Vehicle control method, device, apparatus, readable storage medium and program product

Abstract

The present application relates to a vehicle control method, device, equipment, readable storage medium and program product. The method comprises: when the target vehicle is in an assisted driving state, collecting a road surface image and vehicle dynamics information of the target vehicle; determining a first road surface adhesion coefficient based on the road surface image; determining a second road surface adhesion coefficient based on the vehicle dynamics information; fusing the first road surface adhesion coefficient and the second road surface adhesion coefficient to obtain a fused road surface adhesion coefficient; determining a vehicle control parameter based on the fused road surface adhesion coefficient, and driving and controlling the target vehicle based on the vehicle control parameter. The use of this method can improve the control accuracy of the vehicle.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a vehicle control method, device, equipment, readable storage medium and program product. Background Art

[0002] In autonomous driving and assisted driving systems, the stability and safety of vehicle control is one of the core requirements of system design. These systems usually perceive the surrounding environment through sensors and control algorithms, and dynamically adjust the vehicle's driving strategy based on road conditions.

[0003] To ensure system reliability, system designs usually assume that road conditions are fixed or in a known standard state, such as dry asphalt road or good traffic conditions.

[0004] However, in actual driving, road conditions are often constantly changing, and autonomous driving and assisted driving systems are unable to accurately respond to changes in road conditions, resulting in the inability to effectively control the vehicle. Summary of the invention

[0005] Based on this, it is necessary to provide a vehicle control method, device, computer equipment, computer-readable storage medium and computer program product that can improve control accuracy in response to the above technical problems.

[0006] In a first aspect, the present application provides a vehicle control method, the method comprising:

[0007] When the target vehicle is in the assisted driving state, collecting road images and vehicle dynamics information of the target vehicle;

[0008] determining a first road surface adhesion coefficient based on the road surface image;

[0009] determining a second road adhesion coefficient based on the vehicle dynamics information;

[0010] fusing the first road surface adhesion coefficient and the second road surface adhesion coefficient to obtain a fused road surface adhesion coefficient;

[0011] A vehicle control parameter is determined based on the fused road adhesion coefficient, and driving control is performed on the target vehicle based on the vehicle control parameter.

[0012] In one embodiment, determining the first road surface adhesion coefficient based on the road surface image includes:

[0013] Recognize the road surface image based on a pre-trained road surface recognition model to obtain a road surface category of the road surface image;

[0014] A first road surface adhesion coefficient is determined based on the road surface category.

[0015] In one embodiment, the method further comprises:

[0016] After each recognition, the recognized road surface image is added to a sample set;

[0017] Processing the road surface image in the sample set by using the pre-trained road surface recognition model to obtain a predicted category of the road surface image;

[0018] The parameters of the pre-trained road surface recognition model are optimized based on the predicted category and the real category of the road surface image to obtain an optimized road surface recognition model.

[0019] In one embodiment, determining a second road adhesion coefficient based on the vehicle dynamics information includes:

[0020] determining longitudinal force and lateral force of wheels of the target vehicle based on the vehicle dynamics information;

[0021] determining a second road adhesion coefficient based on a target tire model, a longitudinal force and a lateral force of the wheel;

[0022] The target tire model is a tire model obtained by improving the Dugoff model based on the scaling factor.

[0023] In one embodiment, the fusing the first road surface adhesion coefficient and the second road surface adhesion coefficient to obtain a fused road surface adhesion coefficient includes:

[0024] Obtaining a first gain corresponding to the first road adhesion coefficient and a second gain corresponding to the second road adhesion coefficient;

[0025] determining a fusion coefficient based on the first road adhesion coefficient, the first gain, the second road adhesion coefficient, and the second gain;

[0026] determining a difference between the first road adhesion coefficient and the fusion coefficient, and adjusting the first gain and the second gain based on the difference to obtain an adjusted first gain and an adjusted second gain;

[0027] Return to the step of determining a fusion coefficient based on the first road surface adhesion coefficient, the first gain, the second road surface adhesion coefficient and the second gain until a convergence condition is met, and determine the fusion coefficient at the time of convergence as the fused road surface adhesion coefficient.

[0028] In one embodiment, the vehicle control parameter includes wheel torque; and determining the vehicle control parameter based on the fused road adhesion coefficient includes:

[0029] Determining a plurality of desired control targets of the target vehicle based on the fused road adhesion coefficient and the vehicle dynamics information; the plurality of desired control targets include a following distance deviation target, a desired speed deviation target, a desired acceleration deviation target, and a desired jerk deviation target;

[0030] Determine the weights corresponding to the desired control objectives respectively;

[0031] Determining a current optimal expected longitudinal acceleration based on the weight and the expected control target;

[0032] A wheel torque is determined based on the optimal desired longitudinal acceleration.

[0033] In a second aspect, the present application also provides a vehicle control device, the device comprising:

[0034] An information acquisition module, used to acquire road surface images and vehicle dynamics information of the target vehicle when the target vehicle is in an assisted driving state;

[0035] An image processing module, configured to determine a first road surface adhesion coefficient based on the road surface image;

[0036] a dynamic information processing module, configured to determine a second road adhesion coefficient based on the vehicle dynamic information;

[0037] A coefficient fusion module, used for fusing the first road adhesion coefficient and the second road adhesion coefficient to obtain a fused road adhesion coefficient;

[0038] A control parameter determination module is used to determine a vehicle control parameter based on the fused road adhesion coefficient, and to perform driving control on the target vehicle based on the vehicle control parameter.

