Vehicle control method and device, vehicle and storage medium

Abstract

The embodiment of the invention discloses a vehicle control method and device, a vehicle and a storage medium. The method comprises the following steps: acquiring vehicle data, wherein the vehicle data comprises vehicle operation data, driving environment data and operation information of a driver; the vehicle data are input into a pre-trained large model, various vehicle control instructions corresponding to the vehicle data output by the large model are obtained, and the various vehicle control instructions comprise a power output control instruction, a brake control instruction and a suspension control instruction; and controlling the vehicle to execute operations corresponding to the various vehicle control instructions. According to the method, based on the obtained multi-source vehicle data, the pre-trained large model can output the more accurate vehicle control instruction, and therefore more accurate control over the vehicle can be achieved.

Description

Technical Field

[0001] This application belongs to the technical field of vehicles, and particularly relates to a vehicle control method, device, vehicle, and storage medium. Background Art

[0002] In the field of automotive engineering, the performance of the chassis system plays a key role in the handling, comfort, and safety of vehicles. Traditional automotive chassis systems usually adopt fixed mechanical and electronic control strategies, which are difficult to adapt to complex and changeable road conditions and driving conditions, resulting in inaccurate control of vehicles. For example, in different road surface flatness, slopes, and curve situations, the suspension adjustment, braking distribution, and steering assistance of traditional chassis often cannot reach the optimal state. Summary of the Invention

[0003] In view of the above problems, this application proposes a vehicle control method, device, vehicle, and storage medium to improve the above problems.

[0004] In a first aspect, an embodiment of this application provides a vehicle control method, which includes: obtaining vehicle data, where the vehicle data includes vehicle operation data, driving environment data, and driver operation information; inputting the vehicle data into a pre-trained large model to obtain multiple vehicle control instructions corresponding to the vehicle data output by the large model, where the multiple vehicle control instructions include power output control instructions, braking control instructions, and suspension control instructions; and controlling the vehicle to execute the operations corresponding to the multiple vehicle control instructions.

[0005] In a second aspect, an embodiment of this application provides a vehicle control device, which includes: a data acquisition unit for obtaining vehicle data, where the vehicle data includes vehicle operation data, driving environment data, and driver operation information; an instruction acquisition unit for inputting the vehicle data into a pre-trained large model to obtain multiple vehicle control instructions corresponding to the vehicle data output by the large model, where the multiple vehicle control instructions include power output control instructions, braking control instructions, and suspension control instructions; and a control unit for controlling the vehicle to execute the operations corresponding to the multiple vehicle control instructions.

[0006] In a third aspect, an embodiment of this application provides a vehicle, which includes one or more processors and a memory; one or more programs, where the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs are configured to execute the above method.

[0007] Fourthly, an embodiment of the present application provides a computer-readable storage medium. Program codes are stored in the computer-readable storage medium. When the program codes are running, the above-mentioned method is executed.

[0008] An embodiment of the present application provides a vehicle control method, device, vehicle and storage medium. First, vehicle data is obtained. The vehicle data includes vehicle operation data, driving environment data, and driver operation information. Then, the vehicle data is input into a pre-trained large model to obtain various vehicle control instructions corresponding to the vehicle data output by the large model. The various vehicle control instructions include power output control instructions, braking control instructions, and suspension control instructions. Finally, the vehicle is controlled to execute the operations corresponding to the various vehicle control instructions. Through the above method, based on the obtained multi-source vehicle data, the pre-trained large model can output more accurate vehicle control instructions, thereby enabling more accurate control of the vehicle. Description of the Drawings

[0009] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those skilled in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0010] Figure 1 Shows a flowchart of a vehicle control method proposed by an embodiment of the present application;

[0011] Figure 2 Shows a flowchart of a vehicle control method proposed by another embodiment of the present application;

[0012] Figure 3 Shows a structural block diagram of a vehicle control device proposed by an embodiment of the present application;

[0013] Figure 4 Shows a structural block diagram of a vehicle control device proposed by an embodiment of the present application;

[0014] Figure 5 Shows a structural block diagram of a vehicle for executing the vehicle control method according to an embodiment of the present application;

