Data mining method and device based on automatic driving technology

Abstract

The invention provides a data mining method and device based on an automatic driving technology, and relates to the technical field of data acquisition and processing. The invention discloses a data mining method based on an automatic driving technology. The data mining method comprises the following steps: step S100, registering an automatic driving vehicle and uploading system data by an online cloud platform; and step S200, the online cloud platform sequentially queues and receives the driving data acquired by the acquisition vehicle, the test vehicle and the delivery vehicle. The data mining device based on the automatic driving technology is used for mining, analyzing and optimizing the data of the automatic driving vehicle through the data mining method based on the automatic driving technology. According to the scheme, the operation data is finally collected in real time, the task item list is automatically generated according to the received event information, the task item list is automatically recognized according to the preset system data so as to judge the compliance of the mined event data, and intelligent, efficient and self-adaptive automatic driving data mining is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data acquisition and processing, and particularly relates to a data mining method and device based on autonomous driving technology. Background Art

[0002] With the continuous development of new energy vehicles, the application of autonomous driving technology has become increasingly widespread. During the process of vehicle testing or actual application, undoubtedly many newly discovered problems and difficulties will emerge. Through continuous research and development, the defects of autonomous driving technology are overcome and continuously refined, resulting in the improvement of autonomous driving system technology. During the process of either developing and testing or actual application of autonomous driving technology, it is necessary to accurately mine and analyze the operation data of the autonomous driving system.

[0003] In the related prior art, publication number: CN116010417A; discloses an autonomous driving data mining method and system, including: defining labels for autonomous driving data to obtain data labels; scheduling all data labels at regular intervals, and calculating the values of the data labels corresponding to all autonomous driving data within a set time period; storing the values of the data labels in a columnar manner to construct an analytical database; and querying the autonomous driving data through the analytical database based on the data labels to complete the mining of autonomous driving data.

[0004] Using the above mining method, after mining the autonomous driving data, it can only be automatically stored in the server. The stored data needs to be screened, analyzed, and merged by maintenance personnel before a complete data list can be formed; it is not convenient for automatic identification and merged display of the mined data.

[0005] Therefore, it is necessary to provide a data mining method based on autonomous driving technology to solve the above technical problems. Summary of the Invention

[0006] The present invention provides a data mining method based on autonomous driving technology, which solves the problem in the related technology that it is not convenient for automatic identification and merged display of mined data.

[0007] To solve the above technical problems, the data mining method based on autonomous driving technology provided by the present invention includes the following steps: Step S100, the online cloud platform registers the autonomous driving vehicle and uploads the system data; Step S200, the online cloud platform sequentially queues and receives the driving data collected by the acquisition vehicle, the test vehicle, and the delivery vehicle; Step S300, automatically constructing a task item list for the received driving data and storing the driving data; Step S400, analyzing the driving data in each task item in sequence; Step S410, on the one hand, determine whether the snapshot event triggered by the autonomous vehicle in the operating mode corresponding to the intelligent driving level corresponds to the current driving environment requirements; if it corresponds, proceed to step S420, if not, determine that the driving data is non-compliant; Step S420, on the other hand, determine whether the real-time sensor data of the vehicle when the snapshot event is triggered corresponds to the original sensor data in the corresponding event state; if it corresponds, determine that the driving data is compliant; if not, determine that the driving data is non-compliant; Step S500, after analysis, automatically generate a compliance item list and a non-compliance item list according to the judgment result, and feedback to prompt the non-compliance items.

[0008] Preferably, the driving data includes event data and monitoring data, the event data is the response data of the corresponding sensor in the autonomous vehicle in the triggered event state; the monitoring data is the video information collected by the visual device in the autonomous vehicle in the triggered event state.

[0009] Preferably, the online cloud platform queues and receives the received data according to the priorities of the event data and the monitoring data, and preferentially receives the event data and the monitoring data with higher priorities.

[0010] Preferably, the registered content includes vehicle information and vehicle priority information; the system data includes event script information and driving system information.

[0011] Preferably, the vehicle information includes the vehicle name, vehicle identification number and license plate, which are used to distinguish the collected vehicles, test vehicles and delivered vehicles.

[0012] Preferably, the event script information provides an algorithm script for data processing and analysis for the online cloud platform based on the event snapshot model, and the event data received by the online cloud platform runs according to the event snapshot model to generate a visual display file directory.

