Data processing method and apparatus

Abstract

The present application relates to the technical field of data processing, and discloses a data processing method and apparatus, with the aim of automatically determining whether a first event occurring during an autonomous driving process of a vehicle is a true trigger or a false trigger, so as to eliminate the need for manual confirmation, and thereby improve efficiency. The method comprises: when a first event occurs during an autonomous driving process of a vehicle, performing a determination on the first event by means of a preset rule and / or machine learning according to first data corresponding to the first event to obtain a determination result, and storing, reporting, or deleting the first data according to the determination result. The determination result can indicate whether the first event is a true trigger or a false trigger.

Description

Data processing method and device

[0001] The present application claims priority to the Chinese patent application No. 202411252626.1, filed on September 6, 2024, and entitled "Data processing method and device", the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] The present application relates to the technical field of data processing, in particular to a data processing method and device. BACKGROUND

[0003] With the application of automatic driving technology in vehicles, some problems in the automatic driving process of vehicles also arise. Analyzing the problem data in the automatic driving process can better iterate and upgrade the vehicles. However, in the related art, the problem data in the automatic driving process needs to be confirmed by manual, which consumes a lot of manpower and material resources and has low efficiency. SUMMARY

[0004] The present application provides a data processing method and device, which can automatically process data and improve efficiency.

[0005] To achieve the above purpose, the present application provides the following technical solutions:

[0006] In a first aspect, the present application provides a data processing method, comprising: obtaining first data of a vehicle; the first data comprising data corresponding to a first event, the first event comprising an event occurring in the automatic driving process of the vehicle; determining the first event based on the first data through a preset rule to obtain a first result; the first result indicating that the first event is a correct trigger, or the first result indicating that the first event is an incorrect trigger; determining the first event based on the first data through machine learning to obtain a second result; the second result indicating that the first event is a correct trigger, or the second result indicating that the first event is an incorrect trigger; and processing the first data according to the first result and / or the second result.

[0007] In this way, the vehicle can automatically determine whether the first event in the automatic driving process of the vehicle is a correct trigger or an incorrect trigger based on the preset rule and / or machine learning, without the need for manual confirmation, thereby improving the efficiency.

[0008] In a possible implementation, the first event is determined based on the first data according to a preset rule, and a first result is obtained, including: determining a target object triggering the first event; determining whether the first event has a first problem based on the first data and the target object; the first problem indicates that the target object has an anomaly; the first event has the first problem, and the first result indicates that the first event is a false trigger; the first event does not have the first problem, and the first result indicates that the first event is a correct trigger.

[0009] In this way, the vehicle can automatically determine whether the first event in the automatic driving process of the vehicle is a correct trigger or a false trigger according to a preset rule. The determination process is fast and simple, the data processing amount is reduced, and the efficiency is further improved.

[0010] In a possible implementation, the first problem includes one or more of the following: target object position jump, target object orientation jump, target object speed jump, target object label change, target object false detection, or target object animal mis-detection; the target object false detection indicates that part of the data of the target object in the first data is missing.

[0011] In this way, when the first event has the first problem, the first event can be accurately determined as a false trigger, and the accuracy of the determination result is improved.

[0012] In a possible implementation, after the first event is triggered, an operation of a driver on a brake pedal is received, the first problem is a second problem, and the second problem includes one or more of the following: target object label change, target object false detection, target object position jump, or target object orientation jump.

[0013] In this way, the first event can be an autonomous emergency braking (AEB) event. When the vehicle receives the operation of the driver on the brake pedal, the driver can determine that the target object has no speed jump and is not animal mis-detection, and actively steps on the brake pedal. The second problem does not include target object speed jump and animal mis-detection, which can improve the accuracy of the result of determining that the AEB event is a correct trigger or a false trigger, and improve the efficiency.

[0014] In a possible implementation, in a case where the target object is a pedestrian crossing a road, the first problem is a third problem, and the third problem includes one or more of the following: target object label change, target object false detection, or target object orientation jump.

[0015] The first event can be an AEB event. Since the speed of the pedestrian changes rapidly and the position changes slowly, the target object speed jump and the target object position jump are not included in the third problem, and the accuracy of the result of the vehicle determining that the AEB event is a correct trigger or an AEB event is an incorrect trigger can be improved. Moreover, the target object is determined to be a pedestrian, and the pedestrian crosses the road, which can exclude the animal mis-detection problem, the animal mis-detection is not included in the third problem, and the efficiency and accuracy of the result of the vehicle determining that the AEB event is a correct trigger or an AEB event is an incorrect trigger can be improved.

[0016] In a possible implementation, the second result is obtained by machine learning based on the first data, including: determining a second object, the second object including a first object of each type of the first object having a short predicted collision time with the vehicle, and the predicted collision time of the first object with the vehicle being less than a preset threshold; the type of the first object including a vehicle, a pedestrian, or a static road element; inputting first data corresponding to the second object into a first model; the first data including a label of the second object, a predicted collision time of the second object with the vehicle, and a first event trigger state; the first model being trained by a label of a historical second object, a predicted collision time of the historical second object with the vehicle, and a historical first event trigger state; determining the first event based on the first model to obtain the second result.

[0017] In this way, the vehicle automatically determines whether the first event occurring in the vehicle automatic driving process is a correct trigger or an incorrect trigger through the first model. The determination process is fast and simple, the data processing amount is reduced, and the efficiency is further improved.

[0018] In a possible implementation, the first data is processed according to the first result and / or the second result, including: storing and / or reporting the first data when the first result and / or the second result indicate that the first event is an incorrect trigger; and deleting the first data when the first result and the second result indicate that the first event is a correct trigger.

[0019] In this way, when the first result and the second result indicate that the first event is a correct trigger, the first data is deleted. When the preset rule and the machine learning both determine that the first event occurring in the vehicle automatic driving process is a correct trigger, it is determined that the first event is a correct trigger, and the accuracy of the determination result is improved.

[0020] When the first result and / or the second result indicate that the first event is an incorrect trigger, the first data is stored and / or reported. The vehicle can automatically determine whether the first event occurring in the vehicle automatic driving process is a correct trigger or an incorrect trigger through the complementary manner of the preset rule and the machine learning, the efficiency and the incorrect trigger determination success rate are improved, and the situation of missing the first event being an incorrect trigger is prevented.

[0021] And the vehicle can store and process (such as clustering analysis, etc.) the first data by itself, or the vehicle can also report the first data to a cloud server, and the cloud server processes the first data, and the vehicle can be iteratively upgraded according to the processing result.

[0022] In a possible implementation, the first event includes a through intersection event, the first event is determined based on the first data by a preset rule, and a first result is obtained, including: determining the driving state of the vehicle based on the first data packet, the driving state of the vehicle including an automatic driving state and / or a manual driving state; the driving state of the vehicle is switched from the automatic driving state to the manual driving state, and the first result indicates that the first event is a false trigger; the driving state of the vehicle is the automatic driving state, and the first result indicates that the first event is a correct trigger.

[0023] In this way, when the first event is a through intersection event, the vehicle can determine that the first event is a false trigger or a false trigger by a preset rule.

[0024] In a possible implementation, the type of the target object includes a vehicle, a pedestrian, or a static road element.

[0025] In this way, the target object can include all elements in the road, and the vehicle can classify and process according to different types of target objects to obtain more accurate determination results.

[0026] In a possible implementation, the first event is determined based on the first data and the target object whether there is a first problem, including: determining the scene feature of the first event based on the first data and the target object; the scene feature of the first event indicates whether the target object has a first problem.

[0027] In this way, the vehicle can determine the scene feature of the first event, and subsequent clustering analysis can be based on the scene feature of the first event.

[0028] In a possible implementation, the method further includes: performing clustering analysis based on the first data, the first result and / or the second result; the first result includes the scene feature of the first event; the scene feature of the first event includes one or more of the following: the speed of the vehicle when the first event is triggered, the position of the vehicle when the first event is triggered, the type of the target object, the distance between the target object and the vehicle when the target object first appears, the position of the target object when the first event is triggered, the angle between the target object and the vehicle, the trigger duration of the first event, the operation state of the driver, the road form or the road curvature.

[0029] In this way, the clustering analysis result is more abundant, and the vehicle can be iteratively upgraded according to the clustering analysis result.

[0030] In a possible implementation, the first data further include one or more of the following: a geometric feature of the road, a speed of the vehicle, a position of the vehicle, a first object within a preset distance of the vehicle, a position of the first object, a predicted collision time between the first object and the vehicle, a type of the first object, a first event trigger state, a driving state of the vehicle, or an operation state of a driver.

[0031] In a third aspect, the present application provides a data processing apparatus, comprising: a processor and a memory coupled to the processor, the memory being configured to store computer program codes, the computer program codes comprising computer instructions, and the processor being configured to read the computer instructions from the memory to cause the data processing apparatus to perform the method according to the first aspect.

[0032] In a third aspect, the present application provides a vehicle comprising the apparatus according to the second aspect.

[0033] In a fourth aspect, the present application provides a computer readable storage medium, the computer readable storage medium comprising computer programs or instructions, and the computer programs or instructions being configured to cause the vehicle speed estimation apparatus according to the second aspect to perform the method according to the first aspect when the computer programs or instructions are executed on the vehicle speed estimation apparatus.

[0034] In a fifth aspect, the present application provides a computer program product, the computer program product comprising: computer programs or instructions, and the computer programs or instructions being configured to cause a computer to perform the method according to the first aspect when the computer programs or instructions are executed on the computer.

[0035] In a sixth aspect, the present application provides a chip system, comprising: a processor configured to invoke and execute computer programs stored in a memory to perform any one of the methods according to the implementation manners of the first aspect.

