Vehicle control method and device, computer readable storage medium and vehicle

Abstract

The application discloses a vehicle control method and device, a computer readable storage medium and a vehicle, wherein the vehicle control method comprises the following steps: acquiring a visual image of a road section around the vehicle and an expected driving track; predicting a road surface terrain corresponding to the driving track of the vehicle according to the visual image and the expected driving track; and adjusting a control parameter of the vehicle according to the road surface terrain corresponding to the driving track of the vehicle. The method and device can realize automatic identification of the road surface terrain, and the terrain of the driving track of the vehicle is predicted to adjust the control parameter of the vehicle in advance, so that the adaptability and safety of the vehicle under different road surface conditions are improved.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and in particular to a vehicle control method, a vehicle control device, a computer-readable storage medium, and a vehicle. Background Technology

[0002] In related technologies, there are two main methods for recognizing terrain changes during vehicle operation: terrain recognition using the suspension system and terrain recognition using cameras. The camera-based approach typically relies on image information of the road surface the vehicle is currently traveling on. Image processing algorithms are used to identify the current road terrain and subsequently control the vehicle. However, this approach cannot predict road conditions along the vehicle's path or adjust vehicle parameters in advance to adapt to the terrain. Summary of the Invention

[0003] The present invention aims to at least solve one of the technical problems existing in the prior art. Therefore, one object of the present invention is to provide a vehicle control method that can automatically identify road surface terrain and predict the terrain of the vehicle's driving path to adjust vehicle control parameters in advance, thereby improving the vehicle's adaptability and safety under different road surface conditions.

[0004] The second objective of this invention is to provide a vehicle control device.

[0005] A third objective of this invention is to provide a computer-readable storage medium.

[0006] The fourth objective of this invention is to provide a vehicle.

[0007] To achieve the above objectives, a vehicle control method according to a first aspect of the present invention includes: acquiring visual images of road segments surrounding the vehicle and a predicted driving trajectory; predicting the road surface terrain corresponding to the vehicle's driving trajectory based on the visual images and the predicted driving trajectory; and adjusting the vehicle's control parameters based on the road surface terrain corresponding to the vehicle's driving trajectory.

[0008] According to the vehicle control method of the present invention, by acquiring the visual image of the vehicle and the expected driving trajectory, the road surface terrain corresponding to the current driving trajectory can be automatically identified. This method of comprehensively utilizing the visual image and the expected driving trajectory can avoid the shortcomings of relying on a single sensor for terrain recognition, and can predict the road surface terrain of the vehicle's driving path. Based on the road surface terrain corresponding to the vehicle's driving trajectory, the vehicle's control parameters can be adjusted in advance, so that the vehicle can better adapt to various road surface changes and maintain a stable driving state, thereby improving the vehicle's adaptability and safety under different road conditions.

[0009] In some embodiments, predicting the road surface terrain corresponding to the vehicle's driving trajectory based on the expected driving trajectory and the visual image includes: obtaining the area covered by the expected driving trajectory in the visual image to obtain an image containing the driving trajectory; extracting terrain features from the image containing the driving trajectory; and solving for the terrain data of the road surface at each wheel position along the driving trajectory based on the extracted terrain features.

[0010] In some embodiments, the vehicle control method further includes: determining the confidence level of the terrain data based on the data integrity and identification results of the terrain data.

[0011] In some embodiments, the vehicle control method further includes: acquiring actual road surface terrain data of the current driving road surface of the vehicle; comparing the actual road surface terrain data with road surface terrain data identified in the corresponding driving area; if the deviation between the actual road surface terrain data and the road surface terrain data identified in the corresponding driving area exceeds a preset threshold, marking the identified road surface terrain data as underfitting data; and sending the underfitting data to the cloud to optimize the algorithm for identifying the road surface terrain based on the underfitting data.

[0012] In some embodiments, the vehicle control method further includes: obtaining an algorithm optimized based on the underfitting data for identifying the road surface terrain, so as to perform road surface terrain identification.

[0013] In some embodiments, adjusting the vehicle's control parameters according to the road surface terrain corresponding to the vehicle's driving trajectory includes: pre-controlling at least one of the vehicle's driving force, braking force, steering angle, and suspension system parameters according to the road surface terrain corresponding to the vehicle's driving trajectory and the vehicle's actual driving state.

[0014] In some embodiments, the vehicle control method further includes: adjusting the control intensity of at least one of the vehicle's driving force, braking force, steering angle, and suspension system parameters based on the confidence level of the terrain data.

[0015] In some embodiments, the vehicle control method further includes: acquiring actual road surface terrain data of the current road surface of the vehicle; and performing closed-loop control on at least one of the vehicle's drive system, braking system, steering system, and suspension based on the actual road surface terrain data.

[0016] In some embodiments, the vehicle control method further includes: obtaining feedforward control data based on the road surface terrain corresponding to the vehicle's driving trajectory; acquiring actual road surface terrain data of the road surface where the vehicle is currently driving, and obtaining feedback control data based on the actual road surface terrain data; obtaining ideal target control data based on the feedforward control data and the feedback control data; and controlling the vehicle based on the ideal target control data.

[0017] In some embodiments, obtaining actual road surface terrain data of the current driving surface of the vehicle includes: obtaining vehicle status data and wheel status data; obtaining the actual vehicle body tilt angle and the absolute height of each wheel based on the vehicle status data and the wheel status data; and obtaining the actual road surface terrain data based on the actual vehicle body tilt angle and the absolute height of each wheel.

[0018] To achieve the above objectives, a second aspect of the present invention provides a vehicle control device, comprising: an acquisition module for acquiring visual images of road segments surrounding the vehicle and a predicted driving trajectory; a prediction module for predicting the road surface terrain corresponding to the vehicle's driving trajectory based on the visual images and the predicted driving trajectory; and a control module for adjusting the vehicle's control parameters based on the road surface terrain corresponding to the vehicle's driving trajectory.

[0019] According to the vehicle control device of the present invention, the acquisition module acquires the visual image of the vehicle and the expected driving trajectory, and the prediction module can automatically identify the road surface terrain corresponding to the current driving trajectory. This method of comprehensively utilizing the visual image and the expected driving trajectory can avoid the shortcomings of relying on a single sensor for terrain recognition, and can predict the road surface terrain of the vehicle's driving path. Based on the road surface terrain corresponding to the vehicle's driving trajectory, the vehicle's control parameters can be adjusted in advance, so that the vehicle can better adapt to various road surface changes and maintain a stable driving state, thereby improving the vehicle's adaptability and safety under different road surface conditions.

[0020] To achieve the above objectives, a vehicle control device according to a third aspect of the present invention includes: at least one processor; a memory communicatively connected to the at least one processor; the memory storing a computer program, wherein the at least one processor executes the computer program to implement the vehicle control method described in the above embodiment.

[0021] According to the vehicle control device of the present invention, the vehicle control method described in the above embodiment is implemented by the processor executing a computer program. By comprehensively utilizing visual images and expected driving trajectories, the device achieves the function of automatically identifying road surface terrain and adjusting vehicle control parameters. This comprehensive terrain recognition and parameter adjustment capability enables the vehicle to adapt more accurately to changes in different road surface terrain, thereby improving the vehicle's adaptability and driving safety under different road conditions.

[0022] To achieve the above objectives, a computer-readable storage medium according to a fourth aspect of the present invention stores a computer program thereon, which, when executed, implements the vehicle control method described in the above embodiments.