[0039] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0040] A computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0041] A computer program product comprises a computer program, wherein the computer program implements the steps of the above method when executed by a processor.

[0042] The above-mentioned vehicle control method, device, computer equipment, computer-readable storage medium and computer program product, when the target vehicle is in an assisted driving state, the target vehicle collects road surface images and vehicle dynamics information of the target vehicle, determines a first road surface adhesion coefficient based on the road surface image, and determines a second road surface adhesion coefficient based on the vehicle dynamics information, so that the road surface adhesion coefficient of the current road surface can be estimated from both visual and dynamic aspects, and the road surface adhesion coefficients estimated from these two aspects are fused to obtain an accurate and reliable fused road surface adhesion coefficient, and the vehicle control parameters of the target vehicle are dynamically adjusted according to the fused road surface adhesion coefficient, so as to control the driving of the target vehicle based on the vehicle control parameters, so that the vehicle can optimize the control strategy in real time under different road conditions to ensure driving safety and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the drawings required for use in the embodiments of the present application or related technical descriptions will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A diagram showing an application environment of a vehicle control method in an embodiment;

[0045] Figure 2 A schematic flow chart of a vehicle control method in one embodiment;

[0046] Figure 3 is a flow chart of a vehicle control method in another embodiment;

[0047] Figure 4 A schematic diagram of a flow chart of a step of determining a road adhesion coefficient in one embodiment;

[0048] Figure 5 is a structural block diagram of a vehicle control device in one embodiment;

[0049] Figure 6 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0051] The vehicle control method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. The vehicle control method can be executed by the terminal 102 or the server 104 alone, or it can be executed interactively by the terminal 102 and the server 104. Taking the execution on the terminal 102 as an example, when the target vehicle is in an assisted driving state, the terminal 102 collects a road surface image and vehicle dynamics information of the target vehicle; determines a first road surface adhesion coefficient based on the road surface image; determines a second road surface adhesion coefficient based on the vehicle dynamics information; fuses the first road surface adhesion coefficient and the second road surface adhesion coefficient to obtain a fused road surface adhesion coefficient; determines a vehicle control parameter based on the fused road surface adhesion coefficient, and controls the driving of the target vehicle based on the vehicle control parameter.

[0052] The terminal 102 may be, but is not limited to, various personal computers, vehicle-mounted computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. IoT devices may be smart speakers, smart TVs, smart air conditioners, smart vehicle-mounted devices, projection devices, etc. Portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. Head-mounted devices may be virtual reality (VR) devices, augmented reality (AR) devices, smart glasses, etc. The server 104 may be an independent physical server, or a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides cloud computing services.

[0053] In an exemplary embodiment, Figure 2 As shown, a vehicle control method is provided, which is applied to Figure 1 The terminal 102 in the example is used as an example to illustrate, and the following steps are included:

[0054] S202, when the target vehicle is in an assisted driving state, collecting a road surface image and vehicle dynamics information of the target vehicle.

[0055] The assisted driving state refers to a state in which the system provides a certain degree of automated support for certain driving tasks, but the driver still needs to remain alert and ready to take over control at any time. Driving tasks can specifically include automatic cruise control, automatic emergency braking, lane keeping assist, etc.

[0056] Road surface images are image data collected by vehicle-mounted visual sensors (such as cameras, lidar, or other imaging devices). These images record the conditions of the road surface during vehicle driving.

[0057] Vehicle dynamics information is a parameter that describes the various forces and influences that a vehicle is subject to during its motion. Specifically, it may include vehicle speed, acceleration, steering angle, tire speed, brake pressure, lateral acceleration, etc. Vehicle speed refers to the current speed of the target vehicle. Acceleration describes the current acceleration or deceleration of the target vehicle. The steering angle reflects the current steering wheel operating state of the target vehicle. Lateral acceleration refers to the lateral force applied to the vehicle body when the target vehicle turns, and can also reflect the stability of the vehicle.

[0058] Specifically, when the target vehicle is in a clothing driving state, the road surface image can be collected in real time through the on-board camera, and the vehicle dynamics information of the target vehicle can be collected through the on-board sensor of the target vehicle.

[0059] The on-board cameras include forward-facing cameras and rear-view cameras. The rear-view cameras can also be called electronic rearview mirrors. The on-board sensors can specifically be the vehicle's IMU sensors and steering sensors. The IMU sensor is a device used in the vehicle to measure and report motion parameters such as speed, acceleration, angular velocity, etc. of an object. It is usually composed of multiple sensors. The main function of the IMU sensor is to sense the dynamic state of the vehicle through accelerometers and gyroscopes. The accelerometer measures the acceleration of the vehicle in three axes (longitudinal, lateral, and vertical directions), and the gyroscope measures the angular velocity of the vehicle, especially when the vehicle turns and rotates. It can provide the vehicle's pitch, roll, yaw and other angle change data; the steering sensor is a sensor used to measure the vehicle's steering wheel angle and steering angle changes.

[0060] S204: Determine a first road surface adhesion coefficient based on the road surface image.

