Event triggering type automatic driving data automatic acquisition system based on vehicle cloud return

Abstract

The invention discloses an event-triggered automatic driving data automatic acquisition system based on vehicle cloud return, and belongs to the technical field of automatic driving, the event-triggered automatic driving data automatic acquisition system comprises a cloud end and a vehicle end, and a cloud end algorithm component of the cloud end is used for designing and issuing a trigger algorithm to the vehicle end; the triggering algorithm assembly of the vehicle end is used for receiving a triggering algorithm issued by the cloud end and triggering the automatic driving data assembly to collect data; the data storage component is used for temporarily storing or storing the data acquired by the automatic driving data component for a long time based on the triggering algorithm; and the data return assembly is used for carrying out data interaction between the vehicle end and the cloud end based on the data stored by the data storage assembly. According to the invention, data acquisition and return can be automatically realized, a full-amount data acquisition mode is avoided, and flexible modification and optimization of a vehicle end deployment algorithm can be flexibly realized.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and more specifically, to an event-triggered automated autonomous driving data collection system based on vehicle-to-cloud feedback. Background Art

[0002] At present, there are many methods for collecting data for autonomous driving, but the collection efficiency is generally low. Data collection systems that can truly meet the needs of autonomous driving data collection and have high efficiency, convenient modification and high coverage are relatively scarce.

[0003] The general autonomous driving data collection system adopts the full data collection mode to collect data locally on the real vehicle. This mode deploys the data collection system on the autonomous driving vehicle to comprehensively collect all data related to autonomous driving, and stores this data in the form of data disk on the vehicle hard disk. Subsequently, the data in the hard disk needs to be manually moved and uploaded to the storage server for use in autonomous driving development.

[0004] However, the disadvantages of the existing technology include:

[0005] Large amount of collected data and high cost: Due to the huge amount of collected data, the collection cost, local storage cost of autonomous driving vehicles and data center storage cost are all very high.

[0006] The proportion of valuable data is low: Among the data obtained through full data collection, the proportion of actually valuable data is low, and there is a large amount of useless and duplicate data, which reduces the effective utilization rate of data.

[0007] High dependence on manual labor and low degree of automation: Due to the large amount of data collected and the need for manual work such as moving, uploading and processing hard disk data, the degree of automation is low, resulting in a significant increase in labor costs. Summary of the invention

[0008] One object of the present invention is to provide an event-triggered automatic autonomous driving data collection system based on vehicle-to-cloud feedback to solve the aforementioned problems of huge amount of collected data, high storage cost, less valuable data, low collection efficiency, high proportion of manual labor, and low degree of automation.

[0009] According to the present invention, an event-triggered automatic driving data collection system based on vehicle-to-cloud feedback is provided, including a cloud side and a vehicle side, wherein:

[0010] The cloud comprises:

[0011] A cloud algorithm component, used to design and send a trigger algorithm to the vehicle end;

[0012] The vehicle end comprises:

[0013] Autonomous driving data component, used to collect various data,

[0014] A trigger algorithm component, used to receive the trigger algorithm sent by the cloud and trigger the autonomous driving data component to collect data;

[0015] A data storage component temporarily or long-term stores the data collected by the autonomous driving data component based on the trigger algorithm;

[0016] The data return component performs data interaction between the vehicle and the cloud based on the data stored by the data storage component.

[0017] Optionally, the cloud algorithm component includes:

[0018] Cloud trigger algorithm design module, used to create, optimize and verify trigger algorithms in the cloud;

[0019] The cloud trigger algorithm sending module is used to send the designed trigger algorithm to the vehicle end.

[0020] Optionally, the cloud further includes a cloud data component, and the cloud data component includes:

[0021] A cloud data receiving module, used to receive data sent back by the vehicle-side data return component;

[0022] The cloud data storage module is used to store the data received by the cloud data receiving module in a cloud data center.

