A vehicle autonomous control method and system

Abstract

The application relates to the technical field of vehicle autonomous control, and provides a vehicle autonomous control method and system, the method comprising the following steps: acquiring motion information of a vehicle when a working device of the vehicle is working, wherein the motion information comprises satellite positioning information of the vehicle and wheel speed information of a wheel; judging whether the vehicle is in a lateral slip motion state according to the motion information; stopping the working device from working when the vehicle is in the lateral slip motion state, acquiring a ground visual image of an environment where the vehicle is located and inertial navigation data of the vehicle; determining moving positioning information of the vehicle relative to the ground according to the ground visual image; the moving positioning information comprises longitudinal moving speed, lateral moving speed and positioning information; and generating and enabling the vehicle to execute an obstacle avoidance control instruction for avoiding obstacles according to the moving positioning information when an obstacle is detected. The stability of vehicle control can be improved.

Description

Technical Field

[0001] This application relates to the field of vehicle autonomous control technology, and in particular to a vehicle autonomous control method and system. Background Technology

[0002] In modern smart agriculture and animal husbandry, autonomous agricultural vehicles undertake key tasks such as sowing and fertilization, aiming to improve productivity and reduce human labor input. However, in actual farmland operations, vehicles often encounter complex and changing ground conditions, such as muddy, sandy, or icy roads caused by rainfall. These sudden environmental changes, without map updates, pose a severe challenge to the vehicle's autonomous control system. Traditional control methods often struggle to accurately determine the vehicle's true motion state, leading to command conflicts between multiple control objectives such as path tracking, attitude stabilization, and obstacle avoidance, thus affecting operational safety and efficiency.

[0003] For example, in actual smart agriculture and animal husbandry operations, an autonomous vehicle performing precision fertilization tasks is pre-loaded with a high-precision 3D map of the farmland and relies on high-precision satellite positioning information and inertial navigation unit data to accurately travel along a preset path. However, when the vehicle enters a muddy area formed by heavy rainfall, the ground adhesion decreases sharply, and the vehicle's drive wheels begin to slip noticeably. At this time, the satellite positioning system shows that the vehicle's overall position has deviated from the predetermined route, while the wheel speed sensors indicate that the wheels are still rotating at the correct speed. This inconsistency between information from different sensors causes the control system to make an incorrect judgment on the vehicle's true motion state.

[0004] When the control system detects a deviation from the path, it attempts to correct the vehicle's attitude by increasing the driving torque of the slipping wheel. However, on muddy surfaces with extremely low traction, the increased driving torque exacerbates wheel spin and slippage, causing the vehicle's lateral sliding to become more severe. The actual driving trajectory deviates completely from the control system's expectations, entering an unstable, semi-uncontrolled state. Even more dangerously, when the vehicle is severely slipping and sideslipping in mud, its attitude and actual direction of travel have changed drastically. Even if the control system issues a steering command, the vehicle cannot steer accordingly and instead continues to maintain its previous inertia, sliding laterally towards the obstacle.

[0005] At this point, the vehicle's central control system is in trouble. The path planning module continues to issue commands, attempting to pull the vehicle back onto its original path; meanwhile, the camera- and radar-based proximity perception system has identified a significant risk of an impending collision with an obstacle, triggering the emergency obstacle avoidance module. This module then issues the highest priority command, demanding that the vehicle brake immediately or make a maximum-angle emergency steering maneuver. The commands from these two modules are contradictory. More importantly, the vehicle kinematic model upon which the control system relies for its decisions is based on normal ground adhesion conditions. Under the current severe slippage, this model has completely failed. The system cannot accurately predict the consequences of executing commands such as "emergency braking" or "high-angle steering," potentially leading to a collision with the obstacle in an even more unpredictable manner. Summary of the Invention

[0006] This application provides a vehicle autonomous control method and system, which aims to solve the problem that traditional control methods are unable to accurately determine the actual motion state of agricultural autonomous vehicles in complex and variable terrain environments, resulting in command conflicts between control objectives such as path tracking, attitude stabilization and obstacle avoidance, which affects operational safety and efficiency.

[0007] To achieve the above objectives, this application adopts the following technical solution: In a first aspect, a vehicle autonomous control method is provided, comprising: acquiring vehicle motion information, including vehicle satellite positioning information and wheel speed information, when the vehicle's working device is operating; determining whether the vehicle is in a lateral slip motion state based on the motion information; stopping the working device when the vehicle is in a lateral slip motion state, acquiring ground visual images of the vehicle's environment and the vehicle's inertial navigation data; determining the vehicle's motion positioning information relative to the ground based on the ground visual images; the motion positioning information including longitudinal movement speed, lateral movement speed, and positioning information; and generating and executing obstacle avoidance control commands to avoid obstacles based on the motion positioning information when an obstacle is detected.

[0008] This technical solution enables the effective identification of lateral slippage of vehicles under complex ground conditions and timely cessation of operations. Instead, it utilizes ground visual images and inertial navigation data to obtain more accurate vehicle movement and positioning information relative to the ground. Thus, when an obstacle is detected, it can generate and execute obstacle avoidance control commands based on more reliable positioning information. This solves the problem that traditional methods cannot accurately determine the motion state and effectively avoid obstacles when the vehicle is slipping, significantly improving the vehicle's operational safety and autonomous control capabilities in harsh environments.

[0009] Furthermore, determining whether a vehicle is in a lateral slip motion state based on motion information includes: determining the vehicle's first speed based on satellite positioning information; calculating the vehicle's second speed based on wheel speed information; determining that the vehicle is in a lateral slip motion state when the absolute value of the difference between the first speed and the second speed is greater than a preset speed difference threshold; otherwise, determining that the vehicle is not in a lateral slip motion state.

[0010] By using this technical solution, this application can accurately and objectively determine whether a vehicle is in a lateral slip state by comparing the difference between the vehicle speed determined by satellite positioning information and the vehicle speed calculated by wheel speed information, thus avoiding misjudgment that may be caused by information from a single sensor and improving the accuracy and reliability of lateral slip state judgment.

[0011] Based on this, the vehicle's first speed is determined according to satellite positioning information, including: acquiring the vehicle's inertial navigation data; calculating the vehicle's third speed based on the inertial navigation data; calculating the vehicle's fourth speed based on satellite positioning information; and using the weighted sum of the third and fourth speeds as the vehicle's first speed.

[0012] This technical solution calculates the vehicle's first speed by fusing inertial navigation data and satellite positioning information. It leverages the complementary advantages of two different sensors to improve the accuracy and robustness of vehicle speed estimation. In particular, it can provide more reliable speed information when the signal from a single sensor is limited or contains errors.

[0013] In some preferred embodiments, determining the vehicle's motion positioning information relative to the ground based on ground visual images includes: identifying a visual landmark from the ground visual image; determining the longitudinal displacement pixel distance and lateral displacement pixel distance of the visual landmark in two adjacent ground visual image frames; converting the longitudinal displacement pixel distance and lateral displacement pixel distance to obtain the longitudinal displacement distance and lateral displacement distance, respectively; determining the vehicle's longitudinal movement speed relative to the ground based on the longitudinal displacement distance and the time interval between two adjacent ground visual image frames; determining the vehicle's lateral movement speed relative to the ground based on the lateral displacement distance and the time interval between two adjacent ground visual image frames; inputting the longitudinal movement speed, lateral movement speed, and satellite positioning information of the vehicle before it entered an abnormal motion state into a preset positioning model to obtain the vehicle's motion positioning information; the preset positioning model is used to determine the vehicle's position information.

[0014] This technical solution utilizes ground visual images for visual odometry calculation, which can directly obtain the vehicle's actual speed and displacement relative to the ground. This effectively avoids the problem of traditional wheel- or satellite-based positioning methods failing in lateral slip conditions, thus providing accurate positioning information that matches the actual ground movement for subsequent obstacle avoidance decisions.

[0015] Furthermore, upon detecting an obstacle, an obstacle avoidance control command is generated and executed by the vehicle to avoid the obstacle based on the motion positioning information, including: determining the vehicle's controlled adjustment coefficient based on the longitudinal and lateral movement speeds in the motion positioning information; and generating and executing the obstacle avoidance control command based on the controlled adjustment coefficient, the motion positioning information, and the position information between the obstacle and the vehicle.

