / 2-point Fourier transform result calculated from all odd-numbered points in the original signal. The rotation factor is a complex number.
[0010] In one embodiment of the present invention, step S3, which involves inputting the real-time feature vector into a pre-trained lightweight diagnostic model for parallel inference and outputting diagnostic information, includes: S31. Model preparation and loading: Deploy at least one pre-trained lightweight diagnostic model in advance; S32, Feature Vector Input: The real-time feature vector is synchronously input into the pre-trained lightweight diagnostic model; S33, Parallel Model Inference: Calculate and analyze the real-time feature vector of the input; S34. Multi-dimensional result output: Output diagnostic information, including fault classification, health score, and anomaly probability.
[0011] In one embodiment of the present invention, the trained lightweight diagnostic model includes model pruning and model quantization.
[0012] In one embodiment of the present invention, the model pruning includes: each subsystem of the vehicle.
[0013] In one embodiment of the present invention, the fault classification includes: control failure and sensor failure; the health score includes the current health status score of the vehicle, between 0 and 1; and the anomaly probability includes the confidence level of the abnormal state.
[0014] The present invention also provides a system for analyzing real-time vehicle data, comprising: The multi-source heterogeneous data streaming access and standardization module is used to access the real-time data streams of various vehicle subsystems through the vehicle data acquisition unit, parse and standardize the data with different protocols and formats in the real-time data streams, and encapsulate the data into a standard data object. The real-time feature extraction module is used to set a time memory buffer for each key monitoring parameter in the vehicle, and calculate the real-time feature vector within the buffer. The lightweight parallel fault diagnosis model inference module is used to input the real-time feature vector into the pre-trained lightweight diagnosis model, perform parallel inference, and output diagnostic information.
[0015] The multi-dimensional correlation analysis and root cause inference module is used to activate the correlation analysis engine based on the diagnostic information; The real-time decision-making and hierarchical early warning module is used to generate early warning information of different levels based on the diagnostic information; The diagnostic results and data feedback module is used to package the diagnostic information, early warning information, key feature data in the feature vector, and raw data in the real-time data stream, and transmit them to the ground data center in real time through the communication network.
[0016] The present invention also provides an electronic device, including a processor and a memory, wherein the memory stores program instructions, and the processor executes the program instructions to implement the above-described method for analyzing real-time vehicle data.
[0017] As described above, the method for analyzing real-time vehicle data according to the present invention has the following beneficial effects: (1) The present invention provides a method for analyzing vehicle-mounted real-time data to achieve real-time diagnosis and early warning: by using streaming computing and online analysis, the traditional delayed post-processing mode is changed, and the faults in operation can be intervened in real time.
[0018] (2) The present invention provides a method for analyzing real-time vehicle data to improve the accuracy of fault diagnosis: by combining lightweight model and knowledge graph association analysis, complex faults are identified, the root cause of the fault is located, and the false alarm and false alarm rates of the traditional threshold method are reduced.
[0019] (3) The present invention provides a method for analyzing real-time data in vehicles, which solves the problem of applying complex algorithms under limited vehicle resources. It adopts optimization techniques such as pruning and quantization to enable advanced intelligent algorithms to run efficiently and stably on resource-constrained vehicle platforms, thereby realizing vehicle edge intelligence. Attached Figure Description
[0020] Figure 1 A flowchart illustrating a method for analyzing real-time vehicle data provided in this application embodiment; Figure 2 A flowchart of step S2 of a method for analyzing real-time vehicle data provided in this application embodiment; Figure 3 This is a flowchart of step S3 of a method for analyzing real-time vehicle data provided in an embodiment of this application. Detailed Implementation
[0021] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.
[0022] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the illustrations only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.
[0023] Terms such as "first" or "second" may be used to describe various components, but these components are not limited by the terms described above. The terms described above are used to distinguish one component from another; for example, without departing from the scope of the concept according to this disclosure, a first component may be referred to as a second component, and similarly, a second component may be referred to as a first component.
[0024] Furthermore, "connected / linked" indicates that one component is directly electrically connected to another component or indirectly electrically connected through another component. Unless otherwise explicitly stated in the sentence, the singular form may include the plural form. Additionally, the terms "comprising / including" or "containing / including" as used in this specification indicate the presence or addition of one or more components, steps, operations, and elements. Specific structural or functional descriptions of examples of embodiments of the concepts disclosed in this specification are merely illustrative to describe examples of embodiments of the concepts, and examples of embodiments of the concepts can be implemented in various forms, but these descriptions are not limited to the examples of embodiments described in this specification.
