A data processing method and device for intelligent driving

Abstract

The present invention discloses a data processing method and device for intelligent driving, including: when the driver intervenes in the current driving assistance system, collecting intervention-related information; inputting the intervention-related information into the intention recognition model to obtain the driver's intervention intention information; detecting whether the evaluation information for the driving assistance system after intervention is collected, and if so, determining the target intervention intention information; inputting the intervention-related information, the target intervention intention information and the data information of the driving assistance system into the parameter correction model to obtain the corrected key parameters; performing safety verification on the corrected key parameters, and if the verification passes, adjusting the current driving assistance system based on the corrected key parameters to obtain the adjusted driving assistance system. The present invention performs self-learning of the automatic driving system within the driving safety envelope to improve the performance, safety and user experience of the driving assistance system.

Description

Technical Field

[0001] The present invention relates to the field of intelligent driving technology, and in particular to a data processing method and device applied to intelligent driving. Background Art

[0002] With the development of intelligent driving technology, intelligent driving provides people with many conveniences when driving vehicles. Most of the existing driving assistance systems used in the field of intelligent driving are based on finite state machines and generally do not exceed three levels of sensitivity and fixed parameters. However, in the face of different requirements of different drivers for comfort and safety, there is often a problem that the needs of all users cannot be met.

[0003] With the development of machine learning technology, there are methods for personalizing driving assistance systems based on driver operation data through machine learning. However, the existing methods have the following problems: First, when processing driver data, the impact of improper driver operation or even "dangerous driving" is not considered, which may cause safety hazards in the "personalized" driving assistance system. Secondly, due to the limited trust of drivers in driving assistance systems, the driver's own operating style and the driver's working expectations of ADAS (Advanced Driving Assistance System) are not exactly the same. Directly learning the driver's operation data may aggravate the driver's distrust of the ADAS system; In addition, the huge amount of unlabeled driver operation data has too large a sample size and unclear features for machine learning, which greatly reduces the efficiency and accuracy of self-learning; At the same time, the existing personalization methods do not introduce the driver's evaluation information, and cannot make targeted adjustments and optimizations to the system in a timely manner. All of these will lead to the final personalized intelligent driving system often failing to meet the needs of drivers; Most of the existing personalization solutions use machine learning to directly output control quantities (such as steering wheel angle / torque, accelerator pedal opening, etc.), which cannot be integrated with the original model-based intelligent driving system, making its safety, reliability and robustness problematic.

[0004] It can be seen that the existing intelligent driving machine learning processing cannot meet the personalized needs of users well, and there are problems with poor performance and safety. Summary of the invention

[0005] In response to the above problems, the present invention provides a data processing method and device for intelligent driving, which achieves the purpose of improving the performance, safety and user experience of the driving assistance system.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A data processing method for intelligent driving, comprising:

[0008] When the driver intervenes in the current driving assistance system, intervention-related information is collected, where the intervention-related information includes the driver's intervention operation information, vehicle environment information, and vehicle status information;

[0009] Inputting the intervention-related information into an intention recognition model to obtain the driver's intervention intention information;

[0010] Detect whether evaluation information for the driver assistance system after intervention is collected, and if so, determine target intervention intention information;

[0011] Inputting the intervention association information, target intervention intention information and data information of the driving assistance system into a parameter correction model to obtain corrected key parameters;

[0012] The corrected key parameters are safety verified, and if the verification passes, the current driving assistance system is adjusted based on the corrected key parameters to obtain an adjusted driving assistance system.

[0013] Optionally, the method further comprises:

[0014] The adjusted driving assistance system is virtually operated to obtain an operation result, so as to determine the safety of the adjusted driving assistance system according to the operation result.

[0015] Optionally, the method further comprises:

[0016] If the evaluation information for the post-intervention driving assistance system is not collected, the target intervention intention information is determined based on the intention probability value in the intervention intention information.

