Intelligent driving data generation method and system based on real vehicle test

Abstract

The invention relates to an intelligent driving data generation method based on a real vehicle test, and the method comprises the steps: constructing a scene event library which records the scene data of each scene event affecting the safety of intelligent driving; screening from the real vehicle test data based on the scene data of the scene event library to obtain the vehicle data of the scene event influencing the intelligent driving safety; the vehicle data is data acquired by a sensor and a processor in the vehicle running process; selecting event information elements of the scene event, and converting the vehicle data into a format containing each event information element; classifying and outputting the vehicle data based on the event information elements; through a three-level architecture of vehicle-mounted end real-time screening, edge data calculation preprocessing and cloud closed-loop generation, real vehicle test data is changed from simple passive recording to active generation, high-efficiency extraction, rapid processing and closed-loop utilization of high-value scene events are realized, and a low-cost, high-efficiency and sustainable test data engine is provided for an intelligent driving system.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving technology, and in particular to a method for generating intelligent driving data based on real vehicle testing. Background Technology

[0002] Currently, during the development of intelligent driving systems, real-vehicle testing generates massive amounts of raw data (such as sensor data, vehicle status, surrounding environment status, and driver operations). However, existing technologies have some shortcomings: 1. Low data utilization: In real vehicle testing, more than 90% of the original data cases are invalid cases, such as conventional scenarios like constant speed and straight driving. This makes it impossible to cover high-value edge scenarios, resulting in high data annotation costs and wasted resources.

[0003] 2. Poor processing efficiency: Traditional data processing methods require storing all raw data first, and then cleaning, labeling, and modeling offline. The processing cycle for a single test data may take several days, which seriously slows down the version iteration speed.

[0004] 3. Lack of test loop: Data collected from real vehicles is difficult to quickly convert into reproducible test cases, which leads to a longer development cycle of "testing - problem discovery - problem reproduction" and low testing efficiency. Summary of the Invention

[0005] This invention addresses the technical problems existing in the prior art by providing a method and system for generating intelligent driving data based on real vehicle testing. Through a three-level architecture of "real-time filtering on the vehicle terminal + edge data computation and preprocessing + closed-loop generation in the cloud", it transforms real vehicle test data from simple passive recording into actively generating test assets, realizing efficient extraction, rapid processing and closed-loop utilization of high-value scenario events, and providing a low-cost, high-efficiency and sustainable test data engine for intelligent driving systems.

[0006] According to a first aspect of the present invention, a method for generating intelligent driving data based on real-vehicle testing is provided, comprising: Step 1: Construct a scenario event library, which records scenario data for various scenario events that affect the safety of intelligent driving; Step 2: Based on the scenario data in the scenario event library, filter the real vehicle test data to obtain vehicle data of scenario events that affect intelligent driving safety; the vehicle data is the data acquired by sensors and processors during vehicle operation; Step 3: Select the event information elements of the scene event and convert the vehicle data into a format that includes each event information element; Step 4: Classify the vehicle data based on the event information elements and then output the results.

[0007] Based on the above technical solution, the present invention can also be improved as follows.

[0008] Optionally, the process of obtaining scene data for any scene event in the scene event library in step 1 includes: A training set is constructed by collecting multiple scene data of any scenario event that affects the safety of intelligent driving. The training set is then input into an AI model for scene event recognition, and the AI ​​model for scene event recognition outputs the scene data of the corresponding scene event.

[0009] Optionally, scene events in the scene event library can be added or removed in real time as needed; The scenarios and events that affect the safety of intelligent driving include: collision risk events, lane departure events, solid line lane change events, deceleration exceeding the safety threshold events, abnormal steering wheel angle events, and system and environmental alarm events; The scenario data for the collision risk event includes: the distance to a moving human or vehicle is less than a safety threshold; The scenario data for the lane departure event includes: the vehicle deviating from the current lane without using the turn signal; The scenario data for the solid line lane change event includes: vehicles changing lanes in areas where lane changing is prohibited by solid lines; The scenario data for events where deceleration exceeds a safety threshold includes: vehicle deceleration greater than the threshold; The scenario data for the abnormal steering wheel angle event includes: the steering wheel change speed is greater than a threshold, the steering wheel change angle is greater than a threshold, or the steering wheel angle frequently alternates between positive and negative; The scenario data for system and environmental alarm events includes: vehicle sensor failure or perception capability below a threshold.