[0016] Through this technical solution, this application introduces a controlled adjustment coefficient based on the actual moving speed of the vehicle during the obstacle avoidance command generation process. This enables the obstacle avoidance command to be adaptively adjusted according to the current real motion state of the vehicle, avoiding the problem of inaccurate or invalid obstacle avoidance commands due to the failure of the vehicle kinematic model in the lateral slip state, thereby improving the success rate and safety of obstacle avoidance.

[0017] As a technical improvement, the controlled adjustment coefficient of the vehicle is determined based on the longitudinal and lateral movement speeds in the mobile positioning information, including: determining the lateral slip angle of the vehicle based on the longitudinal and lateral movement speeds; obtaining a first preset correspondence; the first preset correspondence includes a one-to-one correspondence between multiple lateral slip angle ranges and multiple controlled adjustment coefficients; the controlled adjustment coefficient is less than 1, and the controlled adjustment coefficient is positively correlated with the maximum value in the corresponding lateral slip angle range; the controlled adjustment coefficient corresponding to the lateral slip angle range of the vehicle in the first preset correspondence is used as the controlled adjustment coefficient of the vehicle.

[0018] By establishing a preset correspondence between the lateral slip angle and the controlled adjustment coefficient, this application can dynamically adjust the intensity of the control command according to the actual degree of lateral slip of the vehicle, ensuring the effectiveness and safety of obstacle avoidance commands under different slip states, avoiding excessive or insufficient control, and thus optimizing the obstacle avoidance effect.

[0019] To improve the scheme, obstacle avoidance control commands are generated and executed by the vehicle based on the controlled adjustment coefficient, motion positioning information, and the position information between the obstacle and the vehicle. This includes: determining the duration between the current time and the time of collision between the vehicle and the obstacle when the vehicle does not execute the obstacle avoidance control commands, based on the motion positioning information and the position information between the obstacle and the vehicle; multiplying the duration by the controlled adjustment coefficient as the obstacle avoidance duration of the vehicle; and inputting the obstacle avoidance duration, motion positioning information, and the position information between the obstacle and the vehicle into the obstacle avoidance strategy generation model to generate and execute obstacle avoidance control commands.

[0020] By applying the controlled adjustment coefficient to the collision duration calculation, this application obtains an obstacle avoidance duration that is more consistent with the actual lateral slip state, making the generation of obstacle avoidance strategy more accurate and timely, and effectively avoiding obstacle avoidance failure caused by inaccurate traditional collision time prediction when the vehicle is lateral slipping.

[0021] As a further improvement, after the vehicle completes obstacle avoidance, the method also includes: determining whether the vehicle is in a lateral slip recovery state; and resuming the operation of the working device when the vehicle is in a lateral slip recovery state.

[0022] Through this technical solution, after obstacle avoidance is completed, this application adds a judgment on the vehicle's lateral slip recovery state, ensuring that the operation is only resumed after the vehicle's motion state is stable. This avoids the secondary risks that may be caused by blindly resuming the operation when the vehicle is still in an unstable state, and improves the overall safety and continuity of the operation.

[0023] For specific situations, determining whether a vehicle is in a lateral slip recovery state includes: determining the vehicle's longitudinal and lateral movement speeds relative to the ground in real time based on ground visual images; determining the vehicle's real-time lateral slip angle and slip angle change value based on the real-time longitudinal and lateral movement speeds; the slip angle change value is the difference between the lateral slip angle at the current moment and the lateral slip angle at the previous moment; if the lateral slip angle is less than a preset slip angle threshold and the slip angle change value is negative within a preset time period, the vehicle is determined to be in a lateral slip recovery state; otherwise, the vehicle is determined not to be in a lateral slip recovery state.

[0024] This technical solution enables the application to accurately determine whether a vehicle has recovered from a lateral slip state by real-time monitoring of the lateral slip angle and its changing trend. This provides a reliable basis for the recovery of the working device, avoids premature or late recovery of operations, and thus improves the efficiency and safety of operations.

[0025] Secondly, this application also discloses a vehicle autonomous control system, comprising: an acquisition device and a processing device; the acquisition device is used to acquire vehicle motion information when the vehicle's working device is operating, the motion information including the vehicle's satellite positioning information and wheel speed information; the processing device is used to determine whether the vehicle is in a lateral slip motion state based on the motion information; the processing device is used to stop the working device from operating when the vehicle is in a lateral slip motion state, and acquire ground visual images of the vehicle's environment and the vehicle's inertial navigation data; the processing device is used to determine the vehicle's motion positioning information relative to the ground based on the ground visual images; the motion positioning information includes longitudinal movement speed, lateral movement speed, and positioning information; the processing device is used to generate and execute obstacle avoidance control commands to avoid obstacles based on the motion positioning information when an obstacle is detected.

[0026] Beneficial effects This application discloses a vehicle autonomous control method. When the vehicle's working device is operating, the method acquires the vehicle's motion information, including satellite positioning information and wheel speed information. Based on this motion information, it determines whether the vehicle is in a lateral slip motion state. If the vehicle is in a lateral slip motion state, the working device stops operating, and ground visual images of the vehicle's environment and the vehicle's inertial navigation data are acquired. Based on the ground visual images, the method determines the vehicle's motion positioning information relative to the ground, including longitudinal movement speed, lateral movement speed, and positioning information. When an obstacle is detected, an obstacle avoidance control command is generated based on the motion positioning information and the vehicle executes the command to avoid the obstacle.

[0027] Through the above technical solution, this application effectively solves the problem in the prior art where agricultural autonomous vehicles in complex ground environments fail due to abnormal motion states such as slippage and sideslip, leading to command conflicts between control objectives such as path tracking, attitude stabilization, and obstacle avoidance, ultimately affecting operational safety and efficiency. Specifically, this application accurately determines whether the vehicle is in a lateral slip state by acquiring the vehicle's motion information in real time and combining it with satellite positioning and wheel speed information, overcoming the misjudgment that may be caused by information from a single sensor. Once lateral slippage is detected, the system can stop the operating device in time, avoiding the dangers that may be caused by continuing to operate in an unstable state. More importantly, in the lateral slippage state, this application innovatively uses ground visual images and inertial navigation data to determine the vehicle's real movement and positioning information relative to the ground, including longitudinal movement speed, lateral movement speed, and positioning information. This effectively compensates for the deficiency of traditional satellite positioning and inertial navigation systems in terms of positioning accuracy when the vehicle slips, providing the vehicle with a reliable positioning basis that matches the actual ground movement. Based on this, when an obstacle is detected, the system can generate and execute obstacle avoidance control commands based on this accurate motion positioning information, thereby ensuring that the vehicle can effectively avoid risks even in abnormal motion states.

[0028] In summary, this application significantly improves the vehicle's environmental perception and autonomous decision-making capabilities under complex ground conditions such as low adhesion by introducing a lateral slip judgment mechanism based on multi-source information fusion and a precise motion positioning method based on ground vision. It effectively solves the problem that traditional control systems cannot accurately predict the trajectory and effectively avoid obstacles when the vehicle slips, thereby greatly improving the safety and reliability of agricultural autonomous vehicles in actual operations and avoiding property damage and work interruption caused by collisions with obstacles. It has significant technological progress and practical value. Attached Figure Description

[0029] Figure 1 A flowchart illustrating a vehicle autonomous control method provided in this application; Figure 2 A flowchart illustrating a vehicle autonomous control method provided in this application; Figure 3 A flowchart illustrating a vehicle autonomous control method provided in this application; Figure 4 This is a schematic diagram of the architecture of a vehicle autonomous control system provided in this application. Detailed Implementation

[0030] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0031] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0032] In modern smart agricultural and livestock production, autonomous agricultural vehicles often face complex and variable terrain environments, such as muddy, sandy, or icy roads, when performing critical operations like sowing and fertilizing. Traditional control methods struggle to accurately determine the vehicle's true motion state, leading to conflicting commands between control objectives such as path tracking, attitude stabilization, and obstacle avoidance, severely impacting operational safety and efficiency. For example, when a vehicle experiences severe skidding and sideslip in muddy areas, the control system may misjudge the vehicle's state due to inconsistent sensor information, causing conflicting path planning and emergency obstacle avoidance commands, or even causing the vehicle to collide with obstacles in more unpredictable ways.