[0036] The technical effects of the second aspect to the sixth aspect and any one of the implementation manners of the second aspect to the sixth aspect can refer to the technical effects of the first aspect and any one of the implementation manners of the first aspect, which will not be described herein. BRIEF DESCRIPTION OF DRAWINGS

[0037] FIG. 1 is a schematic structural diagram of a vehicle according to an embodiment of the present application;

[0038] FIG. 2 is a schematic diagram of a scenario in which a vehicle interacts with a cloud server according to an embodiment of the present application;

[0039] FIG. 3 is a schematic flow diagram of a data processing method according to an embodiment of the present application;

[0040] FIG. 4 is a schematic flow diagram of another data processing method according to an embodiment of the present application;

[0041] FIG. 5 is a flow diagram of another data processing method according to an embodiment of the present application;

[0042] FIG. 6 is a flow diagram of another data processing method according to an embodiment of the present application;

[0043] FIG. 7 is a flow diagram of another data processing method according to an embodiment of the present application;

[0044] FIG. 8 is a flow diagram of another data processing method according to an embodiment of the present application;

[0045] FIG. 9 is a flow diagram of another data processing method according to an embodiment of the present application;

[0046] FIG. 10 is a data waveform diagram according to an embodiment of the present application;

[0047] FIG. 11 is a vehicle position change scenario diagram according to an embodiment of the present application;

[0048] FIG. 12 is a flow diagram of another data processing method according to an embodiment of the present application;

[0049] FIG. 13 is a road intersection diagram according to an embodiment of the present application;

[0050] FIG. 14 is a data processing apparatus structure diagram according to an embodiment of the present application;

[0051] FIG. 15 is a chip system structure diagram according to an embodiment of the present application. DETAILED DESCRIPTION

[0052] In the embodiments of the present application, the word "exemplary" or "for example" is used to mean serving as an example, instance, or illustration. Any implementation or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being preferred or advantageous over other implementation or design solutions. Rather, the use of the word "exemplary" or "for example" is intended to present concepts in a concrete manner.

[0053] In the embodiments of the present application, the terms "first", "second", etc. are used only for the purpose of description, and should not be interpreted or implied to indicate or suggest relative importance or implicitly indicate the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, unless otherwise specified, the meaning of "a plurality of" is two or more.

[0054] The term "at least one" in the present application means one or more, and the term "multiple" in the present application means two or more than two, for example, multiple second packets means two or more than two second packets. The terms "system" and "network" are often used interchangeably herein.

[0055] It should be understood that the terms used in the description of various described examples herein are merely for the purpose of describing specific examples and are not intended to be limiting.

[0056] It should also be understood that the term "and / or" used herein means and encompasses any and all possible combinations of one or more of the associated listed items. The term "and / or", is a description of an associated relationship with the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent: A exists alone, A and B exist together, and B exists alone. In addition, the character " / " in the present application generally represents an "or" relationship between the associated objects.

[0057] It should also be understood that in various embodiments of the present application, the size of the serial number of each process does not mean the order of execution, and the execution order of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0058] It should be understood that the "one embodiment", "an embodiment", "a possible implementation" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment or implementation are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in an embodiment", "a possible implementation" appearing throughout the specification does not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner.

[0059] The automatic driving technology is applied in a vehicle, and in the related technology, the vehicle can obtain data in the automatic driving process (such as driving data and / or scene data in the automatic driving process), and the obtained data is manually screened to determine the data when the vehicle has a driving problem.

[0060] For example, the vehicle can obtain AEB data when autonomous emergency braking (AEB) is triggered in the automatic driving process. The vehicle may trigger AEB multiple times during driving. Each time AEB is triggered, the vehicle obtains an AEB data. Subsequently, the AEB data when AEB is triggered by mistake needs to be confirmed manually from multiple AEB data, which is low in efficiency.

[0061] To solve the above problems, an embodiment of the present application provides a data processing method. In the method, when a first event occurs in the automatic driving process of a vehicle, the first event can be determined according to first data (such as driving data and / or scene data, etc.) corresponding to the first event, a determination result is obtained, and the first data and / or the determination result is processed according to the determination result. The determination result indicates that the first event is a correct trigger or the first event is an incorrect trigger.

[0062] The first event being a correct trigger can mean that the vehicle triggers the first event in the automatic driving process, and the automatic driving is accurate. The first event being an incorrect trigger can mean that the vehicle triggers the first event by mistake in the automatic driving process, and the automatic driving is wrong.

[0063] In this way, the vehicle can automatically determine whether the first event in the automatic driving process of the vehicle is a correct trigger or an incorrect trigger, without the need for manual confirmation, thereby improving the efficiency.

[0064] In an embodiment, the vehicle can store the determination result when the first event is a correct trigger, the first data and the determination result when the first event is an incorrect trigger, and perform clustering analysis on the stored first data and / or determination result, so as to subsequently iteratively upgrade the vehicle.

[0065] In another embodiment, the determination result when the first event is a correct trigger, the first data and the determination result when the first event is an incorrect trigger can also be reported to a cloud server, and the cloud server performs clustering analysis on the reported first data and / or determination result. The vehicle reports the determination result when the first event is a correct trigger, the first data and the determination result when the first event is an incorrect trigger to the cloud server, thereby reducing the data reporting amount, saving traffic and cloud server resources.

[0066] FIG. 1 is a structural schematic diagram of a vehicle provided by an embodiment of the present application. The vehicle 100 can include various subsystems, such as a travel system 110, a sensor system 120, a control system 130, one or more peripheral devices 140, a power supply 150, a computer system 160, and a user interface 170. Optionally, the vehicle 100 can include more or fewer subsystems, and each subsystem can include multiple elements. In addition, each subsystem and element of the vehicle 100 can be interconnected by wire or wirelessly.

[0067] The travel system 110 can include components that provide powered movement for the vehicle 100. In an embodiment, the travel system 110 can include an engine 111, a transmission 112, an energy source 113, and wheels 114.

[0068] The sensor system 120 can include several sensors that sense information about the environment surrounding the vehicle 100. For example, the sensor system 120 can include a positioning system 121 (which can be a global positioning system (GPS), a Beidou system, or other positioning system), an inertial measurement unit (IMU) 122, a radar 123, a lidar 124, and a camera 125.

[0069] The positioning system 121 can be used to estimate the geographic location of the vehicle 100. The IMU 122 is used to sense changes in position and orientation of the vehicle 100 based on inertial acceleration. In one embodiment, the IMU 122 can be a combination of an accelerometer and a gyroscope.

[0070] The radar 123 can use radio signals to sense objects within the surrounding environment of the vehicle 100. In some embodiments, in addition to sensing objects, the radar 123 can be used to sense the speed and / or heading of the objects. In embodiments of the present application, the radar 123 can detect information such as the speed and / or orientation of first objects around the vehicle, where the first objects can include pedestrians, vehicles, stationary road elements, and the like.

[0071] The lidar 124 can use laser light to sense objects in the environment in which the vehicle 100 is located. In some embodiments, the lidar 124 can include one or more laser sources, a laser scanner, and one or more detectors, among other system components.

[0072] The camera 125 can be used to capture images of the surrounding environment of the vehicle 100, as well as images within the vehicle cabin. The camera 125 can be a still camera or a video camera. In embodiments of the present application, the camera 125 can send captured images to a computer vision system 134 in the control system 130.

[0073] In some examples, the sensor system 120 described above can perform S301 in FIG. 3 described below, and can send data detected by the several sensors to the control system 130 and / or the computer system 160.

[0074] The control system 130 can control the operation of the vehicle 100 and its components. The control system 130 can include various elements, including a steering system 131, a throttle 132, a braking unit 133, a computer vision system 134, a route control system 135, and an obstacle avoidance system 136.

[0075] The braking unit 133 is configured to control the deceleration of the vehicle 100. The braking unit 133 can use friction to slow the wheels 114. In other embodiments, the braking unit 133 can also convert the kinetic energy of the wheels 114 into electrical current. The braking unit 133 can also take other forms to slow the wheels 114 and thereby control the speed of the vehicle 100. In embodiments of the application, the braking unit 133 can be automatically triggered to brake during autonomous driving of the vehicle 100, or can be triggered to brake by a driver of the vehicle 100.

[0076] The computer vision system 134 can process and analyze images captured by the cameras 125 to identify objects and / or features in the environment surrounding the vehicle 100 and the limb and facial features of a driver within the vehicle cabin. The objects and / or features can include traffic signals, road conditions, and obstacles, and the limb and facial features of the driver can include the driver's behavior, gaze, expression, etc. The computer vision system 134 can use object recognition algorithms, structure from motion (SFM) algorithms, video tracking, and other computer vision techniques. In some embodiments, the computer vision system 134 can also be used to map the environment, track objects, estimate the speed of objects, determine driver behavior, facial recognition, etc. In embodiments of the application, the computer vision system 134 can identify the speed, location, etc. of the first object surrounding the vehicle 100.

[0077] The obstacle avoidance system 136 is configured to identify, evaluate, and avoid or otherwise navigate around potential obstacles in the environment of the vehicle 100. During autonomous driving of the vehicle 100, the obstacle avoidance system 136 can identify potential obstacles based on data detected by the sensor system 120, evaluate whether to control the braking unit 133 to brake to avoid a collision with the potential obstacle.

[0078] In embodiments of the application, the obstacle avoidance system 136 can determine the projected time to collision between the first object and the vehicle and / or the target object to trigger the braking unit 133 to brake.

[0079] In some examples, the control system 130 described above can perform S301 in FIG. 3 below and send data from the braking unit 133, the computer vision system 134, and the obstacle avoidance system 136 to the computer system 160.

[0080] Of course, in one instance, the control system 130 can include other components in addition to those shown and described, or some of the components shown can be reduced.