[0015] Figure 6 Shows a storage unit for storing or carrying program codes for implementing the vehicle control method according to an embodiment of the present application. Detailed Embodiments

[0016] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0017] An embodiment of the present application provides a vehicle control method, device, vehicle, and storage medium. First, vehicle data is obtained. The vehicle data includes vehicle operation data, driving environment data, and driver operation information. Then, the vehicle data is input into a pre-trained large model to obtain various vehicle control instructions corresponding to the vehicle data output by the large model. The various vehicle control instructions include power output control instructions, braking control instructions, and suspension control instructions. Finally, the vehicle is controlled to execute the operations corresponding to the various vehicle control instructions. Through the above method, based on the obtained multi-source vehicle data, the pre-trained large model can output more accurate vehicle control instructions, thereby enabling more accurate control of the vehicle.

[0018] Next, each embodiment of the present application will be specifically described in conjunction with the drawings.

[0019] Please refer to Figure 1 , a vehicle control method provided by an embodiment of the present application, the method includes:

[0020] Step S110: Obtain vehicle data, where the vehicle data includes vehicle operation data, driving environment data, and driver operation information.

[0021] In the embodiment of the present application, vehicle data can be understood as data reflecting the current state of the vehicle. Vehicle operation data may include data collected by various sensors installed at key parts of the vehicle (such as acceleration sensors, pressure sensors, displacement sensors, wheel speed sensors, etc.) and vehicle state data in the CAN / LIN bus, such as vehicle speed, acceleration, suspension travel, wheel speed, braking pressure, motor torque, battery SOC, engine speed, etc. Driving environment data may include road environment data and in-vehicle environment data collected by lidar / camera (which needs to be associated with the chassis control logic), such as road bumpiness, slope, obstacle information, road curvature, etc. Driver operation information may include operation information executed by the current driver when controlling the vehicle, such as driver input data such as steering wheel angle, accelerator / brake pedal depth, gear state, driving mode, etc., which is not specifically limited here.

[0022] Among them, the driving environment data and driver operation information can be obtained by calling the in-vehicle network through the data interfaces of the vehicle control system and the driving assistance system.

[0023] As a way, vehicle data can be obtained in real time; or in response to a data acquisition instruction, vehicle data is obtained. Among them, the data acquisition instruction can be an instruction triggered by a specified operation of the driver on the vehicle, or an instruction sent by an electronic device that has established a traffic connection with the vehicle, and specific limitations are not made here. Among them, the specified operation on the vehicle can be a pre-set operation that can trigger a data acquisition instruction. For example, a click operation or a slide operation on the vehicle's display screen, etc.

[0024] Step S120: Input the vehicle data into a pre-trained large model, and obtain various vehicle control instructions corresponding to the vehicle data output by the large model. The various vehicle control instructions include a power output control instruction, a braking control instruction, and a suspension control instruction.

[0025] In the embodiments of the present application, the large model can be a deep neural network, a recurrent neural network, etc. The pre-trained large model can be understood as a lightweight model that can be used to output various vehicle control instructions according to the input vehicle data. The pre-trained large model can be a lightweight model regularly downloaded by the vehicle from the cloud (where the dynamic optimization and iteration of the model can be realized), or can be a lightweight model directly deployed on the vehicle side. According to the resource situation of the vehicle, the large model can be a whole model or a model jointly controlled by multiple distilled small models.

[0026] When the large model is a whole model, the obtained vehicle data is input into the large model, and the large model can directly output various vehicle control instructions; when the large model is a model jointly controlled by multiple distilled small models, the obtained vehicle inputs can be respectively input into multiple distilled small models, and multiple distilled small models can each output a vehicle control instruction.