[0061] Among them, the first road adhesion coefficient is the friction coefficient between a specific road surface and a vehicle tire determined through visual information, also called the adhesion coefficient. The adhesion coefficient determines the ease with which a tire slides on the road surface, and whether the vehicle can effectively perform operations such as acceleration, braking, and steering. It can be understood that the higher the adhesion coefficient, the greater the friction between the tire and the road surface, and the more stable the vehicle control. For example, a dry asphalt road surface has a higher adhesion coefficient, while a wet or icy road surface has a lower adhesion coefficient.

[0062] Specifically, after collecting the road surface image, the terminal may identify the road surface image, thereby predicting the first road surface adhesion coefficient.

[0063] In one embodiment, before the terminal recognizes the road surface image, it can also pre-process the road surface image to obtain a pre-processed road surface image, and then recognize the pre-processed road surface image. The pre-processing can specifically be filtering, denoising, contrast enhancement and other pre-processing operations, so as to extract clearer and more recognizable road surface features from the pre-processed road surface image, and then predict a more accurate first road surface adhesion coefficient based on these road surface features.

[0064] In one embodiment, the process of determining the first road adhesion coefficient based on the road surface image by the terminal specifically includes the following steps: identifying the road surface image based on a pre-trained road surface recognition model to obtain the road surface category of the road surface image; and determining the first road adhesion coefficient based on the road surface category.

[0065] Among them, the road surface recognition model is a model used to automatically recognize road surface images and classify road surface types. Specifically, it can be a deep learning model such as a convolutional neural network (CNN). In an embodiment of the present application, the road surface recognition model can be obtained after pre-training on a large amount of labeled road surface image data.

[0066] Specifically, the terminal can input the acquired road surface image into a pre-trained road surface recognition model, extract features of the road surface image through the feature extraction network of the road surface recognition model to obtain road surface features, and input the road surface features into the classification prediction network of the road surface recognition model, output the predicted road surface category through the classification prediction network, and determine the first road surface adhesion coefficient corresponding to the predicted road surface category based on the pre-established correspondence between the road surface category and the road surface adhesion coefficient. The adhesion coefficient is determined through the road surface image, and combined with the pre-trained model for automatic recognition and estimation, so as to quickly and accurately obtain the road surface state.

[0067] The road surface category may specifically include at least one of dry asphalt road surface, waterlogged road surface, muddy road surface, icy and snowy road surface, cement road surface, etc. For example, the correspondence between the road surface category and the road surface adhesion coefficient refers to the quantitative mapping relationship between each road surface type and its friction force (i.e., adhesion coefficient), for example, the road surface adhesion coefficient corresponding to the dry asphalt road surface is 0.9, the road surface adhesion coefficient corresponding to the cement road surface is 0.6, the road surface adhesion coefficient corresponding to the waterlogged road surface is 0.4, the road surface adhesion coefficient corresponding to the muddy road surface is 0.3, etc.

[0068] S206: Determine a second road adhesion coefficient based on the vehicle dynamics information.

[0069] The second road adhesion coefficient is a friction coefficient between a specific road surface and a vehicle tire determined through vehicle dynamics information.

[0070] Specifically, after obtaining the vehicle dynamics information, the terminal may estimate the second road adhesion coefficient based on the vehicle dynamics information and in combination with a vehicle dynamics model.

[0071] The dynamic model of the vehicle includes a vehicle model and a tire model. The vehicle model may be a three-degree-of-freedom vehicle model, and the tire model may be a Dugoff model or a target tire model obtained by improving the Dugoff model.

[0072] In one embodiment, the Dugoff model can be specifically characterized as follows:

[0073]

[0074]

[0075] in, is the longitudinal force of the tire, represents the tire longitudinal slip stiffness, S is the tire slip rate, is the lateral force of the tire, is the lateral stiffness, represents the tire slip angle, f(λ) is a function used to correct the relationship between the slip angle and the tire characteristics and the adhesion coefficient, λ is a variable related to the slip angle (or slip rate), and represents the influence of the tire slip behavior and road conditions on the mechanical properties, which can be specifically determined according to the estimated longitudinal acceleration information. The estimated longitudinal acceleration information is obtained by correcting the longitudinal acceleration information estimated by the transmission mode based on the road slope information.

[0076] In one embodiment, the process of determining the second road adhesion coefficient by the terminal based on vehicle dynamics information includes the following steps: determining the longitudinal force and lateral force of the wheels of the target vehicle based on the vehicle dynamics information; determining the second road adhesion coefficient based on the target tire model, the longitudinal force and lateral force of the wheels; wherein the target tire model is a tire model obtained by improving the Dugoff model based on a scaling factor.

[0077] Specifically, after obtaining the vehicle dynamics information, the terminal can determine the longitudinal force and lateral force of the target vehicle's wheels based on the vehicle dynamics information and in combination with the three-degree-of-freedom vehicle model, and improve the Dugoff model based on the scaling factor to obtain the target tire model. According to the target tire model combined with the longitudinal force, lateral force and road slope information, the model can be used to calculate the adhesion coefficient between the tire and the road surface, thereby obtaining the second road adhesion coefficient. By determining the second road adhesion coefficient based on vehicle dynamics information and the improved tire model, a more accurate, dynamic and real-time estimation of the road adhesion coefficient can be provided, and the adaptability, stability and safety of the vehicle control system can be improved. This method enables the vehicle to achieve optimal control under various complex road conditions, ensure driving safety, and reduce accident risks. It also provides a reliable foundation for autonomous driving and advanced driver assistance systems (ADAS).