[0081] The vehicle 100 interacts with external sensors, other vehicles, other computer systems, or users through peripherals 140. The peripherals 140 can include a wireless communication system 141, an on-board computer 142, a microphone 143, and / or a speaker 144.

[0082] The wireless communication system 141 can wirelessly communicate with one or more devices, either directly or via a communication network. For example, the wireless communication system 141 can use 3G cellular communication, such as CDMA, EVDO, GSM / GPRS, or 4G cellular communication, such as LTE. Or 5G cellular communication. The wireless communication system 141 can communicate with a wireless local area network (WLAN) using Wi-Fi. In some embodiments, the wireless communication system 141 can communicate directly with devices using an infrared link, Bluetooth, ZigBee, or other wireless protocol. For example, the wireless communication system 141 can include one or more dedicated short range communications (DSRC) devices.

[0083] Some or all of the functionality of the vehicle 100 is controlled by a computer system 160. The computer system 160 can include at least one processor 161 that executes instructions 1621 stored in a non-transitory computer readable medium, such as a data storage device 162. The computer system 160 can also be multiple computing devices that control individual components or subsystems of the vehicle 100 in a distributed manner.

[0084] The processor 161 can be any conventional processor, such as a commercially available central processing unit (CPU). Alternatively, the processor can be a dedicated device such as an application-specific integrated circuit (ASIC) or other hardware-based processor. Although FIG. 1 functionally illustrates the processor, memory, and other elements in the same physical housing, a person having ordinary skill in the art will appreciate that the processor, computer system, or memory can actually include multiple processors, computer systems, or memories that can or can not be stored in the same physical housing. For example, data storage device can be a hard drive, or other storage medium located in a different physical housing than the other storage media. Accordingly, references to the processor computer system are to be understood to include references to a collection of processors or computer systems, or a collection of data storage devices that can or can not be operated in parallel. Unlike using a single processor to perform the steps described herein, such as some of the components of the steering assembly and the deceleration assembly, each can have its own processor that only performs calculations related to the functionality specific to the component. Like the processor 161 can be a non-volatile storage medium in the embodiments of the present application.

[0085] In various aspects described herein, the processor can be located remotely from the vehicle and in wireless communication with the vehicle. In other aspects, some of the processes described herein are performed on a processor disposed within the vehicle while others are performed by a remote processor, including taking the necessary steps to perform a single maneuver.

[0086] In some embodiments, the data storage device 162 can contain instructions 1621 (e.g., program logic) that can be executed by the processor 161 to perform various functions of the vehicle 100, including those described above. The data storage device 162 can also contain additional instructions, including instructions to send data to, receive data from, interact with, and / or control one or more of the travel system 110, the sensor system 120, the control system 130, and the peripherals 140.

[0087] In addition to the instructions 1621, the data storage device 162 can also store data, such as first data of the vehicle's location, direction, speed, and other vehicle data, and other information.

[0088] The computer system 160 can control functions of the vehicle 100 based on inputs received from various subsystems (e.g., the travel system 110, the sensor system 120, and the control system 130) and / or from the user interface 170.

[0089] The computer system 160 can perform S302-S304 in FIG. 3, S401-S404 in FIG. 4, and S501-S503 in FIG. 5. The first event is determined based on a preset rule and / or machine learning, and a first result and / or a second result is obtained. The first data is processed according to the first result and / or the second result.

[0090] Optionally, one or more of the above components can be installed separately from or associated with the vehicle 100. For example, the data storage device 162 can exist partially or entirely separately from the vehicle 100. The above components can be communicatively coupled together in a wired and / or wireless manner.

[0091] Optionally, the above components are only an example, and in actual applications, components in each module can be added or deleted as needed, and FIG. 1 should not be understood as a limitation on the embodiments of the present application.

[0092] The vehicle 100 described above can be a car, a truck, a motorcycle, a bus, a ship, an airplane, a helicopter, a lawnmower, an entertainment vehicle, an amusement park vehicle, construction equipment, a trolley, a golf cart, or a train, etc. with an automatic driving function, and the embodiments of the present application are not particularly limited.

[0093] In the structural schematic diagram of the vehicle described above, the computer system 160 can process the first data. In addition, the computer system 160 can also send the first data to a cloud server for processing. For example, in S502 in FIG. 5 described below, the vehicle reports the first data to the cloud server, or the vehicle reports the first data, the first result, and / or the second result to the cloud server. The cloud server can perform S504 to perform clustering analysis on the first data reported by the vehicle, or perform clustering analysis on the first data, the first result, and / or the second result reported by the vehicle.

[0094] As shown in FIG. 2, it is a schematic diagram of interaction between the vehicle and the cloud server. The cloud server 200 can include a receiving module 210, a processing module 220, and a storage module 230. Optionally, the cloud server 200 can include more or fewer modules, and each module can include multiple elements.

[0095] The receiving module 210 can be configured to receive the first data reported by the vehicle, or receive the first data, the first result and / or the second result reported by the vehicle. The receiving module 210 can also be configured to send the received first data to the processing module 220, or send the received first data, the first result and / or the second result to the processing module 220. The receiving module 210 can also be configured to send the received first data to the storage module 230, or send the received first data, the first result and / or the second result to the storage module 230.

[0096] The processing module 220 can be configured to receive the first data from the receiving module 210 and process the first data, or receive the first data, the first result and / or the second result from the receiving module 210 and process the first data, the first result and / or the second result. For example, the processing module 220 can perform the clustering analysis on the first data reported by the vehicle, as described in S504 of FIG. 5. The processing module 220 can also be configured to send the clustering analysis result to the storage module 230.

[0097] The storage module 230 can be configured to receive the first data from the receiving module 210 and store the first data, or receive the first data, the first result and / or the second result from the receiving module 210 and store the first data, the first result and / or the second result. The storage module 230 can also be configured to receive the clustering analysis result from the processing module 220 and store the clustering analysis result.

[0098] FIG. 3 is a flow diagram of a data processing method according to an embodiment of the present application. It should be noted that the method is not limited to the specific order described in FIG. 3 and the following description. It should be understood that in other embodiments, the order of some steps of the method can be exchanged according to actual needs, or some steps of the method can be omitted or deleted. The method includes the following steps:

[0099] S301, a vehicle obtains first data of the vehicle.

[0100] The first data can include data of the vehicle corresponding to the triggering of the first event. The data of the vehicle can include driving data and / or scene data of the vehicle in the automatic driving process.

[0101] In some examples, the data of the vehicle can include one or more of the following: a geometric feature of a road, a speed of the vehicle, a position of the vehicle, a label of a first object within a preset distance of the vehicle, a position of the first object, an estimated collision time between the first object and the vehicle, a type of the first object, a first event triggering state, a driving state of the vehicle, or an operation state of a driver.

[0102] The geometric features of the road can include information such as whether the road is a straight road, a curved road, a round section of the road or a cross section of the road, road width information, road length information, the number of lanes and the type of lane (such as a general lane, a bus lane, an emergency lane) or other geometric information of the road, etc. The first event triggering state can include whether the first event is triggered. The driving state of the vehicle can include an automatic driving state or a manual driving state. The operation state of the driver can include whether the driver steps on the brake pedal and / or the accelerator pedal, etc.

[0103] The type of the first object can be at least one of a vehicle within a preset distance, a pedestrian or a stationary road element. The stationary road element can include a stationary object in the road, such as one or more of a traffic sign, a traffic signal, a street lamp, a guardrail and a crash barrier, a curb, a drainage system, a bus station and a shelter, a monitoring device, a soundproof wall and a green belt, a structural part of a bridge and a tunnel or other stationary objects present in the road.

[0104] It should be understood that the preset distance can be customized by the user through experience or calculated by the vehicle, and the embodiments of the present application do not make specific limitations. For example, the preset distance is 5m, and the first object can include objects within 5m from the vehicle.

[0105] In some examples, the vehicle can assign a label to the first object when it detects that the first object enters within the preset distance. The label can be a unique identifier, such as an identity document (ID). For example, when pedestrian A enters within the preset distance, the pedestrian can be assigned a label 1; when pedestrian B enters within the preset distance, the pedestrian can be assigned a label 2; when bicycle C enters within the preset distance, the vehicle can be assigned a label 3; when car C enters within the preset distance, the vehicle can be assigned a label 4; when traffic signal D enters within the preset distance, the traffic signal can be assigned a label 5, etc.

[0106] In some examples, the vehicle can also detect the type of the first object and assign a label to the first object according to the type of the first object.

[0107] For example, when pedestrian A enters within the preset distance, the pedestrian can be assigned a pedestrian label 1; when pedestrian B enters within the preset distance, the pedestrian can be assigned a pedestrian label 2; when bicycle C enters within the preset distance, the vehicle can be assigned a vehicle label 1; when car C enters within the preset distance, the vehicle can be assigned a vehicle label 2; when traffic signal D enters within the preset distance, the traffic signal can be assigned a road element label 1, etc.

[0108] In some examples, the vehicle can pre-store the data of the vehicle in real time, or the vehicle can pre-store the data of the vehicle at a fixed frequency. For example, the vehicle can pre-store the data of the vehicle at a frequency of 10 Hz. In this case, the vehicle can pre-store the data of the vehicle in a cache, and delete the pre-stored data of the vehicle when there is no processing of the data of the vehicle at a preset time.

[0109] In some examples, the first data can include the data of the vehicle corresponding to the triggering of the first event. Specifically, the first data can refer to the data of the vehicle acquired within a first time period before and after the triggering of the first event. For example, the vehicle triggers the first event at a first time, and the first time period can include a second time period before the first time, the first time, and a third time period after the first time. The first time period is the sum of the second time period, the first time, and the third time period.

[0110] The first time period, the second time period, and the third time period can be user-defined, or determined by the vehicle according to the specific first event. The lengths of the first time period, the second time period, and the third time period can be different according to different first events, and the embodiments of the present application do not make specific limitations in this regard.

