Vehicle, autonomous driving control method and apparatus for vehicle, and medium

Abstract

A vehicle, an autonomous driving control method and apparatus for the vehicle, and a medium. The method comprises: acquiring a target driving style of a target user, and acquiring a target feature data subset from a preset driving feature data set; and on the basis of the target feature data subset, using a parameter optimization model to iteratively update a target parameter of a preset model predictive controller, such that the model predictive controller synchronously updates a driving control strategy, wherein the target parameter is a weight parameter in a cost function, and the cost function takes the minimization of the difference between the target feature data subset and output feature data as an optimization objective. By means of the technical solution provided in the present application, personalized driving control strategies can be provided for users having different driving styles, thereby improving the user experience.

Description

Vehicles, vehicle autonomous driving control methods, devices and media Technical Field

[0001] This application belongs to the field of driving control technology, and in particular relates to a vehicle, a vehicle automatic driving control method, a device and a medium. Background Technology

[0002] With the development of vehicle driving control technology, methods for automatically controlling vehicle driving behavior based on intelligent control algorithms are gradually being applied. Currently, when using intelligent control algorithms for automatic vehicle driving behavior control, the driving control modes are relatively simple and cannot adequately meet the driving experience needs of different users. Summary of the Invention

[0003] The embodiments of this application provide a vehicle, a vehicle autonomous driving control method, an apparatus, and a medium, which can at least to some extent provide personalized driving control strategies for users with different driving styles, thereby improving the user experience.

[0004] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by practice of the present application.

[0005] According to a first aspect of the embodiments of this application, a vehicle autonomous driving control method is provided, comprising:

[0006] Obtain the target user's target driving style, and retrieve a subset of target feature data that matches the target driving style from a preset driving feature dataset;

[0007] Based on the target feature data subset, a parameter optimization model is used to iteratively update the target parameters of the preset model predictive controller, so that the model predictive controller synchronously updates the driving control strategy based on the target parameters after each iteration.

[0008] The target parameter is the weight parameter in the cost function of the model predictive controller. The cost function aims to minimize the difference between the target feature data subset and the output feature data of the model predictive controller. The output feature data is used to generate the driving control strategy.

[0009] In some embodiments of this application, based on the foregoing scheme, the process of constructing the driving feature dataset includes:

[0010] Acquire historical driving characteristic data of different users in multiple driving environments for the target scenario, wherein the historical driving characteristic data of users with different driving styles are different in the target scenario;

[0011] Several historical driving feature data are classified according to different driving styles, and a feature data subset corresponding to each driving style is constructed.

[0012] The driving feature dataset is constructed based on multiple subsets of the feature data.

[0013] In some embodiments of this application, based on the foregoing scheme, the target scenario includes at least a scenario of accelerating from a first speed and a scenario of following the vehicle in front, wherein the first speed is less than or equal to a preset speed.

[0014] In some embodiments of this application, based on the foregoing scheme, the construction process of the model prediction controller includes:

[0015] The model predictive controller is constructed based on the vehicle longitudinal control model;

[0016] The cost function of the model predictive controller is constructed based on the target parameters used to determine the driving style, wherein the target parameters include multiple weighted sub-parameters;

[0017] The constraints of the model predictive controller are constructed based on the vehicle's input and state constraints.

[0018] In some embodiments of this application, based on the foregoing scheme, the weight sub-parameters include at least speed weight, acceleration weight, bump coefficient weight, and distance from the vehicle in front weight, and the equation of the cost function is as follows:

[0019] ;

[0020] in, Indicates having the target parameter The cost function, , This represents the speed weight. Indicates the estimated speed. This indicates the speed limit value corresponding to the target operating condition. This indicates the weight of the distance from the vehicle in front. Indicates the estimated position of the vehicle. Indicates the estimated position of the vehicle ahead. Indicates the acceleration weight, Indicates longitudinal acceleration. This indicates the weight of the bump coefficient. Indicates the steering angle weight. Indicates the steering wheel angle. Indicates in The prediction step size initialized at time step. Indicates the current calculation time.