[0023] Optionally, the autonomous driving data component includes:

[0024] The sensor data acquisition device collects information about the vehicle's surrounding environment in real time through various sensors installed on the vehicle;

[0025] The automobile CAN data acquisition device acquires the vehicle operation status data interacting between the electronic control units through the vehicle's internal CAN bus;

[0026] The autonomous driving algorithm data acquisition device is used to collect data generated during the operation of the algorithm inside the autonomous driving system.

[0027] Optionally, the trigger algorithm component includes:

[0028] A trigger algorithm receiving module, used to receive trigger algorithm information from the cloud;

[0029] The trigger algorithm trigger module is used to determine the trigger timing according to the preset trigger conditions and start the data collection process of the autonomous driving data component.

[0030] Optionally, the data return component includes:

[0031] A vehicle-to-cloud data uploading module, used to build a data transmission channel from the vehicle to the cloud, and push the data stored by the data storage component;

[0032] The vehicle-to-cloud data retransmission module is used to resend data to the cloud when data upload fails, and to delete the data in the hard disk after resending to leave available space.

[0033] Optionally, the data storage component includes a memory storage module and a hard disk storage module, wherein:

[0034] At the moment of receiving the collected data, the data storage component preferentially activates the memory storage module for quick temporary storage, and synchronously drives the vehicle-to-cloud data upload module of the data return component to push the memory data to the cloud based on the vehicle-to-cloud communication link;

[0035] When the fast temporary storage is interrupted, the data return component immediately instructs the hard disk storage module to migrate the memory data to the hard disk backup;

[0036] The car-to-cloud data retransmission module of the data return component retransmits the hard disk backed-up data. After the retransmission is completed, the successfully returned data in the hard disk is automatically deleted to release hard disk space.

[0037] Optionally, the trigger algorithm component receives the trigger algorithm sent by the cloud and triggers the autonomous driving data component to collect data, including:

[0038] Set the triggering conditions of the triggering algorithm in advance according to the corresponding nodes of the autonomous driving business;

[0039] When the vehicle-side operating state meets the preset trigger condition, the trigger algorithm component is triggered to immediately generate an automatic driving data collection instruction;

[0040] According to the data screening rules built into the trigger algorithm, key data in the autonomous driving data component is located, collected and packaged into data packets and sent to the data storage component.

[0041] Optionally, the data storage component has a built-in intelligent resource management program, which dynamically allocates storage resources of the memory storage module and the hard disk storage module according to real-time data flow, priority, storage duration and hardware performance.

[0042] The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback disclosed in the present invention has the following technical effects:

[0043] (1) The data backhaul component includes the vehicle-to-cloud data upload module and the vehicle-to-cloud data retransmission module. It automates data collection and transmission through vehicle-to-cloud backhaul, avoiding manual data movement and uploading, reducing the proportion of manual labor and costs, and improving the efficiency of automated processing.

[0044] (2) Set the triggering conditions of the triggering algorithm in advance according to the corresponding nodes of the autonomous driving business, and locate the key data in the autonomous driving data component according to the data screening rules built into the triggering algorithm. By designing a triggering algorithm to realize the collection of valuable data, the valuable data is screened out by the algorithm before data collection and transmission, avoiding the full data collection method and reducing the amount and cost of data collection and storage.

[0045] (3) By designing cloud-based algorithm components and sending trigger algorithms to the vehicle side, the vehicle-side deployment algorithm can be flexibly modified and optimized, the definition of valuable data can be continuously optimized, and valuable data collection can be continuously updated.

[0046] Further features and advantages of the present invention will become apparent from the following detailed description of exemplary embodiments of the present invention with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0048] Figure 1 A schematic diagram of the structure of an event-triggered automatic autonomous driving data collection system based on vehicle-to-cloud feedback provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless otherwise specifically stated.

[0050] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0051] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered as part of the specification.