[0033] In this regard, such as Figure 1 As shown, this application proposes a vehicle autonomous control method, including: S101. When the working device of the vehicle is operating, acquire the vehicle's motion information, which includes the vehicle's satellite positioning information and wheel speed information.

[0034] S102. Determine whether the vehicle is in a lateral slip motion state based on the motion information.

[0035] S103. When the vehicle is in a lateral sliding motion state, stop the operation of the working device and acquire ground visual images of the environment where the vehicle is located and the vehicle's inertial navigation data.

[0036] S104. Determine the vehicle's motion and positioning information relative to the ground based on the ground visual image; the motion and positioning information includes longitudinal movement speed, lateral movement speed, and positioning information.

[0037] S105. When an obstacle is detected, generate and execute an obstacle avoidance control command to avoid the obstacle based on the mobile positioning information.

[0038] This application aims to provide a vehicle autonomous control method to address the problem that traditional control systems cannot accurately determine the vehicle's motion state and effectively avoid obstacles when the vehicle is laterally slipping in complex terrain environments. This method comprehensively analyzes the vehicle's motion information to accurately determine whether the vehicle is in a lateral slipping state. Once lateral slipping is detected, the operation of the working device is immediately stopped, and more accurate vehicle movement and positioning information relative to the ground is obtained using ground visual images and inertial navigation data. Therefore, when an obstacle is detected, effective obstacle avoidance control commands can be generated and executed based on this precise positioning information, significantly improving the vehicle's operational safety and control accuracy under abnormal conditions.

[0039] In order to better understand the technical solutions proposed in this application, it is necessary to explain some key terms involved therein.

[0040] "Motion information" refers to comprehensive data describing the current motion state of a vehicle, which may include the vehicle's satellite positioning information and wheel speed information. Among them, "satellite positioning information" is usually provided by a Global Navigation Satellite System (GNSS) receiver to determine the vehicle's absolute geographical location, speed, and direction; "wheel speed information" is acquired by sensors installed on the wheels, reflecting the rotational speed of the wheels, and thus the theoretical driving speed of the vehicle can be calculated.

[0041] "Lateral slip motion" refers to a phenomenon where, during vehicle operation, there is a significant angle between the vehicle's actual direction of travel and the direction its front is pointing, causing the vehicle to drift laterally. This condition typically occurs when there is insufficient traction on the ground or when the vehicle is subjected to lateral forces.

[0042] "Ground visual images" refer to real-time image data of the ground in front of or around a vehicle, obtained through visual sensors such as vehicle-mounted cameras, used to identify ground features, obstacles, and assist in vehicle positioning.

[0043] "Inertial navigation data" refers to data acquired by an inertial navigation system (INS), including information such as the vehicle's angular velocity and acceleration, which is used to independently calculate the vehicle's attitude, velocity, and position changes.

[0044] "Mobile positioning information" refers to the precise motion data of a vehicle relative to the ground, which can include longitudinal movement speed, lateral movement speed, and positioning information. Among them, "longitudinal movement speed" represents the speed of the vehicle along its direction of travel; "lateral movement speed" represents the lateral speed of the vehicle perpendicular to its direction of travel; and "positioning information" refers to the precise position of the vehicle in a specific coordinate system.

[0045] "Obstacle avoidance control commands" refer to a series of operational commands generated and executed by the vehicle control system to avoid collisions with obstacles, which may include steering, braking, acceleration, or deceleration.

[0046] This application proposes a vehicle autonomous control method, the core of which lies in the accurate perception of the vehicle's motion state and intelligent obstacle avoidance in abnormal conditions.

[0047] When the vehicle's work equipment is operating, it is necessary to acquire the vehicle's motion information. Motion information can include the vehicle's satellite positioning information and wheel speed information. For example, the vehicle's latitude, longitude, altitude, speed, and heading can be acquired in real time using an onboard GNSS receiver. Simultaneously, wheel speed sensors installed on each wheel can continuously monitor the wheel's rotational speed and transmit this information to the vehicle's central control unit. Alternatively, point cloud data of the vehicle's surrounding environment can be acquired using onboard radar or lidar systems, and combined with the vehicle's odometer data to calculate the vehicle's motion information.

[0048] Based on the acquired motion information, it is necessary to determine whether the vehicle is in a lateral slip state. For example, this can be determined by comparing the difference between the vehicle's actual speed determined by satellite positioning information and the theoretical speed calculated from wheel speed information. When the difference exceeds a preset threshold, the vehicle can be considered to be in a lateral slip state. As another approach, it is also possible to analyze the vehicle's inertial measurement unit (IMU) data, such as lateral acceleration and angular velocity, and combine this with the vehicle's kinematic model to evaluate the vehicle's lateral slip angle in real time, thereby determining whether lateral slip has occurred.

[0049] When a vehicle is determined to be in a lateral slip motion, the work equipment needs to be stopped, and ground visual images of the vehicle's surroundings and the vehicle's inertial navigation data need to be acquired. For example, once lateral slip is detected, the vehicle's control system will immediately send a stop command to the work equipment to ensure that it ceases operation. Simultaneously, the onboard camera will begin capturing real-time visual images of the ground around the vehicle and transmit them to the image processing module. Furthermore, the vehicle's inertial navigation system will continuously output high-precision inertial navigation data, including the vehicle's attitude, angular velocity, and acceleration. Alternatively, distance information about the vehicle's surroundings can be obtained using an onboard ultrasonic sensor array, combined with the vehicle's attitude sensor data, to assist in determining the vehicle's lateral slip state and, based on this, stop the work equipment operation.

[0050] Based on the acquired ground visual images, it is necessary to determine the vehicle's motion and positioning information relative to the ground. This motion and positioning information can include longitudinal velocity, lateral velocity, and positioning information. For example, visual odometry (VO) technology can be used to analyze the motion trajectories of feature points in consecutive frames of ground visual images, thereby calculating the vehicle's longitudinal and lateral displacements in the image coordinate system, and then converting these into the vehicle's longitudinal and lateral velocities relative to the ground. Simultaneously, fusing this information with inertial navigation data can yield more accurate vehicle positioning information. Alternatively, deep learning-based semantic segmentation techniques can be used to identify usable ground features (such as lane lines and curbs) from the ground visual images, and the relative positional relationship between these features and the vehicle can be combined to determine the vehicle's motion and positioning information.

[0051] When an obstacle is detected, the system needs to generate and execute obstacle avoidance control commands based on the vehicle's motion positioning information. For example, when the vehicle's radar or lidar detects an obstacle ahead, the control system combines the current precise motion positioning information (including longitudinal movement speed, lateral movement speed, and positioning information) with the relative position and speed information between the obstacle and the vehicle. Through a preset obstacle avoidance strategy model, it generates a series of obstacle avoidance control commands, such as adjusting the steering angle, decelerating, or braking, to guide the vehicle to safely avoid the obstacle.

[0052] As another implementation method, an obstacle avoidance algorithm based on reinforcement learning can be used. When an obstacle is detected, real-time motion positioning information is used as input, and the obstacle avoidance control command is directly output through a trained policy network, so that the vehicle can flexibly avoid the obstacle.

[0053] The vehicle autonomous control method proposed in this application works by constructing a perception and decision-making system with high robustness to abnormal vehicle motion states. When the vehicle's operating device is in operation, the system continuously acquires vehicle motion information, including satellite positioning information and wheel speed information. This information is used to initially determine whether the vehicle is in a lateral slip state. For example, by comparing the difference between the actual vehicle speed reflected by the satellite positioning information and the theoretical speed calculated from the wheel speed information, it is possible to effectively identify whether the vehicle has laterally slipped due to skidding.