[0021] In some embodiments of this application, based on the foregoing scheme, the parameter optimization model includes a Bayesian optimization model, and the step of iteratively updating the target parameters of the preset model prediction controller using the parameter optimization model based on the target feature data subset includes:

[0022] The objective function of the Bayesian optimization model is determined based on minimizing the difference between the target feature data subset and the output feature data of the model prediction controller. as follows:

[0023] ;

[0024] in, This represents one or more first feature data points within a subset of the target feature data. This represents the output feature data determined by the model prediction controller based on the target parameters;

[0025] The objective function is estimated using a probabilistic model of Gaussian process regression to determine the probability distribution of the objective function;

[0026] Based on the probability distribution of the objective function, the objective parameters in the Bayesian optimization model are updated using the acquisition function method.

[0027] In some embodiments of this application, based on the foregoing scheme, estimating the objective function using a probabilistic model of Gaussian process regression to determine the probability distribution of the objective function includes:

[0028] For the first feature data of length N, the prior probability distribution of the objective function is estimated using a probabilistic model of Gaussian process regression.

[0029] ;

[0030] in, This represents the objective function based on the first feature data. This represents the mean of the first feature data. Indicates The first kernel function is the input. This represents the target parameters before updating based on the first feature data. This represents the updated target parameters based on the first feature data;

[0031] Based on the prior probability distribution, the posterior probability distribution of the objective function for the target prediction point is determined as follows:

[0032] ;

[0033] Among them, the target prediction point is , The objective function is represented at the target prediction point. This represents the mean of the second feature data at the target prediction point. This represents the second kernel function determined based on the first kernel function at the target prediction point.

[0034] According to a second aspect of the embodiments of this application, a vehicle autonomous driving control device is provided, comprising:

[0035] The acquisition unit is configured to acquire the target driving style of the target user and acquire a subset of target feature data that matches the target driving style from a preset driving feature dataset;

[0036] The update unit is configured to iteratively update the target parameters of the preset model prediction controller based on the target feature data subset using a parameter optimization model, so that the model prediction controller synchronously updates the driving control strategy based on the target parameters after each iteration update.

[0037] The target parameter is the weight parameter in the cost function of the model predictive controller. The cost function aims to minimize the difference between the target feature data subset and the output feature data of the model predictive controller. The output feature data is used to generate the driving control strategy.

[0038] According to a third aspect of the embodiments of this application, a computer-readable storage medium is provided, the computer-readable storage medium storing at least one computer program instruction, the at least one computer program instruction being loaded and executed by a processor to perform the steps performed as described in any of the methods in the first aspect.

[0039] According to a fourth aspect of the present application, a vehicle is provided, which includes one or more processors and one or more memories, wherein at least one piece of program code is stored in the one or more memories, the at least one piece of program code being loaded and executed by the one or more processors to perform the steps of the method described in any of the first aspects.

[0040] The one or more technical solutions provided in the embodiments of the present invention achieve at least the following technical effects or advantages:

[0041] This application establishes a driving style feature dataset for users with different driving styles. During application, it matches a corresponding subset of target feature data based on the target user's target driving style. Based on this subset, a parameter optimization model iteratively updates the target parameters of the preset model predictive controller. This ensures that the model predictive controller synchronously updates the driving control strategy based on the updated target parameters. Since the target parameters are optimized to minimize the difference between the subset of target feature data and the output feature data of the model predictive controller, each updated driving control strategy aligns with the user's preferred driving style. This achieves real-time control of vehicle driving behavior while providing a personalized driving control experience, thereby improving the user's overall driving experience.

[0042] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0043] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings:

[0044] Figure 1 shows a flowchart of a vehicle autonomous driving control method according to an embodiment of this application;

[0045] Figure 2 shows a structural diagram of a vehicle automatic driving control device according to an embodiment of this application;

[0046] Figure 3 shows a schematic diagram of the structure of a computer system suitable for implementing the embodiments of this application in a vehicle. Detailed Implementation

[0047] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this application. However, those skilled in the art will recognize that the technical solutions of this application can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this application.

[0049] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0050] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0051] It should also be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such uses of these terms can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described.

[0052] It should be noted that with the development of autonomous driving technology, the methods for automatically controlling vehicle movement are also constantly evolving and have made significant progress. Thanks to advancements in computer and sensing technologies, such as intelligent sensors, artificial intelligence, and big data analytics, vehicles are able to perceive and respond to the road environment more accurately and efficiently, thereby improving road safety, traffic efficiency, and the sustainability of autonomous driving.

