Extreme scene data acquisition method and device, intelligent vehicle, and readable storage medium

Abstract

The application provides an extreme scene data collection method and device, an intelligent vehicle and a readable storage medium. The collection method comprises the following steps: judging whether an intelligent driving system is activated; when the intelligent driving system is activated, collecting extreme scene data in response to a user instruction. The application solves the problem that the technical solution in the related art cannot collect data of an extreme scene based on the behavior of a user when the vehicle is intelligently driven.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of air conditioning, in particular to an extreme scene data collection method and device, an intelligent vehicle and a readable storage medium. BACKGROUND

[0002] The intelligent driving technology task has been recognized as a task of solving the long-tail effect of data distribution, and the corner case (extreme situation) with a lower occurrence probability is one of the sources of providing data-driven algorithm models for upgrading. The collection of extreme scenes is the key to realizing data closed loop. There are mainly two ways to collect and upload scenes in the current technology:

[0003] 1) During the development process, testers trigger the scenes that need to be collected according to human judgment, record and upload; this method collects limited extreme scene data, and cannot well collect extensive and diverse extreme scene data;

[0004] 2) Collecting extreme scenes on a mass-produced vehicle, which has two types of triggering methods:

[0005] a) During the activation process of automatic driving, the scene data is triggered when the perception results are inconsistent, such as when the results detected by a sensor and the results detected by another sensor are inconsistent, exceeding a certain threshold value will automatically trigger the recording and uploading of scene data;

[0006] b) During the non-activation process of automatic driving, the system behavior is inconsistent with the driver behavior of the ego vehicle, such as when the driver drives the vehicle during driving, the system runs in real time in the back end, but does not control the vehicle, and when the driver's driving behavior is inconsistent with the system's decision planning behavior in a certain driving scene, the scene data will be automatically triggered to record and upload;

[0007] Collecting extreme scenes on a mass-produced vehicle has more extensive scene data and collection sources than method 1, but relying only on the inconsistency of the ego vehicle's perception results during the activation process and the inconsistency of the driver's driving behavior during the non-activation process to trigger scene data collection, many valuable extreme scene data are not collected, such as collecting only according to the inconsistency of the perception results of different sensors during the activation process, which cannot consider the driver's driving behavior to determine whether an extreme scene occurs. If the driver's driving behavior is considered to determine whether an extreme scene occurs at the same time as the activation of automatic driving, it will greatly supplement the data source of the extreme scene, and train and improve the algorithm model through a larger amount of extreme scene data.

[0008] Therefore, the problem in the related art is that the technical solution in the related art cannot collect extreme scene data based on the behavior of the user when the vehicle is intelligently driven. SUMMARY

[0009] The problem solved by the present application is that the technical solutions in the related art cannot collect extreme scene data based on user behavior when the vehicle is intelligently driven.

[0010] To solve the above problems, the first object of the present application is to provide an extreme scene data collection method based on user behavior.

[0011] The second object of the present application is to provide an extreme scene data collection device based on user behavior.

[0012] The third object of the present application is to provide an intelligent vehicle.

[0013] The fourth object of the present application is to provide a readable storage medium.

[0014] To achieve the first object of the present application, an embodiment of the present application provides an extreme scene data collection method based on user behavior, comprising: judging whether an intelligent driving system is activated; when the intelligent driving system is activated, collecting extreme scene data in response to a user instruction.

[0015] The effect that can be achieved by the present embodiment is that the technical solutions in the related art cannot trigger the collection of extreme scene data according to the user behavior in the vehicle when the intelligent driving system is activated during the collection of extreme scene data on mass-produced vehicles, resulting in that many very valuable extreme scene data are not collected. However, the extreme scene data collection method based on user behavior in the present embodiment can trigger the collection of extreme scene data according to the user instruction in the vehicle when the intelligent driving system is activated, greatly supplementing the amount and quality of extreme scene data, so that the algorithm model for providing data driving for the intelligent driving system can be trained and improved through a larger amount of extreme scene data.