[0111] In some examples, the first event can include an event occurring in the automatic driving process of the vehicle, such as an AEB event, a passing intersection event, etc.

[0112] S302, the vehicle determines the first event based on the first data through a preset rule, and obtains a first result.

[0113] The first result indicates that the first event is a correct triggering, or the first result indicates that the first event is an incorrect triggering.

[0114] In some examples, the preset rule can be pre-set, and the rule can verify whether the first event is a correct triggering after the occurrence of the first event.

[0115] In some examples, the preset rule can be different when the first event is different, and the embodiments of the present application do not make specific limitations in this regard. For example, the preset rule can include: when the vehicle triggers the first event due to an automatic driving error of the vehicle, it is determined that the first event is an incorrect triggering; when the vehicle triggers the first event in the automatic driving process of the vehicle, the automatic driving of the vehicle is normal, and it is determined that the first event is a correct triggering. Specific preset rules are exemplified in the corresponding steps of FIG. 6 and FIG. 12 below, and will not be described herein.

[0116] After the occurrence of the first event in the vehicle, the vehicle determines the first event based on the first data corresponding to the first event, and determines that the first event is a correct triggering or an incorrect triggering.

[0117] In some examples, when the first result indicates that the first event is a false trigger, the first result can further include a cause of the false trigger of the first event.

[0118] S303, the vehicle determines the first event based on the first data by machine learning to obtain a second result.

[0119] The second result indicates that the first event is a correct trigger, or the second result indicates that the first event is a false trigger.

[0120] In some examples, the vehicle can input the first data into a machine learning model to obtain the second result.

[0121] In some examples, the machine learning model can be a neural network model, such as a feedforward neural network (FNN) model, a convolutional neural network (CNN) model, a recurrent neural network (RNN) model, or a long short-term memory (LSTM) model, etc. The embodiments of the present application do not make specific limitations.

[0122] For example, the vehicle can input the first data into an LSTM model to obtain an output second result. The LSTM model is used to determine whether the first event is a correct trigger or a false trigger based on the first data. Specific machine learning examples are described in the corresponding steps in FIG. 9 below, and will not be described here.

[0123] In some examples, the order of S303 and S302 is not limited by the embodiments of the present application. The vehicle can execute S303 first and then execute S302, or the vehicle can execute S302 and S303 in parallel.

[0124] S304, the vehicle processes the first data according to the first result and / or the second result.

[0125] In one embodiment, the vehicle can determine whether the first event is a correct trigger or a false trigger based on the first result and the second result, and then process the first data.

[0126] For example, when at least one of the first result and the second result indicates that the first event is a false trigger, it is determined that the first event is a false trigger, and the first data can be stored and / or reported. When the first result and the second result both indicate that the first event is a correct trigger, it is determined that the first event is a correct trigger, and the first data can be deleted.

[0127] In this way, the vehicle can automatically determine whether the first event occurring in the automatic driving process of the vehicle is a correct trigger or an incorrect trigger through the preset rules and the complementary manner of machine learning, without the need for manual confirmation, thereby improving efficiency.

[0128] For example, when the first result and the second result both indicate that the first event is an incorrect trigger, it is determined that the first event is an incorrect trigger, and the first data can be stored and / or reported; when the first result and the second result both indicate that the first event is a correct trigger, it is determined that the first event is a correct trigger, and the first data can be deleted.

[0129] In this way, when the preset rules and the machine learning both determine that the first event occurring in the automatic driving process of the vehicle is a correct trigger, it is determined that the first event is a correct trigger, thereby improving the accuracy of the determination result.

[0130] In another embodiment, the vehicle can perform S304 after performing S302, and the vehicle can determine whether the first event is a correct trigger or an incorrect trigger according to the first result.

[0131] In this way, the vehicle can automatically determine whether the first event occurring in the automatic driving process of the vehicle is a correct trigger or an incorrect trigger according to the preset rules. This can reduce the data processing amount and further improve efficiency.

[0132] In another embodiment, the vehicle can perform S303 after performing S301, and the vehicle can determine whether the first event is a correct trigger or an incorrect trigger according to the second result.

[0133] In this way, the vehicle can automatically determine whether the first event occurring in the automatic driving process of the vehicle is a correct trigger or an incorrect trigger according to the machine learning, which can reduce the data processing amount and further improve efficiency.

[0134] As described above, the vehicle determines whether the first event is a correct trigger or an incorrect trigger according to the first result and / or the second result, and processes the first data according to the first result and / or the second result, such as storing, deleting, and / or reporting the first data. The vehicle can process the first data in the computer system 160 in the vehicle.

[0135] As shown in FIG. 4, the specific process of processing the first data according to the first result and / or the second result in the computer system 160 of the vehicle can include the following steps:

[0136] S401: Whether the first result and / or the second result indicates that the first event is an incorrect trigger. If yes, perform S402; if no, perform S403.

[0137] The first result and / or the second result can be determined according to the method in S304. Details are not described herein again.

[0138] S402, the vehicle stores the first data.

[0139] In S301, the vehicle obtains and pre-stores the first data in real time during automatic driving. When the vehicle triggers the first event, and it is determined that the first event is a false trigger according to S302, S303 and S304, the vehicle can store the first data. During the automatic driving of the vehicle, multiple first events can occur, and the multiple first events can include multiple false trigger first events. The vehicle can store multiple first data corresponding to the false trigger first events.

[0140] In some examples, when the first event is a false trigger, the vehicle can also store the first result and / or the second result.

[0141] S403, the vehicle deletes the first data.

[0142] In S301, the vehicle obtains and pre-stores the first data in real time during automatic driving. When the vehicle triggers the first event, and it is determined that the first event is a true trigger according to S302, S303 and S304, the vehicle can delete the first data. During the automatic driving of the vehicle, multiple first events can occur, and the multiple first events can include multiple true trigger first events. The vehicle can delete multiple first data corresponding to the true trigger first events.

[0143] In some examples, when the first event is a true trigger, the vehicle can also store the first result and / or the second result.

[0144] S404, the vehicle performs clustering analysis on the stored first data.

[0145] In some examples, the vehicle can perform data extraction on the stored multiple first data, and perform clustering analysis on the extracted data, or perform clustering analysis on the extracted data after processing.

[0146] For example, the vehicle can extract speed data of the vehicle from the first data. The first data can include multiple speed data of the vehicle in a first time period. The vehicle can calculate the average value of the multiple speed data to obtain the average speed in the first data. The clustering analysis can be: the first data with the average speed greater than a first threshold is classified into a category, and the first data with the average speed less than or equal to the first threshold is classified into another category.

[0147] In some examples, the clustering analysis manner can include a k-means clustering algorithm, a hierarchical clustering, a density-based spatial clustering of applications with noise (DBSCAN), a mean shift clustering, a Gaussian mixture model clustering, a spectral clustering, and the like, and the embodiments of the present application do not make specific limitations thereto.

[0148] In some examples, the vehicle can pre-annotate the plurality of first data, so as to subsequently iteratively improve the vehicle.

[0149] In some examples, the vehicle can also statistically analyze the stored first result and / or second result. For example, the vehicle can count the number of times of the first event, the number of times of the first event triggered by error, and the number of times of the first event triggered correctly, and the like.

[0150] The above describes the process of the vehicle performing clustering analysis on the first data in the computer system 160 of the vehicle. In an embodiment, the vehicle can also report the first data to a cloud server according to the first result and / or second result, and perform clustering analysis on the cloud server.

[0151] S501, whether the first result and / or second result indicates that the first event is triggered by error. If yes, S502 is performed, and if no, S503 is performed.

[0152] Specifically, whether the first result and / or second result indicates that the first event is triggered by error can be determined according to the above S304. Details are not described herein again.

[0153] S502, the vehicle reports the first data to the cloud server.

[0154] Referring to the above S402, the vehicle can report the first data corresponding to the first event triggered by error in the automatic driving process to the cloud server.

[0155] In some examples, when the first event is triggered by error, the vehicle can also report the first result and / or second result to the cloud server.

[0156] S503, the vehicle deletes the first data.

[0157] Referring to the above S403, the vehicle can delete the first data corresponding to the first event triggered by error in the automatic driving process.

[0158] In some examples, when the first event is triggered correctly, the vehicle can also report the first result and / or second result to the cloud server.

[0159] S504, the cloud server performs clustering analysis on the first data.

[0160] In some examples, the cloud server can receive and store the first data reported by multiple vehicles, and perform clustering analysis on the first data reported by the multiple vehicles. Alternatively, the cloud server can receive and store the first data, the first result and / or the second result reported by multiple vehicles, and perform clustering analysis on the first data reported by the multiple vehicles.

[0161] The specific process of the cloud server performing clustering analysis on the first data can refer to S404 described above. The cloud server can perform data extraction on the reported multiple first data, and perform clustering analysis on the extracted data, or perform clustering analysis after processing the extracted data. Alternatively, the cloud server can perform data pre-labeling on the multiple first data. Alternatively, the cloud server can also perform statistics on the stored determination results of the first event, so as to subsequently iteratively improve the vehicle.

[0162] In this way, compared with all the first data obtained by the reporting vehicle in the related art, the above embodiment reduces the data reporting amount, saves traffic and cloud server resources.

[0163] The above embodiments introduce a process in which the vehicle automatically determines whether a first event occurring in the automatic driving process of the vehicle is a correct trigger or an incorrect trigger by means of a pre-set rule and machine learning complementation, and processes the first data according to the determination result. In an embodiment, the first event can include AEB time, passing through a crossroad event and the like. For example, as shown in FIG. 6, the first event can be an AEB event. When the first event is an AEB event, the specific method in which the vehicle determines the first event to be a correct trigger event or an incorrect trigger event in the above step S302 can include the following steps:

[0164] S601, the vehicle determines a target object triggering the AEB event.