[0053] Currently, the main methods for controlling autonomous driving in vehicles include: first, autonomous driving control systems based on a combination of hardware sensor detection and computer analysis and processing; and second, autonomous driving control methods that employ intelligent control algorithms.

[0054] The first control method relies on numerous onboard sensors for information collection, including but not limited to electronic throttle, electronic brakes, vehicle steering controllers, and distance sensors. It then generates automatic control commands through an onboard computer control system to achieve autonomous driving functionality. However, using multiple sensors for information collection not only increases the overall vehicle hardware and software costs but also introduces instability in information detection, which can affect the accuracy of autonomous driving control command output. Therefore, autonomous driving methods that rely on excessive sensor detection have potential stability risks.

[0055] Currently, the second control method is gradually becoming the mainstream approach for autonomous driving control. For example, related technologies have proposed a model-based path tracking method for autonomous vehicles, which achieves good path tracking performance and ensures stable vehicle operation by establishing a vehicle dynamics model and a tracking error model, using the optimization of the control increment as the objective function. Furthermore, related technologies, based on model predictive control models, use the curvature of the target path trajectory and the actual vehicle speed as reference features for the controller. Under constraints, combined with a dynamic error model, the steering wheel angle of the model predictive control feedback is solved. Additionally, related technologies, based on model predictive control, predict and generate control sequences by detecting road and driving condition types, and automatically adjust the prediction step size to adapt to different road conditions. However, current technologies only achieve autonomous driving control for a single driving mode and cannot adequately meet the autonomous control needs of users with significantly different driving styles.

[0056] Based on the above, the embodiments of this application aim to propose a vehicle autonomous driving control method, which at least partially solves the above problems and can provide personalized driving control strategies for users with different driving styles, thereby improving the user experience.

[0057] Referring to Figure 1, a flowchart of a vehicle autonomous driving control method according to an embodiment of this application is shown.

[0058] As shown in Figure 1, according to a first aspect of the embodiments of this application, a vehicle autonomous driving control method is provided, including but not limited to steps S1 to S2:

[0059] Step S1. Obtain the target user's target driving style, and obtain a subset of target feature data that matches the target driving style from a preset driving feature dataset.

[0060] It should be noted that the driving style in this application embodiment can be divided according to different users' needs for driving speed, such as into sporty style, neutral style or comfort style; it can also be divided according to the age, gender and other characteristics of different users, such as young users, middle-aged and elderly users, etc. The specific implementation of the driving style is not limited here.

[0061] In some embodiments, obtaining the target user's target driving style includes:

[0062] The system obtains the driving style request from the target user and determines the target user's target driving style based on the request. The user can send the driving style request via voice, text, network button, or other means.

[0063] In some embodiments, obtaining the target user's target driving style includes:

[0064] The target user is identified, and the target driving style is determined based on the identification results.

[0065] Understandably, before using the vehicle's autonomous driving system, users can register their identity information and preferred driving style in the system. Thus, when using the autonomous driving system, the system can determine the user's preferred driving style, i.e., the target driving style, based on the user's identity recognition results.

[0066] It should be noted that the preset driving feature dataset includes multiple feature data subsets corresponding to different types of driving styles, and each driving style can correspond to one or more feature data subsets. In order to manage the driving feature data more granularly and improve subsequent driving control strategies, each driving style can set at least one feature data subset for each driving environment (such as road conditions).

[0067] In some embodiments, obtaining a subset of target feature data matching the target driving style from a preset driving feature dataset includes:

[0068] Obtain a preset mapping table, which records each feature data subset of the driving feature dataset and the driving style corresponding to each feature data subset;

[0069] Search the mapping table for a subset of feature data that matches the target driving style.

[0070] In some embodiments, the process of constructing the driving feature dataset includes:

[0071] Step S11. Obtain historical driving characteristic data of different users in multiple driving environments for the target scenario, wherein the historical driving characteristic data of users with different driving styles are different in the target scenario.

[0072] It should be noted that multiple driving environments can be different road conditions, different working conditions, different weather conditions, etc. The specific settings can be made according to needs, and there are no restrictions here.

[0073] In some embodiments, the target scenario includes at least:

[0074] 1) A scenario in which the vehicle accelerates from a first speed, where the first speed is less than or equal to a preset speed, such as a scenario in which the vehicle accelerates from zero speed.