[0078] In one embodiment, the target tire model may be characterized as follows:

[0079]

[0080]

[0081] in, , It represents the scaling factor. It can be understood that the target tire model is closer to the magic formula tire model in the entire range of slip rate and sideslip angle, which is in line with the changing trend of the actual tire force. The improved model also has an obvious peak point near the peak point of the magic formula tire model. The parameters of the target tire model are simple and the road adhesion coefficient in the formula is easy to intuitively separate, which makes the calculation amount relatively small, thus meeting the real-time requirements.

[0082] It should be noted that the second road adhesion coefficient can specifically be the maximum available adhesion coefficient calculated by the tire model, that is, the maximum friction force that can be provided between the tire and the road surface under the current driving state of the vehicle. Specifically, during acceleration or braking, the vehicle tire will slip longitudinally, that is, the slip rate (S) is not zero, and the longitudinal force (Fx) and adhesion coefficient (μ) of the vehicle are in a certain relationship. During steering, the tire will slip laterally, and the lateral force (Fy) of the tire is also related to the adhesion coefficient. In these cases, the adhesion coefficient 𝜇 changes with the change of the slip rate, and usually presents a "peak value", which corresponds to the maximum adhesion that the tire can withstand, that is, the maximum adhesion coefficient.

[0083] S208: Fusing the first road surface adhesion coefficient and the second road surface adhesion coefficient to obtain a fused road surface adhesion coefficient.

[0084] Specifically, after obtaining the first road adhesion coefficient and the second road adhesion coefficient, the terminal can respectively obtain the first gain corresponding to the first road adhesion coefficient and the second gain corresponding to the second road adhesion coefficient, and perform weighted fusion of the first road adhesion coefficient and the second road adhesion coefficient according to the first road adhesion coefficient, the first gain, the second road adhesion coefficient and the second gain to obtain a determined fused road adhesion coefficient.

[0085] Among them, the first gain is used to characterize the importance of the first road adhesion coefficient during fusion, and the second gain is used to characterize the importance of the second road adhesion coefficient during fusion. The first gain and the second gain can be pre-set fixed values, for example, the value of the first gain is 0.4, and the value of the second gain is 0.6. The first gain and the second gain can also be dynamically adjusted according to the size of the first road adhesion coefficient and the second road adhesion coefficient. For example, the value of the first gain is initialized to 0.4, and the value of the second gain is initialized to 0.6. Then, according to the actual size of the first road adhesion coefficient and the second road adhesion coefficient, the preset adjustment strategy is used to adjust the value of the first gain and the value of the second gain, and finally it is determined that the value of the first gain is 0.3 and the value of the second gain is 0.7.

[0086] S210, determining a vehicle control parameter based on the fused road adhesion coefficient, and performing driving control on the target vehicle based on the vehicle control parameter.

[0087] Among them, the vehicle control parameters include the wheel torque on the four wheels of the vehicle. It can be understood that the determined wheel torque is an ideal value calculated based on the vehicle's dynamic model, control objectives (such as acceleration, braking, steering or stability, etc.), road adhesion coefficient and other real-time sensor information (such as wheel speed, vehicle speed, acceleration, etc.).

[0088] Specifically, after determining the wheel torque of the vehicle's four wheels, the terminal can adjust the engine speed or torque output or control the wheel angle based on the wheel torque through the vehicle's engine, electronic control unit (ECU), gearbox and other systems, steering system, etc., thereby achieving driving control of the target vehicle.

[0089] In the above vehicle control method, when the target vehicle is in the assisted driving state, the target vehicle determines the first road adhesion coefficient based on the road image and the second road adhesion coefficient based on the vehicle dynamics information by collecting road surface images and vehicle dynamics information of the target vehicle, so that the road adhesion coefficient of the current road surface can be estimated from both visual and dynamic aspects, and the road adhesion coefficients estimated from these two aspects are fused to obtain an accurate and reliable fused road adhesion coefficient. The vehicle control parameters of the target vehicle are dynamically adjusted according to the fused road adhesion coefficient, so as to control the driving of the target vehicle based on the vehicle control parameters, so that the vehicle can optimize the control strategy in real time under different road conditions to ensure driving safety and stability.

[0090] In one embodiment, the above-mentioned vehicle control method also includes the following steps: after each recognition, adding the recognized road surface image to the sample set; processing the road surface image in the sample set by a pre-trained road surface recognition model to obtain a predicted category of the road surface image; optimizing the parameters of the pre-trained road surface recognition model based on the predicted category and the true category of the road surface image to obtain an optimized road surface recognition model.

[0091] Specifically, after the terminal recognizes the collected road surface image each time, the road surface image can be added to the sample set, and the real category of the road surface image can be manually labeled. When the pre-trained road surface recognition model needs to be further optimized, the road surface image in the sample set is input into the road surface recognition model, and the road surface image in the sample set is processed by the road surface recognition model to obtain the predicted category of the road surface image, and the training loss value is determined based on the predicted category and the real category of the road surface image. The parameters of the pre-trained road surface recognition model are optimized based on the training loss value until the model convergence condition is reached to obtain the optimized road surface recognition model.