[0025] Based on the concept, various modifications and changes can be applied to examples of embodiments, such that examples of embodiments will be illustrated in the accompanying drawings and described in the specification. However, examples of embodiments based on the concept are not limited to specific embodiments, but include all changes, equivalents, or substitutions included within the spirit and scope of this disclosure.
[0026] It should be understood that when describing an element as "connected" or "linked" to another element, the element may be directly connected or linked to the other element, or it may be connected or linked to the other element via a third element. Conversely, it should be understood that when an element is described as "directly connected to" or "directly linked to" another element, no other element is placed between them. Other expressions describing relationships between components (i.e., "between" and "directly between" or "adjacent to" and "directly adjacent to") need to be interpreted in the same way.
[0027] The terminology used in this specification is for the purpose of describing specific examples of implementations only and is not intended to limit this disclosure. The singular form may include the plural form unless there is an explicit contrary meaning in the context. It should be understood in this specification that the terms "comprising" or "having" indicate the presence of the features, quantities, steps, operations, components, parts, or combinations thereof described in the specification, but do not preclude the possibility of the presence or addition of one or more other features, quantities, steps, operations, components, parts, or combinations thereof.
[0028] Unless otherwise defined, all terms used herein (including technical or scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art. If a term is not clearly defined in a common dictionary in this specification, it shall be interpreted as having the same meaning as in the context of the relevant art, and not as an ideal or overly formal meaning.
[0029] Descriptions of known components and processing techniques may be omitted to avoid unnecessarily obscuring the embodiments of this disclosure.
[0030] Throughout this specification, the same reference numerals refer to the same elements. Therefore, even if a reference numeral is not mentioned or described with reference to one drawing, it may be mentioned or described with reference to another drawing. Furthermore, even if a reference numeral is not shown in one drawing, it may be mentioned or described with reference to another drawing.
[0031] Additionally, the logic level of a signal may be different from or opposite to the logic level described. For example, a signal described as having a logic "high" level may optionally have a logic "low" level, and a signal described as having a logic "low" level may optionally have a logic "high" level.
[0032] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been provided in the embodiments of this disclosure to facilitate a better understanding of the disclosure. However, the technical solutions claimed in this disclosure can be implemented even without these technical details and various variations and modifications based on the following embodiments.
[0033] Please see Figure 1 , Figure 1 A flowchart illustrating a method for analyzing real-time vehicle data provided in this application embodiment. This invention provides a method for analyzing real-time vehicle data, comprising: S1. Multi-source heterogeneous data streaming access and standardization: Through the vehicle data acquisition unit, the real-time data streams of various vehicle subsystems are accessed, and the data of different protocols and formats in the real-time data streams are parsed and standardized, and the data is uniformly encapsulated into standard data objects.
[0034] Specifically, the standard data object includes a timestamp, a data source ID, and a data value. Real-time data streams from various vehicle subsystems are accessed through the onboard data acquisition unit. The system uses a protocol adapter to parse and standardize data of different protocols and formats, uniformly encapsulating the data into a standard data object with fields such as timestamp, data source ID, and data value.
[0035] S2. Real-time feature extraction: Set a time memory buffer for each key monitoring parameter in the vehicle, and calculate the real-time feature vector within the buffer.
[0036] Specifically, to avoid processing the entire historical dataset, the system sets a time memory buffer for each key monitoring parameter. Within the buffer, a set of characteristic indicators that can reflect the short-term operating status of the equipment are calculated in real time, including but not limited to: instantaneous value comparison with threshold, mean, variance, peak value, trend of change (slope of linear fitting), frequency domain features (Fast Fourier Transform, FFT), etc.
[0037] S3. Lightweight Parallel Fault Diagnosis Model Inference: Input the real-time feature vector into the pre-trained lightweight diagnosis model, perform parallel inference, and output diagnostic information. Specifically, the real-time feature vectors are input into one or more pre-trained lightweight diagnostic models for parallel inference. The diagnostic models employ pruned and quantized deep learning models (such as one-dimensional convolutional neural networks (1D-CNN) or small long short-term memory networks (LSTM)) to adapt to onboard computing resources. Model outputs include: Fault classification: identifying specific fault types (e.g., control failure, sensor failure, etc.); Health score: providing a current health status score for the device (between 0 and 1); Anomaly probability: providing the confidence level that the current data belongs to an abnormal state.