[0165] In some examples, the vehicle can determine a target object triggering the AEB event according to the first data. For example, the target object can be a first object with the shortest predicted collision time with the vehicle, and the predicted collision time is less than a pre-set threshold.

[0166] Specifically, the vehicle can calculate the predicted collision time (TTC) between each first object and the vehicle according to the first data. The predicted collision time between the first object and the vehicle represents the time required for the first object and the vehicle to continue driving at the current speed and on the same path to collide. The smaller the predicted collision time between the first object and the vehicle, the higher the probability of collision between the first object and the vehicle.

[0167] Specifically, the first data includes the speed of the vehicle, the position of the vehicle, the speed of the first object, and the position of the first object. The vehicle can obtain the relative speed between the vehicle and the first object according to the speed of the vehicle and the speed of the first object, and can obtain the distance between the vehicle and the first object according to the position of the vehicle and the position of the first object. The predicted collision time between the vehicle and the first object can be the quotient of the distance between the vehicle and the first object and the relative speed between the vehicle and the first object.

[0168] In some other examples, the vehicle can obtain the predicted collision time between each first object and the vehicle from the control system 130 of the vehicle, and the embodiments of the present application do not make specific limitations thereto.

[0169] In some examples, the vehicle can classify the first objects, which can include moving pedestrians and vehicles and stationary road elements. The vehicle can determine the first object of which the type has the shortest predicted collision time among each first object. For example, the first objects include at least one pedestrian, and the pedestrian A in the at least one pedestrian has the shortest predicted collision time with the vehicle. The predicted collision time between the pedestrian A and the vehicle can be the pedestrian predicted collision time. The first objects include at least one vehicle, and the vehicle A in the at least one vehicle has the shortest predicted collision time with the vehicle. The predicted collision time between the vehicle A and the vehicle can be the vehicle predicted collision time. The first objects include at least one stationary road element, and the road element A in the at least one stationary road element has the shortest predicted collision time with the vehicle. The predicted collision time between the road element A and the vehicle can be the road element predicted collision time.

[0170] As shown in FIG. 7, a flowchart for the vehicle to determine the target object triggering the first event according to the pedestrian predicted collision time and the vehicle predicted collision time is shown. The method can include the following steps:

[0171] S701, whether the pedestrian predicted collision time and the vehicle predicted collision time are less than a second threshold value.

[0172] The second threshold value is used to distinguish whether the target object is a pedestrian or a vehicle, or the target object is a stationary road element. The second threshold value can be customized by the user, or can be preset in the vehicle. For example, the second threshold value can be 4, and the second threshold value can also be other values, and the embodiments of the present application do not make specific limitations.

[0173] If the pedestrian predicted collision time and the vehicle predicted collision time are less than the second threshold value, it means that the probability of collision between the pedestrian A or the vehicle A and the vehicle is relatively large, and the target object is determined to be the pedestrian A or the vehicle A. S702 is executed.

[0174] If the pedestrian expected collision time and the vehicle expected collision time are greater than or equal to the second threshold, indicating that the probability of collision between the pedestrian A or the vehicle A and the vehicle is small, the target object is determined as the road element A.

[0175] S702, whether the pedestrian expected collision time is less than the vehicle expected collision time.

[0176] If the pedestrian expected collision time is less than the vehicle expected collision time, indicating that the probability of collision between the pedestrian A and the vehicle is greater than the probability of collision between the vehicle A and the vehicle, the target object is determined as the pedestrian A.

[0177] If the pedestrian expected collision time is less than or equal to the vehicle expected collision time, indicating that the probability of collision between the pedestrian A and the vehicle is not greater than the probability of collision between the vehicle A and the vehicle, the target object is determined as the vehicle A.

[0178] In this way, the vehicle can determine the target object triggering the AEB event according to the first data.

[0179] In some examples, the vehicle can also obtain the target object triggering the AEB event from the control system 130 of the vehicle, and the embodiments of the present application do not make specific limitations thereto.

[0180] S602, determining whether the AEB event has a first problem based on the first data and the target object, so as to determine whether the AEB event is a false trigger or a correct trigger.

[0181] If the AEB event has the first problem, the first result indicates that the AEB event is a false trigger; if the AEB event does not have the first problem, the first result indicates that the AEB event is a correct trigger.

[0182] The first problem indicates that the target object has an abnormality.

[0183] In some examples, the vehicle can determine the scene feature of the AEB event according to the first data and the target object, and the scene feature of the AEB event can indicate whether the target object has the first problem.

[0184] The scene feature of the AEB event can include one or more of the following: the speed of the vehicle when the AEB event is triggered, the position of the vehicle when the AEB event is triggered, the type of the target object, the distance between the target object and the vehicle when the target object first appears, the position of the target object when the AEB event is triggered, the included angle between the target object and the vehicle, the trigger duration of the AEB event, the operation state of the driver, the road form or the road curvature, etc. The road form can include a curved road or a straight road.

[0185] The first problem can include one or more of the following: position jump of the target object, orientation jump of the target object, speed jump of the target object, label change of the target object, false detection of the target object in the air, or false detection of the target object as an animal, etc.

[0186] In some examples, the target object position jump can be that, in the first data obtained by the vehicle, the distance between the position of the target object at the first time and the position of the target object at the second time is greater than a third threshold value. The target object orientation jump can be that, in the first data obtained by the vehicle, the difference between the orientation of the target object at the first time and the orientation of the target object at the second time is greater than a fourth threshold value. The target object speed jump can be that, in the first data obtained by the vehicle, the difference between the speed of the target object at the first time and the speed of the target object at the second time is greater than a fifth threshold value. The target object label change can be that, in the first data obtained by the vehicle, the label of the target object at the first time is inconsistent with the label of the target object at the second time, for example, the label of the target object at the first time is label 1, and the label of the target object at the second time is label 2. The target object false detection in the air can be that, in the data of the target object obtained by the vehicle, part of the data is missing, for example, the target object exists at the first time, and the target object does not exist at the second time. The target object false detection as an animal can be that the target object is actually an animal, and the target object recognition result in S601 is a pedestrian.

[0187] The second time is the next time of the first time. The third threshold value, the fourth threshold value, and the fifth threshold value can be user-defined or pre-set in the vehicle. The embodiments of the present application do not make specific limitations on this.

[0188] In some examples, when the target object is different, the values of the third threshold value, the fourth threshold value, or the fifth threshold value can be different, for example, the speed of the pedestrian turning is greater than the speed of the vehicle turning, and the fourth threshold value when the target object is a pedestrian can be greater than the fourth threshold value when the target object is a vehicle. The embodiments of the present application do not make specific limitations on this.

[0189] As shown in FIG. 8, it is a method flow diagram for the vehicle to determine whether the AEB event is a false trigger or a correct trigger. The method can include the following steps:

[0190] S801, the vehicle determines whether to receive the operation of the driver on the brake pedal. If yes, S802 is performed; if no, S803 is performed.

[0191] In some examples, the vehicle can determine whether the vehicle receives the operation of the driver on the brake pedal through an intelligent brake assist (IBA) system. If the operation of the driver on the brake pedal is received, it means that the driver intervenes after the AEB event of the vehicle is triggered.

[0192] S802, determining whether the AEB event has a second problem, thereby determining whether the AEB event is a correct trigger or the AEB event is a false trigger.

[0193] The second problem is a part of the first problem, and the second problem includes one or more of the following: target object label change, target object false detection, target object position jump, or target object orientation jump.

[0194] If the AEB event has the second problem, the AEB event is determined to be a false trigger; if the AEB event does not have the second problem, the AEB event is determined to be a correct trigger.

[0195] When the vehicle receives the operation of the brake pedal by the driver, the driver can determine that the target object has no speed jump and is not an animal false detection, and actively steps on the brake pedal. The second problem does not include target object speed jump and animal false detection, which can improve the accuracy of the result of determining the AEB event to be a correct trigger or the AEB event to be a false trigger by the vehicle, and improve the efficiency.

[0196] S803, whether the target object is a pedestrian crossing the road. If yes, S804 is performed; if no, S805 is performed.

[0197] In some examples, when the pedestrian crosses the road, the vehicle detects a pedestrian in front of the vehicle, and triggers the AEB event to prevent the vehicle from colliding with the pedestrian.

[0198] S804, determining whether the AEB event has a third problem, so as to determine the AEB event to be a correct trigger or the AEB event to be a false trigger.

[0199] The third problem can be a part of the first problem, and the third problem includes one or more of the following: target object label change, target object false detection, or target object orientation jump.

[0200] If the AEB event has the third problem, the AEB event is determined to be a false trigger; if the AEB event does not have the third problem, the AEB event is determined to be a correct trigger.

[0201] Since the speed of the pedestrian changes quickly and the position changes slowly, the third problem does not include target object speed jump and target object position jump, which can improve the accuracy of the result of determining the AEB event to be a correct trigger or the AEB event to be a false trigger by the vehicle. In addition, determining that the target object is a pedestrian crossing the road can exclude the problem of animal false detection, the third problem does not include animal false detection, which improves the efficiency and accuracy of the result of determining the AEB event to be a correct trigger or the AEB event to be a false trigger by the vehicle.

[0202] S805, determining whether the AEB event has the first problem, so as to determine the AEB event to be a correct trigger or the AEB event to be a false trigger.

[0203] When the vehicle does not receive an operation on the brake pedal, and the target object is not a pedestrian, the vehicle can determine whether the AEB event has the first problem, and if the AEB event has the first problem, determine that the AEB event is a false trigger; if the AEB event does not have the first problem, determine that the AEB event is a correct trigger.

[0204] In this way, the vehicle can determine whether the AEB event is a false trigger or a correct trigger, and subsequently process the first data based on this.