[0092] In the above embodiment, the terminal adds the road surface image to the sample set and optimizes the pre-trained model after each recognition, so that the road surface recognition system can have the ability of self-learning and self-optimization, thereby improving the adaptability, accuracy, robustness and processing ability of the model for complex working conditions. This method greatly enhances the driving safety and stability of the vehicle under various road conditions, and enables the autonomous driving or assisted driving system to continue to provide accurate control decisions in a dynamically changing environment.

[0093] In one embodiment, the terminal fuses the first road adhesion coefficient and the second road adhesion coefficient to obtain a fused road adhesion coefficient, and the process includes the following steps: obtaining a first gain corresponding to the first road adhesion coefficient and a second gain corresponding to the second road adhesion coefficient; determining a fusion coefficient based on the first road adhesion coefficient, the first gain, the second road adhesion coefficient and the second gain; determining the difference between the first road adhesion coefficient and the fusion coefficient, and adjusting the first gain and the second gain respectively based on the difference to obtain an adjusted first gain and an adjusted second gain; returning to execute the step of determining the fusion coefficient based on the first road adhesion coefficient, the first gain, the second road adhesion coefficient and the second gain until the convergence condition is met, and determining the fusion coefficient at the time of convergence as the fused road adhesion coefficient.

[0094] The convergence condition refers to a stopping criterion that determines when the fusion process ends, that is, when to stop adjusting the gain and determine the final fusion road adhesion coefficient.

[0095] Specifically, the terminal device can obtain the initialized first gain and the initialized second gain, and perform initial fusion on the first road adhesion coefficient and the second road adhesion coefficient based on the initialized first gain and the initialized second gain to obtain an initial fusion coefficient, and determine the difference between the first road adhesion coefficient and the determined fusion coefficient, and compare the absolute value of the difference with a preset error range. If the absolute value of the difference is greater than the maximum boundary value of the preset range, the initial first gain will be increased and the initial second gain will be reduced to obtain an adjusted first gain and an adjusted second gain. If the absolute value of the difference is less than the minimum boundary value of the preset range, the initial first gain will be reduced. The first gain is calculated, and the initial second gain is increased to obtain an adjusted first gain and an adjusted second gain, and the adjusted first gain and the adjusted second gain are obtained; then the first road adhesion coefficient and the second road adhesion coefficient are re-fused based on the adjusted first gain and the adjusted second gain to obtain a new fusion coefficient, and the difference between the first road adhesion coefficient and the fusion coefficient is re-determined according to the new fusion coefficient, and so on, until the absolute value of the difference is within the preset range, and the adjusted first gain and the second gain obtained at this time are determined as the final gains, and the fusion coefficient obtained at this time is determined as the fused road adhesion coefficient.

[0096] In one embodiment, the terminal may implement the above process of dynamically adjusting the gain through the following fusion estimation function:

[0097]

[0098] in, and denote the first gain and the second gain respectively, and represent the first road adhesion coefficient and the second road adhesion coefficient respectively, is a fusion coefficient obtained by fusing the first road adhesion coefficient and the second road adhesion coefficient based on the first gain and the second gain, yes The derivative of represents the rate of change of the fusion coefficient, which can specifically refer to the magnitude and direction of change of the fusion coefficient. This rate of change is adjusted by combining different estimation methods (such as visual information and dynamic information) and corresponding gains, so that the change of the road adhesion coefficient can be estimated more accurately. Indicates that the fusion coefficient is adjusted according to the first road adhesion coefficient. Indicates that the fusion coefficient is adjusted according to the second road adhesion coefficient.

[0099] It is understandable that in dynamically adjusting the first gain and the second gain In the process, the fusion coefficient also changes accordingly. The change rate of the fusion coefficient is affected by the contribution rate of the first road adhesion coefficient and the contribution rate of the second road adhesion coefficient, that is, the change rate of the fusion coefficient is affected by the first gain and the second gain. When the absolute value of the difference between the first road adhesion coefficient and the fusion coefficient is greater than the maximum boundary value of the preset range, it means that the contribution of the first road adhesion coefficient in the fusion coefficient is too small. Therefore, it is necessary to increase the contribution of the first road adhesion coefficient in the fusion coefficient, that is, increase the first gain and reduce the second gain, so that the fusion coefficient converges quickly to the first road adhesion coefficient. At this time, it can be approximately considered that ,at this time The first road adhesion coefficient The direction of convergence is fast, that is, the fusion coefficient is close to , considering the first road adhesion coefficient Adhesion coefficient with real road surface The error between is a smaller value. When it is approximately 0, that is, when the first road adhesion coefficient is approximated to the true value, it can be considered that at this time Adhesion coefficient along the first road surface The direction converges quickly and can converge to the actual road adhesion coefficient ; When the absolute value of the difference between the first road adhesion coefficient and the fusion coefficient is less than the minimum boundary value of the preset range, it means that the contribution of the first road adhesion coefficient to the fusion coefficient is too large. Therefore, it is necessary to reduce the contribution of the first road adhesion coefficient to the fusion coefficient, that is, reduce the first gain and increase the second gain, so that the fusion coefficient converges quickly to the second road adhesion coefficient. At this time, it can be approximately considered that ,at this time The first road adhesion coefficient The direction converges rapidly until the convergence condition is met.

[0100] In the above embodiment, the terminal integrates the road adhesion coefficients from different sources and optimizes the control parameters by dynamically adjusting the gains, so that the vehicle can adjust the driving strategy in real time and accurately under different road conditions. Ultimately, it can improve the accuracy, adaptability and robustness of the vehicle control system, ensuring a more stable, safe and efficient driving experience in various complex driving environments.