[0017] Optionally, the method further comprises:

[0018] The intervention intention information is stored according to the type of the driving assistance subsystem corresponding to the intervention intention, so that the driving assistance system is adjusted based on the stored information.

[0019] Optionally, the performing safety verification on the corrected key parameters includes:

[0020] Acquire a constraint function corresponding to the corrected key parameter, wherein the constraint function includes boundary information of each parameter;

[0021] Based on the current vehicle environment information, calculate and obtain the safety parameters corresponding to the current vehicle environment;

[0022] Based on the constraint function and the safety parameter, the corrected key parameter is safety verified to obtain a verification result.

[0023] Optionally, the method further comprises:

[0024] generating prompt information corresponding to the adjusted driving assistance system, wherein the prompt information is used to prompt the driver whether to update the current driving assistance system;

[0025] If the received feedback information for the prompt information meets the update condition, the current driving assistance system is updated according to the adjusted driving assistance system.

[0026] Optionally, the method further comprises:

[0027] Based on the evaluation information, the intention recognition model is modified to obtain a modified intention recognition model.

[0028] A data processing device for intelligent driving, comprising:

[0029] A collection unit, used to collect intervention-related information when the driver intervenes in the current driving assistance system, wherein the intervention-related information includes the driver's intervention operation information, vehicle environment information, and vehicle status information;

[0030] an intention recognition unit, configured to input the intervention-related information into an intention recognition model to obtain the driver's intervention intention information;

[0031] a determination unit, configured to detect whether evaluation information for the driver assistance system after intervention is collected, and if so, determine target intervention intention information;

[0032] A parameter correction unit, used for inputting the intervention association information, the target intervention intention information and the data information of the driving assistance system into a parameter correction model to obtain corrected key parameters;

[0033] The adjustment unit is used to perform safety verification on the corrected key parameters, and if the verification passes, adjust the current driving assistance system based on the corrected key parameters to obtain an adjusted driving assistance system.

[0034] Optionally, the device further comprises:

[0035] The virtual operation unit is used to perform a virtual operation on the adjusted driving assistance system to obtain an operation result, so as to determine the safety of the adjusted driving assistance system according to the operation result.

[0036] Optionally, the device further comprises:

[0037] The intention determination unit is used to determine the target intervention intention information based on the intention probability value in the intervention intention information if the evaluation information for the post-intervention driving assistance system is not collected.

[0038] Optionally, the device further comprises:

[0039] The storage unit is used to store the intervention intention information according to the type of the driving assistance subsystem corresponding to the intervention intention, so that the driving assistance system can be adjusted based on the stored information.

[0040] Optionally, the adjustment unit includes:

[0041] A verification subunit is used to perform security verification on the corrected key parameters. The verification subunit is specifically used to:

[0042] Acquire a constraint function corresponding to the corrected key parameter, wherein the constraint function includes boundary information of each parameter;

[0043] Based on the current vehicle environment information, calculate and obtain the safety parameters corresponding to the current vehicle environment;

[0044] Based on the constraint function and the safety parameter, the corrected key parameter is safety verified to obtain a verification result.

[0045] Optionally, the device further comprises:

[0046] a generating unit, configured to generate prompt information corresponding to the adjusted driving assistance system, wherein the prompt information is used to prompt the driver whether to update the current driving assistance system;

[0047] An updating unit is configured to update the current driving assistance system according to the adjusted driving assistance system if the received feedback information for the prompt information meets an updating condition.

[0048] Optionally, the device further comprises:

[0049] The model correction unit is used to correct the intention recognition model based on the evaluation information to obtain a corrected intention recognition model.

[0050] A storage medium stores executable instructions, which, when executed by a processor, implement a data processing method for intelligent driving as described in any one of the above.

[0051] An electronic device, comprising:

[0052] Memory, used to store programs;

[0053] A processor is used to execute the program, and the program is specifically used to implement the data processing method applied to intelligent driving as described in any one of the above.