[0075] It is understandable that users with different driving styles may use different acceleration when accelerating from low speeds. For example, sporty users may use greater acceleration, while comfort users may use less. Therefore, by acquiring driving characteristic data from different users in this scenario, we can obtain different driving characteristic data corresponding to different driving styles. It is also understandable that, in order to better acquire driving characteristic data, there are no obstacles or vehicles in front of the vehicle in the above scenario.

[0076] 2) In the scenario of following the vehicle in front, the first speed is less than or equal to the preset speed.

[0077] Understandably, users with different driving styles may employ different driving strategies when following the vehicle in front. For example, a sporty driver might choose to overtake and continue driving if the vehicle in front is traveling slower than they expect, while a comfort-style driver might choose to follow. Therefore, by acquiring driving characteristic data from different users in this scenario, we can obtain different driving characteristic data corresponding to users with different driving styles.

[0078] Step S12. Classify the historical driving feature data according to different driving styles and construct a feature data subset corresponding to each driving style.

[0079] Therefore, by collecting the historical driving characteristics of several users, the historical driving characteristic data can be classified according to different driving styles, and a subset of characteristic data corresponding to each driving style can be constructed.

[0080] Step S13. Construct the driving feature dataset based on multiple subsets of the feature data.

[0081] In some embodiments, driving characteristic data includes, but is not limited to, speed. acceleration Bump coefficient and distance from the vehicle in front The calculation formulas for the above feature data are as follows:

[0082] speed ;

[0083] acceleration ;

[0084] Bump coefficient ;

[0085] Distance from the car in front ;

[0086] in, This represents the median value of the calculated data. This indicates the number of samples taken by the accelerometer. and These represent the vehicle's speed and acceleration, respectively. and These represent the displacements of the vehicle and the vehicle in front, respectively, thus yielding a set of feature vector data. .

[0087] Step S2. Based on the target feature data subset, the target parameters of the preset model prediction controller are iteratively updated using a parameter optimization model, so that the model prediction controller synchronously updates the driving control strategy based on the target parameters after each iteration.

[0088] The target parameter is the weight parameter in the cost function of the model predictive controller. The cost function aims to minimize the difference between the target feature data subset and the output feature data of the model predictive controller. The output feature data is used to generate the driving control strategy.

[0089] In some embodiments, the process of constructing the model prediction controller includes:

[0090] Step 01. Construct the model predictive controller based on the vehicle longitudinal control model.

[0091] In some embodiments, the vehicle longitudinal control model is represented by the following equation:

[0092] ;

[0093] in, For the car in The state at any given moment, and This provides the vehicle's position coordinates in the global coordinate system. For the vehicle's speed, The input to the autonomous driving control system can be the vehicle's longitudinal acceleration. .

[0094] Furthermore, in scenarios where there are other vehicles in front of the vehicle, the longitudinal control model for the vehicle in front is represented as follows:

[0095] ;

[0096] in, Indicates the status of the vehicle in front. and This indicates the position information of the preceding vehicle in the global coordinate system. Indicates the speed of the vehicle in front. The input to the vehicle's automated driving control system can be the vehicle's longitudinal acceleration. .

[0097] Based on the longitudinal control model, a Model Predictive Control (MPC) is constructed, yielding the vehicle's motion equations as follows:

[0098] ;

[0099] ;

[0100] in, and These represent the estimated position and estimated acceleration, respectively. Indicates in The prediction step size initialized at each time step.

[0101] Step 02. Construct the cost function of the model predictive controller based on the target parameters used to determine the driving style, wherein the target parameters include multiple weight sub-parameters.

[0102] It should be noted that, in order to improve the performance of the model predictive controller and enhance the stability of the controlled vehicle, such as controlling vehicle acceleration and bump coefficient, thereby providing users with personalized driving control strategies in different driving environments, this embodiment constructs the cost function of the model predictive controller based on the target parameters used to determine driving style. The functional equation is as follows:

[0103] ;

[0104] in, Indicates the input to the autonomous driving system. The input acceleration information for predicting the time step k at time t is... For time t, predict vehicle information at time step k, including the vehicle's global coordinate position and vehicle speed information. These are the weight parameters of the cost function.