[0013] Preferably, the task item list includes the snapshot name, event name, vehicle name, trigger location, snapshot data label and trigger event, and operation keys for information expansion and information deletion are set at the end of the task item list for information expansion display and information deletion; The task item list is automatically arranged according to the vehicle priority, and the vehicle priority is delivered vehicle > test vehicle > collected vehicle.

[0014] Preferably, the online cloud platform analyzes the event data recorded in the task item list through the event script information, and automatically determines whether the mined event information meets the preset safety range; When the result is compliant, the display interface of the task item list automatically adds a "Compliant" label; When the result is non-compliant, the display interface of the task item list automatically adds a "Non-compliant" label and pops up a reminder window.

[0015] Preferably, it further includes step S600 of updating the automatic driving rule data according to the non-compliant items, and realizing the synchronous system feedback and update of the autonomous vehicle through the online cloud platform.

[0016] The present invention also provides a data mining device based on autonomous driving technology for mining, analyzing and optimizing the data of autonomous vehicles by the data mining method based on autonomous driving technology; the data mining device based on autonomous driving technology includes: An online cloud platform and an autonomous driving system, the online cloud platform is signal-connected to the autonomous driving system; The autonomous driving system includes a data acquisition module, a data processing module, a data sending module and a system management module. The system management module is respectively signal-connected to the data acquisition module, the data processing module and the data sending module. The data acquisition module is signal-connected to the data processing module, and the data processing module is signal-connected to the data sending module; The online cloud platform includes a data receiving module, a data decompression module, a data analysis module, a safety behavior optimization module and a platform management module. The data receiving module is signal-connected to the data sending module, the data decompression module is signal-connected to the data receiving module, the data analysis module is signal-connected to the data decompression module, the data analysis module is signal-connected to the safety behavior optimization module, and the platform management module is respectively signal-connected to the data receiving module, the data decompression module, the data analysis module, the safety behavior optimization module and the system management module.

[0017] Compared with the related technology, the data mining method based on autonomous driving technology provided by the present invention has the following beneficial effects: Through the online cloud platform, it is convenient to register the collected vehicles, test vehicles and delivered vehicles. During the operation process, the operation data is collected in real time, and the collected information is mined and processed and then received. The received event information automatically generates a task item list, and the task item list is automatically identified according to the preset system data to judge the compliance of the mined event data, realizing intelligent, efficient and adaptive autonomous driving data mining and processing. Brief Description of the Drawings

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on the structures shown in these drawings.

[0019] Figure 1 It is a block diagram of a preferred embodiment of the data mining method based on autonomous driving technology provided by the present invention; Figure 2 It is a block diagram of another preferred embodiment of the data mining method based on autonomous driving technology provided by the present invention; Figure 3 It is a system block diagram of the data mining device based on autonomous driving technology provided by the present invention; Figure 4 is Figure 3 The system block diagram of the platform management module shown; Figure 5 is Figure 4 The system block diagram of the priority management unit shown; Figure 6 It is a data transmission block diagram of the data mining device based on autonomous driving technology provided by the present invention.

[0020] The realization, functional features and advantages of the object of the present invention will be further described in conjunction with the embodiments with reference to the accompanying drawings. Detailed implementation manners

[0021] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0022] The present invention provides a data mining method based on autonomous driving technology.

[0023] Please refer to Figure 1 , in an embodiment of the present invention, the data mining method based on autonomous driving technology includes the following steps: Step S100, the online cloud platform registers the autonomous driving vehicle and uploads the system data; Step S200, the online cloud platform sequentially queues and receives the driving data collected by the acquisition vehicle, the test vehicle and the delivery vehicle; Step S300, automatically construct a task item list for the received driving data and store the driving data; Step S400, analyze the driving data in each task item in sequence; Step S410, on the one hand, determine whether the snapshot event triggered by the autonomous driving vehicle in the operation mode corresponding to the intelligent driving level corresponds to the current driving environment requirements; if it corresponds, proceed to step S420, if not, determine that the driving data is non-compliant; Step S420, on the other hand, determine whether the real-time sensor data of the vehicle when the snapshot event is triggered corresponds to the original sensor data in the corresponding event state; if it corresponds, determine that the driving data is compliant; if not, determine that the driving data is non-compliant; Step S500, after analysis, automatically generate a compliance item list and a non-compliance item list according to the judgment result, and feedback to prompt the non-compliance items.

