CN116353624A

CN116353624A - Automatic driving system takeover processing method

Info

Publication number: CN116353624A
Application number: CN202211663263.1A
Authority: CN
Inventors: 唐烈萍
Original assignee: DeepRoute AI Ltd
Current assignee: DeepRoute AI Ltd
Priority date: 2022-12-23
Filing date: 2022-12-23
Publication date: 2023-06-30

Abstract

The invention belongs to the technical field of automatic driving, and particularly relates to an automatic driving system takeover processing method, which comprises the steps of S1, acquiring event data in the driving process; s2, judging whether the abnormal takeover triggering condition is met, if so, inquiring feedback information, and if not, inquiring; s3, the system records feedback information, analyzes the feedback information and event data of the abnormal take-over time period, and performs optimization training on the system through the model. According to the technical scheme, the existing unknown scene can be optimized into the processable scene, so that the processable scene information range of the automatic driving system is increased, the automatic driving system can be accurately identified and safely controlled, and the occurrence rate of safety accidents is reduced.

Description

Automatic driving system takeover processing method

Technical Field

The invention belongs to the technical field of automatic driving, and particularly relates to an automatic driving system takeover processing method.

Background

Northern sand storm, southern storm, high and low fluctuation topography, various and typical landforms, traffic flow, people flow and the like on daily roads, which lead to the complexity of domestic urban roads, and meanwhile, the urban roads are continuously built and improved, and complex urban scenes become a big subject for the automatic driving technology to land in urban areas. The automatic driving system is an intelligent system for realizing unmanned driving through a computer system, and particularly, the automatic driving mainly relies on cooperation between monitoring devices such as artificial intelligence, visual computation, radar and the like and a global positioning system, so that the computer can automatically and safely operate the motor vehicle without any human being actively involved in operation.

In order to ensure the safety of the automatic driving vehicle and the road, the calculation capability of the automatic driving vehicle to a computer and the stability requirement of a calculation control system are high. The automatic driving system is used for handling the emergency abnormal situation and directly relates to the stability of the automatic driving vehicle. The existing automatic driving technology can treat the normal driving reminding and the taking over reminding sent by the system and the auxiliary driving quitting well, namely the system recognizes the problem and gives out an explicit treatment strategy. However, in urban scenes, besides having multiple functions including unprotected left-turn, narrow road traffic, pedestrian gifts, turning around, automatic obstacle detouring, and the like, the automatic driving system also needs to deal with more unknown scenes, such as missed detection (i.e. missed detection of obstacles obstructing traffic) and false detection (i.e. false detection of obstacles obstructing traffic), and the automatic driving system cannot correctly identify and safely control, thus easily causing safety accidents, which is also the problem existing in the existing automatic driving technology.

Disclosure of Invention

The invention aims to provide a method for taking over and processing an automatic driving system, which aims to solve the technical problems that the system is difficult to identify and process in the face of unknown scenes with safety risks, so that missed detection and false detection exist.

The invention provides a method for processing takeover of an automatic driving system, which comprises the following steps:

s1: acquiring event data in the driving process;

s2: judging whether the abnormal takeover triggering condition is met, if so, inquiring feedback information, and if not, inquiring;

s3: and the system records feedback information, analyzes the feedback information and event data of the abnormal take-over time period, and performs optimization training on the system through a model.

Further, the abnormal takeover triggering conditions include a first condition and a second condition, the first condition is that the vehicle is in emergency takeover under the automatic driving/auxiliary state, the second condition is that the current scene is judged to be in a normal running state under the automatic driving/auxiliary driving state, and the first condition and the second condition are both met, namely the abnormal takeover triggering conditions are met.

Further, the emergency takeover judging condition in the first condition is that the lateral/longitudinal acceleration is greater than a threshold value.

Further, the feedback modes of the feedback information in the step S2 include a first feedback mode and a second feedback mode, where the first feedback mode is active interaction between the system and the user, and the second feedback mode is active reporting by the user.