[0010] Optionally, step 2 further includes: setting a time range and collecting vehicle data within a set time range before and after the occurrence of the scene event.

[0011] Optionally, step 3 further includes: deploying edge computing nodes at the test site or fleet management station; The edge computing node receives the time data uploaded by the vehicle data, aligns various types of data in the vehicle data according to timestamps, and after classifying and labeling the target objects in the video / images acquired by the sensor according to category and behavior using a pre-trained model, it converts the vehicle data into a format containing various event information elements and imports it into the cloud database.

[0012] Optionally, the vehicle data includes data acquired by cameras, lidar, and IMU; The classification of targets in the video / image includes: vehicles, pedestrians, and traffic lights; the classification and labeling of targets in the video / image by behavior includes: turning and crossing. The event information elements include: scene type, timestamp, location, participants, and key parameters.

[0013] Optionally, step 4 includes: the cloud database classifying the vehicle data based on the event information elements to generate: Simulation test cases used for algorithm regression testing; The test vehicles are automatically dispatched to the same road segment based on the geographical location of the scene to reproduce the real vehicle retest test cases. Based on existing scenarios, test cases for data augmentation templates are generated by perturbing parameters to create derivative scenarios and expand test coverage.

[0014] According to a second aspect of the present invention, an intelligent driving data generation system based on real vehicle testing is provided, comprising: an on-board real-time scene recognition and data filtering module, an edge data preprocessing module, and a cloud platform; The in-vehicle real-time scene recognition and data filtering module is used to build a scene event library, which records scene data of various scene events that affect intelligent driving safety; based on the scene data in the scene event library, vehicle data of scene events that affect intelligent driving safety are obtained by filtering from real vehicle test data; the vehicle data is the data acquired by sensors and processors during vehicle operation. The edge data preprocessing module is used to select event information elements of scene events and convert the vehicle data into a format containing each event information element. The cloud platform is used to classify and output the vehicle data based on the event information elements.

[0015] According to a third aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein the processor is configured to execute a computer management program stored in the memory to implement a method for generating intelligent driving data based on real vehicle testing.

[0016] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer management class program is stored, wherein when executed by a processor, the computer management class program implements the steps of a method for generating intelligent driving data based on real vehicle testing.

[0017] This invention provides a method, system, electronic device, and storage medium for generating intelligent driving data based on real-vehicle testing. Through a three-tiered architecture of "real-time filtering at the vehicle end + edge data computation and preprocessing + cloud-based closed-loop generation," it transforms real-vehicle test data from simple passive recording into actively generated test assets. This enables efficient extraction, rapid processing, and closed-loop utilization of high-value scenario events, providing a low-cost, high-efficiency, and sustainable test data engine for intelligent driving systems. Key benefits include: Data utilization increased by over 80%: focusing on high-value scenario data and avoiding the storage and processing of massive amounts of routine, invalid data; significantly improved data processing efficiency: edge preprocessing reduces single-event processing time from hours to minutes, supporting data availability verification on the same day of testing; doubled testing efficiency: the time from problem discovery to generating reproducible test cases is reduced from days to hours, greatly improving testing efficiency and algorithm iteration speed; significantly reduced costs: reducing data storage and annotation costs by 90%, with virtually no need for manual intervention in scenario extraction and annotation; and strong closed-loop capability: achieving a "real vehicle—simulation—real vehicle" testing closed loop, accelerating the maturity of intelligent driving systems. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a method for generating intelligent driving data based on real-vehicle testing, provided by this invention; Figure 2 A structural block diagram of an intelligent driving data generation system based on real vehicle testing provided by the present invention; Figure 3 A schematic diagram of the hardware structure of a possible electronic device provided by the present invention; Figure 4 This is a schematic diagram of the hardware structure of a possible computer-readable storage medium provided by the present invention. Detailed Implementation