[0101] In one embodiment, the vehicle control parameters include wheel torque; the process of determining the vehicle control parameters based on the fused road adhesion coefficient by the terminal includes the following steps: determining multiple expected control targets of the target vehicle based on the fused road adhesion coefficient and vehicle dynamics information; determining the weights corresponding to each expected control target; determining the current optimal expected longitudinal acceleration based on the weights and the expected control targets; and determining the wheel torque based on the optimal expected longitudinal acceleration.

[0102] Among them, the multiple desired control targets include a following distance deviation target, a desired speed deviation target, a desired acceleration deviation target, and a desired jerk deviation target. The following distance deviation target represents the difference between the actual following distance between the current target vehicle and the vehicle ahead and the desired following distance. The desired speed deviation target represents the difference between the actual speed of the current target vehicle and the desired speed. The desired acceleration deviation target represents the difference between the actual acceleration of the target vehicle and the desired acceleration. The desired jerk deviation target represents the difference between the actual jerk (i.e., the rate of change of acceleration) of the current vehicle and the desired jerk. It can be understood that each desired control target can be represented by a corresponding sub-objective function, which can also be called a cost function. The four sub-objective functions corresponding to the above four desired control targets are combined to obtain a total objective function. In order to adjust the priority of different control targets according to their importance, different weights can be assigned to each objective function. It should be noted that the weight assignment is also different under different working conditions. Working conditions refer to various external factors (such as road surface type, weather conditions, traffic conditions, etc.) that affect vehicle driving performance and control requirements, as well as internal factors such as driving mode and driver requirements. Based on these operating conditions, the vehicle control system can dynamically adjust the weights of different objectives to achieve the best driving performance. By reasonably allocating the weights of the objective function under different operating conditions, the vehicle can optimize the control strategy according to real-time conditions to improve safety, stability and comfort.

[0103] Specifically, the terminal can determine the actual following distance, actual speed, actual acceleration, actual jerk of the current target vehicle and the actual speed of the preceding vehicle based on the vehicle dynamics information, and determine the expected following distance based on the actual speed of the target vehicle and the actual speed of the preceding vehicle, and determine the sub-objective functions corresponding to the following distance deviation target, the expected speed deviation target, the expected acceleration deviation target and the expected jerk deviation target respectively based on the actual following distance, the actual speed, the actual acceleration, the actual jerk and the actual speed of the preceding vehicle, and obtain the weight distribution scheme that best matches the current working condition, obtain the weight corresponding to each sub-objective function, and calculate the weight of each sub-objective function. The total objective function is constructed by the number and the weight corresponding to each sub-objective function, and model predictive control is performed based on the workshop kinematic prediction model between the target vehicle and the front vehicle. The goal of model predictive control is to minimize the value of the total objective function, which can be close to 0, so as to obtain the optimal solution. The optimal solution is the optimal expected longitudinal acceleration under the current working condition, and the optimal expected longitudinal acceleration is converted into the expected total longitudinal torque of the target vehicle according to the feedforward control of the inverse longitudinal dynamics model, and the output expected longitudinal torque is feedback adjusted by the PID controller, and finally the total longitudinal torque is converted into the wheel torque of the four wheels of the target vehicle through torque distribution.

[0104] The process of optimizing the overall objective function can be optimized based on constraints, which include at least one of vehicle dynamic constraints (such as vehicle speed and acceleration limits), physical constraints (such as tire adhesion limits), and traffic safety-related constraints. The method of obtaining the weight distribution scheme that best matches the current working condition can be, first, using a genetic algorithm to iteratively optimize the weight distribution schemes under various working conditions offline, and then selecting the weight distribution scheme that best matches the current working condition from the weight distribution schemes under various working conditions. The current working condition can be determined specifically based on the motion states of the target vehicle and the preceding vehicle.

[0105] In the above embodiment, the terminal can effectively improve the control accuracy, safety and adaptability of the vehicle through dynamic adjustment based on the fusion of road adhesion coefficient, vehicle dynamics information and desired control objectives. The optimized control strategy ensures that the vehicle can provide stable driving performance under different road conditions, while improving the driving experience, reducing human intervention, and enhancing the intelligence level of autonomous driving and assisted driving.

[0106] The present application also provides an application scenario, which applies the above-mentioned vehicle control method. The vehicle control method can be specifically implemented through a hierarchical control structure, which includes a control target calculation module, an upper controller and a lower control module. Figure 3 and Figure 4 , the control target calculation module is used to calculate the expected difference, which is implemented by the following steps:

[0107] (1) The main vehicle sensor system is used to detect the actual distance between the two vehicles and the relative speed between the two vehicles. It combines the inertial measurement unit (IMU) sensor on the main vehicle to obtain dynamic information such as the main vehicle's actual speed, actual acceleration, and actual jerk, and collects road images through the front and rear cameras.

[0108] (2) The vehicle speed, yaw rate, longitudinal acceleration, lateral acceleration and other dynamic information are input into the road adhesion coefficient estimator, and the recognition result of the road surface image is also input into the road adhesion coefficient estimator. The road adhesion coefficient estimator estimates the road adhesion coefficient in real time by combining these two parts of information, and calculates the expected longitudinal safety distance by combining the current main vehicle speed. The distance difference is calculated based on the corresponding actual distance at this time. Similarly, the speed difference, acceleration difference and jerk difference are calculated based on the expected value and the actual detection value.