Further, in step S3, the system records feedback information and analyzes the feedback information and event data of the abnormal take-over period, and performs optimization training on the system through a model, which specifically includes the following steps:

s301, record data: determining event data according to the abnormal takeover time point, and recording the event data and feedback information together;

s302, analyzing data: based on the event data, judging whether the event data is credible or not according to the user behavior confidence, and if so, entering the next step; if not, discarding the event data;

s303, training an optimization model: taking the event data as training data, and carrying out optimization training on the model based on user feedback information;

s304, testing and verifying a model: and (3) carrying out real vehicle drive test on the trained model, upgrading the system and continuously observing the application condition of event data after the drive test passes.

Further, the recording time range of the event data is from a time starting point to a time ending point, the time starting point relates to two time points, and the time starting point is the later time point of the two time points; the time end point refers to three time points, and the time end point is an earlier one of the three time points.

Further, the two time points of the time starting point are a first preset time and an automatic driving/auxiliary driving system exit time, and the three time points of the time ending point are a second preset time, an event ending point time and an automatic driving/auxiliary driving system activation time.

Further, the event data includes, but is not limited to, vehicle travel data, vehicle video data, vehicle location data, and vehicle takeover data.

Further, the user confidence judgment criteria include, but are not limited to, a user history of using the habit of driving assistance, a comprehensive duration of using the driving assistance, a comprehensive score of intelligent driving, and a preference mode of the driving assistance.

Further, the step S303 further includes determining whether the model meets the safety standard, and when the number of times of sending the abnormal takeover event is greater than the threshold value or the number of times of occurrence of the abnormal takeover event on the same road section is greater than the threshold value in the preset time period, determining that the model does not meet the safety standard, and retraining the model is required.

An autopilot system take over processing apparatus, the processing apparatus comprising:

the acquisition module is used for acquiring event data in the driving process;

the judging module is used for judging whether the emergency connection triggering condition is met;

the query module is used for querying and reading feedback information;

the optimizing module comprises a recording unit, an analyzing unit, a training optimizing unit and a test verifying unit, wherein the recording unit is used for recording event data and feedback information, the analyzing unit is used for judging whether the event data is credible or not according to user confidence, the training optimizing unit is used for optimizing training on the model according to the event data and the feedback information, and the test verifying unit is used for carrying out real vehicle road test on the trained model.

An autonomous vehicle comprising an electronic device including a memory and a processor coupled to each other for executing program instructions stored in the memory to implement an autonomous system takeover processing method as described above.

A computer readable storage medium having stored thereon a computer program which when executed by a processor implements the steps of an autopilot system take over processing method.

Compared with the prior art, the invention has the beneficial effects that:

1. according to the invention, the event data in the driving process is obtained, the event data belonging to abnormal takeover is screened, and is analyzed by combining with the feedback information of the user, so that the system is optimally trained, the existing 'unknown scene' can be optimized into the 'processable scene', the range of the scene information which can be processed by the automatic driving system is increased, the automatic driving system can be accurately identified and safely controlled, and the occurrence rate of safety accidents is reduced;

2. by setting the abnormal takeover triggering condition and combining the user behavior confidence, the system is promoted to judge the unknown scene with safety risk based on the driver behavior, and after the scene is judged, feedback information of the abnormal takeover is timely obtained through system active interaction or user active feedback, so that the occurrence rate of system missed detection or false detection can be reduced; and meanwhile, the system records the abnormal take-over event data, so that the follow-up engineers can conduct investigation optimization.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical methods in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described, and it is apparent that the drawings in the following description are some embodiments of the present invention and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

FIG. 1 illustrates a flow chart of a method of automated driving system takeover processing in accordance with the present invention;

FIG. 2 is a flowchart showing a step S3 in an autopilot system takeover processing method of the present invention;

FIG. 3 is a block diagram of an autopilot system takeover processing apparatus of the present invention;

fig. 4 shows a specific structure diagram of an optimization module in an autopilot system take-over processing device according to the invention.

Detailed Description

In order to better understand the technical solutions in the present application, the following description will clearly and completely describe the technical solutions in the embodiments of the present application in conjunction with the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.