[0023] According to the computer-readable storage medium of the present invention, by executing the vehicle control method described in the above embodiments, the vehicle's visual image and expected driving trajectory are comprehensively utilized to automatically identify the road surface terrain corresponding to the vehicle's driving trajectory. This method of comprehensively utilizing visual images and expected driving trajectories can avoid the shortcomings of relying solely on a single sensor for terrain recognition, and can predict the road surface terrain of the vehicle's driving path. Based on the road surface terrain corresponding to the vehicle's driving trajectory, the vehicle's control parameters can be adjusted in advance, thereby enabling the vehicle to better adapt to various road surface changes and maintain a stable driving state, improving the vehicle's adaptability and safety under different road surface conditions.

[0024] To achieve the above objectives, a vehicle according to a fifth aspect embodiment of the present invention includes: a vision system for acquiring visual images of the vehicle; and a vehicle control device as described in the above embodiment, wherein the vehicle control device is connected to the vision system.

[0025] According to the vehicle of the present invention, the vehicle control device analyzes and processes the image acquired from the vision system and the expected driving trajectory by adopting the vehicle control method described in the above embodiment. It can automatically identify the road surface terrain corresponding to the current driving trajectory of the vehicle, and automatically adjust the control parameters of the vehicle according to these identification results, so that the vehicle can better adapt to various road surface terrain changes and maintain a stable driving state, thereby improving the adaptability and safety of the vehicle under different road surface conditions.

[0026] In some embodiments, the vehicle further includes a communication device connected to the vehicle control device, for transmitting underfit data and receiving an algorithm for identifying road surface terrain optimized based on the underfit data.

[0027] In some embodiments, the vehicle further includes a wheel height sensor, a wheel speed sensor, and a vehicle body inertial navigation sensor.

[0028] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0029] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0030] Figure 1 This is a schematic diagram of a vehicle data interaction system according to an embodiment of the present invention;

[0031] Figure 2 This is a flowchart of a vehicle control method according to an embodiment of the present invention;

[0032] Figure 3 This is a functional architecture diagram of an automatic terrain control system according to an embodiment of the present invention;

[0033] Figure 4 This is a working scenario diagram of a feature recognition and classification module according to an embodiment of the present invention;

[0034] Figure 5 This is a graph showing the relationship between the longitudinal displacement of the left front wheel and the vertical height of the wheel from the horizontal ground, according to an embodiment of the present invention.

[0035] Figure 6 This is a schematic diagram of the principle of a four-wheel terrain analysis module according to an embodiment of the present invention;

[0036] Figure 7 This is a block diagram of a vehicle control device according to an embodiment of the present invention;

[0037] Figure 8 This is a block diagram of a vehicle control device according to yet another embodiment of the present invention.

[0038] Figure label:

[0039] Terrain automatic control system 100; Vehicle data interaction system 200;

[0040] Vehicle 1; Communication network 2; Cloud server 3;

[0041] Vehicle control device 10; vision system 20; communication device 30; wheel height sensor 40; wheel speed sensor 50; vehicle body inertial navigation sensor 60;

[0042] Automatic terrain recognition module 110; terrain analysis and control module 120; feedback control module 130; four-wheel terrain analysis module 140; acquisition module 11; prediction module 12; control module 13;

[0043] Driving trajectory prediction module 111; visual processing module 112; online data update module 113; feature recognition and classification module 114; feedforward control module 121; ideal target calculation module 122;

[0044] The current ground position of the vehicle is 141; the coordinate system corresponding to the current ground position of the vehicle is 142; the horizontal plane is 143; the coordinate system corresponding to the horizontal plane is 144; the vehicle body is shown 145; the vehicle's center of gravity is 146; the coordinate system corresponding to the plane on which the vehicle chassis is located is 147; the wheels are 148.

[0045] Processor 101; Memory 102. Detailed Implementation

[0046] The embodiments of the present invention are described in detail below. The embodiments described with reference to the accompanying drawings are exemplary. The embodiments of the present invention are described in detail below.

[0047] In related technologies, relying solely on the suspension system for terrain perception or using only visual sensors such as cameras to identify terrain may have limitations in accuracy and comprehensiveness, resulting in vehicles being unable to fully adapt to complex road surface changes. To address this issue, this invention proposes a vehicle control method that provides more comprehensive and accurate terrain recognition capabilities, achieving automatic road surface identification. This allows vehicles to better adapt to different road surface conditions, improving their adaptability and safety under varying road conditions.

[0048] To facilitate the explanation of the technical solution, the vehicle of the embodiment of the present invention will be described first below.

[0049] Figure 1 This is a schematic diagram of the structure of a vehicle data interaction system according to an embodiment of the present invention, as shown below. Figure 1 As shown, the vehicle data interaction system 200 includes: a communication network 2, a cloud server 3, and a vehicle 1.

[0050] The communication network 2 can be composed of various network infrastructures, such as base stations, satellites, and optical cables. This communication network 2 can transmit data with the communication device 30 equipped in the vehicle 1 or the cloud server 3 based on a specific communication protocol.

[0051] Cloud server 3 is a powerful computing resource located in a cloud computing center. It can store all data uploaded by vehicles 1 equipped with the terrain automatic control system and use it to train and iterate visual algorithms. Through cloud server 3, the visual recognition algorithm trained with the latest big data can be updated and optimized online, and the latest version of the software data can be distributed to vehicles 1 through communication network 2.

[0052] Vehicle 1 can refer to a vehicle equipped with an automatic terrain control system. Vehicle 1 can connect to communication network 2 via communication device 30 and ultimately establish communication with cloud server 3.

[0053] In some embodiments, vehicle 1 includes: vision system 20, vehicle control device 10, communication device 30, wheel height sensor 40, wheel speed sensor 50, and vehicle body inertial navigation sensor 60.

[0054] The vision system 20 can be used to acquire visual images of the vehicle 1. In this embodiment, the vision system 20 may include two or more cameras to capture terrain information of the road segment surrounding the vehicle 1. These cameras provide images from different perspectives, enabling the acquisition of multiple visual angles of the same terrain area. Furthermore, the vision system 20 may also include other detection methods, such as lidar or ultrasonic radar, to provide additional physical information. This physical information can assist in calculations during subsequent image processing by the vehicle control device 10 to provide more comprehensive terrain recognition capabilities.

[0055] In some embodiments, the vehicle control device 10 is connected to the vision system 20. By using image information acquired by the vision system 20 and the expected driving trajectory, the road surface terrain corresponding to the vehicle's driving trajectory is identified, and the control parameters of the vehicle 1 are adjusted according to the road surface terrain corresponding to the vehicle's driving trajectory. This enables the vehicle 1 to adapt to different road surface changes, improving its adaptability and safety under different road conditions.

[0056] In some embodiments, the vehicle control device 10 includes a controller, which is the core computing and processing unit of the terrain automatic control system 100. The controller can process data from visual signals and sensor signals, and automatically determine the terrain conditions corresponding to the current terrain and the path ahead based on this data. Finally, the controller sends requests to systems such as drive, braking, steering, and suspension based on these determinations, enabling the vehicle to automatically operate in a control mode adapted to the current terrain. The controller can exchange data with the communication device 30. The controller may internally include any variety of electronic processing devices, storage devices, input / output (I / O) devices, and / or other known components.

[0057] In some embodiments, the controller may include multiple internal modules, such as an internal μC module, an internal storage module, an internal communication module, and an internal sensor module. These internal modules cooperate with each other to enable the controller to process and analyze data, achieve terrain recognition and vehicle control, thereby enabling vehicle 1 to adapt to different road surface terrain changes.

[0058] The internal μC module can include electronic processing devices such as a microprocessor 101, a microcontroller, and an application-specific integrated circuit (ASIC), responsible for executing instructions such as software, firmware, programs, algorithms, and scripts stored in the internal storage module. This module is the core of the controller, enabling it to process and analyze the acquired data.