[0054] Compared with the prior art, the present invention provides a data processing method and device for intelligent driving, including: when the driver intervenes in the current driving assistance system, collecting intervention-related information, the intervention-related information includes the driver's intervention operation information, vehicle environment information and vehicle status information; inputting the intervention-related information into the intention recognition model to obtain the driver's intervention intention information; detecting whether the evaluation information for the driving assistance system after intervention is collected, and if so, determining the target intervention intention information; inputting the intervention-related information, the target intervention intention information and the data information of the driving assistance system into the parameter correction model to obtain the corrected key parameters; performing safety verification on the corrected key parameters, and if the verification passes, adjusting the current driving assistance system based on the corrected key parameters to obtain the adjusted driving assistance system. The present invention performs self-learning of the automatic driving system within the driving safety envelope to improve the performance, safety and user experience of the driving assistance system. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0056] Figure 1 A schematic diagram of a flow chart of a data processing method applied to intelligent driving provided by an embodiment of the present invention;

[0057] Figure 2 A schematic diagram of a learning process of a self-learning intelligent driving assistance system based on a driving safety envelope provided by an embodiment of the present invention;

[0058] Figure 3 A schematic diagram of a driver intention estimation based on a neural network provided by an embodiment of the present invention;

[0059] Figure 4 A schematic diagram of a processing flow of a self-learning module based on a neural network provided by an embodiment of the present invention;

[0060] Figure 5 A schematic diagram of the structure of a data processing device applied to intelligent driving provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] The terms "first" and "second" and the like in the specification and claims of the present invention and the above drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but may include steps or units that are not listed.

[0063] In an embodiment of the present invention, a data processing method for intelligent driving is provided, which is essentially a self-learning method for a driving assistance system, and is applicable to all driving assistance systems that require manual intervention or takeover, and collects driving data during driver intervention or takeover, and performs self-learning of the driving assistance system within a driving safety envelope. It can take into account the driver's performance driving and safety driving requirements of the driving assistance system, thereby improving the user experience.

[0064] See also Figure 1 , is a flow chart of a data processing method for intelligent driving provided by an embodiment of the present invention, the method may include the following steps:

[0065] S101. When the driver intervenes in the current driving assistance system, intervention-related information is collected.

[0066] Generally, the vehicle will perform automatic driving based on the current driving assistance system, but in this process, the driver will intervene or take over the vehicle. When it is detected that the driver intervenes or takes over the vehicle's current driving assistance system, the corresponding intervention-related information is collected. Among them, the intervention-related information includes the driver's intervention operation information, vehicle environment information and vehicle status information. Specifically, recording the driver's intervention operation information is to record the driver's operation when intervening, such as the accelerator pedal opening, brake pedal opening, steering wheel angle and other information; vehicle environment information refers to the environmental information around the vehicle before and after the intervention, such as the distance between the vehicle and obstacles, lane lines, lane boundaries, traffic lights, traffic signal lines and other external environmental information; vehicle status information refers to the vehicle's own status information, such as the vehicle's own speed, posture, yaw angular velocity, acceleration, etc.

[0067] S102: Input the intervention-related information into the intention recognition model to obtain the driver's intervention intention information.

[0068] The intention recognition model is a neural network model trained based on training samples, wherein each training sample includes intervention-related information annotated with driver intention information. The intention recognition model will infer the driver's intention to intervene based on machine learning or preset rules according to the intervention-related information, i.e., the driver's intervention characteristics, environmental information, and vehicle status information. For example, the driver believes that the acceleration of the ACC (Adaptive Cruise Control) system is too slow or too intense, and the ACC is too close or too far from the vehicle; LCC (Lane Centering Control) or TJA (Traffic Jam Assistant) should be driven close to the inside of the curve when turning; the offset of the smart evasion system is too large or too small, LDP (Lane Departure Prevention) or ELK (Emergency Lane Keeping) intervenes too late, and ALCA (Auto Lane ChangAssist) changes lanes too slowly, etc. It should be noted that the intervention intention information includes the driver's possible intention and the probability of the corresponding intention.