The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element. In the present application, "at least one (item)" means one or more, "a plurality" means two or more, and "at least two (items)" means two or three or more, and/or "for describing an association relationship of an association object, three kinds of relationships may exist, for example," a and/or B "may represent: only a, only B and both a and B are present, wherein a, B may be singular or plural. The character "/" generally indicates that the context-dependent object is an "or" relationship. "at least one of (a) or a similar expression thereof means any combination of these items. For example, at least one (one) of a, b or c may represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c".

As shown in fig. 1-2, the present invention provides a method for handling an autopilot system takeover, the method comprising the steps of:

s1: event data during driving is acquired.

The event data includes, but is not limited to, vehicle travel data, vehicle video data, vehicle position data, vehicle takeover data, driver behavior data, and also includes data recorded by installing in-vehicle devices conforming to the related art standards, data acquired based on regulatory requirements.

The event data acquired in the invention necessarily comprises takeover data, wherein the takeover data comprises normal takeover data and abnormal takeover data, and the normal takeover data comprises prompting a user to manually take over when the system recognizes safety risks such as impending collision in the automatic driving process; or the system recognizes the problem, but the system is reminded of manually taking over the problem due to insufficient capacity of the system to deal with the problem; or when the system identifies that the software and hardware of the system have faults, the system reminds the user of taking over the software and hardware manually. Namely, the system recognizes that the events requiring manual take over belong to the normal take over range. And the abnormal take-over is that under the condition that the system does not recognize any event needing to take over, the user takes over subjectively, namely, the event which is unknown and has safety risk to the system is satisfied and all events belong to the abnormal take-over range.

S2: judging whether the abnormal takeover triggering condition is met, if so, inquiring feedback information, and if not, inquiring.

Specifically, the abnormal takeover triggering conditions comprise a first condition and a second condition, wherein the first condition is whether the vehicle is in emergency takeover under an automatic driving/auxiliary state, the emergency takeover condition is that the transverse/longitudinal acceleration is larger than a threshold value, and the threshold value can be set according to the user requirement; the second condition is that the automatic driving/auxiliary driving state judges that the current scene is in a normal running state, namely that the automatic driving/auxiliary driving is in normal automatic driving running.

If the event data simultaneously meets the first condition and the second condition, judging that the event data meets the abnormal takeover triggering condition and belongs to the event data of abnormal takeover, automatically inquiring and calling feedback information by the system, wherein the feedback information refers to the condition description of the abnormal takeover by a user. If the event data only meets one of the conditions or neither of the conditions, judging that the event data does not meet the abnormal takeover triggering condition and does not belong to the event data of the abnormal takeover, and not carrying out subsequent operation.

Specifically, the feedback is responsive to the vehicle experiencing an abnormal take over event, and the system interacts with the user to generate feedback information. The feedback modes of the feedback information can be divided into two modes, wherein the first feedback mode is that the vehicle actively interacts with the user, the event is recorded through on-site interaction, and the triggering condition of the vehicle actively interacts is the same as the triggering condition of the abnormal takeover. When the triggering condition is met, the auxiliary driving system actively inquires the user in a visual (text interaction) +auditory (voice interaction) mode to guide the user to explain the intention of abnormal taking over.

Illustratively, the system actively triggers a voice interaction, guiding the user to say: is there just a sudden situation we have not handled? You can tell me that i will record to let the engineer check the optimization. And provides two options [ Yes ], more of which are [ Yes ].

If the user selects [ yes, and more words ], after the user selects, the abnormal take-over event can be fed back through voice, and the system automatically records the feedback information of the user and packages and records the feedback information together with the event.

If the user selects [ yes ], the system automatically records the event data based on the event recording rule.

Based on the characteristics of human memory capacity, the system is arranged to guide the user to feed back, so that the user can be urged to take over the real-time feedback record for the abnormality, the problem that the user fails to report or report the abnormality due to follow-up forgetting is avoided, and the occurrence rate of system missing detection or false detection can be reduced. The input of the feedback information can help engineers analyze abnormal takeover data, so that the reasons of abnormal takeover can be more clearly analyzed when event data are analyzed, and analysis is performed from two aspects of a user and automatic driving, thereby being beneficial to optimizing and upgrading an automatic driving system.