[0027] In the embodiment of the present application, the vehicle control instruction can be understood as an instruction for controlling the chassis performance of the vehicle, and the multiple vehicle control instructions can be used to adjust the chassis parameters of different categories of vehicles. Among them, the power output control instruction can be understood as an instruction that can be used to adjust the engine power output and the working mode of the electric drive system, wherein the electric drive system mainly refers to the torque distribution mode and power vector control; the brake control instruction can be understood as an instruction that can be used to optimize the brake distribution strategy to achieve a smooth and efficient braking effect; the suspension control instruction can be understood as an instruction that can be used to adjust the hardness and damping parameters of the suspension. Among them, the torque distribution mode refers to the dynamic distribution of the torque output of the front and rear axle motors through the electronic control system in a dual-motor or four-wheel drive system to optimize traction or stability, for example, reducing the output of a single motor on a slippery road to prevent slipping; power vector control refers to the independent control of the speed and torque of the wheel hub motor to achieve precise steering assistance or body stability adjustment; the brake distribution strategy mainly refers to the three dimensions of mechanical and electronic control coordination, energy recovery and safety balance, and dynamic scene adaptation. Optionally, mechanical and electronic control work together: the front and rear wheel braking forces are not evenly distributed. For example, when braking, the load on the front wheels is significantly higher than that on the rear wheels due to the forward shift of the center of gravity, so the front wheel braking force is usually greater. For example, the front and rear braking force ratio of a front-wheel drive vehicle can reach 9:1. The ideal distribution needs to be combined with the vehicle's axle weight ratio and adhesion coefficient to avoid premature locking of the rear wheels and causing skidding; or the front wheels are mainly mechanically braked, and the electric brake force is distributed to the rear wheels to avoid locking of the rear wheels. Energy recovery and safety balance: such as giving priority to electric braking, or dynamically adjusting the motor braking ratio according to vehicle speed, braking intensity, and battery SOC to optimize energy recovery efficiency. Dynamic scene adaptation: such as off-road scenes, such as mechanical differential locks, rigid locking differentials, and forced torque balance to facilitate extreme escape.

[0028] In an embodiment of the present application, the pre-trained large model may include a first sub-model, a second sub-model, and a third sub-model. The first sub-model is used to output a power output control instruction according to the input vehicle data; the second sub-model is used to output a braking control instruction according to the input vehicle data; and the third sub-model is used to output a suspension control instruction according to the input vehicle data. The first sub-model, the second sub-model, and the third sub-model can be understood as distilled small models, which are used to collaboratively output vehicle control instructions to achieve collaborative control of the vehicle.

[0029] As a way, input the vehicle data into the first sub-model of the large model to obtain the power output control instruction corresponding to the vehicle data output by the first sub-model; input the vehicle data into the second sub-model of the large model to obtain the braking control instruction corresponding to the vehicle data output by the second sub-model; input the vehicle data into the third sub-model of the large model to obtain the suspension control instruction corresponding to the vehicle data output by the third sub-model.

[0030] As another way, before step S120, it further includes: obtaining a training data set, where the training data set includes a plurality of training vehicle data and various preset vehicle control instructions corresponding to the plurality of training vehicle data; iteratively training the large model to be trained based on the training data set until the training end condition is met, and obtaining the pre-trained large model.

[0031] Among them, the iteratively training the large model to be trained based on the training data set until the training end condition is met and obtaining the pre-trained large model includes: respectively inputting the plurality of training vehicle data into the large model to be trained, and obtaining various predicted vehicle control instructions corresponding to each training vehicle data output by the large model to be trained, where the first sub-model, the second sub-model, and the third sub-model of the large model to be trained are respectively used to output different predicted vehicle control instructions according to the training vehicle data; determining the corresponding loss value based on the various predicted vehicle control instructions and various preset vehicle control instructions corresponding to each training vehicle data; and iteratively training the large model to be trained based on the loss value until the training end condition is met, and obtaining the pre-trained large model.

[0032] In the embodiments of the present application, the training end condition is a pre-set condition indicating that the model has been trained. For example, the training end condition can be set as the number of training iterations reaching a preset number, or the training end condition can be set as the loss value reaching a preset loss value and no longer decreasing, etc., which is not specifically limited herein.

[0033] A variety of preset vehicle control instructions can be understood as a variety of standard vehicle control instructions corresponding to each piece of training vehicle data set in advance. The variety of preset vehicle control instructions can include a preset power output control instruction, a preset braking control instruction, and a preset suspension control instruction. The variety of predicted vehicle control instructions are vehicle control instructions for controlling the vehicle within a preset time period in the future output by the sub-model based on the input training vehicle data. The variety of predicted vehicle control instructions can include a predicted power output control instruction, a predicted braking control instruction, and a predicted suspension control instruction. Among them, the preset power output control instruction corresponds to the predicted power output control instruction, the preset braking control instruction corresponds to the predicted braking control instruction, and the preset suspension control instruction corresponds to the predicted suspension control instruction.