[0069] S103: Detect whether evaluation information for the driver assistance system after intervention is collected, and if so, determine target intervention intention information.

[0070] After obtaining the driver's intervention intention information through the intention recognition model, corresponding evaluation options can be generated to obtain the driver's evaluation information on the driving assistance system after intervention, so as to facilitate subsequent application in updating and adjusting the driving assistance system.

[0071] Specifically, the driver can be allowed to evaluate the performance of the driving assistance system after intervention through interactive methods such as buttons and voice. The evaluation includes but is not limited to the driver's subjective score, the intention of intervention, and other information. The driver's expectations for the assisted driving system can be directly obtained through the driver's evaluation, so as to conduct targeted learning. On the other hand, the trial run after the training is completed allows the driver to evaluate the effectiveness of the learning. The intention recognition model can also be modified based on the evaluation information to obtain a modified intention recognition model.

[0072] It should be noted that in another implementation of the present invention, if the driver does not make an evaluation, that is, if the evaluation information for the driver assistance system after the intervention is not collected, the target intervention intention information can be determined based on the intention probability value in the intervention intention information. For example, if the driver is not very enthusiastic about participating in the evaluation, for the data with more obvious driver intentions (the probability values ​​corresponding to each intention obtained by the intention recognition model), the intention with a higher probability value output by the intention recognition model can be input into the subsequent self-learning system as the target intervention intention.

[0073] S104: input the intervention association information, the target intervention intention information and the data information of the driving assistance system into a parameter correction model to obtain corrected key parameters.

[0074] S105. Perform safety verification on the corrected key parameters. If the verification passes, adjust the current driving assistance system based on the corrected key parameters to obtain an adjusted driving assistance system.

[0075] The target intervention information determined by the evaluation information, the corresponding intervention-related information, and the data information of the related driving assistance system are used to correct the key parameters of the driving assistance system through machine learning or preset rules. In order to improve the efficiency of self-learning, the self-learning module can be deployed in the cloud, and the trained model can be pushed to the vehicle through the network. The parameter correction model used in self-learning is a neural network model with adjusted parameters obtained through training samples.

[0076] In one implementation, the performing security verification on the corrected key parameters includes:

[0077] Acquire a constraint function corresponding to the corrected key parameter, wherein the constraint function includes boundary information of each parameter;

[0078] Based on the current vehicle environment information, calculate and obtain the safety parameters corresponding to the current vehicle environment;

[0079] Based on the constraint function and the safety parameter, the corrected key parameter is safety verified to obtain a verification result.

[0080] Specifically, after obtaining the adjusted key parameters, in order to ensure the safety of subsequent parameter applications, it is necessary to verify the safety of these parameters. In the embodiment of the present invention, a driving safety envelope verification module is used to implement the safety verification of the parameters. Among them, driving safety includes a module for evaluating whether driving is safe, which is actually a multi-dimensional constraint function. Its expression forms include the boundaries of key parameters, such as the upper and lower limits of the distance between ACC and the vehicle, the upper and lower limits of acceleration or acceleration of ACC acceleration and deceleration, the boundaries of LDW (Lane Departure Warning, lane departure warning) / LDP / ELK alarm or intervention, the bias of the intelligent evasion system, the fastest lane change time of ALCA and the allowed lateral and longitudinal acceleration, etc., as well as the upper and lower limits of the steering wheel angle, accelerator pedal and brake pedal calculated in real time based on environmental information and safety models. These boundary conditions are used to limit the input self-learning module data and the safety evaluation of the self-learning driving assistance system.

[0081] In the embodiment of the present invention, an adjusted driving assistance system is obtained, and the adjusted driving assistance system can be virtually operated and interactively confirmed to determine whether to replace the current driving assistance system with the adjusted driving assistance system.