The second feedback mode is that the user reports actively, the user feeds back afterwards after the journey is finished, the feedback path can be words, voice and the like, and the system records the feedback information.

For on-site feedback, the voice interaction increases the problem feedback verticals, the user can actively input the problem through voice, when the user speaks 'report problem + content', the analysis semantics are successfully divided into the problem feedback verticals, and the system automatically records the event record of the time point when the user initiates the voice based on the event record rule. For post feedback, a user feeds back the problem at the app end of the car machine and the mobile phone through the forms of characters, voice and the like after the journey is finished, the system searches event data of abnormal takeover based on the time period fed back by the user, and packages, records and transmits the event data and feedback information to an engineer for investigation and optimization. Through multiple initiative reporting modes, the feedback information can be timely input through the initiative reporting of the user under the conditions that the automatic driving system does not timely pop up a guide window or pop up the guide window, but the user does not have time to explain the reason of abnormal takeover, and the like, so that more information is provided for the investigation optimization of the automatic driving system.

S3: and the system records feedback information, analyzes the feedback information and event data of the abnormal take-over time period, and performs optimization training on the system through the model.

After inquiring and retrieving feedback information, the system retrieves event data of the abnormal take-over time period, analyzes the feedback information and the event data, screens out training data, and further applies the training data to the system, trains the system through a model, promotes the next time of encountering a similar event, and can automatically identify risks and remind a user to take over manually. The method described by the technical scheme can extract event data, judge whether the scene of the event data belongs to an abnormal takeover scene according to whether the abnormal takeover triggering condition is met, and if so, analyze and train the event data of the abnormal takeover, so as to optimize the system, optimize the unknown scene into a processable scene and improve the processing capacity and the stability of the automatic driving system; meanwhile, the technical scheme can also provide a processing method of the automatic driving system for the unknown scene, solves the problems of missed detection and false detection of automatic driving in the prior art, promotes the automatic driving system to accurately identify and safely control, and reduces the occurrence of safety accidents.

Further, in step S3, the system records feedback information, analyzes the feedback information and event data of the abnormal takeover period, and performs optimization training on the system through a model, and further includes the following steps:

s301, record data: and determining an event data recording time range according to the abnormal taking-over time point, and recording the event data and the feedback information together.

Specifically, the event data recording time range is between a time starting point and a time ending point, wherein the time starting point relates to two time points, namely a first preset time and an automatic driving/auxiliary driving system exit time, and when the time starting point is selected, the later time point of the two time points is taken as the optimal time starting point; the time end point refers to three time points, namely a second preset time, an event end point time and an automatic driving/auxiliary driving system activation time, and when the time end point is selected, the earlier one of the three time points is taken as the optimal time end point. By setting a plurality of time points for the starting point and the ending point of the event data recording time and selectively incorporating the optimal time points, event data in the time starting point and the time ending point range are packaged for further investigation and analysis, so that the event data can be completely recorded and analyzed to the greatest extent, and analysis errors caused by missing part of data are avoided, and the system optimization effect is influenced.

Specifically, the first preset time is a period of time before the start of the event recording, and the preset time length may be set automatically, preferably 30s, 1min or 5min. The second preset time is a period of time after the end point of the event recording, and the preset time length can be set automatically, preferably 30s, 1min or 5min.

S302, analyzing data: based on the event data, combining with the user behavior confidence, judging whether the event data is credible, if so, entering the next step, and if not, discarding the event data.