[0024] In this embodiment, the online cloud platform stores the intelligent driving level corresponding to each vehicle, and the content of the snapshot event allowed to be executed at the corresponding intelligent driving level; the snapshot event includes the event content and driving environment requirements, allowing the autonomous driving vehicle to complete the automatic execution of the event content under the conditions of the corresponding driving environment requirements, and is used for the vehicle to execute the autonomous driving system; The online cloud platform stores the original sensor data corresponding to the event, which is the original data corresponding to all sensors when the autonomous driving vehicle is in the corresponding event state, and provides original data support for subsequent data comparison when the autonomous driving vehicle executes this snapshot event.

[0025] Examples of compliance / non-compliance of snapshot events: Example 1: For example, when the autonomous driving vehicle is in an application scenario without obstacles, and the autonomous driving vehicle executes the event of constantly honking the horn, it does not meet the requirements of the snapshot event authorized to be executed in the current driving environment; then the online cloud platform can automatically analyze the collected and uploaded driving data and independently determine that the driving data is non-compliant; Example 2: For example, when the autonomous driving vehicle is in an application scenario where a U-turn is required, and the autonomous driving vehicle executes the events of gently braking the button and turning the steering wheel, it meets the requirements of the snapshot event authorized to be executed in the current driving environment; then the online cloud platform can automatically analyze the collected and uploaded driving data and independently determine that the driving data is compliant; Example 3: For example, when the autonomous driving vehicle is in an application scenario where the road section is passed but the destination has not been reached, and the autonomous driving vehicle executes the event of pulling the handbrake, it does not meet the requirements of the current driving destination; then the online cloud platform can automatically analyze the collected and uploaded driving data and independently determine that the driving data is non-compliant.

[0026] It is convenient to register the collected vehicles, test vehicles and delivered vehicles on the online cloud platform. During the operation process, the operation data is collected in real time, and the collected information is mined and processed and then received. The received event information automatically generates a task item list, and the task item list is automatically identified according to the preset system data to judge the compliance of the mined event data, realizing intelligent, efficient and adaptive autonomous driving data mining and processing.

[0027] Furthermore, the driving data includes event data and monitoring data. The event data is the response data of the corresponding sensors in the autonomous driving vehicle under the trigger event state; the monitoring data is the video information collected by the visual device in the autonomous driving vehicle under the trigger event state.

[0028] By integrating and transmitting the collected driving data to the storage server of the online cloud platform; The event data is the vehicle data, event ID data and event location data when the autonomous driving vehicle triggers a preset event; and the data received and stored is processed by the online cloud platform to generate a snapshot data directory, so that the snapshot data in the storage server can be directly retrieved through the online cloud platform.

[0029] During the operation of the autonomous driving system, the event data is transmitted to the online cloud platform through the wireless network, so that the event data is integrated and stored in the storage server of the online cloud platform after being received; During the operation of the autonomous driving system, the monitoring data is transmitted to the online cloud platform through the wireless network, so that the monitoring data is integrated and stored in the storage server of the online cloud platform after being received.

[0030] The stored monitoring data has fuzzy search by packet name, vehicle search, date search (retrieving the time range, defaulting to querying yesterday's data and today's data), and after search, it is sorted in descending order according to the start upload time of the data packet; When the monitoring data is transmitted, segmented uploading is adopted to reduce the size of each uploaded data packet, and the monitoring data of each segment is uploaded in turn; for the monitoring data during the emergency event time, it is arranged in the priority upload queue.

[0031] Automatically decompress, automatically identify and automatically combine the monitoring data transmitted back by the same autonomous driving vehicle to generate a complete monitoring data set of the same vehicle.

[0032] Furthermore, the online cloud platform queues and receives the received data according to the priorities of the event data and the monitoring data, and preferentially receives the event data and the monitoring data with higher priorities.

[0033] When faced with the need to transmit multi-batch data, the online cloud platform automatically queues and plans multi-batch tasks, analyzes the priorities of multi-batch tasks, and preferentially transmits the data with the highest priority.

[0034] When there is a data A in the transmission state and a data B with a higher priority appears, data A pauses and data B is preferentially transmitted. After data B is transmitted, data A is transmitted again.