[0034] Since the large model to be trained can include a first sub-model, a second sub-model, and a third sub-model, in order to enable different sub-models to output different vehicle control instructions, therefore, when training the large model to be trained based on the training data set, the training data of the first sub-model, the second sub-model, and the third sub-model are slightly different. Among them, the training data of the first sub-model can include training vehicle data, a preset power output control instruction, and a predicted power output control instruction; the training data of the second sub-model can include training vehicle data, a preset braking control instruction, and a predicted braking control instruction; the training data of the third sub-model can include training vehicle data, a preset suspension control instruction, and a predicted suspension control instruction.

[0035] During the training process, when calculating the corresponding loss value, the loss value corresponding to each piece of training data can include a first loss value, a second loss value, and a third loss value. Among them, the first loss value is the loss value of the first sub-model, and the first loss value can be calculated based on the preset power output control instruction and the predicted power output control instruction; the second loss value is the loss value of the second sub-model, and the second loss value can be calculated based on the preset braking control instruction and the predicted braking control instruction; the third loss value is the loss value of the third sub-model, and the third loss value can be calculated based on the preset suspension control instruction and the predicted suspension control instruction.

[0036] When performing iterative training on the large model to be trained based on the first loss value, the second loss value, and the third loss value, the first loss value, the second loss value, and the third loss value can be weighted and calculated to obtain a total loss value, so that the large model to be trained can be iteratively trained based on the total loss value until the training end condition is met, and the final large model is obtained.

[0037] Similarly, when the large model to be trained is a whole model, the training process of the large model to be trained can also refer to the above model training process, which will not be elaborated here.

[0038] Optionally, in order to improve the adaptability and robustness of the large model, an online learning and model update mechanism can also be adopted. That is, the large model can continuously optimize the model parameters according to new real-time data, so that it can better adapt to different driving conditions and changes in vehicle states.

[0039] Furthermore, when the large model is a whole model, the large model can be deployed in any controller (for example, the central domain controller). When the large model is composed of multiple sub-models, the multiple sub-models can be deployed in the same controller or in different controllers, and no specific limitation is made here.

[0040] Step S130: Control the vehicle to execute the operations corresponding to the multiple vehicle control instructions.

[0041] In the embodiments of the present application, there can be at least one operation corresponding to each vehicle control instruction to achieve the control of different parts of the vehicle chassis. Among them, at least one operation can be an operation for jointly controlling a specific part of the chassis.

[0042] After multiple vehicle control instructions are output by the pre-trained large model, the multiple vehicle control instructions can be sent to the central domain controller. Then, the central domain controller can coordinate the action timings of the suspension, motor, and braking system based on the multiple vehicle control instructions, eliminate control conflicts, and achieve the final optimal execution strategy, so as to realize the precise control of the vehicle chassis.

[0043] As a way, determine the execution order of the multiple vehicle control instructions; based on the execution order, control the vehicle to execute the operations corresponding to the multiple vehicle control instructions.

[0044] Among them, the execution order refers to the time sequence of execution of different types of vehicle control instructions. The execution order of multiple vehicle control instructions can be coordinated by the central domain controller. After the central domain controller coordinates the execution order of multiple vehicle control instructions, it can send the corresponding vehicle control instructions and execution order to the actuators that execute the corresponding operations, so that each actuator executes the multiple vehicle control instructions in sequence according to the execution order.

[0045] A vehicle control method provided by the present application first obtains vehicle data, where the vehicle data includes vehicle operation data, driving environment data, and driver operation information, then inputs the vehicle data into a pre-trained large model to obtain multiple vehicle control instructions corresponding to the vehicle data output by the large model. The multiple vehicle control instructions include power output control instructions, braking control instructions, and suspension control instructions, and finally controls the vehicle to execute operations corresponding to the multiple vehicle control instructions. Through the above method, based on the obtained multi-source vehicle data, the pre-trained large model can output more accurate vehicle control instructions, thereby enabling more accurate control of the vehicle.