[0019] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0020] Figure 1 A flowchart of an intelligent driving data generation method based on real vehicle testing provided by the present invention is shown below. Figure 1 As shown, the intelligent driving data generation method includes: Step 1: Build a scenario event library. The scenario event library records scenario data for various scenario events that affect the safety of intelligent driving.

[0021] Step 2: Based on the scenario event library, the scenario data is filtered from the real vehicle test data to obtain vehicle data of scenario events that affect the safety of intelligent driving; the vehicle data is the data acquired by the sensors and processors during vehicle operation.

[0022] In practice, during real-vehicle testing, the onboard computing unit can analyze sensor and vehicle CAN data in real time.

[0023] Step 3: Select the event information elements of the scene event and convert the vehicle data into a format that includes each event information element.

[0024] Step 4: Classify and output the vehicle data based on event information elements.

[0025] This invention provides a method for generating intelligent driving data based on real vehicle testing. Through a three-level architecture of "real-time filtering on the vehicle side + edge data computation and preprocessing + closed-loop generation in the cloud", it transforms real vehicle test data from simple passive recording into actively generating test assets, realizing efficient extraction, rapid processing and closed-loop utilization of high-value scenario events, and providing a low-cost, high-efficiency and sustainable test data engine for intelligent driving systems.

[0026] Example 1 Embodiment 1 provided by this invention is an embodiment of a method for generating intelligent driving data based on real vehicle testing provided by this invention, combined with... Figure 1 It can be seen that embodiments of this intelligent driving data generation method include: Step 1: Build a scenario event library. The scenario event library records scenario data for various scenario events that affect the safety of intelligent driving.

[0027] In one possible embodiment, the process of obtaining scene data for any scene event in the scene event library in step 1 includes: A training set is constructed by collecting multiple scenario data of any scenario event that affects the safety of intelligent driving. The training set is then input into the AI ​​model for scenario event recognition, and the AI ​​model for scenario event recognition outputs the scenario data of the corresponding scenario event.

[0028] In practice, the AI ​​model for scene event recognition can be a lightweight AI model, such as a decision tree or a small neural network.

[0029] The core architecture of a decision tree includes: a root node, internal nodes, branches, and leaf nodes. The root node contains the starting point for all training data. Internal nodes contain feature test conditions, distributing data to different child nodes based on the test results. Branches are connection nodes, representing the test results. Leaf nodes represent the final decision result of the tree, containing either a class label (classification) or a predicted value (regression).

[0030] A small neural network consists of an input layer, hidden layers, and an output layer. The number of nodes in the input layer equals the feature dimension, and it receives the raw data. Hidden layers can contain 1-3 layers, with the number of nodes in each layer adjusted according to task complexity. Layers are connected by weight matrices, and activation functions are applied to introduce non-linearity. The number of nodes in the output layer is determined by the task; for example, one node may be used for regression, while multiple nodes may be used for multi-class classification. Activation functions such as Softmax or Sigmoid are used.

[0031] In one possible implementation, scene events in the scene event library are added or removed in real time as needed.

[0032] In practice, scene events in the scene event library can be added or removed in real time as needed. For example, when a single scene needs to be tested, other scene events can be removed. Scene events that do not affect the safety of intelligent driving can also be deleted in real time during the test. New scene events that affect the safety of intelligent driving can also be added in real time during the test.

[0033] Scenarios that affect the safety of intelligent driving include: collision risk events, lane departure events, solid line lane change events, deceleration exceeding safety threshold events, abnormal steering wheel angle events, and system and environmental alarm events.

[0034] Collision risk event scenario data includes: the distance to moving humans or vehicles is less than the safety threshold.