[0035] To facilitate the automatic identification, automatic switching, automatic arrangement, and automatic start and stop of preferential data transmission.

[0036] The snapshot events are as follows: ① Event: Long press the horn; Event ID: 0x10001; Priority: 1; Collected data: All sensor raw data; ② Event: Light brake button; Event ID: 0x10002; Priority: 2; Collected data: All sensor raw data; ③ Event: Pull the handbrake; Event ID: 0x10003; Priority: 5; Collected data: All sensor raw data; ④ Event: Emergency stop button; Event ID: 0x10004; Priority: 4; Collected data: All sensor raw data; ⑤ Event: Turn the steering wheel; Event ID: 0x10005; Priority: 3; Collected data: All sensor raw data; ⑥ Event: Node disconnection emergency stop; Event ID: 0x20001; Priority: 6; Collected data: All sensor raw data; ⑦ Event: Abnormal traffic light color; Event ID: 0x30001; Priority: 7; Collected data: All sensor raw data.

[0037] Furthermore, the registered content includes vehicle information and vehicle priority information; the system data includes event script information and driving system information.

[0038] Provide data support and algorithm support for the intelligent operation of the online cloud platform.

[0039] Specifically, the vehicle information includes the vehicle name, vehicle identification number, and license plate, which are used to distinguish collection vehicles, test vehicles, and delivery vehicles.

[0040] The vehicle priority information includes a vehicle priority list and an event priority list, which are used to prioritize the vehicles and events for uploading data.

[0041] Specifically, the event script information provides an algorithm script for data processing and analysis for the online cloud platform based on the event snapshot model. The event data received by the online cloud platform runs according to the event snapshot model to generate a visual display file directory. This facilitates the automatic processing of event snapshot data, generates a corresponding task item list, and arranges them automatically after storage to support the display of data.

[0042] Specifically, the task item list includes snapshot name, event name, vehicle name, trigger location, snapshot data label, and trigger event. Operation keys for information expansion and information deletion are set at the end of the task item list for information expansion display and information deletion. The task item list is automatically arranged according to vehicle priority, where the vehicle priority is delivery vehicle > test vehicle > acquisition vehicle.

[0043] This facilitates the detailed display of event snapshot data for mining, providing support for subsequent data retrieval and review.

[0044] Furthermore, the online cloud platform analyzes the event data recorded in the task item list through the event script information to automatically determine whether the mined event information meets the preset safety range. When the result is compliant, a "compliant" label is automatically added to the display interface of the task item list. When the result is non - compliant, a "non - compliant" label is automatically added to the display interface of the task item list, and a pop - up reminder is given.

[0045] This facilitates the automatic analysis of the mined event data, and automatically reminds the background management personnel according to the analysis result.

[0046] Please refer to Figure 2 , furthermore, a data mining method based on autonomous driving technology further includes step S600, which updates the autonomous driving driving rule data according to non - compliant items, and realizes the synchronous system feedback and update of autonomous driving vehicles through the online cloud platform.

[0047] The online cloud platform is equipped with an artificial intelligence system. The artificial intelligence system combines the recognition result with big data information to automatically search for countermeasures and solutions, and updates the system according to the existing solutions. After the system update, it is allocated to the autonomous driving system to realize the automatic update and maintenance of the autonomous driving system after an abnormal event occurs.

[0048] It realizes the intelligent recognition and processing of data mining. After processing, it can automatically optimize the execution algorithm of the autonomous driving system to automatically solve the problems in non - compliant items.

[0049] When non-compliance items cannot be automatically processed and resolved by the online cloud platform, the analyzed non-compliance items are sent to the back-end management personnel, and manual updates to the autonomous driving system are required by the back-end R & D management personnel.

[0050] The present invention also provides a data mining device based on autonomous driving technology.