[0105] In some embodiments, the weighting sub-parameters include at least speed weight, acceleration weight, bump coefficient weight, and distance from the vehicle in front weight, and the equation of the cost function is as follows:

[0106] ;

[0107] in, Indicates having the target parameter The cost function, , This represents the speed weight. Indicates the estimated speed. This indicates the speed limit value corresponding to the target operating condition. This indicates the weight of the distance from the vehicle in front. Indicates the estimated position of the vehicle. Indicates the estimated position of the vehicle ahead. Indicates the acceleration weight, Indicates longitudinal acceleration. This indicates the weight of the bump coefficient. Indicates the steering angle weight. Indicates the steering wheel angle. Indicates in The prediction step size initialized at time step. Indicates the current calculation time.

[0108] Step 03. Based on the vehicle's input and state constraints, construct the constraints for the model predictive controller.

[0109] For example, the constraints of the model predictive controller are as follows:

[0110] ;

[0111] in, This indicates that the vehicle information at time step k+1 is predicted from the current time t. and These are the input and status information for the autonomous driving system, respectively. and These are input constraints and state constraints, respectively.

[0112] For example, the input U of an autonomous driving system is throttle acceleration. To ensure safe driving, throttle acceleration needs to be limited. Therefore, the input limit here is that the acceleration cannot exceed the preset maximum acceleration. State limitations mainly target vehicle speed; similarly, the vehicle speed can be limited to a safe range.

[0113] In some embodiments, the parameter optimization model includes a Bayesian optimization model, and the step of iteratively updating the target parameters of the preset model prediction controller using the parameter optimization model based on the target feature data subset includes:

[0114] Step S21. Determine the objective function of the Bayesian optimization model based on minimizing the difference between the target feature data subset and the output feature data of the model prediction controller. as follows:

[0115] ;

[0116] in, This represents one or more first feature data points within the subset of target feature data. This represents the output feature data determined by the model prediction controller based on the target parameters.

[0117] It should be noted that in each time step, the model predictive controller synchronously generates an optimized sequence of control actions. The first control action is used as prior information input to the autonomous driving system. After that, the parameter optimization iteration of the model predictive controller and the output of control commands are repeated.

[0118] Therefore, the problem of learning a specific driving mode by using a model predictive controller is transformed into an objective parameter tuning problem, which involves the objective parameters. The iterative optimization problem, namely:

[0119] ;

[0120] Among them, the function It indicates a specific performance metric.

[0121] In this embodiment, minimizing the difference between the target feature data subset and the output feature data of the model prediction controller is taken as the optimization objective, based on the target feature data subset. The output feature data generated by the model predictive controller is Therefore, the target parameter The iterative optimization problem can be expressed as:

[0122] .

[0123] In this embodiment, a Bayesian optimization model is used to solve the iterative update problem of the objective parameters. Therefore, the objective parameters are... Iterative optimization is used as the objective function of the Bayesian optimization model. .

[0124] Step S22. Estimate the objective function using a probabilistic model of Gaussian process regression to determine the probability distribution of the objective function.

[0125] In some embodiments, estimating the objective function using a probabilistic model of Gaussian process regression to determine the probability distribution of the objective function includes:

[0126] For the first feature data of length N, the prior probability distribution of the objective function is estimated using a probabilistic model of Gaussian process regression.

[0127] ;

[0128] in, This represents the objective function based on the first feature data. This represents the mean of the first feature data. Indicates The first kernel function is the input. This represents the target parameters before updating based on the first feature data. This represents the updated target parameters based on the first feature data;

[0129] in, , and This indicates the preset hyperparameters.

[0130] Based on the prior probability distribution, the posterior probability distribution of the objective function for the target prediction point is determined as follows:

[0131] ;

[0132] Among them, the target prediction point is , The objective function is represented at the target prediction point. This represents the mean of the second feature data at the target prediction point. This represents the second kernel function determined based on the first kernel function at the target prediction point.

[0133] in, ;

[0134] Indicates The kernel function calculation result obtained as the input to the first kernel function is similarly... , , It is also obtained based on the calculation formula of the first kernel function mentioned above.

[0135] Step S23. Based on the probability distribution of the objective function, update the objective parameters in the Bayesian optimization model using the acquisition function method. Furthermore, update the objective parameters in the Bayesian optimization model using the acquisition function method. The update process continuously obtains the current parameters through iteration. .