[0035] Lane departure event scenario data includes: a vehicle deviating from its current lane without using its turn signal.

[0036] The scenario data for solid line lane change events includes: vehicles changing lanes in areas where lane changes are prohibited by solid lines.

[0037] Scenario data for events where deceleration exceeds a safety threshold includes: vehicle deceleration greater than a threshold; specifically, the threshold can be: deceleration > 3 m / s². 2。

[0038] The scenario data for abnormal steering wheel angle events include: steering wheel change speed exceeding a threshold, steering wheel change angle exceeding a threshold, or frequent alternation of positive and negative steering wheel angles.

[0039] The scenario data for system and environmental alarm events includes: vehicle sensor failure or perception capability below a threshold.

[0040] In practice, the specific rules or thresholds for driving data in various scenarios that affect the safety of intelligent driving can be based on the data output by the AI ​​model, or can be appropriately adjusted based on the data output by the AI ​​model.

[0041] Step 2: Based on the scenario event library, the scenario data is filtered from the real vehicle test data to obtain vehicle data of scenario events that affect the safety of intelligent driving; the vehicle data is the data acquired by the sensors and processors during vehicle operation.

[0042] In one possible embodiment, step 2 further includes: setting a time range and collecting vehicle data within the set time range before and after the occurrence of the scene event.

[0043] In practice, this time range can be set to 10 seconds before and after the occurrence of the scene event, and only event fragment data can be uploaded, which can compress the amount of data.

[0044] Step 3: Select the event information elements of the scene event and convert the vehicle data into a format that includes each event information element.

[0045] In one possible embodiment, step 3 further includes: deploying edge computing nodes at the test site or fleet management station.

[0046] Edge computing nodes receive the time data uploaded by vehicles and align various types of data in the vehicle data according to timestamps. After classifying and labeling the target objects in the video / images acquired by the sensors according to categories and behaviors using a pre-trained model, the vehicle data is converted into a format containing various event information elements and then imported into the cloud database.

[0047] In one possible embodiment, the vehicle data includes data acquired by cameras, lidar, and IMU.

[0048] The classification of objects in videos / images includes categories such as vehicles, pedestrians, and traffic lights; the classification and labeling of objects in videos / images by behavior includes categories such as turning and crossing.

[0049] Event information elements include: scene type, timestamp, location, participants, and key parameters.

[0050] In practice, automated preprocessing is performed through edge computing nodes to convert raw data into a standard format containing various event information elements: (scene type, timestamp, location, participants, key parameters), processing events with a processing time of less than 30 seconds, and supporting concurrent uploads from multiple vehicles.

[0051] Step 4: After classifying the vehicle data based on event information elements, output the results. In one possible embodiment, step 4 includes: the cloud database classifies the vehicle data based on event information elements and then generates the following: Simulation test cases used for algorithm regression testing.

[0052] This involves automatically dispatching test vehicles to the same road segment based on the scene's geographical location to perform real-vehicle retesting of the scenario.

[0053] Based on existing scenarios, test cases for data augmentation templates are generated by perturbing parameters to create derivative scenarios and expand test coverage.

[0054] It supports test result feedback, forming a closed loop of "real vehicle data collection - scenario generation - simulation / real vehicle verification - model optimization".

[0055] Example 2 Embodiment 2 provided by this invention is an embodiment of an intelligent driving data generation system based on real vehicle testing provided by this invention. Figure 2 This is a structural diagram of an intelligent driving data generation system based on real vehicle testing, provided in an embodiment of the present invention. Figure 2 It is known that the embodiment of the intelligent driving data generation system includes: a vehicle-mounted real-time scene recognition and data filtering module, an edge data preprocessing module, and a cloud platform.

[0056] The vehicle-mounted real-time scene recognition and data filtering module is used to build a scene event library, which records scene data of various scene events that affect intelligent driving safety. Based on the scene data in the scene event library, vehicle data of scene events that affect intelligent driving safety are obtained by filtering from real vehicle test data. The vehicle data is the data acquired by sensors and processors during vehicle operation.