Specifically, the judgment criteria of the user behavior confidence include, but are not limited to, the user history of using the driving assistance behavior habit, the integrated duration of using the driving assistance, the intelligent driving integrated score, and the preference mode of the driving assistance. The driving behavior habit of the user mainly comprises two points, namely, the first point, the user actively takes over frequently, for example, in a good scene state, the user frequently and actively takes over the behavior, and the confidence of the behavior of the user is low. Second, frequent take over of reminders in the expected range, e.g., in good condition of the scene, the system frequently sends out hands-off reminders to the driver, and the confidence of the user behavior can be considered low. The intelligent driving comprehensive score is divided into 100 points, and in the intelligent auxiliary driving state, the wrong operation of the vehicle owner can form a deduction item, and once the score is reduced to below 80 points, the user behavior confidence degree can be considered to be lower. The preference modes of assisted driving include conservative, ordinary or aggressive, where the aggressive user behavior is less confident and conservative and ordinary user behavior is relatively more confident.

Based on the event data recorded in step S301, the data are screened in combination with the user behavior confidence, when the user behavior confidence is greater than the threshold, the event data are considered to be trusted, and when the user behavior confidence is less than the threshold, the event data are considered to be not trusted, the data can be discarded, and the next operation is not performed. The credibility of the event data is analyzed and judged by analyzing the confidence level of the user behavior, so that the active taking over condition of the user under the condition of non-unknown safety risk scene can be screened out, and unnecessary analysis is avoided, and the running cost of the system is increased.

S303, training an optimization model: and taking the event data as training data, and optimally training the model based on the user feedback information.

And (3) taking the event data screened in the step S302 as training data, and analyzing the model types required to be trained by combining with user feedback information to train in a targeted manner. For example, for the false detection/omission problem of the user feedback, the event data in step S302 may be pre-labeled manually, and then the perception model is trained.

Further, during training, whether the model training meets the standard is judged, when the number of times of sending abnormal take-over events in a preset time period is larger than a threshold value, or the number of times of occurrence of abnormal take-over events in the same road section is larger than the threshold value, the system is considered to have a certain risk, and the model needs to be retrained until the model training meets the safety standard. Wherein the threshold value can be set according to the actual driving situation of the user. By combining event data and feedback information, a high-precision unknown scene data set can be provided for optimization training of the system, convenience is provided for engineers to carry out algorithm test iteration on the data set, the system is enabled to accurately and efficiently judge event scenes when encountering the same scenes next time, and an accurate processing scheme is provided.

And (3) carrying out real vehicle road test based on the trained model, upgrading a vehicle system after the road test passes, continuously observing the condition of using data, and continuously finding and optimizing the problem. By combining the theory and practice of the real vehicle drive test, whether the model can better identify the event scene in the actual scene can be better verified, so that an accurate processing scheme is provided, and support is provided for the actual application of the model.

In summary, through re-analyzing the abnormal take-over behavior of the user (namely, the system considers normal automatic driving, the user suddenly takes over emergently), so as to screen out an unknown scene which cannot be processed by the automatic driving system, record the related event data of the abnormal take-over, and conduct investigation optimization algorithm based on the data and the user feedback information, optimize the automatic driving system for training, optimize the unknown scene into a processable scene, and further increase the scene information range which can be processed by the automatic driving system.

As shown in fig. 3-4, the present invention provides an autopilot system takeover processing apparatus, the processing apparatus comprising: the acquisition module is used for acquiring event data in the driving process; the judging module is used for judging whether the event data meets the triggering condition of abnormal takeover or not; the query module is used for querying whether feedback information exists when the event meets the triggering condition of abnormal takeover and retrieving the feedback information based on different feedback modes; the optimizing module comprises a recording unit, an analyzing unit, a training optimizing unit and a test verifying unit, wherein the recording unit is used for recording event data and feedback information, the analyzing unit is used for judging whether the event data is credible according to user behavior confidence, the training optimizing unit is used for optimizing training on the model according to the event data and the feedback information, and the test verifying unit is used for carrying out real vehicle drive test on the trained model.

It should be noted that, the automatic driving system take-over processing device provided by the invention can be realized in a hardware mode or in a software mode, and the implementation mode is not particularly limited.

The invention also provides an autonomous vehicle comprising an electronic device comprising a memory and a processor coupled to each other, the processor being configured to execute program instructions stored in the memory to implement the autonomous system takeover processing method shown in fig. 1-2.