[0136] It should be noted that, because the model predictive controller updates a set of output feature data with each iteration of the target parameters, it generates a new driving control strategy. This allows the driving control strategy to be dynamically updated and adjusted based on the vehicle's driving conditions during operation, achieving path tracking for autonomous driving. For example, if the target path runs from point A to point D, passing through several discrete path points such as points B and C, then in each short segment of the journey, the target parameters can be updated in real-time based on a subset of the target driving feature data, according to the vehicle's real-time road conditions, thereby generating a new driving control strategy. For instance, after the vehicle travels from point A to point B, the path from point B to point C can be optimized.

[0137] Based on the above disclosure, this application embodiment establishes a driving style feature dataset for users with different driving styles. In application, it matches a corresponding subset of target feature data according to the target user's target driving style. Based on this subset, a parameter optimization model iteratively updates the target parameters of the preset model predictive controller. This allows the model predictive controller to synchronously update the driving control strategy based on the updated target parameters. Since the target parameters aim to minimize the difference between the subset of target feature data and the output feature data of the model predictive controller, each updated driving control strategy conforms to the user's preferred driving style. This achieves real-time control of vehicle driving behavior while providing a personalized driving experience, thereby improving the user's driving experience.

[0138] Referring to Figure 2, a structural diagram of a vehicle autonomous driving control device according to an embodiment of this application is shown.

[0139] As shown in Figure 2, according to a second aspect of the embodiments of this application, a vehicle automatic driving control device 200 is provided, comprising:

[0140] The acquisition unit 201 is configured to acquire the target driving style of the target user and acquire a subset of target feature data that matches the target driving style from a preset driving feature dataset;

[0141] Update unit 202 is configured to iteratively update the target parameters of a preset model predictive controller based on the target feature data subset using a parameter optimization model, so that the model predictive controller synchronously updates the driving control strategy based on the target parameters after each iteration update.

[0142] The target parameter is the weight parameter in the cost function of the model predictive controller. The cost function aims to minimize the difference between the target feature data subset and the output feature data of the model predictive controller. The output feature data is used to generate the driving control strategy.

[0143] According to a third aspect of the embodiments of this application, a computer-readable storage medium is provided, which stores at least one computer program instruction, the at least one computer program instruction being loaded and executed by a processor to perform the steps as performed by any of the methods in the first aspect.

[0144] Computer-readable storage media may be portable compact disc read-only memory (CD-ROM) and include program code, and may run on a terminal device, such as a personal computer. However, the computer-readable storage medium of this application is not limited thereto. In this application, the readable storage medium may be any tangible medium that contains or stores a program that may be used by or in conjunction with an instruction execution system, apparatus, or device.

[0145] A readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0146] Program code for performing the operations of this application can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0147] Referring to Figure 3, it is a schematic diagram of the structure of a computer system suitable for implementing the embodiments of this application in a vehicle.

[0148] According to a fourth aspect of the embodiments of this application, a vehicle is provided, including one or more processors and one or more memories, wherein at least one piece of program code is stored in the one or more memories, and the at least one piece of program code is loaded and executed by the one or more processors to perform the steps performed by any of the methods in the first aspect.

[0149] As shown in Figure 3, the vehicle 400 is presented in the form of a general-purpose computing device. The components of the vehicle 400 may include, but are not limited to: at least one processing unit 410, at least one storage unit 420, and a bus 430 connecting different system components (including storage unit 420 and processing unit 410).

[0150] The storage unit stores program code, which can be executed by the processing unit 410, causing the processing unit 410 to perform the steps described in the "Embodiment Method" section above according to various exemplary embodiments of this application.

[0151] Storage unit 420 may include a readable medium in the form of a volatile storage unit, such as random access memory (RAM) 421 and / or cache memory 422, and may further include a read-only memory (ROM) 423.

[0152] Storage unit 420 may also include a program / utility 424 having a set (at least one) of program modules 425, such program modules 425 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0153] Bus 430 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the various bus structures.

[0154] Vehicle 400 can also communicate with one or more external devices 500 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), one or more devices that enable users to interact with vehicle 400, and / or any device that enables vehicle 400 to communicate with one or more other computing devices (e.g., routers, modems, etc.). This communication can be performed via input / output (I / O) interface 450. Furthermore, vehicle 400 can also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via network adapter 460. As shown, network adapter 460 communicates with other modules of vehicle 400 via bus 430. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with vehicle 400, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0155] The functions described herein can be implemented in hardware, software executed by a processor, firmware, or any combination thereof. If implemented in software executed by a processor, the functions can be stored as one or more instructions or codes on or transmitted via a computer-readable medium. Other examples and embodiments are within the scope and spirit of this invention and the appended claims. For example, due to the nature of software, the functions described above can be implemented using software executed by a processor, hardware, firmware, hardwired, or any combination thereof. Furthermore, the functional units can be integrated into a single processing unit, or each unit can exist physically separately, or two or more units can be integrated into a single unit.