[0016] In an embodiment of the present application, the instruction includes at least one of the following: a first instruction, the first instruction including a user's takeover driving behavior, the user transmitting the first instruction to the intelligent driving system through control of the ego vehicle; and a second instruction, the second instruction including the user's expected judgment of the ego vehicle driving behavior, the user transmitting the second instruction to the intelligent driving system through a button or voice.

[0017] The effect that can be achieved by the present embodiment is that the instruction in the present embodiment includes at least one of the first instruction and the second instruction, that is, when either of the two cases of the first instruction and the second instruction occurs, the collection of extreme scene data by the intelligent driving system can be triggered, effectively increasing the amount of extreme scene data collected.

[0018] In one embodiment of the present application, responding to the user instruction comprises: within a first preset time after the intelligent driving system receives the instruction, accepting at least one instruction of the user; when the instruction only includes the first instruction within the first preset time, responding to the first instruction; when the instruction only includes the second instruction within the first preset time, responding to the second instruction; and when the instruction includes the first instruction and the second instruction within the first preset time, responding to the first instruction.

[0019] The effect achieved by the present embodiment is that the first instruction includes the user's takeover driving behavior, the second instruction includes the user's judgment on the driving behavior of the ego vehicle, and the priority of the first instruction is higher than that of the second instruction. In actual driving, the importance of the user's actual takeover driving behavior is obviously higher than the user's judgment on the driving behavior of the ego vehicle. Such a setting can improve the efficiency of extreme scene data collection, that is, the collected extreme scene data is associated with more important instructions, which can help subsequent data analysis and algorithm model upgrading to provide higher quality data.

[0020] In one embodiment of the present application, after collecting the extreme scene data, the collection method comprises: obtaining the predicted driving behavior of the vehicle according to the actual driving behavior of the vehicle; judging whether the extreme scene data needs to be uploaded according to the actual driving behavior, the predicted driving behavior and the instruction; when the instruction is the same as the actual driving behavior or the predicted driving behavior, the extreme scene data does not need to be uploaded; when the instruction is different from any one of the actual driving behavior and the predicted driving behavior, the extreme scene data needs to be uploaded; and feeding back the judgment result to the user.

[0021] The effect achieved by the present embodiment is that the scheme of the present embodiment filters and uploads valuable extreme scene data, effectively improving the quality of extreme scene data and providing high-quality data sources for subsequent product services to diversify experiences. The feedback mechanism improves the intelligent interaction experience between the user and the intelligent driving system.

[0022] In one embodiment of the present application, before collecting the extreme scene data, the collection method comprises: real-time sensing the surrounding environment of the vehicle, and caching the extreme scene data of the surrounding environment to the memory.

[0023] The effect achieved by the present embodiment is that the scheme of the present embodiment can record the surrounding environment of the vehicle completely, and then extract the data in the memory when it is necessary to collect the extreme scene data. Recording the surrounding environment data in the memory can effectively save the memory space and avoid the situation that too much surrounding environment data occupies the memory, causing the intelligent driving system to fail to operate normally.

[0024] In one embodiment of the present application, collecting extreme scene data comprises: recording a time t when extreme scene data needs to be collected; and transferring extreme scene data within a second preset time before and a third preset time after the time t from the memory to the storage.

[0025] The embodiment can achieve the following effects: when extreme scene data needs to be collected, some causes of the extreme scene have appeared in the surrounding environment of the vehicle at a time before the time when the extreme scene occurs, and the data collection method of tracing back a period of time before and after can increase the completeness of the extreme scene data collection, effectively improve the quality of the collected extreme scene data, and provide more valuable scene data for subsequent perception algorithm and prediction algorithm model optimization. The memory stores complete surrounding environment data of the vehicle in real time, and can perform complete data collection on the extreme scene that needs to be collected.

[0026] In one embodiment of the present application, after collecting extreme scene data, the collection method comprises: uploading the extreme scene data from the storage to the cloud when the ego vehicle is turned off or the intelligent driving system is not activated.