[0057] The edge data preprocessing module is used to select event information elements of scene events and convert vehicle data into a format that includes each event information element.

[0058] In the cloud, vehicle data is categorized and output based on event information elements.

[0059] It is understood that the intelligent driving data generation system based on real vehicle testing provided by the present invention corresponds to the intelligent driving data generation method based on real vehicle testing provided in the foregoing embodiments. The relevant technical features of the intelligent driving data generation system based on real vehicle testing can be referred to the relevant technical features of the intelligent driving data generation method based on real vehicle testing, and will not be repeated here.

[0060] Please see Figure 3 , Figure 3 This is a schematic diagram illustrating an embodiment of the electronic device provided in this invention. For example... Figure 3As shown, this embodiment of the invention provides an electronic device, including a memory 1310, a processor 1320, and a computer program 1311 stored in the memory 1310 and executable on the processor 1320. When the processor 1320 executes the computer program 1311, it performs the following steps: constructing a scene event library, which records scene data of various scene events affecting intelligent driving safety; filtering vehicle data of scene events affecting intelligent driving safety from real vehicle test data based on the scene data in the scene event library; the vehicle data being data acquired by sensors and the processor during vehicle operation; selecting event information elements of scene events and converting the vehicle data into a format containing various event information elements; and outputting the vehicle data after classifying it based on the event information elements.

[0061] Please see Figure 4 , Figure 4 This is a schematic diagram illustrating an embodiment of a computer-readable storage medium provided by the present invention. (See diagram below.) Figure 4 As shown, this embodiment provides a computer-readable storage medium 1400, on which a computer program 1411 is stored. When the computer program 1411 is executed by a processor, it performs the following steps: constructing a scene event library, which records scene data of various scene events affecting intelligent driving safety; filtering vehicle data of scene events affecting intelligent driving safety from real vehicle test data based on the scene data in the scene event library; the vehicle data being data acquired by sensors and processors during vehicle operation; selecting event information elements of scene events and converting the vehicle data into a format containing various event information elements; and outputting the vehicle data after classifying it based on the event information elements.

[0062] This invention provides a method, system, electronic device, and storage medium for generating intelligent driving data based on real-vehicle testing. Through a three-tiered architecture of "real-time filtering at the vehicle end + edge data computation and preprocessing + cloud-based closed-loop generation," it transforms real-vehicle test data from simple passive recording into actively generated test assets. This enables efficient extraction, rapid processing, and closed-loop utilization of high-value scenario events, providing a low-cost, high-efficiency, and sustainable test data engine for intelligent driving systems. Key benefits include: Data utilization increased by over 80%: focusing on high-value scenario data and avoiding the storage and processing of massive amounts of routine, invalid data; significantly improved data processing efficiency: edge preprocessing reduces single-event processing time from hours to minutes, supporting data availability verification on the same day of testing; doubled testing efficiency: the time from problem discovery to generating reproducible test cases is reduced from days to hours, greatly improving testing efficiency and algorithm iteration speed; significantly reduced costs: reducing data storage and annotation costs by 90%, with virtually no need for manual intervention in scenario extraction and annotation; and strong closed-loop capability: achieving a "real vehicle—simulation—real vehicle" testing closed loop, accelerating the maturity of intelligent driving systems.

[0063] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0064] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0065] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0066] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0067] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0068] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0069] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for generating intelligent driving data based on real-vehicle testing, characterized in that, The intelligent driving data generation method includes: Step 1: Construct a scenario event library, which records scenario data for various scenario events that affect the safety of intelligent driving; Step 2: Based on the scenario data in the scenario event library, filter the real vehicle test data to obtain vehicle data of scenario events that affect intelligent driving safety; the vehicle data is the data acquired by sensors and processors during vehicle operation; Step 3: Select the event information elements of the scene event and convert the vehicle data into a format that includes each event information element; Step 4: Classify the vehicle data based on the event information elements and then output the results.