[0156] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0157] The units described as separate components may or may not be physically separate. Similarly, the components of the control device may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0158] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0159] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of the claims of the present application.

Claims

1. A method for controlling autonomous driving of a vehicle, characterized in that, include: Obtain the target user's target driving style, and retrieve a subset of target feature data that matches the target driving style from a preset driving feature dataset; Based on the target feature data subset, a parameter optimization model is used to iteratively update the target parameters of the preset model predictive controller, so that the model predictive controller synchronously updates the driving control strategy based on the target parameters after each iteration. The target parameter is the weight parameter in the cost function of the model predictive controller. The cost function aims to minimize the difference between the target feature data subset and the output feature data of the model predictive controller. The output feature data is used to generate the driving control strategy.

2. The method according to claim 1, characterized in that, The process of constructing the driving feature dataset includes: Acquire historical driving characteristic data of different users in multiple driving environments for the target scenario, wherein the historical driving characteristic data of users with different driving styles are different in the target scenario; Several historical driving feature data are classified according to different driving styles, and a feature data subset corresponding to each driving style is constructed. The driving feature dataset is constructed based on multiple subsets of the feature data.

3. The method according to claim 2, characterized in that, The target scenario includes at least a scenario of accelerating from a first speed and a scenario of following the vehicle in front, wherein the first speed is less than or equal to a preset speed.

4. The method according to claim 1, characterized in that, The construction process of the model prediction controller includes: The model predictive controller is constructed based on the vehicle longitudinal control model; The cost function of the model predictive controller is constructed based on the target parameters used to determine the driving style, wherein the target parameters include multiple weighted sub-parameters; The constraints of the model predictive controller are constructed based on the vehicle's input and state constraints.

5. The method according to claim 4, characterized in that, The weighted sub-parameters include at least speed weight, acceleration weight, bump coefficient weight, and distance from the vehicle in front weight. The equation for the cost function is as follows: ； in, Indicates having the target parameter The cost function, ， This represents the speed weight. Indicates the estimated speed. This indicates the speed limit value corresponding to the target operating condition. This indicates the weight of the distance from the vehicle in front. Indicates the estimated position of the vehicle. Indicates the estimated position of the vehicle ahead. This represents the acceleration weight. Indicates longitudinal acceleration. This indicates the weight of the bump coefficient. Indicates the steering angle weight. Indicates the steering wheel angle. Indicates in The prediction step size initialized at time step. Indicates the current calculation time.

6. The method according to claim 1, characterized in that, The parameter optimization model includes a Bayesian optimization model. The step of iteratively updating the target parameters of the preset model prediction controller using the parameter optimization model based on the target feature data subset includes: The objective function of the Bayesian optimization model is determined based on minimizing the difference between the target feature data subset and the output feature data of the model prediction controller. as follows: ； in, This represents one or more first feature data points within a subset of the target feature data. This represents the output feature data determined by the model prediction controller based on the target parameters; The objective function is estimated using a probabilistic model of Gaussian process regression to determine the probability distribution of the objective function; Based on the probability distribution of the objective function, the objective parameters in the Bayesian optimization model are updated using the acquisition function method.

7. The method according to claim 6, characterized in that, The estimation of the objective function using a probabilistic model employing Gaussian process regression to determine the probability distribution of the objective function includes: For the first feature data of length N, the prior probability distribution of the objective function is estimated using a probabilistic model of Gaussian process regression. ； in, This represents the objective function based on the first feature data. This represents the mean of the first feature data. Indicated by The first kernel function is the input. This represents the target parameters before updating based on the first feature data. This represents the updated target parameters based on the first feature data; Based on the prior probability distribution, the posterior probability distribution of the objective function for the target prediction point is determined as follows: ； Among them, the target prediction point is ， The objective function is represented at the target prediction point. This represents the mean of the second feature data at the target prediction point. This represents the second kernel function determined based on the first kernel function at the target prediction point.