[0054] Once the system determines that the vehicle is in a lateral slip state, it indicates that the vehicle may be in an unstable and abnormal operating condition that is difficult to control precisely using traditional methods. At this point, to avoid decreased operational accuracy or potential hazards caused by the slippage, the system will immediately stop the operation of the working device. Simultaneously, to obtain more accurate vehicle motion information, the system will activate the onboard camera to acquire ground visual images of the vehicle's environment and simultaneously acquire the vehicle's inertial navigation data. This is because, in a lateral slip state, traditional satellite positioning and wheel speed information may be distorted due to signal drift or wheel spin, while ground visual images and inertial navigation data can provide more direct and reliable information about the vehicle's motion relative to the ground.

[0055] Next, the system uses the acquired ground visual images, combined with inertial navigation data, to determine the vehicle's movement and positioning information relative to the ground. This information includes longitudinal movement speed, lateral movement speed, and precise positioning information. Through technologies such as visual odometry, the actual displacement and velocity of the vehicle can be extracted from continuous ground visual images and fused with inertial navigation data, thereby overcoming the limitations of a single sensor under sideslip conditions and obtaining a high-precision picture of the vehicle's true motion.

[0056] Ultimately, when the system detects an obstacle while the vehicle is in a lateral slip state, it uses this precise motion and positioning information to generate and execute obstacle avoidance control commands to prevent the vehicle from colliding with the obstacle. Since the system now possesses the vehicle's actual longitudinal and lateral speeds and precise position, it can more accurately predict the vehicle's trajectory and formulate effective obstacle avoidance strategies accordingly, such as adjusting the steering angle, applying emergency braking, or accelerating to avoid a collision. This entire process forms a closed-loop control system, ensuring the vehicle maintains safe operation even under abnormal conditions.

[0057] The vehicle autonomous control method proposed in this application has significant innovation and advantages compared to existing technologies. Traditional methods often rely on single or limited sensor information to make judgments when vehicles encounter complex terrain such as mud or sand and experience lateral slippage. This is prone to misjudgment, leading to conflicting control commands and even exacerbating the risk of loss of vehicle control. For example, existing technologies may only use satellite positioning information to determine if the vehicle has deviated from its path and attempt to correct it by increasing driving torque, but this may actually exacerbate slippage on low-traction surfaces.

[0058] The core innovation of this application lies in its establishment of a multi-source information fusion-based lateral slip determination mechanism, which immediately switches to a more robust positioning and obstacle avoidance strategy upon confirmation of lateral slip. Specifically, this method first determines whether the vehicle is in a lateral slip state by comprehensively analyzing satellite positioning information and wheel speed information, which is more accurate than relying on a single type of information. Once lateral slip is confirmed, the system immediately stops the operation of the working device to avoid potential losses caused by operating in an unstable state. More importantly, the system actively acquires ground visual images and inertial navigation data at this time, and uses this information to determine the vehicle's precise movement and positioning information relative to the ground, including longitudinal movement speed, lateral movement speed, and positioning information. This strategy effectively solves the problem that traditional methods cannot accurately obtain the vehicle's true motion state during sideslip due to sensor information distortion.

[0059] By acquiring precise motion positioning information, this application can generate and execute obstacle avoidance control commands based on this high-precision real-time data when an obstacle is detected. This contrasts sharply with the potential conflicts between path planning and emergency obstacle avoidance commands in existing technologies. This method ensures effective obstacle avoidance even when the vehicle is in abnormal motion, significantly improving the operational safety and control accuracy of the vehicle under complex conditions. Therefore, this application not only solves the challenges of perception and decision-making in lateral slip conditions using traditional methods, but also provides a more intelligent and safer autonomous vehicle control solution.

[0060] like Figure 2 As shown, this application further proposes a step for determining whether a vehicle is in a lateral slip motion state based on motion information, which includes: S201. Determine the vehicle's initial speed based on satellite positioning information.

[0061] S202. Calculate the vehicle's second speed based on the wheel speed information.

[0062] S203. When the absolute value of the difference between the first vehicle speed and the second vehicle speed is greater than the preset vehicle speed difference threshold, the vehicle is determined to be in a lateral slip motion state; otherwise, the vehicle is determined not to be in a lateral slip motion state.

[0063] Specifically, the first vehicle speed refers to the vehicle's actual ground speed determined by its satellite positioning information. Satellite positioning information can be provided by the Global Positioning System (GPS), BeiDou Navigation Satellite System (BDS), or other satellite navigation systems, reflecting the vehicle's true speed on the ground. The second vehicle speed refers to the vehicle speed calculated from wheel speed information. Wheel speed information is typically acquired by wheel speed sensors installed on the wheels, reflecting the wheel's rotational speed. The theoretical vehicle speed can be calculated using parameters such as wheel radius. The preset speed difference threshold is a pre-set value used to measure the degree of difference between the first and second vehicle speeds. This threshold can be calibrated and adjusted based on factors such as vehicle type, operating environment, and road conditions.

[0064] The solution proposed in this application determines the lateral slip motion state by comparing the vehicle's actual ground speed (first vehicle speed) with the theoretical wheel speed (second vehicle speed). When a vehicle experiences lateral slip, a difference arises between the wheel rotation speed and the vehicle's actual speed on the ground. For example, on a slippery surface or during a sharp turn, the vehicle may slide laterally. In this case, the wheels may still be rotating normally or at a lower speed, but the vehicle's actual ground speed (especially the lateral velocity component) will deviate significantly from the speed reflected by the wheel rotation speed. It is precisely because of this difference that when the absolute value of the difference between the first and second vehicle speeds exceeds a preset speed difference threshold, it can be accurately determined that the vehicle is in a lateral slip motion state.

[0065] The above technical solution provides a method for determining lateral slip motion state based on multi-source motion information fusion, effectively improving the accuracy and reliability of lateral slip detection. Compared to methods relying on a single sensor or empirical judgment, this solution can more sensitively capture abnormal vehicle motion states, thus providing timely and accurate decision-making basis for subsequent work device stopping and obstacle avoidance control, significantly enhancing the safety and robustness of the vehicle's autonomous control system.

[0066] This application proposes a method for improving the accuracy of first vehicle speed determination by combining multiple data sources.

[0067] The vehicle's initial speed is determined based on the aforementioned satellite positioning information, specifically including: Acquire the vehicle's inertial navigation data; calculate the vehicle's third speed based on the inertial navigation data; calculate the vehicle's fourth speed based on satellite positioning information; and use the weighted sum of the third and fourth speeds as the vehicle's first speed.

[0068] Specifically, inertial navigation data refers to vehicle motion parameters acquired by the inertial navigation system (INS) or inertial measurement unit (IMU) installed on the vehicle, such as the vehicle's angular velocity and linear acceleration. Inertial navigation data has a high update frequency and good short-term accuracy, but its errors will drift over time.

[0069] The calculation of the vehicle's third speed based on inertial navigation data involves integrating and solving the inertial navigation data using an inertial navigation algorithm to obtain the vehicle's instantaneous speed over a short period of time. This third speed can reflect the vehicle's motion state in real time, but its long-term accuracy is limited by the drift characteristics of the inertial navigation system.

[0070] In practical applications, calculating a vehicle's fourth speed based on satellite positioning information involves performing differential or filtering processing on continuous satellite positioning points to calculate the vehicle's average or instantaneous speed over a longer time scale. This fourth speed exhibits good long-term stability, but it may contain instantaneous errors when satellite signals are unstable or the update frequency is low.

[0071] Satellite positioning information is crucial data contained in the radio signals that navigation satellites continuously broadcast back to Earth. A complete satellite positioning system typically consists of three parts: Space segment (navigation satellites): Satellites continuously transmit radio signals containing specific information. This information primarily includes the satellite's position (ephemeris) and the signal transmission time. The satellite's position is its precise spatial coordinates at a given moment. The signal transmission time is the precise time indicated by a high-precision atomic clock on the satellite.

[0072] Ground control section: The main control stations and monitoring stations distributed on the ground are responsible for monitoring the satellite's operating status, calculating the satellite's orbit and clock errors, and periodically "injecting" this correction information into the satellite to ensure the accuracy of its broadcast information.