2. The intelligent driving data generation method according to claim 1, characterized in that, The process of obtaining scene data for any scene event in the scene event library in step 1 includes: A training set is constructed by collecting multiple scene data of any scenario event that affects the safety of intelligent driving. The training set is then input into an AI model for scene event recognition, and the AI ​​model for scene event recognition outputs the scene data of the corresponding scene event.

3. The intelligent driving data generation method according to claim 1, characterized in that, The scene events in the scene event library can be deleted or added in real time as needed; The scenarios and events that affect the safety of intelligent driving include: collision risk events, lane departure events, solid line lane change events, deceleration exceeding the safety threshold events, abnormal steering wheel angle events, and system and environmental alarm events; The scenario data for the collision risk event includes: the distance to a moving human or vehicle is less than a safety threshold; The scenario data for the lane departure event includes: the vehicle deviating from the current lane without using the turn signal; The scenario data for the solid line lane change event includes: vehicles changing lanes in areas where lane changing is prohibited by solid lines; The scenario data for events where deceleration exceeds a safety threshold includes: vehicle deceleration greater than the threshold; The scenario data for the abnormal steering wheel angle event includes: the steering wheel change speed is greater than a threshold, the steering wheel change angle is greater than a threshold, or the steering wheel angle frequently alternates between positive and negative; The scenario data for system and environmental alarm events includes: vehicle sensor failure or perception capability below a threshold.

4. The intelligent driving data generation method according to claim 1, characterized in that, Step 2 further includes: setting a time range and collecting vehicle data within a set time range before and after the occurrence of the scene event.

5. The intelligent driving data generation method according to claim 1, characterized in that, Step 3 further includes: deploying edge computing nodes at the test site or fleet management station; The edge computing node receives the time data uploaded by the vehicle data, aligns various types of data in the vehicle data according to timestamps, and after classifying and labeling the target objects in the video / images acquired by the sensor according to category and behavior using a pre-trained model, it converts the vehicle data into a format containing various event information elements and imports it into the cloud database.

6. The intelligent driving data generation method according to claim 5, characterized in that, The vehicle data includes data acquired by cameras, lidar, and IMU. The classification of targets in the video / image includes: vehicles, pedestrians, and traffic lights; the classification and labeling of targets in the video / image by behavior includes: turning and crossing. The event information elements include: scene type, timestamp, location, participants, and key parameters.

7. The intelligent driving data generation method according to claim 1, characterized in that, Step 4 includes: the cloud database classifies the vehicle data based on the event information elements to generate: Simulation test cases used for algorithm regression testing; The test vehicles are automatically dispatched to the same road segment based on the geographical location of the scene to reproduce the real vehicle retest test cases. Based on existing scenarios, test cases for data augmentation templates are generated by perturbing parameters to create derivative scenarios and expand test coverage.

8. A smart driving data generation system based on real vehicle testing, characterized in that, The intelligent driving data generation system includes: an on-board real-time scene recognition and data filtering module, an edge data preprocessing module, and a cloud platform; The in-vehicle real-time scene recognition and data filtering module is used to build a scene event library, which records scene data of various scene events that affect intelligent driving safety; based on the scene data in the scene event library, vehicle data of scene events that affect intelligent driving safety are obtained by filtering from real vehicle test data; the vehicle data is the data acquired by sensors and processors during vehicle operation. The edge data preprocessing module is used to select event information elements of scene events and convert the vehicle data into a format containing each event information element. The cloud platform is used to classify and output the vehicle data based on the event information elements.

9. An electronic device, characterized in that, It includes a memory and a processor, wherein the processor is used to execute computer management programs stored in the memory to implement the steps of the intelligent driving data generation method based on real vehicle testing as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, It stores a computer management program, which, when executed by a processor, implements the steps of the intelligent driving data generation method based on real vehicle testing as described in any one of claims 1-7.

Images