8. A vehicle automatic driving control device, characterized in that, include: The acquisition unit is configured to acquire the target driving style of the target user and acquire a subset of target feature data that matches the target driving style from a preset driving feature dataset; The update unit is configured to iteratively update the target parameters of the preset model prediction controller based on the target feature data subset using a parameter optimization model, so that the model prediction controller synchronously updates the driving control strategy based on the target parameters after each iteration update. The target parameter is the weight parameter in the cost function of the model predictive controller. The cost function aims to minimize the difference between the target feature data subset and the output feature data of the model predictive controller. The output feature data is used to generate the driving control strategy.

9. The apparatus according to claim 8, characterized in that, The device further includes a driving feature dataset construction unit, which is configured to: Acquire historical driving characteristic data of different users in multiple driving environments for the target scenario, wherein the historical driving characteristic data of users with different driving styles are different in the target scenario; Several historical driving feature data are classified according to different driving styles, and a feature data subset corresponding to each driving style is constructed. The driving feature dataset is constructed based on multiple subsets of the feature data.

10. The apparatus according to claim 9, characterized in that, The target scenario includes at least a scenario of accelerating from a first speed and a scenario of following the vehicle in front, wherein the first speed is less than or equal to a preset speed.

11. The apparatus according to claim 8, characterized in that, The apparatus further includes a model prediction controller construction unit, which is configured to: The model predictive controller is constructed based on the vehicle longitudinal control model; The cost function of the model predictive controller is constructed based on the target parameters used to determine the driving style, wherein the target parameters include multiple weighted sub-parameters; The constraints of the model predictive controller are constructed based on the vehicle's input and state constraints.

12. The apparatus according to claim 11, characterized in that, The weighted sub-parameters include at least speed weight, acceleration weight, bump coefficient weight, and distance from the vehicle in front weight. The equation for the cost function is as follows: ； in, Indicates having the target parameter The cost function, ， This represents the speed weight. Indicates the estimated speed. This indicates the speed limit value corresponding to the target operating condition. This indicates the weight of the distance from the vehicle in front. Indicates the estimated position of the vehicle. Indicates the estimated position of the vehicle ahead. This represents the acceleration weight. Indicates longitudinal acceleration. This indicates the weight of the bump coefficient. Indicates the steering angle weight. Indicates the steering wheel angle. Indicates in The prediction step size initialized at time step. Indicates the current calculation time.

13. The apparatus according to claim 8, characterized in that, The parameter optimization model includes a Bayesian optimization model. The step of iteratively updating the target parameters of the preset model prediction controller using the parameter optimization model based on the target feature data subset includes: The objective function of the Bayesian optimization model is determined based on minimizing the difference between the target feature data subset and the output feature data of the model prediction controller. as follows: ； in, This represents one or more first feature data points within a subset of the target feature data. This represents the output feature data determined by the model prediction controller based on the target parameters; The objective function is estimated using a probabilistic model of Gaussian process regression to determine the probability distribution of the objective function; Based on the probability distribution of the objective function, the objective parameters in the Bayesian optimization model are updated using the acquisition function method.

14. The apparatus according to claim 13, characterized in that, The estimation of the objective function using a probabilistic model employing Gaussian process regression to determine the probability distribution of the objective function includes: For the first feature data of length N, the prior probability distribution of the objective function is estimated using a probabilistic model of Gaussian process regression. ； in, This represents the objective function based on the first feature data. This represents the mean of the first feature data. Indicated by The first kernel function is the input. This represents the target parameters before updating based on the first feature data. This represents the updated target parameters based on the first feature data; Based on the prior probability distribution, the posterior probability distribution of the objective function for the target prediction point is determined as follows: ； Among them, the target prediction point is ， The objective function is represented at the target prediction point. This represents the mean of the second feature data at the target prediction point. This represents the second kernel function determined based on the first kernel function at the target prediction point.

15. A computer-readable storage medium storing at least one computer program instruction, the at least one computer program instruction being loaded and executed by a processor to perform the steps of the method as described in any one of claims 1-7.

16. A vehicle, characterized in that, It includes one or more processors and one or more memories, wherein at least one piece of program code is stored in the one or more memories, and the at least one piece of program code is loaded and executed by the one or more processors to perform the steps of the method as described in any one of claims 1-7.

Images