[0073] The user unit (receiver): This is the navigation chip in the vehicle. It receives satellite signals, decodes the location and time information from them, and calculates its own real-time location, speed, and time using built-in algorithms.

[0074] The receiver primarily determines the vehicle's speed using the following two methods: 1. Velocity calculation based on position change The receiver can continuously calculate its position at extremely short time intervals (such as once per second). By calculating the change in position per unit time, the speed of movement can be determined.

[0075] 2. Velocity measurement based on Doppler frequency shift The radio signals transmitted by navigation satellites have a fixed frequency. However, because both the satellite and the receiver are moving at high speeds (the satellite is especially fast), the actual frequency of the signal received by the receiver will change slightly: the frequency increases when the satellite is closer and decreases when the satellite is farther away. This change in frequency is called the Doppler shift, and its magnitude accurately reflects the radial velocity (i.e., the velocity component along the line connecting the two) of the relative motion between the satellite and the receiver.

[0076] The receiver simultaneously measures the Doppler shift from at least four satellites. By combining the known velocity and position of each satellite, it can accurately calculate its own velocity in three-dimensional space, including direction and magnitude, through a complex set of equations.

[0077] As a preferred implementation, the weighted sum of the third and fourth vehicle speeds is used as the vehicle's first speed. This aims to leverage the advantages of combining inertial navigation data and satellite positioning information. By assigning different weights to the third and fourth speeds, the weights can be dynamically adjusted based on actual conditions (such as satellite signal quality and inertial navigation system accuracy) to obtain a first speed that combines short-term accuracy with long-term stability. For example, when the satellite signal is good, the weight of the fourth speed can be appropriately increased; when the satellite signal is poor, the weight of the third speed can be increased.

[0078] This application's solution determines the vehicle's initial speed by fusing inertial navigation data and satellite positioning information, effectively addressing the accuracy and reliability issues that may arise with relying solely on satellite positioning information in complex environments. Specifically, inertial navigation data provides high-frequency, high-precision short-term motion information, compensating for potential jumps or delays in satellite positioning information over short periods; while satellite positioning information provides long-term, stable absolute position and velocity information, correcting the inherent defects of long-term integral drift in inertial navigation systems. By weighted and fused, the two technologies complement each other, resulting in a determined initial speed with higher accuracy and robustness under various operating conditions.

[0079] Through the above technical solution, the accuracy and reliability of the determined first vehicle speed are significantly improved. This not only more accurately reflects the actual motion state of the vehicle, but also provides a more reliable input for subsequent judgments on whether the vehicle is in a lateral slip motion state, thereby improving the accuracy of lateral slip motion state judgment, avoiding misjudgments or omissions caused by inaccurate speed judgment, and further enhancing the safety and operational efficiency of the vehicle's autonomous control system.

[0080] Specifically, determining the vehicle's movement and positioning information relative to the ground based on ground visual images can include the following steps: A visual landmark is identified from a ground visual image; the longitudinal and lateral displacement pixel distances of the visual landmark in two adjacent ground visual images are determined; the longitudinal and lateral displacement pixel distances are converted to obtain the longitudinal and lateral displacement distances, respectively; the longitudinal speed of the vehicle relative to the ground is determined based on the longitudinal displacement distance and the time interval between two adjacent ground visual images; the lateral speed of the vehicle relative to the ground is determined based on the lateral displacement distance and the time interval between two adjacent ground visual images; the longitudinal speed, lateral speed, and satellite positioning information of the vehicle before it entered the abnormal motion state are input into a preset positioning model to obtain the vehicle's motion positioning information; the preset positioning model is used to determine the vehicle's position information.

[0081] Specifically, visual landmarks refer to feature points in ground visual images that possess unique texture, color, or geometric shape, such as corner points, edge features, or high-gradient regions. These visual landmarks can be identified and extracted using various feature extraction algorithms, such as Scale Invariant Feature Transform (SIFT), Speed-Up Robust Feature Transform (SURF), or Oriented Fast and Rotationally Shortened (ORB). After identifying visual landmarks, techniques such as feature matching or optical flow can be used to track the positional changes of these landmarks in two consecutive frames of ground visual images, thereby determining their vertical and horizontal displacement pixel distances in the image coordinate system.

[0082] The conversion between longitudinal and lateral displacement pixel distances typically requires using camera intrinsic parameters (such as focal length and principal point coordinates) and information about the camera's relative height to the ground. Using perspective projection, pixel displacements on the image plane can be converted into longitudinal and lateral displacement distances in the actual physical world. This conversion process ensures that the motion information obtained from the image data has real physical meaning. Once the actual longitudinal and lateral displacement distances are obtained, combined with the time interval between two adjacent frames of ground visual images, the vehicle's longitudinal and lateral speeds relative to the ground can be calculated.

[0083] In practical applications, the preset positioning model can be a fusion positioning algorithm, such as an extended Kalman filter (EKF), an unscented Kalman filter (UKF), or a particle filter. This model fuses the vehicle's longitudinal and lateral velocities relative to the ground, obtained through visual odometry, with satellite positioning information prior to the vehicle entering an abnormal motion state (i.e., lateral slippage). The satellite positioning information prior to the abnormal motion state can serve as an initial state or reference to correct for potential accumulated errors in visual odometry and provide a global positioning reference. This fusion yields more robust and accurate vehicle motion positioning information, including the vehicle's real-time position, velocity, and attitude.

[0084] This application's solution directly perceives the relative motion between the vehicle and the ground using ground-based visual images. Even when the vehicle is laterally slipping, it can accurately obtain the vehicle's longitudinal and lateral speeds relative to the ground. Traditionally, when a vehicle laterally slips, motion estimation based on wheel speed information and satellite positioning information can have significant errors because the wheel rotation speed no longer fully represents the vehicle's actual forward speed, and satellite positioning information may lose accuracy during rapid attitude changes or when signal is limited. By introducing ground-based visual images for motion estimation, these shortcomings can be effectively compensated for, providing high-precision relative motion information. Furthermore, fusing this visual motion information with satellite positioning information prior to the lateral slip ensures that the accuracy of local motion estimation is improved while maintaining global positioning consistency.

[0085] The above technical solution overcomes the limitations of traditional sensors in abnormal motion states when a vehicle is in a lateral slip state, accurately determining the vehicle's position relative to the ground. This significantly improves the vehicle's positioning accuracy and robustness under complex conditions, especially when lateral slip occurs. Consequently, it provides reliable and high-precision input data for the generation of subsequent obstacle avoidance control commands, effectively enhancing the safety and reliability of the vehicle's autonomous control system and avoiding potential collision risks caused by inaccurate positioning.

[0086] like Figure 3 As shown, this application further proposes a method for generating and executing obstacle avoidance control commands for a vehicle to avoid obstacles based on motion positioning information when an obstacle is detected, which includes: S301. Determine the controlled adjustment coefficient of the vehicle based on the longitudinal and lateral movement speeds in the mobile positioning information.

[0087] S302. Generate and execute obstacle avoidance control commands for the vehicle based on the controlled adjustment coefficient, motion positioning information, and the position information between the obstacle and the vehicle.

[0088] Specifically, mobile positioning information refers to the vehicle's motion state information relative to the ground, including longitudinal velocity, lateral velocity, and positioning information. Longitudinal velocity represents the vehicle's velocity component along its direction of travel, lateral velocity represents the velocity component perpendicular to its direction of travel, and positioning information refers to the vehicle's position in a specific coordinate system. This information can be obtained by processing ground visual images, for example, through visual odometry technology, or by fusion calculations combining inertial navigation data.

[0089] The controlled adjustment coefficient is a parameter used to correct obstacle avoidance control commands. Its value is typically less than 1, and it's used to make appropriate conservative adjustments to the obstacle avoidance strategy when the vehicle is in a lateral slip state. This coefficient can be determined based on the vehicle's longitudinal and lateral speeds, as these two speed components reflect the degree of lateral slip. For example, the greater the lateral slip, the smaller the controlled adjustment coefficient can be set to reduce the aggressiveness of the obstacle avoidance maneuver, thereby improving the stability of the obstacle avoidance process.