[0052] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0053] The present invention proposes an embodiment of an event-triggered automatic driving data collection system based on vehicle-to-cloud feedback. Specifically, Figure 1 As shown, it includes the cloud and the vehicle, where:

[0054] The cloud includes:

[0055] Cloud algorithm component, used to design and send trigger algorithms to the vehicle side;

[0056] It should be noted that the cloud algorithm component is based on the cloud trigger algorithm design module to implement the design, development, and verification of the trigger algorithm in the cloud, and then the designed trigger algorithm is sent to the autonomous driving vehicle through the cloud trigger algorithm sending module of the cloud algorithm component; the autonomous driving vehicle receives and deploys the trigger algorithm sent from the cloud based on the trigger algorithm receiving module. Through the trigger algorithm design and sending deployment link, the trigger algorithm deployed on the autonomous driving vehicle can be optimized and updated at any time, and the definition of valuable data can be optimized to achieve continuous collection of valuable data.

[0057] The vehicle side includes an autonomous driving data component, which is used to collect a variety of data, including vehicle surrounding environment information, vehicle operating status data, and data generated during the operation of the internal algorithm of the autonomous driving system.

[0058] It should be noted that the autonomous driving data components include:

[0059] The sensor data acquisition device collects the vehicle's surrounding environment information in real time through various sensors installed on the vehicle. Specifically, the vehicle's surrounding environment information is obtained in real time by various sensors such as cameras and radars equipped on the vehicle in the form of images, distance detection values, etc.

[0060] The vehicle CAN data acquisition device obtains the vehicle operation status data exchanged between the electronic control units through the vehicle's internal CAN bus, covering vehicle speed, engine speed, gear information, and is used to reflect the overall operating conditions of the vehicle;

[0061] The autonomous driving algorithm data acquisition device is used to collect data generated during the operation of the internal algorithm of the autonomous driving system, including data generated by path planning and decision-making algorithms, and to display the system's operating logic and decision results.

[0062] The trigger algorithm component is used to receive the trigger algorithm sent from the cloud and trigger the autonomous driving data component to collect data.

[0063] It should be noted that the trigger algorithm component includes: a trigger algorithm receiving module, which is used to receive trigger algorithm information from the cloud; a trigger algorithm triggering module, which is used to determine the trigger timing according to preset trigger conditions and start the data collection process of the autonomous driving data component. The preset trigger conditions include but are not limited to entering a specific road section, driving mode matching, etc.

[0064] Specifically, the trigger algorithm triggers the autonomous driving data collection link through the trigger algorithm receiving module and the trigger algorithm triggering module: when the preset trigger condition of the trigger algorithm is met, the autonomous driving data collection instruction is triggered, and the relevant data in the autonomous driving data component defined by the trigger algorithm is collected, and the collected data packets are sent to the data storage component. Through this link, valuable data can be screened and collected, avoiding the collection of full data, and only collecting defined valuable data, reducing storage costs and data mining costs.

[0065] A data storage component that temporarily or long-term stores the data collected by the autonomous driving data component based on a trigger algorithm;

[0066] The data return component performs data exchange between the vehicle and the cloud based on the data stored by the data storage component.

[0067] It should be noted that when the data storage component receives the data packet collected by the autonomous driving, it first stores the data in the memory (not on disk) through the memory storage module of the data storage component, and uploads the data to the cloud based on the vehicle-to-cloud data upload module of the data return component; if the data return component cannot upload the data normally, the data return component sends an instruction to make the hard disk storage module of the data storage component store the data packet collected in the memory storage module to the hard disk, and retransmit the autonomous driving data packet in the hard disk storage module through the vehicle-to-cloud data retransmission module of the data return component, and delete the data in the hard disk after the retransmission is completed to leave available space. Through this data storage-return-retransmission link, it is possible to achieve reasonable utilization of resources. In general scenarios, the autonomous driving data is stored in the memory and not on disk, which reduces the use of storage resources and costs. When encountering certain special scenarios, the data can also be returned normally.