[0082] In an embodiment of the present invention, a data processing method for intelligent driving is provided, including: when the driver intervenes in the current driving assistance system, collecting intervention-related information, the intervention-related information includes the driver's intervention operation information, vehicle environment information and vehicle status information; inputting the intervention-related information into the intention recognition model to obtain the driver's intervention intention information; detecting whether the evaluation information for the driving assistance system after intervention is collected, and if so, determining the target intervention intention information; inputting the intervention-related information, the target intervention intention information and the data information of the driving assistance system into the parameter correction model to obtain the corrected key parameters; performing safety verification on the corrected key parameters, and if the verification passes, adjusting the current driving assistance system based on the corrected key parameters to obtain the adjusted driving assistance system. The present invention performs self-learning of the automatic driving system within the driving safety envelope to improve the performance, safety and user experience of the driving assistance system.

[0083] In one embodiment of the present invention, the method further includes: performing a virtual operation on the adjusted driving assistance system to obtain an operation result, so as to determine the safety of the adjusted driving assistance system according to the operation result.

[0084] Specifically, considering the need for safety, the adjusted driving assistance system after training will not directly intervene in the control. Instead, virtual operation is carried out with the help of a virtual operation platform, that is, the calculation results are output in real time on the car through the virtual operation platform, and the safety envelope is used to check whether the output is safe at this time. Usually, only driving assistance systems that have been virtually operated for a period of time without "dangerous driving" are allowed to be updated to the existing driving assistance system. Of course, if the virtual operation module is not performed, but the key parameters of the driving assistance system are checked through the safety envelope and the real-time output results are limited, the safety of the system can also be guaranteed. However, this safety strictly depends on the reliability of the upper and lower limits of the original calibration parameters in the safety envelope and the reliability of the output safety assessment module. In order to prevent the risk of failure during the operation of this module, the corresponding failure logic needs to be added.

[0085] In one implementation of the present invention, the method further includes:

[0086] The intervention intention information is stored according to the type of the driving assistance subsystem corresponding to the intervention intention, so that the driving assistance system is adjusted based on the stored information.

[0087] Specifically, the recorded data is classified according to the intention label information generated by the driver evaluation system or the intention recognition model, and stored in the corresponding storage area (such as ROM). When the data volume of a certain label meets the requirements, it can be input to the self-learning module for training. Among them, the classification is based on the intervention intention and the corresponding subsystem classification. For example, in the ACC system, the driver expects the following distance to be farther / closer; the driver expects faster / slower acceleration; in the smart evasion system, the evasion distance is farther / closer, etc.

[0088] In another implementation, after obtaining the adjusted driving assistance system, it is possible to determine whether to apply the adjusted driving assistance system according to the feedback information generated by the user on whether to make adjustments. The process may include: generating prompt information corresponding to the adjusted driving assistance system, the prompt information being used to prompt the driver whether to update the current driving assistance system; if the received feedback information for the prompt information meets the update condition, updating the current driving assistance system according to the adjusted driving assistance system.

[0089] The embodiment of the present invention can be applied to a driving assistance system that allows manual intervention. For LDW / LDP / ELK, the driver can be reminded in advance to intervene in the vehicle earlier or later. For LCC / TJA / intelligent evasion system, the vehicle can be controlled to reasonably offset driving relative to lane boundaries / vehicles, etc., to reduce the psychological burden of using assisted driving. For ACC, the following distance, acceleration and deceleration intensity and time can be optimized according to the driver's intervention. For ALCA, lane change decisions, trajectories and speeds can be optimized according to the driver's intervention. By introducing the concept of driving safety envelope, from the processing of raw data to the verification of the model after self-learning, the driving safety envelope must be tested, which effectively improves the safety of self-learning driving assistance. In addition, a neural network is introduced to speculate the driver's intervention intention and correct the key parameters of the model, and a driver evaluation system is introduced in terms of intention. The driving safety envelope is used to limit the correction of key parameters, giving full play to the advantages of neural networks in personalized parameters and model-based development in stability and reliability, thereby realizing a safe and reliable personalized self-learning intelligent driving assistance system.