[0090] In practical applications, the positional information between an obstacle and a vehicle refers to the relative position and distance of the obstacle to the vehicle. This information can be perceived and calculated using sensors on the vehicle, such as radar, lidar, or vision sensors. This information forms the basis for generating obstacle avoidance control commands, which are used to plan the vehicle's obstacle avoidance path and actions.

[0091] The solution proposed in this application effectively addresses the issues of insufficient accuracy and safety in obstacle avoidance control command generation during vehicle lateral slippage by introducing a controlled adjustment coefficient. Specifically, when a vehicle is in a lateral slippage state, the adhesion characteristics between its tires and the ground change, affecting the vehicle's handling. If conventional obstacle avoidance control commands are executed under these conditions, obstacle avoidance failure or secondary accidents may occur due to delayed vehicle response or oversteering.

[0092] By determining controlled adjustment coefficients based on longitudinal and lateral movement speeds, the degree of lateral slip of the vehicle can be quantified. These controlled adjustment coefficients are then used to refine the generation process of obstacle avoidance control commands. For example, the desired obstacle avoidance speed can be reduced, the radius of curvature of the obstacle avoidance path increased, or the obstacle avoidance response time extended, thereby making the generated obstacle avoidance control commands more consistent with the vehicle's actual motion capabilities and stability requirements under lateral slip conditions. It is precisely this thorough consideration of the vehicle's dynamic characteristics that makes the obstacle avoidance process safer and more reliable.

[0093] Through the above technical solution, this application can generate safer and more precise obstacle avoidance control commands when the vehicle is in a lateral slip motion state and detects an obstacle. Specifically, by introducing and determining controlled adjustment coefficients based on the vehicle's longitudinal and lateral movement speeds, the degree of lateral slip can be quantitatively assessed, and the obstacle avoidance strategy can be adaptively adjusted accordingly. This avoids executing overly aggressive or inappropriate obstacle avoidance actions when the vehicle is dynamically unstable, significantly improving the success rate and safety of the obstacle avoidance process, effectively reducing the risk of collision when the vehicle is in a lateral slip state, thereby improving the overall reliability of the vehicle's autonomous control system.

[0094] This application further proposes a method for determining the controlled adjustment coefficient of a vehicle based on the longitudinal and lateral movement speeds in the mobile positioning information, specifically including: The lateral slip angle of the vehicle is determined based on the longitudinal and lateral movement speeds; a first preset correspondence is obtained; the first preset correspondence includes a one-to-one correspondence between multiple lateral slip angle ranges and multiple controlled adjustment coefficients; the controlled adjustment coefficient is less than 1, and the controlled adjustment coefficient is positively correlated with the maximum value in the corresponding lateral slip angle range; the controlled adjustment coefficient corresponding to the lateral slip angle range in the first preset correspondence is used as the controlled adjustment coefficient of the vehicle.

[0095] Specifically, the lateral slip angle refers to the angle between the vehicle's actual direction of movement and its longitudinal axis. It directly reflects whether the vehicle is prone to lateral drift or instability during movement. This lateral slip angle can be calculated based on the vehicle's longitudinal and lateral speeds; for example, it can be obtained using the arctangent function of the ratio of lateral speed to longitudinal speed. The first preset correspondence can be understood as a pre-established mapping table or functional relationship. This correspondence divides the possible lateral slip angles of the vehicle into multiple discrete ranges, and presets a corresponding controlled adjustment coefficient for each range.

[0096] The controlled adjustment coefficient is a value less than 1. Its purpose is to adjust the obstacle avoidance strategy or duration during obstacle avoidance maneuvers to adapt to the vehicle's current lateral slip state. In practical applications, the controlled adjustment coefficient is positively correlated with the maximum value within the corresponding lateral slip angle range. This means that the larger the lateral slip angle, the higher the vehicle's lateral instability, and the larger the required controlled adjustment coefficient. This results in a more conservative obstacle avoidance strategy or a larger adjustment range to ensure obstacle avoidance safety.

[0097] In practical applications, after determining the lateral slip angle of the vehicle, the system will query or calculate a specific range in the first preset correspondence where the lateral slip angle is located, and use the controlled adjustment coefficient corresponding to the range as the controlled adjustment coefficient of the current vehicle.

[0098] The proposed solution uses the vehicle's lateral slip angle as a key criterion for determining the controlled adjustment coefficient, thus more accurately reflecting the vehicle's actual dynamic characteristics during lateral slip motion. The lateral slip angle directly quantifies the vehicle's lateral instability; as the lateral slip angle increases, the vehicle's handling decreases, and the space and time required for obstacle avoidance also increase accordingly.

[0099] By acquiring a first preset correspondence and selecting or calculating the corresponding controlled adjustment coefficient based on the vehicle's real-time lateral slip angle, this coefficient can dynamically adapt to the degree of lateral slip. For example, when the lateral slip angle is small, the controlled adjustment coefficient may be close to 1, indicating that the vehicle still has good handling; while when the lateral slip angle is large, the controlled adjustment coefficient will increase accordingly, thereby introducing a larger safety margin or a more conservative adjustment in the subsequent obstacle avoidance command generation to compensate for the decrease in handling caused by lateral slip. Thus, this scheme ensures that the generated obstacle avoidance control commands are more reasonable and safer when the vehicle is in different lateral slip states.

[0100] Through the above technical solution, this application can dynamically and accurately determine the controlled adjustment coefficient based on the actual lateral slip of the vehicle. This allows the generated obstacle avoidance control commands to better adapt to the vehicle's current dynamic characteristics when the vehicle is in a lateral slip motion state, avoiding problems such as poor obstacle avoidance performance or reduced safety due to inaccurate controlled adjustment coefficients. Specifically, by introducing the lateral slip angle and establishing its correspondence with the controlled adjustment coefficient, this application can provide a more refined obstacle avoidance strategy adjustment mechanism, significantly improving the reliability and safety of obstacle avoidance in abnormal motion states, thereby effectively reducing collision risk.

[0101] This application further proposes generating and executing obstacle avoidance control commands for the vehicle to avoid obstacles based on the aforementioned controlled adjustment coefficients, motion positioning information, and position information between the obstacle and the vehicle, including: Based on the mobile positioning information and the position information between the obstacle and the vehicle, determine the duration between the current time and the time of collision between the vehicle and the obstacle when the vehicle does not execute the obstacle avoidance control command; multiply the duration by the controlled adjustment coefficient as the obstacle avoidance duration of the vehicle; input the obstacle avoidance duration, mobile positioning information and the position information between the obstacle and the vehicle into the obstacle avoidance strategy generation model to generate and enable the vehicle to execute the obstacle avoidance control command to avoid the obstacle.

[0102] Specifically, when a vehicle is in a lateral sliding motion and detects an obstacle, it is first necessary to predict the time required for a collision with the obstacle without taking any obstacle avoidance measures, based on the vehicle's motion and positioning information (including longitudinal movement speed, lateral movement speed, and positioning information) and the position information between the obstacle and the vehicle. This time is called the "duration between the current moment and the moment of collision between the vehicle and the obstacle," and it reflects the urgency of the potential collision. For example, this can be predicted using a kinematic model based on the vehicle's current speed and direction, as well as the relative position and speed of the obstacle.

[0103] The aforementioned "controlled adjustment coefficient" is a value less than 1, and its magnitude is positively correlated with the vehicle's lateral slip angle, reflecting the vehicle's handling and stability under lateral slip conditions. The more severe the lateral slip, the smaller the controlled adjustment coefficient, indicating poorer vehicle handling and requiring earlier or more conservative obstacle avoidance. Multiplying the "duration between the current moment and the moment of collision with the obstacle" by the "controlled adjustment coefficient" yields the "obstacle avoidance duration." This obstacle avoidance duration is an adjusted effective obstacle avoidance time window, taking into account the vehicle's actual response capability under the current lateral slip condition.

[0104] In practical applications, the aforementioned "obstacle avoidance strategy generation model" can be a pre-trained machine learning model, a rule-based expert system, or an optimization algorithm. This model takes "obstacle avoidance duration," the vehicle's "motion and positioning information," and "positional information between the obstacle and the vehicle" as input. By integrating this information, the model can generate an optimal obstacle avoidance control command, such as steering angle, braking intensity, and acceleration commands, and instruct the vehicle to execute these commands, thereby safely avoiding the obstacle.