[0068] In an embodiment of the present invention, the cloud algorithm component includes: a cloud trigger algorithm design module, which is used to create, optimize and verify the trigger algorithm in the cloud; a cloud trigger algorithm sending module, which is used to send the designed trigger algorithm to the vehicle end. The accurate trigger algorithm not only guides more effective data collection, but also the collected high-quality data can in turn provide a basis for further optimization of the cloud algorithm. For example, data such as vehicle status (vehicle speed information, vehicle steering angle, steering angle and speed information output by the automatic driving system) and driver status (driver's operation information on the vehicle at this time) collected around the inconsistency of human-machine actions can be used through big data analysis technology to discover the rules and potential problems hidden in them, helping technicians to better understand the characteristics and difficulties of human-machine interaction, thereby providing a more targeted direction for the next round of trigger algorithm optimization, and continuously improving the intelligence level of the entire automatic driving system.

[0069] In an embodiment of the present invention, the cloud also includes a cloud data component, which includes: a cloud data receiving module for receiving data transmitted back by the data transmission component on the vehicle side; and a cloud data storage module for storing the data received by the cloud data receiving module in a cloud data center.

[0070] It should be noted that the data backhaul component transmits the autonomous driving data packet back to the cloud data receiving module of the cloud data component through the vehicle-cloud data backhaul-receiving-storage link. After the data reception is completed, the cloud data storage module stores the collected autonomous driving data packet in the cloud data center. Through this link, the data packets collected from various autonomous driving vehicles can be transmitted back to the cloud data center in an orderly manner, realizing the centralized aggregation of data. Regardless of the geographical location of the vehicle or the road conditions it is operating under, the autonomous driving related data generated by it are collected in a unified manner. It helps to fully understand the operating conditions of different vehicle groups and provide data resources for subsequent large-scale data analysis, system performance evaluation, and strategy formulation.

[0071] In an embodiment of the present invention, the data return component includes:

[0072] The vehicle-to-cloud data upload module is used to build a data transmission channel from the vehicle to the cloud and push the data stored in the data storage component;

[0073] The car-cloud data retransmission module is used to resend data to the cloud when data upload fails, and delete the data in the hard disk after resending to leave free space.

[0074] In an embodiment of the present invention, the data storage component includes a memory storage module and a hard disk storage module, wherein:

[0075] When the data storage component receives the collected data, it first enables the memory storage module for quick temporary storage, and simultaneously drives the vehicle-to-cloud data upload module of the data return component to push the memory data to the cloud based on the vehicle-to-cloud communication link;

[0076] When the fast temporary storage is interrupted, the data return component immediately instructs the hard disk storage module to migrate the memory data to the hard disk backup;

[0077] The car-to-cloud data retransmission module of the data return component retransmits the hard disk backed-up data. After the retransmission is completed, the successfully returned data in the hard disk is automatically deleted to free up hard disk space.

[0078] It should be noted that the data storage component gives priority to enabling the memory storage module for rapid temporary storage, taking advantage of the extremely fast memory read and write speed. At the same time, the synchronously driven car-cloud data upload module transmits memory data based on the car-cloud communication link, realizing the seamless connection between data collection and transmission operations, reducing the time interval for data waiting for storage and transmission, and speeding up the overall efficiency of the process from data generation to transmission back to the cloud. Even if there is an interruption during the rapid temporary storage process (for example, due to a sudden electronic system failure of the vehicle, a short network fluctuation that affects the reading and writing of memory data, etc.), the data return component can immediately instruct the hard disk storage module to intervene and migrate the memory data to the hard disk backup. This flexible switching mechanism ensures that data will not be lost, and can continue the subsequent data processing process after the problem is solved, maintaining the continuity of the entire data storage and transmission work. Therefore, most of the data does not need to be written to the hard disk, which reduces the number of frequent reads and writes of the hard disk, reduces the loss of the hard disk, and extends the service life of the hard disk. By automatically deleting the data that has been successfully transmitted back from the hard disk, the hard disk space is dynamically released. The hard disk can always maintain a certain amount of available space to prevent the inability to store important data due to the exhaustion of hard disk storage space. This ensures the sustainable operation of the vehicle-side storage system and can flexibly respond to various complex data storage needs.