[0038] S4. Multi-dimensional correlation analysis and root cause inference: Based on the diagnostic information, start the correlation analysis engine; Specifically, upon identifying a fault code or anomaly, the system activates its correlation analysis engine. This engine, based on a predefined vehicle system knowledge graph, analyzes other parameters that are physically, functionally, or logically related to the fault parameter. By analyzing the coordinated changes in these related parameters, the root cause of the fault is inferred.
[0039] S5. Real-time decision-making and tiered early warning: Based on the diagnostic information, generate early warning information of different levels; Specifically, based on the diagnostic and analysis results, the system generates different levels of early warning information. Level 1 warning: for minor deviations in equipment performance or situations requiring attention. Level 2 warning: for situations where clear fault symptoms have been identified and inspection is required in the next maintenance cycle. Level 3 warning: for immediate faults that seriously affect driving safety; the system will directly display the alarm to staff through the vehicle's safety network and suggest or trigger emergency response measures.
[0040] S6. Diagnostic Results and Data Feedback: The diagnostic information, early warning information, key feature data in the feature vector, and raw data in the real-time data stream are packaged and transmitted to the ground data center in real time through the communication network.
[0041] Specifically, diagnostic conclusions, early warning information, key characteristic data, and related raw data are packaged and transmitted in real time to the ground data center via a communication network for further in-depth analysis and historical archiving to provide decision support for operation and maintenance.
[0042] Please see Figure 2 , Figure 2 The flowchart below illustrates step S2 of a method for analyzing real-time vehicle data provided in this application embodiment. Step S2 involves setting a time memory buffer for each key monitoring parameter in the vehicle, and calculating the real-time feature vector within the buffer, including: S21. Determine monitoring parameters: Select the key vehicle parameters to be monitored and set a buffer in memory; S22. Calculate based on the set trigger each time a new data point arrives; S23. After receiving the real-time data stream, the computational processing unit performs unit computation according to the stream processing framework. S24. After receiving the real-time data stream according to the aforementioned stream processing framework, perform calculations using the mean method, variance method, standard deviation method, peak value method, linear fitting slope method, and frequency domain feature method. In step S24: The formula for the mean method is: , The formula for the variance method is: , The formula for the standard deviation method is: , The peak method formula is: , The formula for the linear fitting slope method is: , Where n is the number of data points. It is the sum of the products of x and y. and It is the sum of x and y. It is the sum of x squared; The formula for the frequency domain feature method is: , Where N is the total length of the original signal, and k is the frequency index, with a value range of 0, 1, ..., N / 2. 10,1,...,N / 2 1. X(k) is the k-th frequency component of the N-point Fourier transform result, G(k) is the N / 2-point Fourier transform result calculated from all even-numbered points in the original signal, and H(k) is the N / 2-point Fourier transform result calculated from all odd-numbered points in the original signal. The rotation factor is a complex number that acts as the "glue" in the algorithm, responsible for phase adjustment of the transformation result H(k) of the odd sequence.
[0043] S25. After all the computing processing units have completed their calculations, they assemble all the features into the real-time feature vector.
[0044] Please see Figure 3 , Figure 3 This is a flowchart of step S3 of a method for analyzing real-time vehicle data provided in an embodiment of this application. Step S3 involves inputting the real-time feature vector into a pre-trained lightweight diagnostic model for parallel inference, outputting diagnostic information, including: S31. Model preparation and loading: Deploy at least one pre-trained lightweight diagnostic model in advance; S32. Feature Vector Input: The real-time feature vector is synchronously input into the pre-trained lightweight diagnostic model; the "feature vector" extracted from the real-time sensor data is synchronously input into the diagnostic model (one or more) loaded in step 3-1. "One" model: This can be a comprehensive model that simultaneously performs fault classification, health scoring, and anomaly probability calculation. "Multiple" models: These are multiple specialized models, for example: Model A: specifically responsible for fault classification, Model B: specifically responsible for health scoring, and Model C: specifically responsible for anomaly overview calculation.
[0045] S33, Parallel Model Inference: Calculate and analyze the real-time feature vector of the input; S34. Multi-dimensional result output: Output diagnostic information, including fault classification, health score, and anomaly probability.