[0051] Please refer to Figure 3 , the data mining device based on autonomous driving technology is used to mine, analyze and optimize the data of autonomous driving vehicles by the data mining method based on autonomous driving technology; the data mining device based on autonomous driving technology includes: An online cloud platform and an autonomous driving system, the online cloud platform is signal-connected to the autonomous driving system; The autonomous driving system includes a data acquisition module, a data processing module, a data sending module and a system management module. The system management module is respectively signal-connected to the data acquisition module, the data processing module and the data sending module. The data acquisition module is signal-connected to the data processing module, and the data processing module is signal-connected to the data sending module; The online cloud platform includes a data receiving module, a data decompression module, a data analysis module, a safety behavior optimization module and a platform management module. The data receiving module is signal-connected to the data sending module, the data decompression module is signal-connected to the data receiving module, the data analysis module is signal-connected to the data decompression module, the data analysis module is signal-connected to the safety behavior optimization module, and the platform management module is respectively signal-connected to the data receiving module, the data decompression module, the data analysis module, the safety behavior optimization module and the system management module.

[0052] The data acquisition module integrates vehicle sensors, a V2X communication unit, and an environment perception module; supports multi-modal data acquisition and preprocessing; is used to obtain vehicle sensor data (cameras, radars, IMUs, etc.), V2X traffic signals, weather data, and high-precision map information; through spatio-temporal alignment technology, unify multi-source data to the same spatio-temporal coordinate system.

[0053] The data processing module integrates an edge computing engine, an intelligent feedback controller and a redundancy and fault tolerance module; Edge-end real-time status detection: Deploy a lightweight vehicle status detection model in the in-vehicle edge computing unit. The input of the model is multi-source fusion data, and the output is the predicted vehicle status; Dynamic weight adaptive mechanism: Dynamically adjust the weights of online and offline results according to the real-time scene complexity (such as traffic flow, weather conditions); Low-complexity scenario: High online weight ratio (80%), fast response; High-complexity scenarios (such as heavy rain and congestion): Increase the weight of offline historical data (60%) to enhance robustness.

[0054] Implement online incremental learning at the edge: If the difference between the online and offline results exceeds the threshold, trigger fine-tuning of the model parameters; the updated model parameters are encrypted and synchronized to the cloud for global model optimization.

[0055] The dynamic weight adjustment includes: Judge the current environmental complexity through a scene classifier, and allocate online weights and offline weights to satisfy: Online weight value + offline weight value = 1; In high-complexity scenarios, the offline weight value is increased to 0.6 - 0.8; in low-complexity scenarios, the online weight value is increased to 0.7 - 0.9.

[0056] Redundant design of multiple sensors to support fault switching; in abnormal scenarios, trigger dual backups of emergency backhaul and local storage.

[0057] Collect multi-source data, including vehicle sensor data, V2X traffic signals, and environmental information; perform multi-source data fusion and real-time status detection at the edge computing node, and output the predicted vehicle status; dynamically adjust the weights of online detection results and offline historical data according to the real-time scene complexity; update the local model based on incremental learning, and intelligently determine the data backhaul priority and compression strategy.

[0058] Reduce the state detection error caused by multi-source data fusion; the edge computing compresses and reduces the decision-making delay; the intelligent backhaul strategy reduces the redundant data transmission volume; improve the determination accuracy of the dynamic weight mechanism in complex scenarios; the redundant design ensures the continuous operation of the system in case of sensor failures; finally, through edge intelligence and multi-modal collaboration, efficient, robust, and adaptive autonomous driving data mining is achieved.

[0059] Please refer to Figure 4 , the platform management unit includes a vehicle registration unit, an upload management unit, a system image management unit, and a priority management unit; The vehicle registration unit is used to register the vehicle information that needs data mining in advance on the online cloud platform, and one vehicle corresponds to one vehicle identification number; The upload management unit is used to manage the registered vehicle information, facilitating addition, modification, deletion, etc. Once the registration is successful, the vehicle identification number of the vehicle cannot be modified; The system image management unit is used to upload the driving system image data of the autonomous driving system, providing data support for the update of the autonomous driving system; it stores at least two versions of the driving system image data, and when an abnormality occurs in the current version, it can provide a stable version to support the continuous operation of the autonomous driving system. Please refer to Figure 5 , the priority management unit includes vehicle intelligence classification, data priority upload classification, and mirror data classification. The vehicle intelligence classification is used to register vehicle priority data; the data priority upload classification is used to register event priority data; the mirror data classification is used to register mirror system priority data, providing stable data support for the classification logic.

[0060] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention. Any equivalent structural transformation made under the concept of the present invention using the content of the specification and drawings of the present invention, or direct / indirect application in other related technical fields, is included in the patent protection scope of the present invention.