[0205] The above describes that when the first event is an AEB event, the vehicle determines the first event by the preset rule to determine that the first event is a correct trigger event or a false trigger event. When the first event is an AEB event, the specific method of determining that the first event is a correct trigger event or a false trigger event by the vehicle through machine learning in step S303 can be as shown in FIG. 9, including the following steps:

[0206] S901, the vehicle determines a second object.

[0207] In some examples, the second object includes a first object of each type of first object that has a short predicted collision time with the vehicle, and the predicted collision time of the first object with the vehicle is less than a preset threshold.

[0208] In some examples, the vehicle can determine the second object according to the first data. The vehicle can determine a pedestrian and / or a vehicle and / or a stationary road element that has a short predicted collision time with the vehicle from the predicted collision time of at least one first object with the vehicle. The predicted collision time of at least one first object with the vehicle can be obtained from the control system 130, or the predicted collision time of at least one first object with the vehicle is calculated according to the first data, which is specifically described in S601 above and will not be repeated here.

[0209] For example, the first object includes at least one pedestrian, a pedestrian A in the at least one pedestrian has the shortest predicted collision time with the vehicle, and the predicted collision time between the pedestrian A and the vehicle is less than a preset threshold, then the pedestrian A is determined as the second object; the first object includes at least one vehicle, a vehicle A in the at least one vehicle has the shortest predicted collision time with the vehicle, and the predicted collision time between the vehicle A and the vehicle is less than a preset threshold, then the vehicle A is determined as the second object. The first object includes at least one stationary road element, a road element A in the at least one stationary road element has the shortest predicted collision time with the vehicle, and the predicted collision time between the road element A and the vehicle is less than a preset threshold, then the road element A is determined as the second object.

[0210] Or, the first object includes at least one pedestrian and at least one vehicle, a pedestrian A in the at least one pedestrian has the shortest predicted collision time with the vehicle, a vehicle A in the at least one vehicle has the shortest predicted collision time with the vehicle, the predicted collision time between the pedestrian A and the vehicle is less than a preset threshold, and the predicted collision time between the vehicle A and the vehicle is less than the preset threshold, and the pedestrian A and the vehicle A are determined as the second object; the first object includes at least one pedestrian and at least one static road element, a pedestrian A in the at least one pedestrian has the shortest predicted collision time with the vehicle, a road element A in the at least one static road element has the shortest predicted collision time with the vehicle, the predicted collision time between the pedestrian A and the vehicle is less than a preset threshold, and the predicted collision time between the road element A and the vehicle is less than the preset threshold, and the pedestrian A and the road element A are determined as the second object; the first object includes at least one vehicle and at least one static road element, a vehicle A in the at least one vehicle has the shortest predicted collision time with the vehicle, a road element A in the at least one static road element has the shortest predicted collision time with the vehicle, the predicted collision time between the vehicle A and the vehicle is less than a preset threshold, and the predicted collision time between the road element A and the vehicle is less than the preset threshold, and the pedestrian A and the road element A are determined as the second object.

[0211] Or, the first object includes at least one pedestrian, at least one vehicle and at least one static road element, a pedestrian A in the at least one pedestrian has the shortest predicted collision time with the vehicle, a vehicle A in the at least one vehicle has the shortest predicted collision time with the vehicle, a road element A in the at least one static road element has the shortest predicted collision time with the vehicle, the predicted collision time between the pedestrian A and the vehicle is less than a preset threshold, the predicted collision time between the vehicle A and the vehicle is less than the preset threshold, and the predicted collision time between the vehicle A and the vehicle is less than the preset threshold, and the pedestrian A, the vehicle A and the road element A are determined as the second object.

[0212] In some examples, the vehicle can obtain the second object from the control system 130 of the vehicle, and the embodiments of the present application do not make specific limitations thereto.

[0213] S902, the vehicle inputs first data corresponding to the second object into the first model.

[0214] The first data corresponding to the second object can include a label of the second object, a predicted collision time between the second object and the vehicle, and an AEB event triggering state.

[0215] In some examples, the input of the first model can include a label of the pedestrian, a predicted time to collision between the pedestrian and the vehicle, a label of the vehicle, a predicted time to collision between the vehicle and the vehicle, and an AEB event trigger state. Alternatively, the input of the first model can include a label of the pedestrian, a predicted time to collision between the pedestrian and the vehicle, a label of the vehicle, a predicted time to collision between the vehicle and the vehicle, a label of a stationary road element, a predicted time to collision between the stationary road element and the vehicle, and an AEB event trigger state. The embodiments of the present application do not make a specific limitation in this regard.

[0216] In the embodiments of the present application, the input of the first model can include a label of the pedestrian, a predicted time to collision between the pedestrian and the vehicle, a label of the vehicle, a predicted time to collision between the vehicle and the vehicle, and an AEB event trigger state.

[0217] In some examples, the label of the second object, the predicted time to collision between the second object and the vehicle, and the AEB event trigger state can be input into the first model in the form of a waveform.

[0218] As shown in FIG. 10, the input of the first model can be a waveform of the label of the pedestrian, a waveform of the predicted time to collision between the pedestrian and the vehicle, a waveform of the label of the vehicle, a waveform of the predicted time to collision between the vehicle and the vehicle, and a waveform of the AEB event trigger state. Taking the example that the second object includes the vehicle A.

[0219] In the time period corresponding to the waveforms shown in FIG. 10, the predicted time to collision between the vehicle A and the vehicle within the preset distance of the vehicle is the shortest and is less than a preset threshold, and the vehicle triggers the AEB event. The label of the vehicle A can be 2.

[0220] As shown in FIG. 10(a), it is a waveform diagram of the label of the pedestrian. In the waveform diagram of the label of the pedestrian, the horizontal coordinate is time, and the vertical coordinate is the state of the pedestrian (whether the second object includes the pedestrian). If the second object includes the pedestrian, the state value of the waveform is 1, and if the second object does not include the pedestrian, the state value of the waveform is 0. As shown in FIG. 10(a), the second object does not include the pedestrian, and the state value of the waveform is 0.

[0221] As shown in FIG. 10(b), it is a waveform diagram of the predicted time to collision between the pedestrian and the vehicle. In the waveform diagram of the predicted time to collision between the pedestrian and the vehicle, the horizontal coordinate is time, and the vertical coordinate is the predicted time to collision between the pedestrian and the vehicle. As shown in FIG. 10(b), the second object does not include the pedestrian, and the predicted time to collision between the pedestrian and the vehicle is 0.

[0222] As shown in FIG. 10(c), a waveform diagram of the vehicle 2 (the tag of the vehicle A) is shown, where the horizontal axis of the waveform diagram of the vehicle 2 is time, and the vertical axis is the state of the vehicle 2 (whether the second object includes the vehicle 2). If the second object includes the vehicle 2, the state value of the waveform is 1, and if the second object does not include the vehicle 2, the state value of the waveform is 0. The tag of the vehicle A is detected within the preset distance of the vehicle, and the vehicle A enters the preset distance of the vehicle at time t1, and travels out of the preset distance of the vehicle at time t2. As shown in FIG. 10(c), the second object is detected to include the vehicle 2 at time t1, that is, the vehicle 2 (vehicle A) satisfies the vehicle with the shortest predicted collision time with the vehicle at time t1, and the predicted collision time between the vehicle 2 and the vehicle is less than the preset threshold. The second object is detected to not include the vehicle 2 at time t2, that is, the vehicle 2 (vehicle A) does not satisfy the vehicle with the shortest predicted collision time with the vehicle at time t1, and the predicted collision time between the vehicle 2 and the vehicle is less than the preset threshold. The waveform state value of the vehicle 2 is 1 at time t1-t2, and the waveform state value of the vehicle 2 is 0 at other times.

[0223] As shown in FIG. 10(d), a waveform diagram of the predicted collision time between the vehicle 2 and the vehicle is shown, where the horizontal axis of the waveform diagram of the predicted collision time between the vehicle 2 and the vehicle is time, and the vertical axis is the predicted collision time between the vehicle 2 and the vehicle. As shown in FIG. 10(d), the second object is detected to include the vehicle 2 at time t1, and the predicted collision time between the vehicle 2 and the vehicle gradually increases at time t1, and the predicted collision time between the vehicle 2 and the vehicle is greater than the second threshold at time t3, which may trigger an AEB event. The predicted collision time between the vehicle 2 and the vehicle is less than the second threshold after time t4, which does not trigger an AEB event. The second object is detected to not include the vehicle 2 at time t2, and the waveform returns to the original value. The waveform changes at time t1-t2, and the value of the waveform at time t1-t2 is the predicted collision time between the vehicle A and the vehicle.

[0224] As shown in FIG. 10(e), a waveform diagram of the AEB event triggering state is shown, where the horizontal axis of the waveform diagram of the AEB event triggering state is time, and the vertical axis is the AEB event triggering state (whether the AEB event is triggered). The predicted collision time between the vehicle 2 and the vehicle is less than the second threshold at time t3, and the AEB event is triggered at time t3; the predicted collision time between the vehicle 2 and the vehicle is greater than the second threshold at time t4, and the AEB event is triggered at time t4. The triggering duration of the AEB event is t3-t4. As shown in FIG. 10(e), the waveform state value of the AEB event triggering state is 1 at time t3-t4, and the waveform state value of the AEB event triggering state is 0 at other times.

[0225] In some examples, the vehicle can input the above-mentioned waveforms in FIG. 10 into the first model.

[0226] S903, the first model outputs a second result.

[0227] The first model can be a machine learning model.

[0228] For example, the first model can be a long short-term memory network model, which processes the waveform in FIG. 10 as input and outputs the second result. It is determined whether the AEB event is a correct trigger or an incorrect trigger.