[0027] The embodiment can achieve the following effects: data storage needs to occupy a large memory, and timely uploading the collected extreme scene data to the cloud can avoid the situation that the memory of the vehicle is full, and the intelligent driving system of the vehicle cannot normally operate. At the same time, since a large network bandwidth is needed when uploading, and a large network bandwidth is also needed when the intelligent driving system is running, in the embodiment, the extreme scene data is uploaded from the storage to the cloud when the ego vehicle is turned off or the intelligent driving system is not activated, which can avoid occupying too much network bandwidth of the vehicle when uploading the extreme scene data, and affect the normal use of the intelligent driving system of the vehicle.

[0028] To achieve the second object of the present application, an embodiment of the present application provides an extreme scene data collection device based on user behavior, which comprises: a judgment module, the judgment module being configured to judge whether the intelligent driving system is activated; and a collection module, the collection module being configured to collect extreme scene data in response to a user instruction when the intelligent driving system is activated.

[0029] The extreme scene data collection device based on user behavior of the embodiment of the present application implements the steps of the extreme scene data collection method based on user behavior of any embodiment of the present application, and thus has all the beneficial effects of the extreme scene data collection method based on user behavior of any embodiment of the present application, which will not be described herein again.

[0030] To achieve the third object of the present application, the embodiments of the present application provide a smart vehicle, comprising a processor, a memory, and a program or instructions stored in the memory and executable on the processor, the program or instructions being executed by the processor to implement the steps of the user behavior based extreme scenario data collection method according to any of the embodiments of the present application.

[0031] The smart vehicle according to the embodiments of the present application implements the steps of the user behavior based extreme scenario data collection method according to any of the embodiments of the present application, and thus has all the beneficial effects of the user behavior based extreme scenario data collection method according to any of the embodiments of the present application, which will not be repeated here.

[0032] To achieve the fourth object of the present application, the embodiments of the present application provide a readable storage medium, the readable storage medium storing a program or instructions, the program or instructions being executed by a processor to implement the steps of the user behavior based extreme scenario data collection method according to any of the embodiments of the present application.

[0033] The readable storage medium according to the embodiments of the present application implements the steps of the user behavior based extreme scenario data collection method according to any of the embodiments of the present application, and thus has all the beneficial effects of the user behavior based extreme scenario data collection method according to any of the embodiments of the present application, which will not be repeated here. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 One of the step flowcharts of the user behavior based extreme scenario data collection method according to some embodiments of the present application;

[0035] Figure 2 The second of the step flowcharts of the user behavior based extreme scenario data collection method according to some embodiments of the present application. DETAILED DESCRIPTION

[0036] To make the above objects, features and advantages of the present application more obvious and understandable, specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0037]

First Embodiment

[0038] Referring to Figure 1 , the present embodiment provides a user behavior based extreme scenario data collection method, comprising:

[0039] S100: judging whether a smart driving system is activated;

[0040] S200: when the smart driving system is activated, collecting extreme scenario data in response to a user instruction.

[0041] In this embodiment, the extreme scenario refers to some special situations that may occur during the driving of the vehicle. Exemplarily, the driving behavior of the vehicle during driving is sudden acceleration, sudden deceleration or sudden turning, the road condition suddenly appears an obstacle or a large pit, the sensing and recognition system of the intelligent driving system of the vehicle cannot recognize the road condition, etc. The extreme scenario data refers to the environmental data detected by the vehicle in the extreme scenario, the driving data of the vehicle, etc. A large amount of extreme scenario data collection can provide a large amount of data for the upgrade of the data-driven algorithm model in the intelligent driving system.

[0042] It should be noted that the vehicle containing the intelligent driving system can maintain an activated state or an inactivated state during driving. The user behavior-based extreme scenario data collection method of the present embodiment is applied when the intelligent driving system is in the activated state. Only when the intelligent driving system of the vehicle is in the activated state, the user behavior-based extreme scenario data collection method of the present embodiment can be applied. When the intelligent driving system of the vehicle is in the activated state, the user does not need to perform driving operation, and the intelligent driving system of the vehicle automatically controls the vehicle to drive on the road.

[0043] In this embodiment, when the intelligent driving system is activated, the intelligent driving system collects extreme scenario data in response to the user's instruction. The method of conveying the user's instruction to the intelligent driving system includes but is not limited to voice triggering, key triggering and taking over the driving behavior of the vehicle, etc.