[0059] The internal storage module can be used to store sensor readings, visual images, lookup tables, or other data structures and algorithms. This internal storage module provides the controller with the ability to store and retrieve data for terrain recognition and control parameter adjustment.

[0060] The internal communication module can connect to other vehicle devices, modules, and systems, and exchange data when needed. This module enables the controller to coordinate and exchange information with other vehicle components, achieving a higher level of functionality.

[0061] The internal sensor module can be a sensor integrated within the controller, or it can include other sensing circuits such as power supply voltage and current monitoring. The internal sensor module is used to collect various parameters and status information inside vehicle 1, so that the controller can make corresponding decisions and adjustments.

[0062] In some embodiments, the communication device 30 is connected to the vehicle control device 10, enabling information transmission between the vehicle data and the communication network 2. The communication device 30 has various communication interfaces, such as CAN, Ethernet, WIFI, 4G / 5G, etc. The communication device 30 obtains updated data from the cloud server 3 through the communication network 2. This data may include map information, new software versions, etc., for updating the software modules inside the vehicle 1. The communication device 30 is responsible for transmitting this data to the internal system of the vehicle 1 for corresponding iterations and updates.

[0063] Furthermore, the communication device 30 is also used to send underfit data and receive an algorithm for recognizing road terrain optimized based on the underfit data. Specifically, when the vehicle control device 10 detects a significant discrepancy between visual and sensor data, the communication device 30 uploads the corresponding underfit data to the cloud server 3. This data can be used for algorithm optimization training in the cloud to improve the terrain recognition function of the vehicle 1.

[0064] In some embodiments, wheel height sensors 40 are used to sense the relative height between each wheel 148 and the vehicle body. Generally, there are at least four sensors, which are typically arranged on the suspension between the wheels 148 and the vehicle body. The controller can acquire signals from these sensors to better understand the attitude and suspension status of the vehicle 1, thereby making corresponding adjustments and controls.

[0065] In some embodiments, wheel speed sensors 50 are used to monitor the rotational speed of each wheel 148. Generally, there are at least four sensors, typically arranged on the wheel hub. The controller can acquire signals from these sensors to monitor the rotational speed of the wheels 148 and the motion state of the vehicle 1 in real time, providing the system with necessary information.

[0066] In some embodiments, the vehicle inertial navigation sensor 60 is used to detect the vehicle's motion information in various directions, including six dimensions: longitudinal acceleration, lateral acceleration, vertical acceleration, yaw rate, pitch rate, and roll rate. These sensors can directly obtain pitch and roll angle information, which can be used to simplify the algorithms of some modules, provide more accurate vehicle attitude information, and reference data for the vehicle's motion state.

[0067] The following is for reference. Figures 2-6 A vehicle control method according to an embodiment of the present invention is described.

[0068] Figure 2 This is a flowchart of a vehicle control method according to an embodiment of the present invention, such as... Figure 2 As shown, the vehicle control method includes at least steps S1-S3, as detailed below.

[0069] S1, acquire visual images of the road sections surrounding the vehicle and the expected driving trajectory.

[0070] In some embodiments, a vehicle can continuously capture visual images of its surroundings using an equipped vision system, such as a camera, radar, or other sensing device. These devices can be positioned at different locations on the vehicle, such as the front, rear, left, right, and top, to obtain an all-around field of view. The visual images can include elements such as roads, the surrounding environment, obstacles, traffic signs, and other vehicles. These visual images are presented in the form of digital images and may contain information such as color, brightness, and texture.

[0071] The expected driving trajectory can be a theoretical driving path derived from vehicle navigation systems, map data, and planning algorithms. This path describes the route the vehicle is expected to follow, including actions such as turning, going straight, accelerating, and decelerating. The expected driving trajectory can take into account factors such as traffic rules, destination, and route optimization.

[0072] S2 predicts the road surface terrain corresponding to the vehicle's driving trajectory based on visual images and expected driving trajectories.

[0073] Specifically, vehicles acquire surrounding image information through a vision system. Using image processing techniques and visual recognition algorithms based on large-scale image data, features such as road texture, edges, and colors can be extracted from the images. Simultaneously, the expected driving trajectory provides the vehicle's anticipated path. By comparing and analyzing the expected driving trajectory with road features in the image, the system can predict the road surface terrain corresponding to the vehicle's trajectory. Pre-trained visual recognition algorithms can improve recognition accuracy and efficiency.

[0074] In some embodiments, road surface terrain can be of various types, including but not limited to: paved roads (such as asphalt roads, concrete roads, etc.), unpaved roads (such as dirt roads, gravel roads, etc.), uneven roads (such as bumpy, potholed, and uneven sections), flat roads, and slopes (such as uphill, downhill, and side slopes with different gradients). By analyzing images, the system can identify features such as road surface tilt, texture, and curvature. For example, for unpaved roads, the color and texture of gravel or soil may be visible in the image. For slopes, tilt may be determined by changes in perspective in the image. Furthermore, information about the expected driving trajectory can also help the system infer the terrain conditions the vehicle is about to traverse. For example, if the expected driving trajectory indicates that the vehicle will enter a curve, the system can identify the road surface terrain based on the curve features in the image.

[0075] In some embodiments, the system can also use sensors such as lidar or ultrasonic radar to determine changes in ground height by taking advantage of the intensity and time of reflected signals, thereby determining the undulation of the road surface and predicting different types of road terrain.

[0076] S3 adjusts the vehicle's control parameters based on the road surface terrain corresponding to the vehicle's driving trajectory.

[0077] Specifically, based on road surface information corresponding to the vehicle's driving trajectory, the vehicle control unit automatically adjusts various control parameters of the vehicle in advance, including driving force, braking force, steering angle or assist strength, and suspension stiffness or damping. These adjustments can achieve optimal control for different terrains, providing better vehicle stability, suspension adjustment, and brake distribution. For example, when the system detects uneven road surface or obstacles, it may reduce vehicle speed and increase suspension damping to reduce body roll. If the road surface is flat and open, the system may optimize the vehicle's power distribution to improve driving performance.

[0078] For example, on unpaved roads, the system can reduce suspension stiffness to mitigate vibrations and provide a more comfortable ride. On steep inclines, the system can adjust braking force and power distribution to ensure the vehicle maintains a safe speed when going uphill or downhill.

[0079] In some embodiments, the system may also consider the current driving mode, such as cruise mode or sport mode, to adjust the control parameters. Additionally, the system may incorporate actual vehicle dynamic data, such as vehicle speed and acceleration, to further optimize the parameter adjustments.

[0080] According to the vehicle control method of the present invention, by acquiring the visual image of the vehicle and the expected driving trajectory, the road surface terrain corresponding to the current driving trajectory can be automatically identified. This method of comprehensively utilizing the visual image and the expected driving trajectory can avoid the shortcomings of relying on a single sensor for terrain recognition, and can predict the terrain of the vehicle's driving path. Based on the road surface terrain corresponding to the vehicle's driving trajectory, the vehicle's control parameters can be adjusted in advance, so that the vehicle can better adapt to various road surface changes and maintain a stable driving state, thereby improving the vehicle's adaptability and safety under different road surface conditions.

[0081] In some embodiments, predicting the road surface terrain corresponding to a vehicle's driving trajectory based on a predicted driving trajectory and a visual image includes: obtaining the area covered by the predicted driving trajectory in the visual image to obtain an image containing the driving trajectory; extracting terrain features from the image containing the driving trajectory; and solving for the terrain data of the road surface at each wheel position along the driving trajectory based on the extracted terrain features.