[0229] In some examples, the first model can be trained by the label of the historical first object, the predicted collision time of the historical first object and the vehicle, and the historical first event trigger state.

[0230] It should be understood that in the above embodiments, the label of the first object, the predicted collision time of the first object and the vehicle, and the first event trigger state are taken as examples to obtain the second result. In the embodiments of the present application, the input of the first model also includes other, such as the speed of the vehicle, etc., which is not limited in the embodiments of the present application.

[0231] In this way, the vehicle can determine the second result through machine learning.

[0232] The above introduces that when the first event is an AEB event, the vehicle determines the first event to be a correct trigger event or an incorrect trigger event through machine learning. In this way, when the first event is an AEB event, the vehicle can obtain the first result and the second result, and process the first data according to the first result and the second result.

[0233] In an embodiment, the vehicle can also store or report the first result and / or the second result. The first result can include the reason for the incorrect trigger of the AEB event, such as the first problem (such as target object position jump) causing the incorrect trigger of the AEB event. The first result can also include the scene characteristics of the first event. In this way, the vehicle or the cloud server can perform clustering analysis, etc. according to the first data, the first result and / or the second result,

[0234] For example, as shown in FIG. 11, a schematic diagram of a scenario in which the position jump of vehicle A causes the vehicle to incorrectly trigger the AEB event, where vehicle A is the target object.

[0235] The vehicle detects that vehicle A is located at position A at a first time, and the speed is 50 m / s; the vehicle detects that vehicle A is located at position B at a second time, and the speed is 35 m / s. The distance between position A and position B is greater than a third threshold. The vehicle determines that the AEB event is an incorrect trigger according to a preset rule and / or machine learning.

[0236] The data stored or reported by the vehicle can be that the first result and / or the second result indicate that the AEB event is a false trigger. The first result: the AEB event is a false trigger; target object position jump: target object label: vehicle 2, AEB event triggering time: the third time, jump distance: 25 m, jump time: the first time->the second time, jump point: position A->position B; scene characteristics of the first event: speed of the vehicle when the AEB event is triggered: 30 m / s, type of the target object: vehicle, distance between the target object and the vehicle when the target object first appears: 4.1 m, position of the target object when the AEB event is triggered: position 3, included angle between the target object and the vehicle: 10 degrees, triggering time of the AEB event: 0.6 s, operation state of the driver: not stepping on the brake pedal and the accelerator pedal, the road is a curve, road curvature: 0.003; the second result: the AEB event is a false trigger.

[0237] In some examples, the first result can further include other data, for example, the first result can further include: whether the target object is a stationary road element: no, no stationary road element triggering the AEB event is found within a preset distance. The embodiments of the present application do not make specific limitations on this.

[0238] In the above-mentioned stored or reported data, the position data can be represented by latitude and longitude data, or can also be represented by other ways, and the embodiments of the present application do not make specific limitations on this.

[0239] In some examples, the first problem causing the vehicle to falsely trigger the AEB event can be multiple, for example, the first problem can include target object position jump and target object speed jump,

[0240] For example, the vehicle detects vehicle A at position A at the first time, and the speed is 50 m / s; the vehicle detects vehicle A at position B at the second time, and the speed is 35 m / s. The distance between position A and position B is greater than a third threshold, and the difference between the speed of vehicle A at the first time and the speed of vehicle A at the second time is greater than a fifth threshold. The vehicle determines that the AEB event is a false trigger according to a preset rule and / or machine learning.

[0241] The vehicle can also store or report the following content: the first result and / or the second result indicating that the AEB event is a false trigger. The first result: the AEB event is a false trigger; target object position jump: target object label: vehicle 2, AEB event triggering time: third time, jump distance: 5 m, jump time: first time->second time, jump point: position A->position B; target object speed jump: target object label: vehicle 2, AEB event triggering time: third time, jump speed: 15 m / s, jump time: first time->second time, jump point: position A->position B; scene characteristics of the first event: speed of the vehicle when the AEB event is triggered: 30 m / s, type of the target object: vehicle, distance between the target object and the vehicle when the target object first appears: 4.1 m, position of the target object when the AEB event is triggered: position 3, angle between the target object and the vehicle: 10 degrees, triggering duration of the AEB event: 0.6 s, operation state of the driver: not stepping on the brake pedal and the accelerator pedal, the road is a curve, road curvature: 0.003; the second result: the AEB event is a false trigger.

[0242] In some examples, the vehicle or the cloud server can perform cluster analysis on the first data, the first result and / or the second result corresponding to the AEB event.

[0243] For example, the data (such as vehicle accelerator value data) in the first data is extracted to generate a data table of the AEB event together with the first result and / or the second result, so as to facilitate subsequent analysis according to the content in the classification table.

[0244] For example, as shown in Table 1 below, the data table of the AEB event can include.

[0245] Table 1

[0246] As shown in Table 1, the vehicle accelerator amount corresponding to AEB event A is 0.61, the triggering duration of the AEB event is 0.6 s, the type of the target object is a vehicle, the speed of the target object when the AEB event is triggered is 30 km / h, and the position of the target object when the AEB event is triggered is position C. The vehicle accelerator amount corresponding to AEB event B is 0.4, the triggering duration of the AEB event is 1.52 s, the type of the target object is a pedestrian, the speed of the target object when the AEB event is triggered is 5 km / h, and the position of the target object when the AEB event is triggered is position D.

[0247] It should be understood that the first data, the first result and / or the second result corresponding to the AEB event in Table 1 above can also include other more or less data, and the AEB event can also include other more or less AEB events, which are not specifically limited by the embodiments of the application.

[0248] In some examples, the vehicle or the cloud server can pre-label the data according to the clustering analysis result, improve the analysis efficiency of the AEB event, and be used for iterative improvement of the vehicle to reduce the probability of false triggering of the AEB event of the vehicle.

[0249] The above introduces that when the first event is an AEB event, the vehicle automatically determines whether the first event occurring in the automatic driving process of the vehicle is a correct trigger or a false trigger through a preset rule and a machine learning complementary manner, and processes the first data according to the determination result. In another example, the first event can also be a through intersection event. When the first event can also be a through intersection event, the specific method for the vehicle to determine the first event as a correct trigger event or a false trigger event in step S302 of the above can be as shown in FIG. 12, including the following steps:

[0250] S1201, the vehicle determines the driving state of the vehicle based on the first data.

[0251] In some examples, the driving state of the vehicle includes an automatic driving state and / or a manual driving state.

[0252] The road can be divided into an intersection area and a non-intersection area, wherein the intersection area can include a pre-intersection area, a middle-intersection area, and a post-intersection area. The pre-intersection area refers to an area within a certain distance range in front of the intersection from the driving direction of the vehicle. The specific range of this area can be adjusted according to actual needs, such as including the area from the beginning of the solid line prohibiting lane changing to the area before the stop line of the intersection. The middle-intersection area refers to an area within the middle range of the intersection. The specific range of this area can be adjusted according to actual needs, such as including the area before the stop line of the intersection. The post-intersection area refers to an area within a certain distance range in front of the intersection from the driving direction of the vehicle. The specific range of this area can be adjusted according to actual needs, such as including the area from the stop line of the intersection to the end of the solid line prohibiting lane changing. The non-intersection area refers to other areas of the road except the intersection area.

[0253] For example, as shown in FIG. 13, a schematic diagram of an intersection in a road is shown, area a is a pre-intersection area, area b is a middle-intersection area, area c is a post-intersection area, and area d is a non-intersection area.

[0254] When the vehicle straightly passes through the intersection, the driving state of the vehicle can include: the vehicle automatically drives through the intersection area, and the vehicle automatically drives into the intersection area and switches to manually drive through the intersection area.

[0255] For example, the vehicle switches to manual driving to pass through the remaining intersection front area a, intersection middle area b and exit intersection area c after the vehicle automatically drives into the intersection front area a; or the vehicle automatically drives through the area a after the vehicle automatically drives into the intersection front area a, switches to manual driving to pass through the remaining intersection middle area b and exit intersection area c after the vehicle automatically drives into the intersection middle area b; or the vehicle automatically drives through the area a and intersection middle area b after the vehicle automatically drives into the intersection front area a, switches to manual driving to pass through the remaining exit intersection area c after the vehicle automatically drives into the exit intersection area c.

[0256] In S1202, the vehicle determines a first result based on a driving state of the vehicle.

[0257] In some examples, the driving state of the vehicle is that the vehicle switches to manual driving to pass through the intersection area after the vehicle automatically drives into the intersection area, and the first result indicates that the intersection passing event is a false trigger; the driving state of the vehicle is an automatic driving state, and the first result indicates that the intersection passing event is a true trigger.

[0258] The above describes that when the first event is the intersection passing event, the vehicle can automatically determine whether the first event is a true trigger or a false trigger in the automatic driving process of the vehicle according to a preset rule, and the vehicle can process the first data according to the first result of the determination.

[0259] For example, when the first result indicates that the intersection passing event is a false trigger, the vehicle can store or report the first data corresponding to the intersection passing event; or the vehicle can store or report the first data corresponding to the intersection passing event and the first result. When the first result indicates that the intersection passing event is a true trigger, the vehicle can delete the first data corresponding to the intersection passing event; or the vehicle can delete the first data corresponding to the intersection passing event and store or report the first result.

[0260] Subsequently, the vehicle or the cloud server can perform clustering analysis and the like on the first data and / or the first result corresponding to the intersection passing event.

[0261] For example, the vehicle or the cloud server can determine the number of times of passing through the intersection of the vehicle according to the first data and / or the first result; or the vehicle or the cloud server can determine the intersection passing rate of the vehicle according to the first data and / or the first result, wherein the intersection passing rate can be a ratio of the number of times of the intersection passing event being a true trigger to the number of times of the intersection passing event. Or, the vehicle or the cloud server can obtain other clustering analysis results according to the first data and / or the first result, which are not limited in the embodiments of the application.