[0044] It can be understood that the technical solution in the related art cannot trigger the data collection of the extreme scenario according to the user behavior in the vehicle when the intelligent driving system is activated during the data collection of the extreme scenario on the mass production vehicle, resulting in that a lot of very valuable extreme scenario data is not collected. The user behavior-based extreme scenario data collection method of the present embodiment can trigger the data collection of the extreme scenario according to the user instruction in the vehicle when the intelligent driving system is activated, greatly supplementing the data quantity and quality of the extreme scenario, so that the data-driven algorithm model for the intelligent driving system can be trained and improved by a larger scale of extreme scenario data.

[0045]

Second embodiment

[0046] In one specific embodiment, the instruction includes at least one of the following: a first instruction, the first instruction includes a user's driving behavior of taking over, the user transmits the first instruction to the intelligent driving system through the control of the ego vehicle; a second instruction, the second instruction includes the user's expected judgment of the ego vehicle driving behavior, the user transmits the second instruction to the intelligent driving system through a button or voice.

[0047] In the embodiment, the instructions include the first instructions and the second instructions, when the intelligent driving system is activated, the user instructions include at least one of the first instructions and the second instructions, that is, the instructions conveyed by the user to the intelligent driving system can include the first instructions, can include the second instructions, or can include both the first instructions and the second instructions.

[0048] It should be noted that the first instructions include the user's takeover driving behavior, that is, when the intelligent driving system is activated, the vehicle automatically drives on the road, when the driving behavior of the intelligent driving system does not conform to the user's expectation at a certain moment, the user can take over the driving of the vehicle, at this time, the driving operation of the user to the vehicle is the first instructions, the control behavior of the user to the vehicle is recognized by the intelligent driving system, the intelligent driving system of the vehicle responds to the first instructions of the user, and collects the extreme scene data. The second instructions include the user's expected judgment on the driving behavior of the vehicle, that is, when the intelligent driving system is activated, the vehicle automatically drives on the road, when the driving behavior of the intelligent driving system does not conform to the user's expectation at a certain moment, the user transmits the correct driving behavior considered by the user to the intelligent driving system in the form of a button or voice, at this time, the information transmitted by the user to the intelligent driving system is the second instructions, at this time, the vehicle still automatically drives under the control of the intelligent driving system, the second instructions transmitted by the user to the intelligent driving system are recognized by the intelligent driving system, and the intelligent driving system responds to the second instructions of the user to collect the extreme scene data.

[0049] It should be noted that the vehicle in the embodiment refers to the vehicle driven by the user.

[0050] It can be understood that the instructions in the embodiment include at least one of the first instructions and the second instructions, that is, when any one of the two cases of the first instructions and the second instructions occurs, the intelligent driving system can be triggered to collect the extreme scene data, and the amount of data collected in the extreme scene is effectively increased.

[0051]

Third Embodiment

[0052] In a specific embodiment, responding to the user instructions includes:

[0053] S210: accepting at least one instruction of the user within a first preset time when the intelligent driving system receives the instructions;

[0054] S220: when the instructions only include the first instructions within the first preset time, responding to the first instructions;

[0055] S230: when the instructions only include the second instructions within the first preset time, responding to the second instructions;

[0056] S240: when the instruction includes the first instruction and the second instruction within the first preset time, responding to the first instruction.

[0057] In the embodiment, in the state that the intelligent driving system is activated, the vehicle is in the state of automatic driving, when the actual driving behavior of the vehicle is inconsistent with the expected driving behavior of the user, the user can feed back the intelligent driving system through the first instruction or the second instruction, when the user only issues the first instruction within the first preset time, the intelligent driving system only responds to the first instruction; when the user only issues the second instruction within the first preset time, the intelligent driving system only responds to the second instruction; when the user issues the first instruction and the second instruction within the first preset time, the intelligent driving system only responds to the first instruction, that is, the priority of the first instruction is higher than that of the second instruction.

[0058] Exemplarily, when the actual driving behavior of the vehicle is inconsistent with the expected driving behavior of the user, the user issues the second instruction when the vehicle is intelligently driving, the user takes over the driving within the first preset time after issuing the second instruction, that is, issues the first instruction, at this time, the intelligent driving system of the vehicle only responds to the first instruction of the user, and collects the extreme scene data, that is, the collected extreme scene is associated with the first instruction of the user.