[0090] See also Figure 2 , which shows a learning process diagram of a self-learning intelligent driving assistance system based on a driving safety envelope provided by an embodiment of the present invention, including:

[0091] Check whether the self-learning switch is on. When the self-learning switch is on, the self-learning system will start working.

[0092] Record the driver's intervention data and surrounding environmental information when the smart evasion system / TJA / LCC is working. The intervention information includes the opening of the accelerator or brake pedal, the steering wheel angle, the turn signal, etc.; and the environmental information includes the lane line at this time. The driver's identity information needs to be obtained when recording data. The driver's identity information can be obtained through the user's network account or various biometric monitoring functions such as FACE ID, fingerprint and voiceprint recognition. This system does not directly involve related detection functions, but only obtains the identity information output given by other systems of the car through the gateway. If there is no relevant information input, the driver can also set the corresponding user through the switch or large screen input.

[0093] See also Figure 3 , which is a schematic diagram of driver intention inference based on a neural network provided in an embodiment of the present invention, inputs the driver's intervention data, the environment around the vehicle, the vehicle's own status information and the output stream of the auxiliary driving system control into the driver intention inference module to obtain the driver's possible intentions and output the corresponding probability.

[0094] The driver's intention to intervene is confirmed through interaction. If the driver does not provide relevant information, the confidence level of the inference module output determines whether the data will be included in subsequent training. In addition, the driver's intention confirmed through interaction will also be recorded for correction by the driver's intention inference module.

[0095] The intervention data and the driver's intention are input into the driving safety envelope check module for data verification. The data that does not comply with safety assessments and regulations is processed accordingly and re-input into the safety envelope check module.

[0096] The processed data is classified according to the intervention intention and stored in the system's ROM.

[0097] Until the amount of data meets the training requirements, then input Figure 4 The self-learning module of the driver assistance system is used to correct the relevant parameters of the function. For different subsystems of the driver assistance system, the input and output of the neural network are different. For example, in the ACC function, there is no need to input the information of the road line and boundary.

[0098] Since the correction of some parameters may lead to unpredictable consequences in other scenarios, it is also necessary to conduct a virtual state test run of the self-learning assisted driving system. During the operation, the system will output the calculation results in real time and input them into the driving safety envelope for inspection, but the output at this time is not directly used for system control. Generally speaking, only after a certain time and distance of evaluation based on the safety envelope will the upgrade request be sent to the driver through the interactive device for update. Of course, the safety envelope can also be used to check the key parameters of the driving assistance system, limit the real-time output results, and ensure the safety of the system. However, this safety strictly depends on the reliability of the upper and lower limits of the original calibration parameters in the safety envelope and the reliability of the output safety assessment module. In order to prevent the risk of failure during the operation of this module, the corresponding failure logic needs to be added.

[0099] Self-learning is completed through driver satisfaction evaluation or a significant decrease in the intervention frequency of the function over a period of time.

[0100] Based on the above embodiments, an embodiment of the present invention further provides a data processing device for intelligent driving, see Figure 5 ,include:

[0101] A collection unit 10 is used to collect intervention-related information when the driver intervenes in the current driving assistance system, wherein the intervention-related information includes the driver's intervention operation information, vehicle environment information, and vehicle status information;

[0102] an intention recognition unit 20, configured to input the intervention-related information into an intention recognition model to obtain the driver's intervention intention information;

[0103] a determination unit 30, configured to detect whether evaluation information for the driver assistance system after intervention is collected, and if so, determine target intervention intention information;

[0104] A parameter correction unit 40, used to input the intervention association information, the target intervention intention information and the data information of the driving assistance system into a parameter correction model to obtain corrected key parameters;

[0105] The adjustment unit 50 is used to perform safety verification on the corrected key parameters. If the verification passes, the current driving assistance system is adjusted based on the corrected key parameters to obtain an adjusted driving assistance system.