[0105] This application's solution effectively addresses the problem of accurately and timely generating obstacle avoidance commands when a vehicle is in a lateral slip state by introducing the concept of "obstacle avoidance duration" as an important input to the obstacle avoidance strategy generation model. Specifically, when a vehicle is in a lateral slip motion, its handling is affected, and traditional obstacle avoidance strategies may not be able to fully consider this dynamic change.

[0106] By multiplying the "duration between the current moment and the moment of collision between the vehicle and the obstacle" by the "controlled adjustment coefficient," a more conservative "obstacle avoidance duration" that better reflects the vehicle's actual dynamic capabilities can be obtained. When the controlled adjustment coefficient is small (i.e., when lateral slippage is severe), the calculated obstacle avoidance duration will be correspondingly shorter. This prompts the obstacle avoidance strategy generation model to initiate obstacle avoidance actions earlier or generate a smoother and safer obstacle avoidance path, thereby avoiding obstacle avoidance failure or secondary risks caused by limited vehicle handling.

[0107] Through the above technical solution, this application can adaptively adjust the timing and intensity of obstacle avoidance strategy generation based on the actual dynamic state of the vehicle (especially the degree of lateral slip). This makes the generation of obstacle avoidance commands more accurate and reliable, effectively improving the vehicle's obstacle avoidance capability and driving safety under complex conditions (such as lateral slip). This solution not only avoids collisions between the vehicle and obstacles but also maintains the vehicle's stability to the greatest extent possible during obstacle avoidance, reducing the possibility of other dangers caused by improper obstacle avoidance operations.

[0108] This application further proposes a vehicle autonomous control method, which, after the vehicle completes obstacle avoidance, also includes: Determine if the vehicle is in a lateral slip recovery state; if the vehicle is in a lateral slip recovery state, resume operation of the recovery device.

[0109] Specifically, after the vehicle completes the obstacle avoidance maneuver, its motion status needs to be continuously monitored to determine whether it has recovered from the lateral slippage state. Here, "lateral slippage recovery state" means that the vehicle's lateral slippage trend has been effectively controlled, and the vehicle's driving posture has stabilized, allowing for a safe resumption of normal work activities. Once it is determined that the vehicle is indeed in the lateral slippage recovery state, a command to resume work can be sent to the work device, causing it to restart and continue executing the predetermined work task.

[0110] This application's solution effectively addresses the issue of prolonged downtime for the working device in the basic solution by introducing a mechanism to determine the vehicle's lateral slip recovery state after obstacle avoidance. Precise judgment of the vehicle's recovery state allows the working device to resume operation promptly once the vehicle is safely and stably stabilized, avoiding unnecessary work interruptions and efficiency losses. By determining whether the vehicle is in a lateral slip recovery state and resuming operation in that state, the continuity of operation after handling abnormal situations is ensured, improving overall work efficiency.

[0111] Through the above technical solution, this application ensures that the working device can resume operation promptly and safely after the vehicle has undergone lateral slippage and obstacle avoidance operations, thereby significantly improving the continuity of vehicle operations and overall operational efficiency. This solution avoids resource waste and time delays caused by unnecessary work stoppages, enabling the vehicle to complete its work tasks more efficiently and improving the practicality and economic benefits of the automated operating system.

[0112] This application further proposes steps for determining whether a vehicle is in a lateral slip recovery state, including: The vehicle's longitudinal and lateral movement speeds relative to the ground are determined in real time based on ground visual images. The vehicle's real-time lateral slip angle and slip angle change value are determined based on the real-time longitudinal and lateral movement speeds. The slip angle change value is the difference between the lateral slip angle at the current moment and the lateral slip angle at the previous moment. If the lateral slip angle is less than a preset slip angle threshold and the slip angle change value is negative within a preset time period, the vehicle is determined to be in a lateral slip recovery state; otherwise, the vehicle is determined not to be in a lateral slip recovery state.

[0113] Specifically, determining the vehicle's longitudinal and lateral movement speeds relative to the ground in real time based on ground visual images can be achieved using visual odometry technology. For example, by matching and tracking feature points in consecutive frames of ground visual images, the displacement of these feature points on the image plane is calculated. Then, by combining camera parameters and a vehicle motion model, the vehicle's longitudinal and lateral movement speeds in the ground coordinate system can be estimated. This method can provide accurate motion information of the vehicle relative to the ground, and is particularly suitable for scenarios where satellite positioning information may be limited or inaccurate.

[0114] The real-time lateral slip angle of a vehicle can be understood as the angle between the vehicle's actual direction of motion and its longitudinal axis. This lateral slip angle can be calculated based on the real-time determined longitudinal and lateral movement speeds, for example, using the arctangent function arctan(lateral movement speed / longitudinal movement speed). The slip angle change value refers to the difference between the current lateral slip angle and the previous lateral slip angle, and its purpose is to reflect the trend of change in the lateral slip angle.

[0115] In practical applications, the conditions for determining whether a vehicle is in a lateral slip recovery state include two aspects: first, the lateral slip angle is less than a preset slip angle threshold, indicating that the degree of lateral slip has been reduced to an acceptable safe range; second, the change in slip angle within a preset time period is all negative, indicating that the vehicle's lateral slip angle is continuously decreasing, meaning the vehicle is steadily recovering from a lateral slip state. Using these two conditions together allows for a more accurate determination of whether the vehicle has truly stabilized, avoiding misjudgments caused by instantaneous data fluctuations. The preset slip angle threshold and preset time period can be empirically set or calibrated experimentally based on vehicle type, operating environment, and safety requirements.

[0116] This application's solution addresses the aforementioned issue of insufficient accuracy in determining the recovery state of a vehicle's lateral slip by introducing real-time monitoring of the vehicle's lateral slip angle and its changing trend. Specifically, as the vehicle recovers from lateral slip motion, its lateral slip angle gradually decreases. By acquiring real-time ground visual images and determining the vehicle's longitudinal and lateral movement speeds relative to the ground, the vehicle's current lateral slip angle can be accurately calculated.

[0117] Furthermore, by calculating the change in the lateral slip angle, it can be determined whether the lateral slip angle is decreasing or increasing. When the lateral slip angle is not only less than a preset safety threshold, but also continues to decrease over a preset period of time (i.e., the change in slip angle is always negative), this indicates that the vehicle has stably returned to a normal or near-normal driving state, thus reliably confirming that the vehicle is in a lateral slip recovery state. This judgment mechanism, based on a combination of dynamic trends and thresholds, effectively avoids misjudgments caused by fluctuations in a single indicator or instantaneous data errors, ensuring that the timing of the recovery operation is safe and reasonable.

[0118] The above technical solution enables accurate and robust assessment of the vehicle's lateral slip recovery status. This solution not only considers the absolute degree of lateral slip (lateral slip angle less than a preset slip angle threshold), but more importantly, it incorporates the assessment of the lateral slip trend (slip angle changes are all negative within a preset time period). This avoids the risk of resuming operation before the vehicle's lateral slip has fully stabilized, significantly improving operational safety. Simultaneously, the increased accuracy of the assessment also avoids unnecessary downtime for the operating equipment, thereby improving operational efficiency.

[0119] This application proposes a vehicle autonomous control system, comprising: an acquisition device and a processing device; the acquisition device is used to acquire vehicle motion information, including vehicle satellite positioning information and wheel speed information, when the vehicle's working device is operating; the processing device is used to determine whether the vehicle is in a lateral slip motion state based on the motion information; the processing device is used to stop the working device from operating when the vehicle is in a lateral slip motion state, and acquire ground visual images of the vehicle's environment and the vehicle's inertial navigation data; the processing device is used to determine the vehicle's motion positioning information relative to the ground based on the ground visual images; the motion positioning information includes longitudinal movement speed, lateral movement speed, and positioning information; the processing device is used to generate and execute obstacle avoidance control commands to avoid obstacles based on the motion positioning information when an obstacle is detected.