[0082] Specifically, based on the vehicle's expected travel trajectory and its current location, the system can determine the area covered by the expected trajectory in the visual image. Then, from the visual images acquired from the vehicle, images covering the area of ​​the expected travel trajectory are selected. This step ensures that the selected images include the road segment the vehicle is about to travel on.

[0083] Furthermore, the selected image will undergo terrain feature extraction processing. This can be achieved using image processing and computer vision techniques to extract terrain-related information from the image. For example, edge detection, texture analysis, and other methods can be used to identify road boundaries, color variations, texture features, and so on.

[0084] Furthermore, by mapping these features to the expected driving trajectory, terrain data at different locations along the vehicle's trajectory can be calculated. This terrain data can include altitude change curves, driving trajectory, vectors, coordinates, terrain pitch angle changes, terrain roll angle changes, yaw angle changes, etc.

[0085] The elevation change curve describes the undulations of the road surface. The driving trajectory describes the vehicle's path. Vectors represent the vehicle's direction and speed. Coordinates identify the vehicle's position at different points along its trajectory. Terrain pitch angle changes describe the vertical tilt of the road surface. Terrain roll angle changes describe the lateral tilt of the road surface. Yaw angle changes describe the vehicle's lateral sway.

[0086] In summary, in practical applications, these steps work together to enable the system to acquire detailed road surface information based on the expected driving trajectory and terrain features extracted from visual images. This information is then output to the cruise control system or the driver to assist in planning more suitable driving routes and speeds, setting more appropriate drive torque distribution methods, etc., to adapt to different terrain conditions and improve vehicle stability and performance.

[0087] In some embodiments, the system not only identifies the road surface terrain corresponding to the vehicle's driving trajectory, but also evaluates the acquired terrain data to determine its credibility or confidence level. Specifically, the system can evaluate the data completeness of the terrain data obtained from visual images and the expected driving trajectory. This can include checking for missing data, noise, or other inaccuracies. The system then evaluates the previous identification steps to determine the degree of matching between the identified terrain and the actual terrain. This can be achieved by comparing the extracted terrain features with the actual terrain features.

[0088] Furthermore, based on the completeness of the data and the recognition results, the system can calculate the confidence level of the terrain data. If the data is complete and the recognition results match the actual terrain height well, the confidence level will be high. Conversely, if the data is incomplete or the recognition results are inaccurate, the confidence level will be low.

[0089] In some embodiments, the system can also comprehensively consider multiple other factors to determine the confidence level of the terrain data. For example, the system can adjust the confidence level calculation based on factors such as vehicle speed, driving status (acceleration, braking, steering, etc.), and sensor stability. Additionally, the system can incorporate historical data, such as previous identification results of similar terrain, to further improve the confidence level of the terrain data.

[0090] In some embodiments, the system not only focuses on identifying the road surface terrain corresponding to the expected driving trajectory, but also compares the actual road surface terrain data with the road surface terrain data identified in the corresponding driving area to detect whether there is underfitting, that is, the actual terrain is inconsistent with the identified terrain.

[0091] Specifically, during vehicle operation, the system can use various sensors to acquire actual terrain data of the road surface the vehicle is currently traveling on. This data can include information such as changes in ground elevation, slope, and roll angle.

[0092] Furthermore, by comparing the acquired actual road surface terrain data with the road surface terrain data identified in the corresponding driving area, the system can determine whether the deviation between them exceeds a preset threshold. If the deviation between the actual road surface terrain data and the identified terrain data exceeds the preset threshold, the system can mark these identified terrain data as underfitting data. This indicates that the identification algorithm performs poorly in these areas and needs further optimization.

[0093] Furthermore, road surface terrain data labeled as underfitting data can be sent to the cloud to optimize the road surface terrain recognition algorithm. In the cloud, the algorithm can analyze this underfitting data, identify the problems, and optimize it. This optimization process may involve adjusting algorithm parameters, training machine learning or deep learning models, etc. The accuracy of the algorithm is improved through continuous iteration and training.

[0094] In some embodiments, the system can record the difference between the identified terrain data and the actual terrain data while labeling underfitting data. This information can be used to further analyze the direction of algorithm improvement. Furthermore, the system can dynamically adjust preset thresholds based on factors such as different road surface types and driving conditions to adapt to different real-world situations.

[0095] Furthermore, when optimizing in the cloud, the system can analyze the common features of underfitting data to improve algorithm performance. For example, if multiple vehicles report underfitting data on the same road segment, the system can centrally process these cases in the cloud to further optimize the algorithm for better performance in road surface recognition in that area.

[0096] In some embodiments, after optimization in the cloud, the system can obtain an algorithm for identifying road surface terrain optimized based on underfitting data. This optimized algorithm is the result of training and improvement based on previously collected underfitting data. By using the optimized algorithm, the system can more accurately identify the road surface terrain corresponding to the vehicle's driving trajectory based on visual images and the expected driving trajectory during actual vehicle operation.

[0097] Through the above steps, the system can more accurately identify road terrain based on the optimized algorithm, improving the vehicle's adaptability and safety under different terrain conditions. This real-time application optimization method can continuously improve the system's performance, enabling vehicles to better adapt to changing road environments.

[0098] In some embodiments, adjusting vehicle control parameters based on the road surface terrain corresponding to the vehicle's driving trajectory includes: pre-controlling at least one of the vehicle's driving force, braking force, steering angle, and suspension system parameters based on the road surface terrain corresponding to the vehicle's driving trajectory and the vehicle's actual driving state. The suspension system parameters may include suspension damping, suspension stiffness, and suspension height, etc. These adjustments aim to optimize the vehicle's stability, comfort, and performance.

[0099] In this context, driving force refers to the power output of the vehicle's engine, which propels the vehicle forward. In pre-control, the system can adjust the engine's output power based on the road surface and the vehicle's actual driving conditions, thereby altering the vehicle's acceleration performance. For example, on an uphill slope, the system can increase driving force in advance to ensure the vehicle can smoothly ascend the hill.

[0100] Braking force refers to the decelerating force generated by the braking system, used to slow down or stop a vehicle. In pre-control, the system can adjust the braking force distribution in advance based on the expected driving trajectory and road surface information. For example, when a vehicle is about to pass through a steep downhill section, the system can increase the braking force to control the vehicle speed and prevent excessive speed.

[0101] Steering angle refers to the angle of deflection of the wheels relative to the vehicle's direction of travel. In pre-control, the system can adjust the steering system in advance based on the expected driving trajectory and curve angle, making the vehicle more adaptable to curves. For example, when approaching a curve, the system can pre-adjust the steering angle to allow the vehicle to steer more smoothly.

[0102] Suspension system parameters refer to the adjustments made to the suspension's damping, stiffness, and height. In pre-control mode, the system can adjust these parameters in advance based on changes in road surface and the vehicle's actual driving conditions. For example, on uneven roads, the system can pre-reduce the suspension damping to provide better shock absorption.

[0103] In some embodiments, the vehicle control method further includes adjusting the control intensity of at least one of the vehicle's driving force, braking force, steering angle, and suspension system parameters based on the confidence level of the terrain data. This adjustment can achieve more refined vehicle control based on the confidence level of the terrain data to adapt to terrain recognition results with different confidence levels.

[0104] The confidence level of terrain data refers to the degree of trust in the identified road surface terrain information. This can be assessed based on factors such as the reliability of the identification algorithm and the accuracy of the sensors. Confidence level can be expressed as a value, such as a percentage or a range of 0 to 1, reflecting the reliability of the data. Based on the confidence level of the terrain data, the weights of control parameters such as driving force, braking force, steering angle, and suspension system can be adjusted. For high-confidence terrain data, the system can increase the weights of the corresponding control parameters to respond more actively to terrain changes. Conversely, for low-confidence terrain data, the system can decrease the weights of the corresponding control parameters to reduce the impact on vehicle control.