[0059] It should be noted that the first preset time is set to increase the efficiency of the intelligent driving system in processing instructions. That is, after the user issues the first instruction, the intelligent driving system responds to the first instruction and collects the extreme scene data at the time when the user issues the first instruction, after a period of time, the second instruction is issued, if the period of time exceeds the first preset time, the intelligent driving system will respond to the second instruction issued by the user at this time, and collect the extreme scene data again. If there is no first preset time, the intelligent driving system will ignore the second instruction issued later and prefer to respond to the first instruction, resulting in the loss of part of the valuable extreme scene data.

[0060] It can be understood that the first instruction includes the user's takeover driving behavior, the second instruction includes the user's judgment on the driving behavior of the vehicle, and the priority of the first instruction is higher than that of the second instruction. In actual driving, the importance of the user's actual takeover driving behavior is obviously higher than that of the user's judgment on the driving behavior of the vehicle, and such setting can improve the efficiency of extreme scene data collection, that is, the collected extreme scene data is associated with more important instructions, which can help subsequent data analysis and algorithm model upgrading to provide higher quality data.

[0061]

Fourth embodiment

[0062] In a specific embodiment, after collecting the extreme scene data, the collection method comprises:

[0063] S300: obtaining a predicted driving behavior of the vehicle according to an actual driving behavior of the vehicle;

[0064] S310: judging whether the extreme scene data needs to be uploaded according to the actual driving behavior, the predicted driving behavior and the instruction;

[0065] S320: when the instruction is the same as the actual driving behavior or the predicted driving behavior, the extreme scene data does not need to be uploaded;

[0066] S330: when the instruction is different from any one of the actual driving behavior and the predicted driving behavior, the extreme scene data needs to be uploaded;

[0067] S340: feeding back the judgment result to the user.

[0068] In the embodiment, before collecting the extreme scene data, the intelligent driving system of the vehicle can perceive the surrounding environment of the vehicle in real time. The surrounding environment information detected by the vehicle includes road information, static obstacle information and dynamic target information, etc. The intelligent driving system can calculate the optimal driving behavior of the vehicle according to the road environment where the vehicle is located, and take the optimal driving behavior as the actual driving behavior of the vehicle.

[0069] It should be noted that, after responding to the instruction of the user, the extreme scene data needs to be collected. At this time, the intelligent driving system can calculate the predicted driving behavior of the vehicle according to the actual driving behavior of the vehicle. The predicted driving behavior refers to other driving behaviors that can be adapted to the current road environment in addition to the actual driving behavior of the vehicle. After obtaining the predicted driving behavior of the vehicle, the intelligent driving system can compare and judge the actual driving behavior, the predicted driving behavior of the vehicle and the instruction of the user. When the instruction of the user is the same as one of the actual driving behavior and the predicted driving behavior of the vehicle, it indicates that the intelligent driving system of the vehicle has calculated the driving behavior indicated in the instruction of the user. At this time, the extreme scene data does not have training value and does not need to be uploaded. When the instruction of the user is different from any one of the actual driving behavior and the predicted driving behavior of the vehicle, it indicates that the intelligent driving system of the vehicle has not calculated the driving behavior indicated in the instruction of the user. At this time, the extreme scene data has training value and needs to be uploaded. After the judgment is completed, the judgment result is fed back to the user.

[0070] It should be noted that the feedback mode includes but is not limited to the mode of feeding back through voice prompt or displaying information on the display screen.

[0071] It should be noted that the actual driving behavior of the vehicle is only one, and the predicted driving behavior of the vehicle can be multiple or none. Illustratively, when the vehicle detects that there is an obstacle in front of the vehicle under the current road condition, the intelligent driving system of the vehicle calculates and obtains that the optimal way is to brake, at this time the vehicle makes the driving behavior of braking, which is the actual driving behavior of the vehicle; after collecting extreme scene data in response to the user instruction, the intelligent driving system obtains the predicted driving behavior of the vehicle according to the actual driving behavior of the vehicle, at this time the predicted driving behavior of the vehicle can include two predicted driving behaviors of left lane changing and right lane changing, when the user instruction is right lane changing, it indicates that the intelligent driving system has calculated the driving behavior of right lane changing, so the extreme scene data does not need to be uploaded; when the user instruction is to accelerate, it indicates that the intelligent driving system has not calculated this driving behavior, so the extreme scene data needs to be uploaded.