[0079] In an embodiment of the present invention, the trigger algorithm component receives the trigger algorithm sent by the cloud and triggers the autonomous driving data component to collect data, including:

[0080] The triggering conditions of the triggering algorithm are set in advance according to the corresponding nodes of the autonomous driving business; wherein the corresponding nodes include but are not limited to complex road conditions, time period characteristics, etc.;

[0081] When the vehicle-side operating status meets the preset trigger conditions, the trigger algorithm component immediately generates an autonomous driving data collection instruction;

[0082] According to the data screening rules built into the trigger algorithm, the key data in the autonomous driving data component is located, collected and packaged into data packets and sent to the data storage component. Among them, the key data includes but is not limited to the vehicle status (vehicle speed information, vehicle steering angle, steering angle and speed information output by the autonomous driving system) and the driver status (driver's operation information on the vehicle at this time). Since only the key data defined by the trigger algorithm is collected, a large amount of redundant information caused by the full data collection is avoided, and the amount of data that needs to be stored is reduced from the source.

[0083] In an embodiment of the present invention, the data storage component has a built-in intelligent resource management program, which dynamically allocates storage resources of the memory storage module and the hard disk storage module according to real-time data flow, priority, storage duration and hardware performance.

[0084] Specifically, in one embodiment of the present invention, a trigger algorithm is designed with the inconsistency of human-machine actions in autonomous driving as the design idea, and the trigger scenario of the trigger algorithm is set as "the vehicle enters a curve, travels at 10km / h, and when the vehicle turns left on the predetermined track, the person turns right and keeps the opposite direction for 5 seconds or more". Through the trigger algorithm design and delivery deployment link, the cloud completes the trigger algorithm design, development, verification and delivery process, and the vehicle completes the algorithm reception and deployment process. The scenario design in this example aims to keenly capture the contradictions in human-machine operations, lay the foundation for subsequent accurate data collection, deeply analyze the abnormal conditions of human-machine collaboration, and help optimize and upgrade the autonomous driving system.

[0085] The vehicle in motion monitors its own operating status and the driver's operation dynamics in real time. When the system determines that the current situation accurately matches the pre-set trigger algorithm trigger conditions, the cloud trigger algorithm dispatch module immediately issues an automatic driving data collection instruction. The valuable data that needs to be collected at this time is set as the vehicle status (vehicle speed information, vehicle steering angle, steering angle and speed information output by the automatic driving system) and the driver status (driver's operation information on the vehicle at this time). The data is screened and collected through the built-in intelligent screening mechanism of the automatic driving data collection link, and only the defined valuable data is collected to avoid full data collection.

[0086] After data collection is completed, the data is packaged and the data packets are stored in the memory through the memory storage module. Then the data is transmitted back to determine if the cloud can receive the data. Otherwise, the data is stored in the hard disk (stored in the hard disk storage module). After the cloud can receive the data, the data packets stored in the hard disk are transmitted back. After the data transmission is completed, the data packets in the hard disk are deleted to leave free space.

[0087] After the cloud completes data reception, it stores the data packets in the cloud data center.

[0088] In summary, the present invention realizes the automated processing of the entire business process based on the vehicle-cloud link, without any manual processing on the vehicle side, reducing labor costs, simplifying work processes, and improving work efficiency. By triggering the autonomous driving data collection link through the trigger algorithm, valuable data can be screened and collected, avoiding the collection of full data, and only collecting defined valuable data, reducing storage costs and data mining costs. Through the data storage-return-retransmission link, the rational use of resources can be achieved. In general scenarios, the autonomous driving data is stored in the memory and not on the disk, which reduces the use and cost of storage resources. When encountering certain special scenarios, the data can also be guaranteed to be returned normally. By designing and issuing the deployment link through the trigger algorithm, the trigger algorithm deployed on the autonomous driving vehicle side can be optimized and updated at any time, and the definition of valuable data can be optimized to achieve the continuous collection of valuable data.