[0120] This system aims to address the problem that traditional control systems cannot accurately determine the vehicle's motion state and effectively avoid obstacles when the vehicle laterally slips in complex terrain environments. By configuring acquisition and processing devices, this system can achieve precise perception of the vehicle's motion state and intelligent obstacle avoidance in abnormal conditions. The acquisition device is responsible for collecting the vehicle's motion information in real time, while the processing device makes intelligent decisions based on this information, including determining lateral slippage, stopping operation, acquiring auxiliary positioning data, and ultimately generating obstacle avoidance control commands. This collaborative mechanism ensures that the vehicle can maintain safe operation under abnormal conditions, significantly improving operational safety and control accuracy.

[0121] Specifically, the acquisition device can be configured to include one or more sensor units for collecting vehicle motion information. For example, the acquisition device can integrate a Global Navigation Satellite System (GNSS) receiver and multiple wheel speed sensors, which are used to acquire the vehicle's satellite positioning information and wheel speed information, respectively. Alternatively, the acquisition device can include a multi-sensor fusion module capable of receiving raw data from different types of sensors (e.g., in addition to GNSS and wheel speed sensors, it can also include lidar or millimeter-wave radar) and performing preliminary data preprocessing to provide more comprehensive motion information.

[0122] The processing unit can be configured to include one or more computing units for performing various processing tasks in the vehicle autonomous control method. For example, the processing unit can be a high-performance embedded controller integrating a central processing unit (CPU), a graphics processing unit (GPU), and memory, capable of running complex algorithm models, such as algorithms for determining lateral slip states, visual odometry, and obstacle avoidance strategy generation models. As a preferred implementation, the processing unit can employ a distributed architecture, where different computing modules (e.g., one for motion state determination, another for visual image processing, and yet another for obstacle avoidance decision-making) are interconnected via a high-speed communication bus to improve the system's parallel processing capabilities and response speed. The processing unit can also be configured to communicate with the vehicle's operating devices, steering system, braking system, and other actuators to achieve precise vehicle control.

[0123] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for autonomous vehicle control, characterized in that, include: When the vehicle's working device is in operation, the vehicle's motion information is acquired, including the vehicle's satellite positioning information and wheel speed information. Determine whether the vehicle is in a lateral slip motion state based on the motion information; When the vehicle is in a lateral slip motion, stop the operation of the working device and acquire ground visual images of the vehicle's environment and the vehicle's inertial navigation data. The vehicle's movement and positioning information relative to the ground is determined based on the ground visual image; the movement and positioning information includes longitudinal movement speed, lateral movement speed, and positioning information; When an obstacle is detected, the vehicle generates and executes obstacle avoidance control commands based on the mobile positioning information to avoid the obstacle.

2. The vehicle autonomous control method according to claim 1, characterized in that, Determining whether the vehicle is in a lateral slip motion state based on the motion information includes: The vehicle's first speed is determined based on the satellite positioning information; Calculate the vehicle's second speed based on the wheel speed information; If the absolute value of the difference between the first vehicle speed and the second vehicle speed is greater than a preset vehicle speed difference threshold, the vehicle is determined to be in a lateral slip motion state; otherwise, the vehicle is determined not to be in a lateral slip motion state.

3. The vehicle autonomous control method according to claim 2, characterized in that, Determining the vehicle's first speed based on the satellite positioning information includes: Acquire the inertial navigation data of the vehicle; The third speed of the vehicle is calculated based on the inertial navigation data; The fourth speed of the vehicle is calculated based on the satellite positioning information; The weighted sum of the third speed and the fourth speed is taken as the first speed of the vehicle.

4. The vehicle autonomous control method according to claim 1, characterized in that, Determining the vehicle's movement and positioning information relative to the ground based on the ground visual image includes: Identify a visual landmark from the ground visual image; Determine the vertical and horizontal displacement pixel distances of visual markers in two adjacent ground visual images; The vertical displacement pixel distance and the horizontal displacement pixel distance are converted respectively to obtain the vertical displacement distance and the horizontal displacement distance; The longitudinal speed of the vehicle relative to the ground is determined based on the longitudinal displacement distance and the time interval between two adjacent frames of ground visual images; The lateral movement speed of the vehicle relative to the ground is determined based on the time interval between two adjacent frames of ground visual images of the lateral displacement distance. The vehicle's longitudinal and lateral movement speeds, as well as its satellite positioning information before it entered the abnormal movement state, are input into a preset positioning model to obtain the vehicle's motion positioning information; the preset positioning model is used to determine the vehicle's location information.

5. The vehicle autonomous control method according to claim 1, characterized in that, Upon detecting an obstacle, the system generates and executes obstacle avoidance control commands based on the vehicle's location information, including: The controlled adjustment coefficient of the vehicle is determined based on the longitudinal and lateral movement speeds in the mobile positioning information; Based on the controlled adjustment coefficient, motion positioning information, and the position information between the obstacle and the vehicle, an obstacle avoidance control command is generated and the vehicle executes to avoid the obstacle.

6. The vehicle autonomous control method according to claim 5, characterized in that, The controlled adjustment coefficients of the vehicle are determined based on the longitudinal and lateral movement speeds in the mobile positioning information, including: The lateral slip angle of the vehicle is determined based on the longitudinal and lateral movement speeds. Obtain a first preset correspondence; the first preset correspondence includes a one-to-one correspondence between multiple lateral slip angle ranges and multiple controlled adjustment coefficients; the controlled adjustment coefficient is less than 1, and the controlled adjustment coefficient is positively correlated with the maximum value in the corresponding lateral slip angle range; The controlled adjustment coefficient corresponding to the lateral slip angle range of the vehicle in the first preset correspondence is taken as the controlled adjustment coefficient of the vehicle.

7. The vehicle autonomous control method according to claim 5, characterized in that, Generate and execute obstacle avoidance control commands for the vehicle to avoid obstacles based on the controlled adjustment coefficient, motion positioning information, and position information between the obstacle and the vehicle, including: The duration between the current moment and the moment of collision between the vehicle and the obstacle is determined based on the mobile positioning information and the position information between the obstacle and the vehicle when the vehicle does not execute the obstacle avoidance control command; The product of the duration and the controlled adjustment coefficient is taken as the obstacle avoidance duration of the vehicle. The obstacle avoidance duration, motion positioning information, and position information between the obstacle and the vehicle are input into the obstacle avoidance strategy generation model to generate and enable the vehicle to execute obstacle avoidance control commands to avoid obstacles.

8. A vehicle autonomous control method according to any one of claims 1-7, characterized in that, After the vehicle completes obstacle avoidance, the methods also include: Determine if the vehicle is in a state of lateral slip recovery; The recovery device operates when the vehicle is in a lateral slip recovery state.

9. A vehicle autonomous control method according to claim 8, characterized in that, Determining whether a vehicle is in a lateral slip recovery state includes: The vehicle's longitudinal and lateral speeds relative to the ground are determined in real time based on ground visual images. The real-time lateral slip angle and slip angle change value of the vehicle are determined based on the real-time longitudinal and lateral movement speeds; the slip angle change value is the difference between the lateral slip angle at the current moment and the lateral slip angle at the previous moment. If the lateral slip angle is less than the preset slip angle threshold and the slip angle change value is negative within the preset time period, the vehicle is determined to be in the lateral slip recovery state; otherwise, the vehicle is determined not to be in the lateral slip recovery state.

10. A vehicle autonomous control system, characterized in that, include: Acquisition device and processing device; The acquisition device is used to acquire the vehicle's motion information when the vehicle's working device is in operation. The motion information includes the vehicle's satellite positioning information and wheel speed information. The processing device is used to determine whether the vehicle is in a lateral slip motion state based on the motion information; The processing device is used to stop the operation of the working device and acquire ground visual images of the vehicle's environment and the vehicle's inertial navigation data when the vehicle is in a lateral sliding motion state. The processing device is used to determine the vehicle's movement and positioning information relative to the ground based on the ground visual image; the movement and positioning information includes longitudinal movement speed, lateral movement speed, and positioning information; the processing device is used to generate and execute obstacle avoidance control commands to avoid obstacles based on the movement and positioning information when an obstacle is detected.

Images