[0046] Please refer to Figure 2 , a vehicle control method provided by an embodiment of the present application, the method includes:

[0047] Step S210: Obtain vehicle data, where the vehicle data includes vehicle operation data, driving environment data, and driver operation information.

[0048] Step S220: Obtain the driving habit data of the driver.

[0049] In the embodiment of the present application, the driving habit data of the driver may include data such as the driver's habitual driving time, driving route, and driving operation actions, and no specific limitation is made here.

[0050] As a way, the driving habit data of the driver can be obtained simultaneously with the vehicle data, or can be obtained before the vehicle data is obtained, or can be obtained after the vehicle data is obtained, and no specific limitation is made here.

[0051] As another way, the driving habit data of the driver can also be obtained when the driver needs to optimize or adjust multiple vehicle control instructions output by the large model based on the vehicle data. Optionally, the driving habit data of the driver can be obtained through vehicle-cloud big data for personal customized calibration; it has the ability to generate predictive control instructions. For example, the large model can output the suspension height curve, four-wheel torque distribution matrix, and steering angle increment within the next 5 seconds based on the driving habit data of the driver.

[0052] When obtaining the driving habit data of the driver, the legitimacy of the driver can be identified first, so that the driving habit data of the legal driver can be obtained from the preset storage area. Among them, when identifying the legitimacy of the driver, the face image of the current driver can be collected through an image acquisition device (such as a camera) set in the vehicle to identify the legitimacy of the current driver.

[0053] Specifically, the facial images or personal feature information of legal drivers can be pre-stored. After the personal image of the current driver is collected by the image acquisition device, it can be compared with the pre-stored facial images or personal feature information of legal drivers to determine whether the current driver is a legal driver.

[0054] Step S230: Input the driving habit data and the vehicle data into the large model to obtain various optimized vehicle control instructions corresponding to the vehicle data output by the large model. The optimized vehicle control instructions are vehicle control instructions obtained by combining the driving habit data and the vehicle data.

[0055] In an embodiment of the present application, when the driving habit data of the driver and the current vehicle data are obtained, the driving habit data and the vehicle data can be input into a pre-trained large model together. Thus, the large model can optimize and adjust various vehicle control instructions corresponding to the current vehicle data in combination with the driving habit data of the driver to output various optimized vehicle control instructions. Among them, the various optimized vehicle control instructions can include optimized power output control instructions, optimized braking control instructions, and optimized suspension control instructions; the optimized power output control instructions can be understood as instructions obtained by optimizing and adjusting the power output control instructions based on the driving habit data of the driver; the optimized braking control instructions can be understood as instructions obtained by optimizing and adjusting the braking control instructions based on the driving habit data of the driver; the optimized suspension control instructions can be instructions obtained by optimizing and adjusting the suspension control instructions based on the driving habit data of the driver.

[0056] As a way, before step S230, it can also include: iteratively training the large model based on historical driving habit data, historical vehicle data, and historical various vehicle control instructions, so that the trained large model can have the ability to generate optimized vehicle control instructions by combining driving habit data and vehicle data.

[0057] Step S240: Control the vehicle to perform operations corresponding to the various optimized vehicle control instructions.

[0058] In an embodiment of the present application, similarly, when various optimized vehicle control instructions are obtained, the various optimized vehicle control instructions can also be sent to the central domain controller. Furthermore, the central domain controller can coordinate the action timings of the suspension, motor, and braking system based on the various optimized vehicle control instructions, eliminate control conflicts, and achieve the final optimal execution strategy, thereby realizing precise control of the vehicle chassis.

[0059] As a way, determine the execution order of the various optimized vehicle control instructions; based on the execution order, control the vehicle to perform operations corresponding to the various optimized vehicle control instructions.

[0060] Among them, the execution order refers to the time sequence of executing different types of optimized vehicle control instructions. The execution order of multiple optimized vehicle control instructions can be coordinated by the central domain controller. After the central domain controller coordinates the execution order of multiple optimized vehicle control instructions, it can send the corresponding optimized vehicle control instructions and the execution order to the actuators that perform corresponding operations, so that each actuator executes multiple optimized vehicle control instructions in sequence according to the execution order.