[0109] The upper controller is used to plan the expected longitudinal acceleration with multi-objective characteristics based on various difference targets. The process specifically includes the following steps:

[0110] (3) Establish a model predictive control based on the relative motion of the two vehicles. The goal of the model predictive control is to make the deviation in step (2) approach 0. However, since different goals will restrict each other, it is necessary to prioritize the weights of the goals according to the motion status of the two vehicles. Therefore, the genetic algorithm is first used to iteratively optimize the optimal weights under various working conditions offline. Then, the optimal weights under different working conditions are switched dynamically based on the online real-time status of the two vehicles to ensure the expected longitudinal acceleration under various working conditions.

[0111] The lower control module is used to control the vehicle based on the desired longitudinal acceleration. The process specifically includes the following steps:

[0112] (4) The desired longitudinal acceleration is input into the lower-level controller and converted into the desired total longitudinal torque of the main vehicle through the feedforward control of the inverse longitudinal dynamics model. The output desired longitudinal torque is feedback-adjusted using the PID controller. Finally, the total longitudinal torque is converted into the desired torque on the four wheels through torque distribution, forming a closed-loop control.

[0113] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0114] Based on the same inventive concept, the embodiment of the present application also provides a vehicle control device for implementing the vehicle control method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more vehicle control device embodiments provided below can refer to the limitations on the vehicle control method above, and will not be repeated here.

[0115] In an exemplary embodiment, Figure 5 As shown, a vehicle control device is provided, including: an information acquisition module 502, an image processing module 504, a dynamic information processing module 506, a coefficient fusion module 508 and a control parameter determination module 510, wherein:

[0116] The information acquisition module 502 is used to acquire road surface images and vehicle dynamics information of the target vehicle when the target vehicle is in an assisted driving state;

[0117] An image processing module 504, configured to determine a first road surface adhesion coefficient based on the road surface image;

[0118] A dynamic information processing module 506, configured to determine a second road adhesion coefficient based on the vehicle dynamic information;

[0119] A coefficient fusion module 508, used for fusing the first road adhesion coefficient and the second road adhesion coefficient to obtain a fused road adhesion coefficient;

[0120] The control parameter determination module 510 is used to determine the vehicle control parameters based on the fused road adhesion coefficient, and to perform driving control on the target vehicle based on the vehicle control parameters.

[0121] In the above embodiment, when the target vehicle is in the assisted driving state, the target vehicle collects road surface images and vehicle dynamics information of the target vehicle, determines a first road surface adhesion coefficient based on the road surface images, and determines a second road surface adhesion coefficient based on the vehicle dynamics information, so that the road surface adhesion coefficient of the current road surface can be estimated from both visual and dynamic aspects, and the road surface adhesion coefficients estimated from these two aspects are fused to obtain an accurate and reliable fused road surface adhesion coefficient. The vehicle control parameters of the target vehicle are dynamically adjusted according to the fused road surface adhesion coefficient, so that the target vehicle is driven and controlled based on the vehicle control parameters, so that the vehicle can optimize the control strategy in real time under different road conditions to ensure driving safety and stability.

[0122] In one embodiment, the image processing module 504 is further used to: identify the road surface image based on a pre-trained road surface recognition model to obtain the road surface category of the road surface image; and determine the first road surface adhesion coefficient based on the road surface category.

[0123] In one embodiment, the image processing module 504 is also used to: add the recognized road surface image to the sample set after each recognition; process the road surface image in the sample set by using a pre-trained road surface recognition model to obtain a predicted category of the road surface image; and optimize the parameters of the pre-trained road surface recognition model based on the predicted category and the true category of the road surface image to obtain an optimized road surface recognition model.

[0124] In one embodiment, the dynamic information processing module 506 is also used to: determine the longitudinal force and lateral force of the wheels of the target vehicle based on the vehicle dynamics information; determine the second road adhesion coefficient based on the target tire model, the longitudinal force and lateral force of the wheels; wherein the target tire model is a tire model obtained by improving the Dugoff model based on a scaling factor.

[0125] In one embodiment, the coefficient fusion module 508 is also used to: obtain a first gain corresponding to the first road adhesion coefficient and a second gain corresponding to the second road adhesion coefficient; determine a fusion coefficient based on the first road adhesion coefficient, the first gain, the second road adhesion coefficient and the second gain; determine the difference between the first road adhesion coefficient and the fusion coefficient, and adjust the first gain and the second gain respectively based on the difference to obtain an adjusted first gain and an adjusted second gain; return to the step of determining the fusion coefficient based on the first road adhesion coefficient, the first gain, the second road adhesion coefficient and the second gain until the convergence condition is met, and determine the fusion coefficient at the time of convergence as the fused road adhesion coefficient.

[0126] In one embodiment, the vehicle control parameters include wheel torque; the control parameter determination module 510 is also used to: determine multiple expected control targets of the target vehicle based on the fusion of road adhesion coefficient and vehicle dynamics information; the multiple expected control targets include a following distance deviation target, an expected speed deviation target, an expected acceleration deviation target and an expected jerk deviation target; determine the weights corresponding to each expected control target; determine the current optimal expected longitudinal acceleration based on the weight and the expected control target; determine the wheel torque based on the optimal expected longitudinal acceleration.