[0106] Furthermore, the device also includes:

[0107] The virtual operation unit is used to perform a virtual operation on the adjusted driving assistance system to obtain an operation result, so as to determine the safety of the adjusted driving assistance system according to the operation result.

[0108] Furthermore, the device also includes:

[0109] The intention determination unit is used to determine the target intervention intention information based on the intention probability value in the intervention intention information if the evaluation information for the post-intervention driving assistance system is not collected.

[0110] Furthermore, the device also includes:

[0111] The storage unit is used to store the intervention intention information according to the type of the driving assistance subsystem corresponding to the intervention intention, so that the driving assistance system can be adjusted based on the stored information.

[0112] Correspondingly, the adjustment unit includes:

[0113] A verification subunit is used to perform security verification on the corrected key parameters. The verification subunit is specifically used to:

[0114] Acquire a constraint function corresponding to the corrected key parameter, wherein the constraint function includes boundary information of each parameter;

[0115] Based on the current vehicle environment information, calculate and obtain the safety parameters corresponding to the current vehicle environment;

[0116] Based on the constraint function and the safety parameter, the corrected key parameter is safety verified to obtain a verification result.

[0117] Correspondingly, the device further includes:

[0118] a generating unit, configured to generate prompt information corresponding to the adjusted driving assistance system, wherein the prompt information is used to prompt the driver whether to update the current driving assistance system;

[0119] An updating unit is configured to update the current driving assistance system according to the adjusted driving assistance system if the received feedback information for the prompt information meets an updating condition.

[0120] Optionally, the device further comprises:

[0121] The model correction unit is used to correct the intention recognition model based on the evaluation information to obtain a corrected intention recognition model.

[0122] The embodiment of the present invention provides a data processing device for intelligent driving, comprising: when the driver intervenes in the current driving assistance system, the collection unit collects intervention-related information, the intervention-related information includes the driver's intervention operation information, vehicle environment information and vehicle status information; the intention recognition unit inputs the intervention-related information into the intention recognition model to obtain the driver's intervention intention information; the determination unit detects whether the evaluation information for the driving assistance system after intervention is collected, and if so, determines the target intervention intention information; the parameter correction unit inputs the intervention-related information, the target intervention intention information and the data information of the driving assistance system into the parameter correction model to obtain the corrected key parameters; the adjustment unit performs safety verification on the corrected key parameters, and if the verification passes, adjusts the current driving assistance system based on the corrected key parameters to obtain the adjusted driving assistance system. The present invention performs self-learning of the automatic driving system within the driving safety envelope to improve the performance, safety and user experience of the driving assistance system.

[0123] Based on the foregoing embodiments, an embodiment of the present invention provides a storage medium, which stores executable instructions. When the instructions are executed by a processor, a data processing method for intelligent driving as described in any one of the above is implemented.

[0124] Based on the foregoing embodiments, an embodiment of the present invention further provides an electronic device, comprising a memory for storing a program; and a processor for executing the program, wherein the program is specifically used to implement a data processing method for intelligent driving as described in any one of the above.

[0125] It should be noted that the processor or CPU may be at least one of an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a Central Processing Unit (CPU), a controller, a microcontroller, and a microprocessor. It is understandable that the electronic device that implements the functions of the processor may also be other, which is not specifically limited in the embodiments of the present invention.

[0126] It should be noted that the above-mentioned computer storage medium / memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory (Flash Memory), a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM) and other memories; it can also be various terminals including one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc.

[0127] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0128] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0129] In addition, all functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional units. A person of ordinary skill in the art may understand that all or part of the steps of the above-mentioned method embodiments may be completed by hardware related to program instructions, and the above-mentioned program may be stored in a computer-readable storage medium, and when the program is executed, the steps of the above-mentioned method embodiments are executed; and the above-mentioned storage medium includes: mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks or optical disks, and other media that can store program codes.