[0061] Step S250: Obtain historical vehicle data, where the historical vehicle data is the vehicle data obtained before the current moment.

[0062] In the embodiment of the present application, the historical vehicle data can be the vehicle data obtained within a period of time before the current moment, or it can also be the vehicle data at a certain moment before the current moment. If the historical vehicle data is used to determine whether the current vehicle has a fault, then the historical vehicle data needs to be the vehicle data obtained to determine that the vehicle has no fault, or the vehicle data obtained to determine that the vehicle has a fault.

[0063] Step S260: Based on the vehicle data and the historical vehicle data, determine whether the vehicle has a fault.

[0064] In the embodiment of the present application, when determining whether the vehicle has a fault based on the vehicle data and the historical vehicle data, the vehicle data and the historical vehicle data can be compared one by one to see if they are the same, so as to determine whether the vehicle has a fault.

[0065] Step S270: If the vehicle has a fault, give a fault reminder.

[0066] In the embodiment of the present application, when it is determined that the vehicle has a fault, a warning can be sent to the driver and passengers through the in-vehicle language system, and the abnormal data can be uploaded to the cloud. For example, when it is determined that the vehicle has abnormal vibration frequency, overheating, etc., a warning can be sent to the passengers through the in-vehicle language system, reminding them not to move around casually, fasten their seat belts, etc., and upload the abnormal data to the cloud.

[0067] A vehicle control method provided by the present application, the large model combines the driving habit data of the driver and the vehicle data, and can output personalized multiple vehicle control instructions. For the driver, the control of the vehicle can be more in line with the driver's habits. Further, a fault reminder can also be given when the vehicle has a fault, improving the driving safety of the driver.

[0068] Please refer to Figure 3 , a vehicle control device 300 provided by the embodiment of the present application, the device 300 includes:

[0069] A data acquisition unit 310, configured to acquire vehicle data, where the vehicle data includes vehicle operation data, driving environment data, and driver operation information.

[0070] An instruction acquisition unit 320, configured to input the vehicle data into a pre-trained large model, and acquire a variety of vehicle control instructions corresponding to the vehicle data output by the large model, where the variety of vehicle control instructions includes a power output control instruction, a braking control instruction, and a suspension control instruction.

[0071] As one way, the instruction acquisition unit 320 is specifically configured to input the vehicle data into the first sub-model of the large model to acquire a power output control instruction corresponding to the vehicle data output by the first sub-model; input the vehicle data into the second sub-model of the large model to acquire a braking control instruction corresponding to the vehicle data output by the second sub-model; input the vehicle data into the third sub-model of the large model to acquire a suspension control instruction corresponding to the vehicle data output by the third sub-model.

[0072] As another way, the instruction acquisition unit 320 is further specifically configured to acquire the driving habit data of the driver; input the driving habit data and the vehicle data into the large model to acquire a variety of optimized vehicle control instructions corresponding to the vehicle data output by the large model, where the optimized vehicle control instructions are vehicle control instructions obtained by combining the driving habit data and the vehicle data.

[0073] A control unit 330, configured to control the vehicle to perform operations corresponding to the variety of vehicle control instructions.

[0074] As one way, the control unit 330 is specifically configured to determine the execution order of the variety of vehicle control instructions; based on the execution order, control the vehicle to perform operations corresponding to the variety of vehicle control instructions.

[0075] Optionally, as Figure 4 shown, the device 300 further includes:

[0076] A model training unit 340, configured to acquire a training data set, where the training data set includes a plurality of training vehicle data and a variety of preset vehicle control instructions corresponding to the plurality of training vehicle data respectively; perform iterative training on the large model to be trained based on the training data set until a training end condition is met, and obtain the pre-trained large model.

[0077] Specifically, the model training unit 340 is specifically configured to input the multiple pieces of training vehicle data into the large model to be trained respectively, and obtain multiple predicted vehicle control instructions corresponding to each piece of training vehicle data output by the large model to be trained. Among them, the first sub-model, the second sub-model, and the third sub-model of the large model to be trained are respectively configured to output different predicted vehicle control instructions according to the training vehicle data; determine the corresponding loss value based on the multiple predicted vehicle control instructions and multiple preset vehicle control instructions corresponding to each piece of training vehicle data; perform iterative training on the large model to be trained based on the loss value until the training end condition is satisfied, and obtain the pre-trained large model.