[0105] In some embodiments, a smooth approach can be considered when adjusting the intensity of control parameters to avoid the impact of sudden control changes on the vehicle. For example, the control parameters can be adjusted gradually according to the level of confidence to make the control transition of the vehicle smoother.

[0106] Therefore, the actual vehicle response is compared with the expected response, and the intensity of the control parameters is adjusted based on the actual feedback. Through continuous real-time adjustments, the vehicle's control becomes more adaptable to the current terrain and driving conditions. This method fully utilizes the confidence information of terrain data to adjust the vehicle's control parameters in a more intelligent and adaptive way, thereby improving vehicle stability, suspension adjustment, brake distribution, and overall driving performance.

[0107] In some embodiments, the vehicle control method further includes: acquiring actual road surface terrain data of the road surface where the vehicle is currently traveling; and performing closed-loop control on at least one of the vehicle's drive system, braking system, steering system, and suspension system based on the actual road surface terrain data.

[0108] Closed-loop control can be a feedback control mechanism where the system continuously measures the actual state, compares this information with the desired state, and then adjusts the control input to gradually bring the actual state closer to the desired state. In this embodiment of the invention, the vehicle's drive system, braking system, steering system, and suspension system can be subject to closed-loop control to make appropriate adjustments based on changes in the actual road surface. Such closed-loop control can adjust the vehicle's control parameters in real time according to the actual road surface conditions to ensure the vehicle's stability and performance under different terrain conditions.

[0109] Specifically, vehicles can acquire actual terrain data of the road surface they are currently driving on through various sensors (such as lidar, cameras, ultrasonic sensors, etc.). This data can include information such as changes in road height, bumps, potholes, curvature, and road type (paved road, unpaved road, etc.).

[0110] Furthermore, the acquired actual road surface terrain data is processed and analyzed to extract key terrain features. This may involve filtering, smoothing, and feature extraction of the terrain data to obtain effective information about the terrain. Based on the analyzed actual road surface terrain data, the vehicle's control system can adjust the control parameters of at least one of the drive system, braking system, steering system, and suspension system in real time. For example, when driving on uneven road sections, the control system can increase the damping of the suspension to reduce vehicle bumps. When driving uphill, the output of the drive system can be adjusted to ensure sufficient power.

[0111] Furthermore, based on actual road surface terrain data, the control system implements closed-loop control. This means that the vehicle's sensors are not only used to acquire actual road surface terrain data, but also to monitor the vehicle's status in real time. The control system can compare the actual state with the desired state and adjust the control parameters according to changes in the actual terrain to maintain the vehicle's stability and performance.

[0112] In this way, the vehicle can perform closed-loop control based on actual road terrain data, enabling it to maintain stable and efficient operation under different terrain conditions. This helps improve driving comfort, safety, and overall driving performance.

[0113] In some embodiments, the vehicle control method further includes: obtaining feedforward control data based on the road surface terrain corresponding to the vehicle's driving trajectory; acquiring actual road surface terrain data of the current driving road surface and obtaining feedback control data based on the actual road surface terrain data; obtaining ideal target control data based on the feedforward and feedback control data; and controlling the vehicle based on the ideal target control data.

[0114] Specifically, based on the road surface terrain corresponding to the vehicle's driving trajectory, the system can predict potential terrain changes the vehicle may encounter in the future. Based on these predictions, the system generates feedforward control data. This data includes adjustments to control parameters such as driving force, braking force, and steering angle based on anticipated terrain changes, so that the vehicle can remain stable under expected terrain variations.

[0115] Furthermore, the system acquires actual terrain data of the current road surface using sensors equipped in the vehicle. This data may include information such as ground elevation, slope, and degree of unevenness. Based on this actual terrain data, the system generates feedback control data. This feedback control data is used to dynamically adjust the vehicle's control parameters according to changes in the actual terrain to maintain vehicle stability and performance.

[0116] Furthermore, the feedforward and feedback control data are combined to generate ideal target control data. This data defines the optimal control parameters for the vehicle under both anticipated and actual terrain changes. Ideal target control data can be generated using mathematical models, control algorithms, and other methods to achieve smooth vehicle operation under various terrain conditions.

[0117] Furthermore, based on the generated ideal target control data, the vehicle's control system adjusts various control parameters of the vehicle. This may include adjusting driving force, braking force, steering angle, suspension system parameters, etc. Through the synergy of feedforward and feedback control, the vehicle can maintain good stability, comfort, and performance in various terrains.

[0118] In summary, combining feedforward and feedback control fully leverages the advantages of both control methods. Feedforward control uses visual prediction of terrain to enable the system to make adjustments in advance, enhancing the timeliness of control. Feedback control, on the other hand, ensures the accuracy of the final control and the robustness of the entire system. The combination of the two results in a system with better control performance and stability.

[0119] In some embodiments, acquiring actual road surface terrain data of the current driving surface includes: acquiring vehicle state data and wheel state data; obtaining the actual vehicle body tilt angle and the absolute height of each wheel based on the vehicle state data and wheel state data; and obtaining actual road surface terrain data based on the actual vehicle body tilt angle and the absolute height of each wheel.

[0120] Specifically, data related to the vehicle's status and wheel status are acquired through sensors installed on the vehicle. The vehicle status data can include information such as the vehicle's speed, acceleration, and attitude, while the wheel status data can include information such as the position and rotational speed of each wheel and the displacement of the suspension system.

[0121] Furthermore, by utilizing vehicle status data and wheel status data, the actual vehicle camber angle and the absolute height of each wheel can be calculated. The actual camber angle can refer to the angle of inclination between the vehicle body and the horizontal plane, while the absolute height of the wheel can refer to the vertical distance between the wheel and a reference position.

[0122] Furthermore, based on the calculated actual vehicle body tilt angle and the absolute height of the wheels, the current road surface conditions of the vehicle can be inferred. For example, by comparing the absolute height of each wheel, it can be determined whether the vehicle is traveling on an uneven road surface. The actual vehicle body tilt angle can be used to estimate whether the vehicle is traveling on a slope, and the magnitude of the slope.

[0123] Figure 3 This is a functional architecture diagram of an automatic terrain control system according to an embodiment of the present invention, such as... Figure 3As shown, the automatic terrain control system 100 includes an automatic terrain recognition module 110, a terrain analysis control module 120, a feedback control module 130, and a four-wheel terrain analysis module 140.

[0124] The automatic terrain recognition module 110 includes a driving trajectory prediction module 111, a visual processing module 112, an online data update module 113, and a feature recognition and classification module 114.

[0125] In some embodiments, the driving trajectory prediction module 111 may input signals including gear position signal, driver throttle signal, driver brake signal, steering wheel angle signal, and cruise command signal. The gear position signal indicates the vehicle's current gear, such as drive or reverse. The driver throttle signal reflects the driver's current accelerator pedal input, i.e., the driver's desired acceleration intensity of vehicle 1. The driver brake signal reflects the driver's current brake pedal input, i.e., the driver's desired deceleration or stopping intensity of vehicle 1. The steering wheel angle signal indicates the current steering wheel angle, i.e., the driver's desired steering angle of vehicle 1. The cruise command signal may include information such as virtual throttle, target braking intensity, target acceleration / deceleration, target vehicle speed, and target steering angle, used to indicate the vehicle 1's cruise target.