[0130] The methods disclosed in the several method embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments.

[0131] The features disclosed in several product embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new product embodiments.

[0132] The features disclosed in several method or device embodiments provided by the present invention can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0133] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0134] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0135] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A data processing method for intelligent driving, It is characterized in that include: When the driver intervenes in the current driving assistance system, intervention-related information is collected, where the intervention-related information includes the driver's intervention operation information, vehicle environment information, and vehicle status information; Inputting the intervention-related information into an intention recognition model to obtain the driver's intervention intention information; Detect whether evaluation information for the driver assistance system after intervention is collected, and if so, determine target intervention intention information; Inputting the intervention association information, target intervention intention information and data information of the driving assistance system into a parameter correction model to obtain corrected key parameters; The corrected key parameters are safety verified, and if the verification passes, the current driving assistance system is adjusted based on the corrected key parameters to obtain an adjusted driving assistance system; wherein, the safety verification of the corrected key parameters comprises: obtaining a constraint function corresponding to the corrected key parameters, wherein the constraint function includes boundary information of each parameter; based on current vehicle environment information, calculating safety parameters corresponding to the current vehicle environment; based on the constraint function and the safety parameters, the corrected key parameters are safety verified to obtain a verification result.

2. The method according to claim 1, It is characterized in that The method further comprises: The adjusted driving assistance system is virtually operated to obtain an operation result, so as to determine the safety of the adjusted driving assistance system according to the operation result.

3. The method according to claim 1, It is characterized in that The method further comprises: If the evaluation information for the post-intervention driving assistance system is not collected, the target intervention intention information is determined based on the intention probability value in the intervention intention information.

4. The method according to claim 1, It is characterized in that The method further comprises: The intervention intention information is stored according to the type of the driving assistance subsystem corresponding to the intervention intention, so that the driving assistance system is adjusted based on the stored information.

5. The method according to claim 1, It is characterized in that The method further comprises: generating prompt information corresponding to the adjusted driving assistance system, wherein the prompt information is used to prompt the driver whether to update the current driving assistance system; If the received feedback information for the prompt information meets the update condition, the current driving assistance system is updated according to the adjusted driving assistance system.

6. The method according to claim 1, It is characterized in that The method further comprises: Based on the evaluation information, the intention recognition model is modified to obtain a modified intention recognition model.

7. A data processing device for intelligent driving, It is characterized in that include: A collection unit, used for collecting intervention-related information when the driver intervenes in the current driving assistance system, wherein the intervention-related information includes the driver's intervention operation information, vehicle environment information and vehicle status information; an intention recognition unit, configured to input the intervention-related information into an intention recognition model to obtain the driver's intervention intention information; a determination unit, configured to detect whether evaluation information for the driver assistance system after intervention is collected, and if so, determine target intervention intention information; A parameter correction unit, used for inputting the intervention association information, the target intervention intention information and the data information of the driving assistance system into a parameter correction model to obtain corrected key parameters; An adjustment unit is used to perform safety verification on the corrected key parameters. If the verification passes, the current driving assistance system is adjusted based on the corrected key parameters to obtain an adjusted driving assistance system; wherein, the safety verification on the corrected key parameters includes: obtaining a constraint function corresponding to the corrected key parameters, wherein the constraint function includes boundary information of each parameter; based on current vehicle environment information, calculating safety parameters corresponding to the current vehicle environment; based on the constraint function and the safety parameters, the corrected key parameters are safety verified to obtain a verification result.

8. A storage medium, It is characterized in that The storage medium stores executable instructions, which, when executed by a processor, implement the data processing method for intelligent driving as described in any one of claims 1 to 6.

9. An electronic device, It is characterized in that include: Memory, used to store programs; A processor is used to execute the program, wherein the program is specifically used to implement the data processing method applied to intelligent driving as described in any one of claims 1 to 6.

Images