[0078] The reminder unit 350 is configured to obtain historical vehicle data, where the historical vehicle data is vehicle data obtained before the current moment; determine whether the vehicle has a fault based on the vehicle data and the historical vehicle data; if the vehicle has a fault, give a fault reminder.

[0079] It should be noted that the device embodiments in this application correspond to the foregoing method embodiments. The specific principles in the device embodiments can be referred to the content in the foregoing method embodiments, and will not be elaborated here.

[0080] Next, a vehicle provided by this application will be described in conjunction with Figure 5 a vehicle provided by this application will be described.

[0081] Please refer to Figure 5 , based on the above vehicle control method and device, another vehicle 800 that can execute the foregoing vehicle control method is further provided in an embodiment of this application. The vehicle 800 includes one or more (only one is shown in the figure) processors 802, a memory 804, and a network module 806 that are coupled to each other. Among them, a program that can execute the content in the foregoing embodiments is stored in the memory 804, and the processor 802 can execute the program stored in the memory 804.

[0082] Among them, the processor 802 may include one or more processing cores. The processor 802 connects various parts within the entire vehicle 800 through various interfaces and lines, and executes various functions of the vehicle 800 and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 804, and by calling the data stored in the memory 804. Optionally, the processor 802 may be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), or programmable logic array (PLA). The processor 802 may integrate a combination of one or several of a central processing unit (CPU), a graphics processing unit (GPU), and a modem, etc. Among them, the CPU mainly processes the operating system, user interface, application programs, etc.; the GPU is responsible for rendering and drawing the displayed content; the modem is used to process wireless communication. It can be understood that the above-mentioned modem may not be integrated into the processor 802 and may be implemented separately through a communication chip.

[0083] The memory 804 may include random access memory (RAM) and may also include read-only memory. The memory 804 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 804 may include a program storage area and a data storage area. Among them, the program storage area may store instructions for implementing the operating system, instructions for implementing at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the following various method embodiments, etc. The data storage area may also store data created during the use of the vehicle 800 (such as phone books, audio and video data, chat record data), etc.

[0084] The network module 806 is used to receive and transmit electromagnetic waves, realizing the mutual conversion between electromagnetic waves and electrical signals, so as to communicate with a communication network or other devices, such as communicating with a vehicle. The network module 806 may include various existing circuit elements for performing these functions, such as antennas, radio frequency transceivers, digital signal processors, encryption / decryption chips, subscriber identity module (SIM) cards, memories, and so on. The network module 806 can communicate with various networks such as the Internet, enterprise intranets, wireless networks or communicate with other devices through a wireless network. The above-mentioned wireless network may include a cellular phone network, a wireless local area network or a metropolitan area network. For example, the network module 806 can interact with a base station.

[0085] Please refer to Figure 6 , which shows a structural block diagram of a computer-readable storage medium provided by an embodiment of the present application. Program code is stored in the computer-readable storage medium 900, and the program code can be called by a processor to execute the method described in the above method embodiment.

[0086] The computer-readable storage medium 900 can be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk or a ROM. Optionally, the computer-readable storage medium 900 includes a non-transitory computer-readable storage medium. The computer-readable storage medium 900 has a storage space for the program code 910 that executes any method step in the above method. These program codes can be read out from or written into one or more computer program products. The program code 910 can be compressed in an appropriate form, for example.

[0087] A vehicle control method, device, vehicle and storage medium provided by the present application first obtain vehicle data, where the vehicle data includes vehicle operation data, driving environment data and driver operation information, and then input the vehicle data into a pre-trained large model to obtain various vehicle control instructions corresponding to the vehicle data output by the large model. The various vehicle control instructions include power output control instructions, braking control instructions and suspension control instructions, and finally control the vehicle to execute operations corresponding to the various vehicle control instructions. Through the above method, based on the obtained multi-source vehicle data, the pre-trained large model can output more accurate vehicle control instructions, so as to achieve more accurate control of the vehicle.