[0089] The above describes in detail the structure, features and effects of the present invention based on the embodiments shown in the drawings, but the above is only a preferred embodiment of the present invention. It should be noted that the technical features involved in the above embodiments and their preferred methods can be reasonably combined and matched into a variety of equivalent schemes by those skilled in the art without departing from or changing the design ideas and technical effects of the present invention; therefore, the present invention is not limited to the scope of implementation shown in the drawings, and all changes made in accordance with the concept of the present invention, or modifications to equivalent embodiments with equivalent changes, which still do not exceed the spirit covered by the specification and drawings, should be within the protection scope of the present invention.

Claims

1. An event-triggered automatic driving data collection system based on vehicle-to-cloud feedback, characterized in that: Including the cloud and the car side, including: The cloud comprises: A cloud algorithm component, used to design and send a trigger algorithm to the vehicle end; The vehicle end comprises: Autonomous driving data component, used to collect various data, A trigger algorithm component, used to receive the trigger algorithm sent by the cloud and trigger the autonomous driving data component to collect data; A data storage component temporarily or long-term stores the data collected by the autonomous driving data component based on the trigger algorithm; The data return component performs data interaction between the vehicle and the cloud based on the data stored by the data storage component.

2. The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback according to claim 1 is characterized in that: The cloud algorithm components include: Cloud trigger algorithm design module, used to create, optimize and verify trigger algorithms in the cloud; The cloud trigger algorithm sending module is used to send the designed trigger algorithm to the vehicle end.

3. The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback according to claim 2 is characterized in that: The cloud also includes a cloud data component, and the cloud data component includes: A cloud data receiving module, used to receive data sent back by the vehicle-side data return component; The cloud data storage module is used to store the data received by the cloud data receiving module in a cloud data center.

4. The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback according to claim 1, characterized in that: The autonomous driving data component includes: The sensor data acquisition device collects information about the vehicle's surrounding environment in real time through various sensors installed on the vehicle; The automobile CAN data acquisition device acquires the vehicle operation status data interacting between the electronic control units through the vehicle's internal CAN bus; The autonomous driving algorithm data acquisition device is used to collect data generated during the operation of the algorithm inside the autonomous driving system.

5. The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback according to claim 4 is characterized in that: The trigger algorithm components include: A trigger algorithm receiving module, used to receive trigger algorithm information from the cloud; The trigger algorithm trigger module is used to determine the trigger timing according to the preset trigger conditions and start the data collection process of the autonomous driving data component.

6. The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback according to claim 1, characterized in that: The data return component includes: A vehicle-to-cloud data uploading module, used to build a data transmission channel from the vehicle to the cloud, and push the data stored by the data storage component; The vehicle-to-cloud data retransmission module is used to resend data to the cloud when data upload fails, and to delete the data in the hard disk after resending to leave available space.

7. The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback according to claim 6 is characterized in that: The data storage component includes a memory storage module and a hard disk storage module, wherein: At the moment of receiving the collected data, the data storage component preferentially activates the memory storage module for quick temporary storage, and synchronously drives the vehicle-to-cloud data upload module of the data return component to push the memory data to the cloud based on the vehicle-to-cloud communication link; When the fast temporary storage is interrupted, the data return component immediately instructs the hard disk storage module to migrate the memory data to the hard disk backup; The car-to-cloud data retransmission module of the data return component retransmits the hard disk backed-up data. After the retransmission is completed, the successfully returned data in the hard disk is automatically deleted to release hard disk space.

8. The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback according to claim 1, characterized in that: The trigger algorithm component receives the trigger algorithm sent by the cloud and triggers the autonomous driving data component to collect data, including: Set the triggering conditions of the triggering algorithm in advance according to the corresponding nodes of the autonomous driving business; When the operating state of the vehicle meets the preset trigger condition, the trigger algorithm component is triggered to immediately generate an automatic driving data collection instruction; According to the data screening rules built into the trigger algorithm, key data in the autonomous driving data component is located, collected and packaged into data packets and sent to the data storage component.

9. The event-triggered automatic driving data collection system based on vehicle-to-cloud feedback according to claim 7, characterized in that: The data storage component has a built-in intelligent resource management program, which dynamically allocates storage resources of the memory storage module and the hard disk storage module according to real-time data flow, priority, storage duration and hardware performance.

Images