[0046] Specifically, the trained lightweight diagnostic model includes model pruning and model quantization, as detailed below: Detailed steps for pruning the model: Step 1) Initial training: First, train a complete, high-precision original model on a powerful cloud server using a large dataset.
[0047] Step 2) Importance Assessment: Analyze the importance of each parameter, neuron, or connection in the model to the final training result. Identify important parameters.
[0048] Step 3) Remove redundancy: Remove unimportant parameters from the model.
[0049] Step 4) Fine-tuning recovery: Train or fine-tune the streamlined model using a smaller dataset.
[0050] Final result: We obtained a model with a sparser structure and smaller size. It retains the core decision-making capabilities, but the computational cost and model file size are significantly reduced.
[0051] Detailed steps for model quantization: Step 1) Precision Analysis: First, in the original model, analyze the precision of the data, such as the fact that numbers are usually represented using high precision (e.g., 32-bit floating-point numbers, float32).
[0052] Step 2) Precision Conversion: Reduce the precision of the numbers in the model from high precision to low precision (such as 16-bit floating-point number float16, or even 8-bit integer int8).
[0053] The present invention also provides a system for analyzing real-time vehicle data, comprising: The multi-source heterogeneous data streaming access and standardization module is used to access the real-time data streams of various vehicle subsystems through the vehicle data acquisition unit, parse and standardize the data with different protocols and formats in the real-time data streams, and encapsulate the data into a standard data object. The real-time feature extraction module is used to set a time memory buffer for each key monitoring parameter in the vehicle, and calculate the real-time feature vector within the buffer. The lightweight parallel fault diagnosis model inference module is used to input the real-time feature vector into the pre-trained lightweight diagnosis model, perform parallel inference, and output diagnostic information.
[0054] The multi-dimensional correlation analysis and root cause inference module is used to activate the correlation analysis engine based on the diagnostic information; The real-time decision-making and hierarchical early warning module is used to generate early warning information of different levels based on the diagnostic information; The diagnostic results and data feedback module is used to package the diagnostic information, early warning information, key feature data in the feature vector, and raw data in the real-time data stream, and transmit them to the ground data center in real time through the communication network.
[0055] This invention also proposes an electronic device comprising a processor and a memory. The memory stores program instructions, and the processor executes these program instructions to implement the aforementioned method for analyzing real-time vehicle data. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The memory may include Random Access Memory (RAM) and may also include Non-Volatile Memory, such as at least one disk storage device. The memory can also be internal memory of the Random Access Memory (RAM) type. The processor and memory can be integrated into one or more independent circuits or hardware, such as an Application Specific Integrated Circuit (ASIC). It should be noted that the computer program in the aforementioned memory, when implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, an electronic device, or a network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention.
[0056] This invention also proposes a computer-readable storage medium storing computer instructions for instructing a computer to execute the aforementioned method for analyzing real-time vehicle data. The computer-readable storage medium can be an electronic medium, magnetic medium, optical medium, electromagnetic medium, infrared medium, or a semiconductor system or propagation medium. The computer-readable storage medium may also include semiconductor or solid-state memory, magnetic tape, removable computer disk, random access memory (RAM), read-only memory (ROM), hard disk, and optical disk. Optical disks may include optical disc-read-only memory (CD-ROM), optical disc-read / write (CD-RW), and DVD.
[0057] In summary, the present invention provides a method for analyzing real-time vehicle data, enabling real-time diagnosis and early warning: by using streaming computing and online analysis, it changes the traditional reactive approach and allows for immediate intervention in operational faults. The invention also improves fault diagnosis accuracy: by combining a lightweight model with knowledge graph association analysis, it identifies complex faults, locates the root cause of the fault, and reduces the false alarm and false negative rates of traditional threshold methods.
[0058] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.
Claims
1. A method for analyzing real-time vehicle data, characterized in that, include: S1. Multi-source heterogeneous data streaming access and standardization: Through the vehicle data acquisition unit, the real-time data streams of various vehicle subsystems are accessed, and the different protocols and formats of the real-time data streams are parsed and standardized, and the data is uniformly encapsulated into standard data objects. S2. Real-time feature extraction: Set a time memory buffer for each key monitoring parameter in the vehicle, and calculate the real-time feature vector within the buffer; S3. Lightweight Parallel Fault Diagnosis Model Inference: Input the real-time feature vector into the pre-trained lightweight diagnosis model, perform parallel inference, and output diagnostic information. S4. Multi-dimensional correlation analysis and root cause inference: Based on the diagnostic information, start the correlation analysis engine; S5. Real-time decision-making and tiered early warning: Based on the diagnostic information, generate early warning information of different levels; S6. Diagnostic Results and Data Feedback: The diagnostic information, early warning information, key feature data in the feature vector, and raw data in the real-time data stream are packaged and transmitted to the ground data center in real time through the communication network.