[0088] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative rather than restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the purpose of the present invention and the scope protected by the claims, and all of them fall within the protection scope of the present invention.

Claims

1. A vehicle control method, characterized in that, The method includes: Obtaining vehicle data, where the vehicle data includes vehicle operation data, driving environment data, and the operation information of the driver; Inputting the vehicle data into a pre-trained large model to obtain various vehicle control instructions corresponding to the vehicle data output by the large model, where the various vehicle control instructions include power output control instructions, braking control instructions, and suspension control instructions; Controlling the vehicle to perform operations corresponding to the various vehicle control instructions.

2. The method according to claim 1, characterized in that, The step of inputting the vehicle data into a pre-trained large model to obtain various vehicle control instructions corresponding to the vehicle data output by the large model includes: Inputting the vehicle data into the first sub-model of the large model to obtain the power output control instructions corresponding to the vehicle data output by the first sub-model; Inputting the vehicle data into the second sub-model of the large model to obtain the braking control instructions corresponding to the vehicle data output by the second sub-model; Inputting the vehicle data into the third sub-model of the large model to obtain the suspension control instructions corresponding to the vehicle data output by the third sub-model.

3. The method according to claim 2, characterized in that Before the step of inputting the vehicle data into a pre-trained large model to obtain various vehicle control instructions corresponding to the vehicle data output by the large model, it further includes: Obtaining a training data set, where the training data set includes multiple training vehicle data and various preset vehicle control instructions corresponding to the multiple training vehicle data respectively; Performing iterative training on the large model to be trained based on the training data set until the training end condition is met, to obtain the pre-trained large model.

4. The method according to claim 3, characterized in that, The step of performing iterative training on the large model to be trained based on the training data set until the training end condition is met, to obtain the pre-trained large model, includes: Respectively inputting the multiple training vehicle data into the large model to be trained to obtain various predicted vehicle control instructions corresponding to each training vehicle data output by the large model to be trained, where the first sub-model, the second sub-model, and the third sub-model of the large model to be trained are respectively used to output different predicted vehicle control instructions according to the training vehicle data; Determining the corresponding loss value based on the various predicted vehicle control instructions and various preset vehicle control instructions corresponding to each training vehicle data; Performing iterative training on the large model to be trained based on the loss value until the training end condition is met, to obtain the pre-trained large model.

5. The method according to claim 1, wherein The step of inputting the vehicle data into a pre-trained large model to obtain various vehicle control instructions corresponding to the vehicle data output by the large model includes: Obtaining the driving habit data of the driver; Inputting the driving habit data and the vehicle data into the large model to obtain various optimized vehicle control instructions corresponding to the vehicle data output by the large model, where the optimized vehicle control instructions are vehicle control instructions obtained by combining the driving habit data and the vehicle data.

6. The method according to claim 1, characterized in that The step of controlling the vehicle to perform operations corresponding to the various vehicle control instructions includes: Determining the execution order of the various vehicle control instructions; Based on the execution order, control the vehicle to perform operations corresponding to the multiple vehicle control instructions.

7. The method according to claim 1, wherein The method further includes: Obtain historical vehicle data, where the historical vehicle data is vehicle data obtained before the current moment; Based on the vehicle data and the historical vehicle data, determine whether the vehicle has a fault; If the vehicle has a fault, give a fault reminder.

8. A vehicle control device, characterized in that, The device includes: A data acquisition unit for acquiring vehicle data, where the vehicle data includes vehicle operation data, driving environment data, and driver operation information; An instruction acquisition unit for inputting the vehicle data into a pre-trained large model to obtain multiple vehicle control instructions corresponding to the vehicle data output by the large model, where the multiple vehicle control instructions include power output control instructions, braking control instructions, and suspension control instructions; A control unit for controlling the vehicle to perform operations corresponding to the multiple vehicle control instructions.

9. A vehicle, characterized in that, Includes one or more processors and a memory; one or more programs are stored in the memory and are configured to be executed by the one or more processors to perform the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, Program code is stored in the computer-readable storage medium, where the method according to any one of claims 1-7 is executed when the program code is run by a processor.

Images