[0126] The trajectory prediction module 111 can also combine vehicle parameters, such as the vehicle's mass, suspension stiffness, and power performance, to estimate the expected trajectory of the vehicle within a future driving range. The output can be the movement vector curve of the vehicle's center of gravity, or information such as the current trajectory radius of the vehicle. This predicted trajectory information will be used for subsequent terrain recognition and control strategy formulation.

[0127] In some embodiments, the online data update module 113 is a functional module used to upload underfitting data identified by the system to the cloud server 3 and obtain the latest version of the visual algorithm from the cloud server 3 to update the feature learning module. When the automatic terrain recognition module 110 finds a large deviation between the visual image recognition result and the actual terrain, it marks this underfitting data and uploads it to the cloud server 3 via the communication network 2. On the cloud server 3, this underfitting data will be used to optimize the training of the visual algorithm and improve the recognition accuracy. At the same time, the cloud server 3 can periodically update the feature learning module and transmit the latest version of the visual algorithm to the vehicle control device 10 of the terrain automatic control system 100. Through online data updates, the system can continuously improve the terrain recognition algorithm and maintain its accuracy and robustness.

[0128] In some embodiments, the feature recognition and classification module 114 can emit tagged data when needed. This data can be captured by the online data update module 113, packaged into data packets of a fixed size according to the data volume, and then stored in the erasable non-volatile portion of the data storage module. The online data update module 113 can select one or more of the following methods to upload these data packets to the cloud backend according to different strategies:

[0129] First, periodic updates: the system can be set to upload data every certain period of time (i.e. every few minutes or hours).

[0130] Second, event-based updates: whenever a visual and measurement discrepancy occurs, the online data update module 113 can immediately start uploading data until the discrepancy no longer exceeds the specified range.

[0131] Third, resource planning-based updates: when the utilization rate of the storage space allocated to the marked data reaches a certain level (e.g., greater than 90%), the online data update module 113 starts uploading.

[0132] Fourth, manual updates: drivers or testers can manually request the system to upload data through specific operations.

[0133] Furthermore, once each data packet is successfully uploaded to the cloud server 3, the online data update module 113 can remove the uploaded data from the storage space to free up storage space for subsequent data recording.

[0134] Furthermore, the cloud-based backend uses a large amount of underfitting data to train and enhance the visual algorithm, confirming the improvement in recognition efficiency. After the visual algorithm optimization is completed in the cloud-based backend, a new version of the algorithm software can be pushed to the terminal. After receiving a request from the cloud server 3 or the vehicle operator, the online data update module 113 can update the new version of the software unit to the feature recognition and classification module 114. These updates may include decision trees, random forests, deep convolutional neural networks, combinations of deep convolutional neural networks and recurrent neural networks and their variants, or combinations of deep convolutional neural networks and conditional random fields, etc. These algorithms will be used for terrain feature recognition and classification, thereby assisting the terrain automatic control system 100 in making corresponding control strategies and improving the driving performance and safety of vehicle 1 on unpaved roads. Through continuous updating and optimization of algorithms, the terrain automatic control system 100 can continuously improve and enhance its recognition and control performance.

[0135] In some embodiments, the vision processing module 112 can preprocess the images transmitted from the camera. Preprocessing includes sharpness checks, camera selection, depth information extraction from binocular images, and image frame extraction to obtain an image library of the terrain of interest. Based on the trajectory information input by the driving trajectory prediction module 111, the module selects the predicted trajectory coverage area in the image, combines the trajectory information with the depth image to obtain image data containing both trajectory and depth information, performs image cropping according to the trajectory coverage module, further processes the image into a form more suitable for feature recognition and classification, and then inputs it to the feature recognition and classification module 114.

[0136] In some embodiments, the feature recognition and classification module 114 uses a pre-trained visual recognition algorithm to extract features from the depth image containing the trajectory. It then combines this with the terrain where the vehicle 1 is located to correct the information in the visual image, offsetting the influence of the vehicle's angle on the camera's shooting angle, and thus outputs the predicted height change curves of each wheel on the ground.

[0137] Furthermore, during feature recognition, the feature recognition and classification module 114 can compare features in the depth image with pre-trained visual algorithms to identify terrain features, such as changes in ground elevation, terrain color, and texture. In addition to image recognition, distance or depth data obtained from radar can also be used to assist in feature recognition and classification. This information helps to more accurately identify the terrain where the vehicle 1 is currently located, providing important information for subsequent terrain-based automatic control strategies.

[0138] The following is for reference. Figure 4 The operation of the feature recognition and classification module in this embodiment of the invention will be described in detail.

[0139] Figure 4 This is a working scenario diagram of a feature recognition and classification module according to an embodiment of the present invention, such as... Figure 4 As shown, a scene terrain model can be constructed based on the calculated expected driving trajectory and information such as ground distance / depth / height / object size provided by visual recognition. The absolute height of each wheel on the ground at different locations along the driving trajectory is calculated. For example, Figure 4 Taking the left front wheel of vehicle 1 as an example, the relationship curve between the longitudinal displacement of the left front wheel and the vertical height of wheel 148 from the horizontal ground was obtained, as shown in the appendix. Figure 5 .

[0140] In addition to altitude curves, the feature recognition and classification module 114 outputs various terrain information, such as driving trajectory, vectors, coordinates, terrain pitch angle changes, terrain roll angle changes, and yaw angle changes. Furthermore, the feature recognition and classification module 114 can output the confidence level of this information based on the completeness of the data and the recognition results; the confidence level will determine the intervention intensity of the feedforward control.

[0141] In some embodiments, the feature recognition and classification module 114 can retain the results of the visual prediction until the vehicle 1 travels to the predicted area for comparison and inspection. When the prediction result of the visual recognition algorithm, i.e., the terrain model constructed visually, is compared with the measurement data obtained after the vehicle 1 has traveled, and the deviation exceeds a certain range, for example, the longitudinal slope deviation of the vehicle 1 is greater than 3%; the lateral slope deviation is greater than 3%; the longitudinal slope deviation is greater than 2% and the lateral slope deviation is greater than 2%; the absolute height deviation of a single wheel exceeds 0.1m; or the feature recognition and classification module 114 cannot recognize the terrain ahead, the feature recognition and classification module 114 can mark the vehicle driving terrain image and the measured terrain data in the time domain corresponding to the data exceeding the deviation as underfitting data.

[0142] In some embodiments, the terrain analysis control module 120 is an important module of the terrain automatic control system 100, including a feedforward control module 121 and an ideal target calculation module 122. These functions can be manually selected by the driver or turned off as needed.

[0143] Among them, the feedforward control module 121 intervenes in the driving, braking, steering and suspension systems of the vehicle 1 in advance based on the terrain information output by the feature recognition and classification module 114 and the current driving state of the vehicle 1, so as to adapt to the characteristics of the terrain ahead, optimize the control strategy of the vehicle 1, and ensure the stability and safety of the vehicle 1 on specific terrain.

[0144] In some embodiments, the feedforward control module 121 selects an appropriate intervention strategy based on the terrain information output by the feature recognition and classification module 114. These intervention strategies include, but are not limited to, the following:

[0145] First, the feedforward control module 121 can issue an alarm to the driver or limit power output. If the terrain ahead may prevent vehicle 1 from passing safely, such as mud, deep water, obstacles, pits, cliffs, etc., the feedforward control module 121 can issue an alarm to the driver or limit power output to prevent vehicle 1 from getting into dangerous situations.

[0146] Second, request the active suspension system to raise the vehicle height. If the terrain ahead may scrape the vehicle chassis at its current height, the feedforward control module 121 can request the active suspension system to raise the vehicle height to ensure the safe passage of the vehicle chassis.