[0262] In this way, the vehicle or the cloud server can perform iterative upgrading and the like on the vehicle according to the clustering analysis results.

[0263] It should be appreciated that some of the operations in the flowcharts above can be optional depending on the embodiment. And / or the order of some operations can be changed. Also, the order of execution or performance of the operations of the flowcharts can be changed. Not all of the operations are required, and some of the operations can be rearranged. Further, some of the operations can be performed simultaneously. Examples of these and other alternatives are also within the scope of the present disclosure. Moreover, while the procedures of the methods are described in a particular sequential order, other procedures can be performed in parallel or concurrently. Also, some or all of the procedures can be repeated. The methods can also incorporate adaptivity or flexibility, such that some operations can be performed more than once, others can be performed only once, others can be omitted, and others can be performed in an order that is different than the order described.

[0264] Further, some of the steps of the method embodiments can be replaced by other possible steps. Or, some of the steps of the method embodiments can be optional, and can be deleted in some use scenarios. Or, other possible steps can be added in the method embodiments.

[0265] Also, the method embodiments above can be implemented individually, or in combination.

[0266] Some other embodiments of the present disclosure provide a device, which can be the data processing or a component (such as a chip system) in the data processing.

[0267] The device can include a memory and one or more processors. The memory and the processor are coupled. The memory is configured to store computer program code, the computer program code including computer instructions. When the processor executes the computer instructions, the data processing device can perform each function or step performed by the mobile phone in the method embodiments above.

[0268] The core structure of the data processing device can be represented as the structure shown in FIG. 14, which includes a processing module 151, a storage module 152, and a communication module 153.

[0269] The processing module 151 can include at least one of a central processing unit (CPU), an application processor (AP), or a communication processor (CP). The processing module 151 can perform operations or data processing related to the control and / or communication of at least one of the other elements of the data processing device. Optionally, the processing module 151 is configured to determine a first result and / or a second result based on the first data, and to process the first data according to the first result and / or the second result.

[0270] The storage module 152 can include volatile memory and / or non-volatile memory. The storage module is configured to store at least one relevant instruction or data of other modules of the user terminal device. Optionally, the storage module 152 is configured to store the first data, the first result and / or the second result.

[0271] The communication module 153 is configured to support the personal terminal to communicate with other personal terminals (through a communication network). For example, the communication module can be connected to a network via wireless communication or wired communication to communicate with other personal terminals or network servers. The wireless communication can employ at least one of cellular communication protocols, such as Long Term Evolution (LTE), Long Term Evolution-Advanced (LTE-A), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), Universal Mobile Telecommunications System (UMTS), Wireless Broadband (WiBro), or Global System for Mobile Communications (GSM). The wireless communication can include, for example, short-range communication. The short-range communication can include at least one of Wireless-Fidelity (Wi-Fi), Bluetooth, Near Field Communication (NFC), Magnetic Stripe Transmission (MST), or GNSS. Optionally, the communication module 154 is configured to communicate with a cloud server, and report the first data, the first result and / or the second result to the cloud server.

[0272] The apparatus shown in FIG. 14 can further include more, fewer, or different components, or split components, or have other arrangements of components, which are not limited in the embodiments of the present application.

[0273] The embodiments of the present application also provide a chip system, as shown in FIG. 15, which includes at least one processor 161 and at least one interface circuit 162. The processor 161 and the interface circuit 162 can be interconnected by a line. For example, the interface circuit 162 can be configured to receive a signal from another device (e.g., a memory of a data processing device). For another example, the interface circuit 162 can be configured to send a signal to another device (e.g., the processor 161). For example, the interface circuit 162 can read an instruction stored in the memory and send the instruction to the processor 161. When the instruction is executed by the processor 161, the data processing device can perform various steps in the above-described embodiments. Of course, the chip system can also include other discrete devices, which are not limited in the embodiments of the present application.

[0274] The embodiments of the present application also provide a computer storage medium, which includes computer instructions, when the computer instructions are run on the above-described data processing device, the data processing device performs various functions or steps performed by the mobile phone in the above-described method embodiments.

[0275] The embodiment of the present application further provides a computer program product, which, when running on a computer, enables the computer to perform each function or step of the method embodiment described above.

[0276] Through the above description of the embodiments, those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional modules is taken as an example for illustration, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device and module described above can refer to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0277] In several embodiments provided in the present application, it should be understood that the disclosed device and method can be implemented by other ways. For example, the device embodiment described above is only schematic, for example, the division of the modules or units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another device, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some interface, device or unit, and can be electrical, mechanical or other forms.

[0278] The units described as separate components can or can not be physically separate, and the components shown as units can be one physical unit or multiple physical units, that is, can be located in one place or distributed to multiple different places. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0279] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0280] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application essentially or say the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The software product is stored in a storage medium, including a plurality of instructions to make a device (which can be a single-chip microcomputer, a chip, etc.) or a processor execute all or part of the steps of the method described in various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0281] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any change or replacement within the technical scope disclosed in the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A data processing method, characterized by, The method comprises: acquiring first data of a vehicle; the first data comprises data corresponding to a first event, the first event comprising an event occurring in an automatic driving process of the vehicle; determining the first event based on the first data by a preset rule to obtain a first result; the first result indicates that the first event is a correct trigger, or the first result indicates that the first event is an incorrect trigger; determining the first event based on the first data by machine learning to obtain a second result; the second result indicates that the first event is a correct trigger, or the second result indicates that the first event is an incorrect trigger; processing the first data according to the first result and / or the second result.

2. The method of claim 1, wherein, The method comprises: determining a target object triggering the first event; determining whether the first event has a first problem based on the first data and the target object; the first problem indicates that the target object has an anomaly; if the first event has the first problem, the first result indicates that the first event is an incorrect trigger; if the first event does not have the first problem, the first result indicates that the first event is a correct trigger.

3. The method of claim 2, wherein, The first problem comprises one or more of the following: position jump of the target object, orientation jump of the target object, speed jump of the target object, label change of the target object, empty mis-detection of the target object, or animal mis-detection of the target object; the empty mis-detection of the target object indicates that part of the data of the target object in the first data is missing.

4. The method of claim 2, wherein, After the first event is triggered, an operation of a driver on a brake pedal is received, the first problem is a second problem, and the second problem comprises one or more of the following: label change of the target object, empty mis-detection of the target object, position jump of the target object, or orientation jump of the target object.

5. The method of claim 2, wherein, In a case where the target object is a pedestrian and the pedestrian crosses a road, the first problem is a third problem, and the third problem comprises one or more of the following: label change of the target object, empty mis-detection of the target object, or orientation jump of the target object.

6. The method according to any one of claims 1 to 5, characterized in that, The method comprises: determining a second object, the second object comprising a first object of a type of the first object and having a predicted collision time with the vehicle less than a preset threshold; the type of the first object comprises a vehicle, a pedestrian, or a static road element; inputting first data corresponding to the second object into a first model; the first data comprises a label of the second object, a predicted collision time of the second object with the vehicle, and a triggering state of the first event; the first model is trained based on a label of a historical second object, a predicted collision time of the historical second object with the vehicle, and a triggering state of a historical first event; determining the first event based on the first model to obtain the second result.

7. The method according to any one of claims 1 to 6, characterized in that, The processing of the first data according to the first result and / or the second result comprises: storing and / or reporting the first data when the first result and / or the second result indicates that the first event is a false trigger; deleting the first data when the first result and the second result indicates that the first event is a correct trigger.

8. The method of claim 1, wherein, The first event comprises a pass-through intersection event, and the first event is determined based on the first data according to a preset rule to obtain a first result, which comprises: determining a driving state of the vehicle based on the first data packet, wherein the driving state of the vehicle comprises an automatic driving state and / or a manual driving state; when the driving state of the vehicle is switched from the automatic driving state to the manual driving state, the first result indicates that the first event is a false trigger; and when the driving state of the vehicle is the automatic driving state, the first result indicates that the first event is a correct trigger.

9. The method according to any one of claims 2-5, characterized in that, The type of the target object comprises a vehicle, a pedestrian, or a static road element.

10. The method of claims 2-7, wherein, determining whether the first event has a first problem based on the first data and the target object, which comprises: determining a scene feature of the first event based on the first data and the target object, wherein the scene feature of the first event indicates whether the target object has a first problem.

11. The method of claim 10, wherein, The method further comprises: performing clustering analysis based on the first data, the first result, and / or the second result, wherein the first result comprises a scene feature of the first event, and the scene feature of the first event comprises one or more of the following: a speed of the vehicle when the first event is triggered, a position of the vehicle when the first event is triggered, a type of the target object, a distance between the target object and the vehicle when the target object first appears, a position of the target object when the first event is triggered, an included angle between the target object and the vehicle, a trigger duration of the first event, an operation state of a driver, a road shape, or a road curvature.

12. The method according to any one of claims 1 to 11, characterized in that, The first data further comprises one or more of the following: a geometric feature of a road, a speed of the vehicle, a position of the vehicle, a first object within a preset distance of the vehicle, a position of the first object, an estimated collision time between the first object and the vehicle, a type of the first object, a trigger state of the first event, a driving state of the vehicle, or an operation state of a driver.

13. A data processing apparatus, characterized by: The data processing apparatus comprises: a processor and a memory coupled to the processor, wherein the memory is configured to store computer program codes, and the computer program codes comprise computer instructions, and when the processor reads the computer instructions from the memory, the data processing apparatus is enabled to perform the method according to any one of claims 1-12.

14. A vehicle characterized by comprising: The data processing apparatus is configured to perform the method according to any one of claims 1-12.

15. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1-12 is implemented.

Images