[0072] Understandably, the scheme of the embodiment filters and uploads valuable extreme scene data, effectively improving the quality of the extreme scene data, and providing a high-quality data source for subsequent product service diversification experience. The feedback mechanism improves the intelligent interaction experience between the user and the intelligent driving system.

[0073]

Fifth Embodiment

[0074] Referring to Figure 2 In one specific embodiment, before collecting the extreme scene data, the collection method includes:

[0075] S110: Real-time perception of the surrounding environment of the vehicle, and buffering the extreme scene data of the surrounding environment to the memory.

[0076] In this embodiment, S110 includes: real-time perception of the surrounding environment of the vehicle, that is, the vehicle continuously records its surrounding environment, which includes road information, static obstacle information and dynamic target information.

[0077] Understandably, the scheme of the embodiment can record the surrounding environment of the vehicle completely, and then extract the data in the memory when it is necessary to collect the extreme scene data. Recording the surrounding environment data in the memory can effectively save the storage space, and avoid the situation that too much surrounding environment data occupies the memory, causing the intelligent driving system to be unable to operate normally.

[0078]

Sixth Embodiment

[0079] Referring to Figure 2 In one specific embodiment, collecting the extreme scene data includes:

[0080] S250: Record the time t when the extreme scene data needs to be collected;

[0081] S260: Moving the extreme scene data within the first preset time before time t and the third preset time after time t from the memory to the storage.

[0082] Preferably, the second preset time is 30s, and the third preset time is 10s.

[0083] Illustratively, record the time t when extreme scene data needs to be collected, and move the extreme scene data recorded by the vehicle within 30s before time t and 10s after time t from the memory to the storage.

[0084] It can be understood that when extreme scene data needs to be collected, some causes of the extreme scene have often appeared in the vehicle's surrounding environment a period of time before the time of the extreme scene. The data collection method of tracing back a period of time before and after can increase the completeness of the extreme scene data collection, effectively improve the quality of the collected extreme scene data, and provide more valuable scene data for the optimization of subsequent perception algorithms and prediction algorithm models. The memory stores the complete surrounding environment data of the vehicle in real time, which can perform complete data collection on the extreme scene that needs to be collected.

[0085]

Seventh Embodiment

[0086] Referring to Figure 2 In one specific embodiment, after collecting the extreme scene data, the collection method comprises:

[0087] S400: uploading the extreme scene data from the storage to the cloud after the ego vehicle is turned off or the intelligent driving system is not activated.

[0088] In this embodiment, after collecting the extreme scene data, the extreme scene data needs to be uploaded to the cloud. In this embodiment, uploading to the cloud means uploading the data to the cloud for data collection through the network.

[0089] It can be understood that data storage needs to occupy a large memory. Timely uploading the collected extreme scene data to the cloud can avoid the situation that the memory of the vehicle is full, causing the intelligent driving system of the vehicle to malfunction. At the same time, since a large network bandwidth is needed when uploading, and a large network bandwidth is also needed when the intelligent driving system is running, in this embodiment, the extreme scene data is uploaded from the storage to the cloud after the ego vehicle is turned off or the intelligent driving system is not activated. Such a setting can avoid occupying too much network bandwidth of the vehicle when uploading the extreme scene data, affecting the normal use of the intelligent driving system of the vehicle.

[0090]

Eighth Embodiment

[0091] The embodiment provides a user behavior-based extreme scene data acquisition device, the acquisition device comprising: a judgment module, the judgment module being used for judging whether an intelligent driving system is activated; and an acquisition module, the acquisition module being used for acquiring extreme scene data in response to a user instruction when the intelligent driving system is activated.