[0147] Third, request a reduction in intensity or pre-braking. If there are foreign objects or terrain unfavorable for high-speed driving ahead, the feedforward control module 121 can request the drive system to reduce intensity or the braking system to pre-brake to ensure that the vehicle 1 can smoothly pass through the specific terrain.

[0148] Fourth, request increased steering assist or adjustment of suspension stiffness. If the road ahead is rocky or has significant undulations, the feedforward control module 121 can request the steering system to increase the steering assist or enhance the return-to-center stability control to improve the handling and stability of vehicle 1. Simultaneously, it can request the active suspension system to raise the vehicle height or reduce the damping of the shock absorbers to ensure vehicle 1 can adapt to complex terrain.

[0149] Fifth, request the powertrain system to perform pre-processing. If the road ahead may cause some wheels 148 to lift off the ground, the feedforward control module 121 can request the powertrain system to perform pre-processing, such as switching to four-wheel drive or entering a traction control mode, to increase the vehicle's passability. At the same time, it will also request optimization of the recognition thresholds of functions such as ABS (Anti-lock Braking System) or TCS (Traction Control System) to improve the vehicle 1's control performance on specific terrains.

[0150] Furthermore, the feedforward control module 121 can acquire the confidence level of terrain information along with the terrain information. The confidence level reflects the reliability of the terrain information by the feature recognition and classification module 114. Based on the confidence level, the feedforward control module 121 can adjust the intervention intensity of the control. For example, if the visual information ahead is blurry and the terrain cannot be determined, resulting in a confidence level of 0, the intervention variable output by the feedforward control module 121 will also be adjusted to no intervention to avoid erroneous control interference.

[0151] In some embodiments, the ideal target calculation module 122 calculates the ideal operating target parameters of vehicle 1 based on the current state of the vehicle, the driver's request, and the terrain conditions. These target parameters include the target range of vehicle speed, the target slip range of wheel speed 148, the target height of the vehicle body, the target pitch angle / target roll angle of the vehicle body, and the target yaw rate of the vehicle body.

[0152] In some embodiments, the ideal target calculation module 122 includes two modes corresponding to different calculation results. One mode is based on the terrain recognized by vision, which is used as a reference for the feedforward control module 121. The other mode is based on the measured terrain obtained after the vehicle 1 has passed, which is used as a reference for the feedback control module 130. Based on the calculation results of the feedforward and feedback modes, the control logic is calibrated / trained with a large amount of data to improve the accuracy of the judgment and output the ideal control target parameters.

[0153] Furthermore, the target parameters output by the ideal target calculation module 122 will be used for closed-loop control. In closed-loop control, the targets of feedforward control and feedback control will be calculated by weighting their confidence levels, i.e., control target = feedforward control confidence level + feedback control confidence level. These are then used to adjust the drive, braking, steering, and suspension systems of vehicle 1 to achieve more precise control.

[0154] In some embodiments, the feedback control module 130 performs closed-loop intervention on the vehicle's driving, braking, steering, and suspension systems based on the actual terrain where the vehicle 1 is located, as output by the four-wheel terrain analysis module 140. In this way, even if the feedforward control suffers from pre-control deviations due to visual recognition errors, the feedback control module 130 can adjust the control strategy in a timely manner according to the actual terrain conditions, avoiding the accumulation of control deviations and ensuring the vehicle's stability on complex terrain.

[0155] In some embodiments, the four-wheel terrain analysis module 140 is based on various sensor signals and vehicle status, including longitudinal acceleration, lateral acceleration, vertical acceleration signals, pitch angular velocity, roll angular velocity, yaw angular velocity signals, wheel speeds, wheel heights, pitch angles, roll angles, yaw angles, and other information. Combined with overall vehicle parameters, this module can accurately measure and solve for the actual tilt angle of the vehicle body and the absolute height of each wheel, and perform terrain calculation based on this information. The principle of terrain calculation can be found in [reference needed]. Figure 5 Please provide an explanation.

[0156] Figure 6 This is a schematic diagram illustrating the principle of a four-wheel terrain analysis module according to an embodiment of the present invention, as shown below. Figure 6 As shown in the figure, this figure describes the position of vehicle 1 on the current ground and the relationship between the coordinate systems. Through this information, the coordinate system corresponding to the ground where vehicle 1 is located can be derived, thereby realizing the analysis and measurement of the terrain.

[0157] The text describes a vehicle's location within a plane. The first part, "Ground 141," represents the actual ground surface where vehicle 1 is located, which can be an uneven road surface or other terrain. The second part, "Coordinate System 144," represents a plane relative to vehicle 1, with its origin at the vehicle's center of mass 146, and has a transformation relationship with the vehicle's body coordinate system. The third part, "Horizontal Plane 143" (parallel to sea level), represents a horizontal plane, typically parallel to the Earth's sea level. The fourth part, "Coordinate System 144," represents a plane relative to vehicle 1, with its origin at the vehicle's center of mass 146, and is described on horizontal plane 143. The fifth part, "Vehicle Body 145," represents the overall shape of the vehicle. The sixth part, "Vehicle Center of Mass 146," represents the location of vehicle 1's center of gravity. The seventh part, "Coordinate System 147," represents a plane with its chassis on it, with its origin at the vehicle's center of mass 146, and describes the position of the chassis. The eighth part, "Wheels 148," represents the four wheels of vehicle 1.

[0158] In some embodiments, by knowing the relationship between the vehicle body coordinate system and the horizontal coordinate system, as well as the relationship between the plane on which the vehicle chassis is located and the heights (Hfl, Hfr, Hrl, Hrr) of the wheels 148, and combining this with the overall vehicle parameters, the coordinate system corresponding to the ground 141 where the vehicle is currently located can be solved. In this way, terrain information such as the height and slope of the ground where the vehicle 1 is located can be accurately measured.

[0159] Based on this terrain analysis principle, the terrain automatic control system 100 can acquire real-time information about the terrain where vehicle 1 is located, providing accurate data to the feedforward control module 121 and the feedback control module 130 to achieve precise adjustment and optimization of the vehicle control strategy. In this way, vehicle 1 can more intelligently adapt to driving needs under different terrain conditions, improving driving safety and adaptability. Based on the vehicle control method of the above embodiment, the following refers to... Figures 7-8 A vehicle control device according to an embodiment of the present invention is described.

[0160] Figure 7 This is a block diagram of a vehicle control device according to an embodiment of the present invention, such as... Figure 7 As shown, the vehicle control device 10 includes: an acquisition module 11; a prediction module 12; and a control module 13.

[0161] The acquisition module is used to acquire visual images of the road surrounding the vehicle and the expected driving trajectory. This module may include perception devices such as cameras, sensors, and radar, which are responsible for capturing information about the vehicle's surrounding environment in real time. The visual images may include elements such as roads, the surrounding environment, obstacles, traffic signs, and other vehicles. The expected driving trajectory can be a theoretical driving path derived from the vehicle's navigation system, map data, and planning algorithms.

[0162] The prediction module predicts the road surface terrain corresponding to the vehicle's trajectory based on visual images and the expected driving trajectory. Road surface terrain can be of various types, including but not limited to: paved roads (such as asphalt roads, concrete roads, etc.), unpaved roads (such as dirt roads, gravel roads, etc.), uneven road surfaces (such as bumpy, potholed, and uneven sections), flat roads, and slopes (such as uphill, downhill, and side slopes with different gradients). The prediction module's task is to match image information and driving trajectory with terrain type and features for further control and adjustment.