The present invention also provides a computer readable storage medium having a computer program stored thereon, which when executed by a processor, implements the steps of an autopilot system take over processing method as shown in fig. 1-2.

The computer readable medium described above in the present disclosure may be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this disclosure, a computer-readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present disclosure, however, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with the computer-readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, fiber optic cables, RF (radio frequency), and the like, or any suitable combination of the foregoing.

The computer readable medium may be contained in the electronic device; or may exist alone without being incorporated into the electronic device.

Computer program code for carrying out operations of the present disclosure may be written in one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any kind of network, including a local Area Network (AN) or a Wide Area Network (WAN), or can be connected to AN external computer (for example, through the Internet using AN Internet service provider).

The foregoing descriptions of specific exemplary embodiments of the present invention are presented for purposes of illustration and description. It is not intended to limit the invention to the precise form disclosed, and obviously many modifications and variations are possible in light of the above teaching. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application to thereby enable one skilled in the art to make and utilize the invention in various exemplary embodiments and with various modifications as are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the claims and their equivalents.

Claims

1. An automatic driving system takeover processing method is characterized in that: the processing method comprises the following steps:

s1: acquiring event data in the driving process;

2. An autopilot system takeover processing method as set forth in claim 1 wherein: the abnormal takeover triggering conditions comprise a first condition and a second condition, wherein the first condition is that the vehicle is in emergency takeover under the automatic driving/auxiliary driving state, the second condition is that the current scene is judged to be in a normal running state under the automatic driving/auxiliary driving state, and the first condition and the second condition are both met, namely the abnormal takeover triggering conditions are met.

3. An autopilot system takeover processing method as set forth in claim 2 wherein: the emergency takeover judging condition in the first condition is that the transverse/longitudinal acceleration is larger than a threshold value.

4. An autopilot system takeover processing method as set forth in claim 1 wherein: the feedback modes of the feedback information in the step S2 comprise a first feedback mode and a second feedback mode, wherein the first feedback mode is that the system actively interacts with the user, and the second feedback mode is that the user actively reports the feedback information.

5. An autopilot system takeover processing method as set forth in claim 1 wherein: in step S3, the system records feedback information and analyzes the feedback information and event data of the abnormal take-over time period, and performs optimization training on the system through a model, specifically comprising the following steps:

6. An autopilot system takeover processing method as set forth in claim 5 wherein: the time range of the event data is from a time starting point to a time ending point, the time starting point relates to two time points, and the time starting point is the later time point of the two time points; the time end point refers to three time points, and the time end point is an earlier one of the three time points.

7. An autopilot system takeover processing method as set forth in claim 6 wherein: the two time points of the time starting point are first preset time and automatic driving/auxiliary driving system exit time, and the three time points of the time ending point are second preset time, event ending point time and automatic driving/auxiliary driving system activation time.

8. An autopilot system takeover process as claimed in claim 1, 5 or 6 wherein: the event data includes, but is not limited to, vehicle travel data, vehicle video data, vehicle location data, and vehicle takeover data.

9. An autopilot system takeover processing method as set forth in claim 5 wherein: the user confidence judgment criteria include, but are not limited to, historical driving assistance behavior habit of the user, comprehensive duration of driving assistance, comprehensive score of intelligent driving, and preference mode of driving assistance.

10. An autopilot system takeover processing method as set forth in claim 5 wherein: step S303 further includes determining whether the model meets the safety standard, and when the number of times of sending the abnormal takeover event is greater than a threshold value or the number of times of occurrence of the abnormal takeover event on the same road section is greater than a threshold value in a preset period of time, considering that the model does not meet the safety standard, and retraining the model is required.

11. An autopilot system take over processing apparatus, the processing apparatus comprising:

the acquisition module is used for acquiring event data in the driving process;

the query module is used for querying and reading feedback information;

12. An autonomous vehicle comprising an electronic device including a memory and a processor coupled to each other, the processor for executing program instructions stored in the memory to implement the autonomous system takeover processing method of any of claims 1-10.

13. A computer-readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the autopilot system take over processing method according to any one of claims 1 to 10.