Claims

1. A data mining method based on autonomous driving technology, characterized in that: The steps include: Step S100, the online cloud platform registers the autonomous driving vehicle and uploads system data; Step S200, the online cloud platform queues and receives the driving data collected by the collection vehicle, the test vehicle and the delivery vehicle in sequence; Step S300, automatically constructing a task item list from the received driving data and storing the driving data; Step S400, analyzing the driving data in each task item in turn; Step S410, on the one hand, determines whether the triggered snapshot event of the autonomous driving vehicle corresponds to the current driving environment requirements in the operation mode corresponding to the intelligent driving level; if so, proceeds to step S420; if not, determines that the driving data is non-compliant; Step S420, on the other hand, determines whether the real-time sensor data of the vehicle when the snapshot event is triggered corresponds to the original sensor data under the corresponding event state; if so, determines that the driving data is compliant; If they do not correspond, the driving data is determined to be non-compliant; Step S500: After analysis, a compliance project list and a non-compliance project list are automatically generated according to the judgment result, and feedback is given to the non-compliance projects.

2. The data mining method based on autonomous driving technology according to claim 1, characterized in that: The driving data includes event data and monitoring data, wherein the event data is the response data of the corresponding sensor in the autonomous driving vehicle under the triggering event state; The monitoring data is video information collected by the visual equipment in the autonomous driving vehicle under the trigger event state.

3. The data mining method based on autonomous driving technology according to claim 2, characterized in that: The online cloud platform queues the received data according to the priorities of the event data and the monitoring data, and gives priority to receiving the event data and monitoring data with high priority.

4. The data mining method based on autonomous driving technology according to claim 3, characterized in that: The registered content includes vehicle information and vehicle priority information; the system data includes event script information and driving system information.

5. The data mining method based on autonomous driving technology according to claim 4, characterized in that: The vehicle information includes vehicle name, frame number and license plate, which are used to distinguish collection vehicles, test vehicles and delivery vehicles.

6. The data mining method based on autonomous driving technology according to claim 5, characterized in that: The event script information provides the online cloud platform with an algorithm script for data processing and analysis based on the event snapshot model. The event data received by the online cloud platform is run according to the event snapshot model to generate a visual display file directory.

7. The data mining method based on autonomous driving technology according to claim 6, characterized in that: The task item list includes a snapshot name, an event name, a vehicle name, a trigger location, a snapshot data tag, and a trigger event. At the end of the task item list, operation keys for information expansion and information deletion are provided for information expansion display and information deletion; The task item list is automatically arranged according to vehicle priority, and the vehicle priority is delivery vehicle>test vehicle>collection vehicle.

8. The data mining method based on autonomous driving technology according to claim 7, characterized in that: The online cloud platform analyzes the event data recorded in the task item list through the event script information, and automatically determines whether the mined event information meets the preset safety range; When the result is compliant, the display interface of the task item list automatically adds a "compliant" label; When the result is non-compliant, the display interface of the task item list automatically adds a "non-compliant" label and pops up a reminder.

9. The data mining method based on autonomous driving technology according to claim 8, characterized in that: It also includes step S600, updating the autonomous driving driving rule data according to the non-compliant items, and realizing synchronous system feedback and update of the autonomous driving vehicle through the online cloud platform.

10. A data mining device based on autonomous driving technology, used for mining, analyzing and optimizing data of an autonomous driving vehicle according to the data mining method based on autonomous driving technology as claimed in any one of claims 1 to 9; characterized in that: The data mining device based on the autonomous driving technology includes: An online cloud platform and an autonomous driving system, wherein the online cloud platform is signal-connected to the autonomous driving system; The automatic driving system includes a data acquisition module, a data processing module, a data sending module and a system management module, wherein the system management module is signal-connected to the data acquisition module, the data processing module and the data sending module respectively, the data acquisition module is signal-connected to the data processing module, and the data processing module is signal-connected to the data sending module; The online cloud platform includes a data receiving module, a data decompression module, a data analysis module, a safety behavior optimization module and a platform management module. The data receiving module is signal-connected to the data sending module, the data decompression module is signal-connected to the data receiving module, the data analysis module is signal-connected to the data decompression module, the data analysis module is signal-connected to the safety behavior optimization module, and the platform management module is signal-connected to the data receiving module, the data decompression module, the data analysis module, the safety behavior optimization module and the system management module respectively.

Images