[0127] Each module in the above vehicle control device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.

[0128] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 6 As shown. The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, near field communication (Near Field Communication, NFC) or other technologies. When the computer program is executed by the processor, a terminal control method is implemented. The display unit of the computer device is used to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse.

[0129] Those skilled in the art will understand that Figure 6The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0130] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0131] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0132] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0133] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0134] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, an artificial intelligence (AI) processor, etc., but are not limited to this.

[0135] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0136] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be construed as limiting the scope of the present application. It should be noted that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A vehicle control method, characterized in that: The method comprises: When the target vehicle is in the assisted driving state, collecting road images and vehicle dynamics information of the target vehicle; determining a first road surface adhesion coefficient based on the road surface image; determining a second road adhesion coefficient based on the vehicle dynamics information; Obtaining a first gain corresponding to the first road adhesion coefficient and a second gain corresponding to the second road adhesion coefficient; the first gain is used to characterize the importance of the first road adhesion coefficient during fusion, and the second gain is used to characterize the importance of the second road adhesion coefficient during fusion; determining a fusion coefficient based on the first road adhesion coefficient, the first gain, the second road adhesion coefficient, and the second gain; determining a difference between the first road adhesion coefficient and the fusion coefficient, and adjusting the first gain and the second gain based on the difference to obtain an adjusted first gain and an adjusted second gain; Returning to the step of determining a fusion coefficient based on the first road surface adhesion coefficient, the first gain, the second road surface adhesion coefficient, and the second gain until a convergence condition is satisfied, and determining the fusion coefficient at the time of convergence as a fusion road surface adhesion coefficient; the convergence condition includes that an absolute value of a difference between the first road surface adhesion coefficient and the fusion coefficient is within a preset range; A vehicle control parameter is determined based on the fused road adhesion coefficient, and driving control is performed on the target vehicle based on the vehicle control parameter.

2. The method according to claim 1, characterized in that The determining a first road surface adhesion coefficient based on the road surface image includes: Recognize the road surface image based on a pre-trained road surface recognition model to obtain a road surface category of the road surface image; A first road surface adhesion coefficient is determined based on the road surface category.

3. The method according to claim 2, characterized in that The method further comprises: After each recognition, the recognized road surface image is added to a sample set; Processing the road surface image in the sample set by using the pre-trained road surface recognition model to obtain a predicted category of the road surface image; The parameters of the pre-trained road surface recognition model are optimized based on the predicted category and the real category of the road surface image to obtain an optimized road surface recognition model.

4. The method according to claim 1, characterized in that: The determining a second road adhesion coefficient based on the vehicle dynamics information includes: determining longitudinal force and lateral force of wheels of the target vehicle based on the vehicle dynamics information; determining a second road adhesion coefficient based on a target tire model, a longitudinal force and a lateral force of the wheel; The target tire model is a tire model obtained by improving the Dugoff model based on the scaling factor.

5. The method according to claim 1, characterized in that The vehicle control parameter includes wheel torque; the determining of the vehicle control parameter based on the fused road adhesion coefficient includes: Determining multiple desired control targets of the target vehicle based on the fused road adhesion coefficient and the vehicle dynamics information; the multiple desired control targets include a following distance deviation target, an expected speed deviation target, an expected acceleration deviation target, and an expected jerk deviation target; the expected jerk deviation target represents the difference between the actual jerk of the target vehicle and the expected jerk, wherein jerk refers to the rate of change of acceleration; Determine the weights corresponding to the desired control objectives respectively; Determining a current optimal expected longitudinal acceleration based on the weight and the expected control target; A wheel torque is determined based on the optimal desired longitudinal acceleration.

6. A vehicle control device, characterized in that: The device comprises: An information acquisition module, used to acquire road surface images and vehicle dynamics information of the target vehicle when the target vehicle is in an assisted driving state; An image processing module, configured to determine a first road surface adhesion coefficient based on the road surface image; a dynamic information processing module, configured to determine a second road adhesion coefficient based on the vehicle dynamic information; A coefficient fusion module, used for obtaining a first gain corresponding to the first road adhesion coefficient and a second gain corresponding to the second road adhesion coefficient; the first gain is used to characterize the importance of the first road adhesion coefficient during fusion, and the second gain is used to characterize the importance of the second road adhesion coefficient during fusion; determining a fusion coefficient based on the first road adhesion coefficient, the first gain, the second road adhesion coefficient and the second gain; determining a difference between the first road adhesion coefficient and the fusion coefficient, and adjusting the first gain and the second gain respectively based on the difference to obtain an adjusted first gain and an adjusted second gain; returning to execute the step of determining a fusion coefficient based on the first road adhesion coefficient, the first gain, the second road adhesion coefficient and the second gain until a convergence condition is met, and determining the fusion coefficient at the time of convergence as the fused road adhesion coefficient; the convergence condition includes that the absolute value of the difference between the first road adhesion coefficient and the fusion coefficient is within a preset range; A control parameter determination module is used to determine a vehicle control parameter based on the fused road adhesion coefficient, and to perform driving control on the target vehicle based on the vehicle control parameter.

7. The device according to claim 6, characterized in that The image processing module is further used for: Recognize the road surface image based on a pre-trained road surface recognition model to obtain a road surface category of the road surface image; A first road surface adhesion coefficient is determined based on the road surface category.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images