[0092] The user behavior-based extreme scene data acquisition device provided by the embodiment of the application realizes the steps of the user behavior-based extreme scene data acquisition method of any embodiment of the application, and thus has all the beneficial effects of the user behavior-based extreme scene data acquisition method of any embodiment of the application, which will not be repeated here.

[0093]

Ninth Embodiment

[0094] The embodiment provides an intelligent vehicle, which comprises a processor, a memory, and a program or instruction stored in the memory and executable on the processor, and the program or instruction is executed by the processor to realize the steps of the user behavior-based extreme scene data acquisition method of any embodiment of the application.

[0095] The intelligent vehicle provided by the embodiment of the application realizes the steps of the user behavior-based extreme scene data acquisition method of any embodiment of the application, and thus has all the beneficial effects of the user behavior-based extreme scene data acquisition method of any embodiment of the application, which will not be repeated here.

[0096]

Tenth Embodiment

[0097] The embodiment provides a readable storage medium, and the readable storage medium stores a program or instruction, and the program or instruction is executed by a processor to realize the steps of the user behavior-based extreme scene data acquisition method of any embodiment of the application.

[0098] The readable storage medium provided by the embodiment of the application realizes the steps of the user behavior-based extreme scene data acquisition method of any embodiment of the application, and thus has all the beneficial effects of the user behavior-based extreme scene data acquisition method of any embodiment of the application, which will not be repeated here.

[0099] Although the application is disclosed as above, the application is not limited to this. Any person skilled in the art, without departing from the spirit and scope of the application, can make various changes and modifications, and therefore the protection scope of the application should be subject to the range defined by the claims.

Claims

1. A user behavior based extreme scenario data collection method, characterized in that, The collection method comprises: determining whether the intelligent driving system is activated; when the intelligent driving system is activated, collecting extreme scene data in response to a user instruction; the instruction comprises at least one of the following: a first instruction, the first instruction comprising a user's takeover driving behavior, the user delivering the first instruction to the intelligent driving system through control of the ego vehicle; a second instruction, the second instruction comprising a user's expected judgment of ego vehicle driving behavior, the user delivering the second instruction to the intelligent driving system through a button or voice; the response to the user instruction comprises: within a first preset time when the intelligent driving system receives the instruction, accepting at least one instruction of the user; when, within the first preset time, the instruction only comprises the first instruction, responding to the first instruction; when, within the first preset time, the instruction only comprises the second instruction, responding to the second instruction; when, within the first preset time, the instruction comprises the first instruction and the second instruction, responding to the first instruction; after the collection of the extreme scene data, obtaining a predicted driving behavior of the vehicle according to an actual driving behavior of the vehicle; determining whether the extreme scene data needs to be uploaded according to the actual driving behavior, the predicted driving behavior and the instruction; when the instruction is the same as the actual driving behavior or the predicted driving behavior, the extreme scene data does not need to be uploaded; when the instruction is different from any one of the actual driving behavior and the predicted driving behavior, the extreme scene data needs to be uploaded; feeding back the determination result to the user; after the collection of the extreme scene data, uploading the extreme scene data from the storage to the cloud when the ego vehicle is turned off or the intelligent driving system is not activated.

2. The method of claim 1, wherein, Before the collection of the extreme scene data, the collection method comprises: real-time perception of a surrounding environment of the vehicle, and caching extreme scene data of the surrounding environment to an internal memory.

3. The method of claim 2, wherein, The collection of the extreme scene data comprises: recording a time t when the extreme scene data needs to be collected; carrying extreme scene data within a second preset time before the time t and a third preset time after the time t from the internal memory to the storage.

4. A user behavior based extreme scenario data collection apparatus, characterized by, The collection device comprises: a determination module for determining whether the intelligent driving system is activated; a collection module for collecting extreme scene data in response to a user instruction when the intelligent driving system is activated.

5. An intelligent vehicle, characterized by The intelligent vehicle comprises a processor, a memory, and a program or instruction stored on the memory and executable on the processor, the program or instruction being executed by the processor to implement the steps of the collection method according to any one of claims 1 to 3.

6. A readable storage medium characterized by, The readable storage medium stores a program or instruction, the program or instruction being executed by the processor to implement the steps of the collection method according to any one of claims 1 to 3.

Images