2. The method for analyzing real-time vehicle data according to claim 1, characterized in that: The standard data object includes a timestamp, a data source ID, and a data value.
3. The method for analyzing real-time vehicle data according to claim 1, characterized in that: Step S2 involves setting a time memory buffer for each key monitoring parameter in the vehicle, and calculating the real-time feature vector within the buffer, including: S21. Determine monitoring parameters: Select the key vehicle parameters to be monitored and set a buffer in memory; S22. Calculate based on the set trigger each time a new data point arrives; S23. After receiving the real-time data stream, the computational processing unit performs unit computation according to the stream processing framework. S24. After receiving the real-time data stream according to the aforementioned stream processing framework, perform calculations using the mean method, variance method, standard deviation method, peak value method, linear fitting slope method, and frequency domain feature method. S25. After all the computing processing units have completed their calculations, they assemble all the features into the real-time feature vector.
4. The method for analyzing real-time vehicle data according to claim 3, characterized in that, In step S24: The formula for the mean method is: ， The formula for the variance method is: ， The formula for the standard deviation method is: ， The peak method formula is: ， The formula for the linear fitting slope method is: ， Where n is the number of data points. It is the sum of the products of x and y. and It is the sum of x and y. It is the sum of x squared; The formula for the frequency domain feature method is: ， Where N is the total length of the original signal, and k is the frequency index, with a value range of 0, 1, ..., N / 2. 10,1,...,N / 2 1. X(k) is the k-th frequency component of the N-point Fourier transform result, G(k) is the N / 2-point Fourier transform result calculated from all even-numbered points in the original signal, and H(k) is the N / 2-point Fourier transform result calculated from all odd-numbered points in the original signal. The rotation factor is a complex number.
5. The method for analyzing real-time vehicle data according to claim 4, characterized in that: In step S3, the real-time feature vector is input into a pre-trained lightweight diagnostic model for parallel inference, and the output diagnostic information includes: S31. Model preparation and loading: Deploy at least one pre-trained lightweight diagnostic model in advance; S32, Feature Vector Input: The real-time feature vector is synchronously input into the pre-trained lightweight diagnostic model; S33, Parallel Model Inference: Calculate and analyze the real-time feature vector of the input; S34. Multi-dimensional result output: Output diagnostic information, including fault classification, health score, and anomaly probability.
6. The method for analyzing real-time vehicle data according to claim 5, characterized in that: The lightweight diagnostic model trained includes model pruning and model quantization.
7. The method for analyzing real-time vehicle data according to claim 1, characterized in that, The model pruning includes: each subsystem of the vehicle.
8. The method for analyzing real-time vehicle data according to claim 5, characterized in that, The fault classification includes: control failure and sensor failure; the health score includes the current health status score of the vehicle, between 0 and 1; the anomaly probability includes the confidence level of the abnormal state.
9. A system for analyzing real-time vehicle data, characterized in that, include: The multi-source heterogeneous data streaming access and standardization module is used to access the real-time data streams of various vehicle subsystems through the vehicle data acquisition unit, parse and standardize the data with different protocols and formats in the real-time data streams, and encapsulate the data into a standard data object. The real-time feature extraction module is used to set a time memory buffer for each key monitoring parameter in the vehicle, and calculate the real-time feature vector within the buffer. The lightweight parallel fault diagnosis model inference module is used to input the real-time feature vector into the pre-trained lightweight diagnosis model, perform parallel inference, and output diagnostic information. The multi-dimensional correlation analysis and root cause inference module is used to activate the correlation analysis engine based on the diagnostic information; The real-time decision-making and hierarchical early warning module is used to generate early warning information of different levels based on the diagnostic information; The diagnostic results and data feedback module is used to package the diagnostic information, early warning information, key feature data in the feature vector, and raw data in the real-time data stream, and transmit them to the ground data center in real time through the communication network.
10. An electronic device comprising a processor and a memory, the memory storing program instructions, characterized in that: The processor executes program instructions to implement a method for analyzing real-time vehicle data as described in any one of claims 1 to 8.