[0163] The control module adjusts the vehicle's control parameters based on the road surface terrain corresponding to the vehicle's driving trajectory. These parameters include driving force, braking force, steering force, and suspension system parameters. Depending on the different characteristics of the road surface, the control module can automatically adjust these parameters to achieve optimal vehicle performance and stability. For example, on rough roads, it may increase the damping of the suspension system to reduce bumps; while on smooth roads, it can optimize power distribution to improve driving performance.

[0164] According to the vehicle control device of the present invention, the acquisition module acquires the visual image and expected driving trajectory of vehicle 1, and the prediction module can automatically identify the road surface terrain corresponding to the current driving trajectory of the vehicle. This method of comprehensively utilizing the visual image and the expected driving trajectory can avoid the shortcomings of relying on a single sensor for terrain recognition, and can predict the road surface terrain of the vehicle's driving path. Based on the road surface terrain corresponding to the vehicle's driving trajectory, the control parameters of vehicle 1 can be adjusted in advance, so that vehicle 1 can better adapt to various road surface changes and maintain a stable driving state, thereby improving the adaptability and safety of vehicle 1 under different road surface conditions.

[0165] Figure 8 This is a block diagram of a vehicle control device according to yet another embodiment of the present invention, such as Figure 8 As shown, the vehicle control device 10 includes at least one processor 101 and a memory 102.

[0166] The processor 101 is used to execute instructions in a computer program. This processor 101 may be a central processing unit (CPU) or other computing device.

[0167] The memory 102 is communicatively connected to the processor 101 and is used to store computer programs and related data. This memory 102 may include random access memory (RAM), read-only memory (ROM), flash memory, etc.

[0168] The computer program stored in memory 102 is a series of instructions that implement the vehicle control method described in the above embodiment. These instructions are executed by processor 101 to implement the various steps and functions described in the vehicle control method.

[0169] According to the vehicle control device 10 of the present invention, when the processor 101 executes the computer program, it implements the vehicle control method described in the above embodiment. By comprehensively utilizing visual images and expected driving trajectories, it realizes the function of automatically identifying road terrain and adjusting vehicle control parameters. This comprehensive terrain recognition and parameter adjustment capability enables the vehicle 1 to more accurately adapt to changes in different road terrain, thereby improving the adaptability and driving safety of the vehicle 1 under different road conditions.

[0170] In some embodiments of the present invention, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed, implements the vehicle control method of any of the above embodiments.

[0171] According to the computer-readable storage medium of the present invention, by executing the vehicle control method described in the above embodiments, the vehicle's visual image and expected driving trajectory are comprehensively utilized to automatically identify the road surface terrain corresponding to the vehicle's driving trajectory. This method of comprehensively utilizing visual images and expected driving trajectories can avoid the shortcomings of relying solely on a single sensor for terrain recognition, and can predict the road surface terrain of the vehicle's driving path. Based on the road surface terrain corresponding to the vehicle's driving trajectory, the control parameters of vehicle 1 can be adjusted in advance, thereby enabling vehicle 1 to better adapt to various road surface changes and maintain a stable driving state, thus improving the adaptability and safety of vehicle 1 under different road surface conditions.

[0172] The computer-readable storage medium in the embodiments of the present invention may include, but is not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be described in detail here.

[0173] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example.

[0174] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A vehicle control method, characterized in that, include: Acquire visual images of the road surrounding the vehicle and the expected driving trajectory; Predicting the road surface terrain corresponding to the vehicle's driving trajectory based on the visual image and the expected driving trajectory includes: obtaining the area covered by the expected driving trajectory in the visual image to obtain an image containing the driving trajectory; extracting terrain features from the image containing the driving trajectory; and solving the terrain data of the road surface at each wheel position along the driving trajectory based on the extracted terrain features. Adjusting the vehicle's control parameters according to the road surface terrain corresponding to the vehicle's driving trajectory includes: obtaining feedforward control data according to the road surface terrain corresponding to the vehicle's driving trajectory; The vehicle control method further includes: acquiring actual road surface terrain data of the current driving road surface of the vehicle, obtaining feedback control data based on the actual road surface terrain data, obtaining ideal target control data based on the feedforward control data and the feedback control data, and controlling the vehicle based on the ideal target control data.

2. The vehicle control method according to claim 1, characterized in that, The vehicle control method further includes: The confidence level of the terrain data is determined based on the data completeness and identification results.

3. The vehicle control method according to claim 1, characterized in that, The vehicle control method further includes: Obtain the actual road terrain data of the road surface where the vehicle is currently traveling; The actual road surface terrain data is compared with the road surface terrain data identified in the corresponding driving area; If the deviation between the actual road surface terrain data and the road surface terrain data identified in the corresponding driving area exceeds a preset threshold, the identified road surface terrain data is marked as underfitting data. The underfit data is sent to the cloud to optimize the algorithm for identifying the road surface terrain based on the underfit data.

4. The vehicle control method according to claim 3, characterized in that, The vehicle control method further includes: Obtain an algorithm optimized based on the underfitting data to identify the road surface terrain, and then perform road surface terrain identification.

5. The vehicle control method according to claim 1, characterized in that, Adjusting the vehicle's control parameters based on the road surface terrain corresponding to the vehicle's driving trajectory includes: Based on the road surface terrain corresponding to the vehicle's driving trajectory and the vehicle's actual driving state, at least one of the following parameters—driving force, braking force, steering angle, and suspension system parameters—is pre-controlled.

6. The vehicle control method according to claim 5, characterized in that, The vehicle control method further includes: The control intensity of at least one of the vehicle's driving force, braking force, steering angle, and suspension system parameters is adjusted based on the confidence level of the terrain data.

7. The vehicle control method according to claim 1, characterized in that, Obtain the actual road terrain data of the road surface where the vehicle is currently traveling, including: Acquire vehicle status data and wheel status data; The actual vehicle body tilt angle and the absolute height of each wheel are obtained based on the vehicle status data and the wheel status data. The actual road surface terrain data is obtained based on the actual vehicle body tilt angle and the absolute height of each wheel.

8. A vehicle control device, characterized in that, include: The acquisition module is used to acquire visual images of the road sections surrounding the vehicle and the expected driving trajectory; The prediction module is used to predict the road surface terrain corresponding to the vehicle's driving trajectory based on the visual image and the expected driving trajectory, including: obtaining the area covered by the expected driving trajectory in the visual image to obtain an image containing the driving trajectory; extracting terrain features from the image containing the driving trajectory; and solving the terrain data of the road surface at each wheel position along the driving trajectory based on the extracted terrain features. The control module is used to adjust the control parameters of the vehicle according to the road surface terrain corresponding to the vehicle's driving trajectory, including: obtaining feedforward control data according to the road surface terrain corresponding to the vehicle's driving trajectory, obtaining actual road surface terrain data of the current driving road surface of the vehicle, obtaining feedback control data according to the actual road surface terrain data, obtaining ideal target control data according to the feedforward control data and the feedback control data, and controlling the vehicle according to the ideal target control data.

9. A vehicle control device, characterized in that, include: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores a computer program, and when the at least one processor executes the computer program, it implements the vehicle control method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, it implements the vehicle control method according to any one of claims 1-7.

11. A vehicle, characterized in that, include: A vision system used to acquire visual images of the vehicle; The vehicle control device of claim 9, wherein the vehicle control device is connected to the vision system.

12. The vehicle according to claim 11, characterized in that, The vehicle also includes: A communication device, connected to the vehicle control device, is used to send underfit data and receive an algorithm for identifying road surface terrain optimized based on the underfit data.

13. The vehicle according to claim 11, characterized in that, The vehicle also includes wheel height sensors, wheel speed sensors, and vehicle body inertial navigation sensors.

Images