Vehicle curve driving control method and related equipment

Abstract

The invention discloses a vehicle curve driving control method and related equipment, and relates to the technical field of vehicle control, and the method comprises the steps: obtaining the vehicle speed information of a target vehicle, a vehicle positioning point and the curvature radius of a lane where the target vehicle is located; determining a lane center line of a lane where the target vehicle is located based on the curvature radius; based on the vehicle positioning point and the lane center line, determining a transverse deviation distance between a historical planning track of the previous planning period and the lane center line; determining a target planning track of the current planning period based on the transverse deviation distance and the vehicle speed information; and controlling the target vehicle to run on the curve based on the target planning track. According to the method, the lane center line is determined by combining the vehicle speed, the vehicle positioning point and the lane curvature, and the planning track is dynamically optimized based on the historical transverse deviation, so that the vehicle can more accurately, stably and safely run along the corrected target track in the curve.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicle control, and more particularly, to a vehicle curve driving control method and related equipment. BACKGROUND

[0002] With the continuous development of intelligent transportation technology and intelligent driving system, the safe driving control of vehicles in curve scenarios has become one of the important basic capabilities to realize high-order assisted driving. When a vehicle passes through complex road environments such as high-speed curves, ramps, and roads with continuous curvature changes, its driving stability and path following capability directly affect the driving safety and user driving experience.

[0003] In related technologies, the control of vehicles in curve scenarios usually relies on trajectory planning based directly on vehicle positioning points or local path information, but such schemes often fail to fully consider the curvature characteristics of the lane and the dynamic deviation between the vehicle and the lane center line. In particular, for curves with large curvature changes or unstable visual perception, the vehicle trajectory may continuously follow the biased planning results, causing the vehicle to gradually deviate to the outside in the curve, thereby affecting the stability and centering accuracy of curve driving, and resulting in problems such as lateral deviation and slow trajectory convergence during vehicle driving. That is, there is a technical problem of insufficient trajectory planning accuracy and difficulty for the vehicle to stably maintain the lane center during curve driving in related technologies. SUMMARY

[0004] A series of simplified concepts are introduced in the summary part of the present application, which will be further described in detail in the specific embodiment part. The summary part of the present application does not mean to try to limit the key features and necessary technical features of the claimed technical solution, and even less to determine the protection scope of the claimed technical solution.

[0005] The vehicle curve driving control method and related equipment provided by the present application can determine the lane center line by combining the vehicle speed, vehicle positioning point and lane curvature, and dynamically optimize the planned trajectory based on the historical lateral deviation, so that the vehicle can drive more accurately, smoothly and safely along the corrected target trajectory in the curve.

[0006] In a first aspect, the present application provides a vehicle curve driving control method applied to a target vehicle, comprising: obtaining vehicle speed information of the target vehicle, a vehicle positioning point and a curvature radius of a lane where the target vehicle is located; determining a lane center line of the lane where the target vehicle is located based on the curvature radius; determining a lateral deviation distance between a historical planning trajectory of a previous planning period and the lane center line based on the vehicle positioning point and the lane center line; determining a target planning trajectory of a current planning period based on the lateral deviation distance and the vehicle speed information; and controlling the target vehicle to drive on a curve based on the target planning trajectory.

[0007] In some embodiments, the determining of the lateral deviation distance between the historical planning trajectory of the previous planning period and the lane center line based on the vehicle positioning point and the lane center line comprises: vertically projecting the vehicle positioning point onto the historical planning trajectory to obtain a first projection point; vertically projecting the first projection point onto the lane center line to obtain a second projection point; and determining the lateral deviation distance based on a distance between the first projection point and the second projection point.

[0008] In some embodiments, the determining of the target planning trajectory of the current planning period based on the lateral deviation distance and the vehicle speed information comprises: determining a first translation distance corresponding to the current planning period based on the vehicle speed information; determining a second translation distance based on a product of the lateral deviation distance and a first coefficient; determining a target translation distance as a smaller one of the first translation distance and the second translation distance; and determining the target planning trajectory based on the target translation distance and the lateral deviation distance.

[0009] In some embodiments, the determining of the first translation distance corresponding to the current planning period based on the vehicle speed information comprises: determining a third translation distance based on a product of the vehicle speed information and a second coefficient; and determining the first translation distance as a boundary value in a preset distance range that matches the third translation distance, if the third translation distance is within the preset distance range, otherwise.

[0010] In some embodiments, before the determining of the third translation distance based on the product of the vehicle speed information and the second coefficient, the vehicle curve driving control method further comprises: obtaining a calibration coefficient value corresponding to a driving mode of the target vehicle from a preset parameter mapping table as the second coefficient, wherein the calibration mapping table stores calibration coefficient values corresponding to a plurality of driving modes, and the calibration coefficient values are used to calculate the third translation distance.

[0011] In some embodiments, before determining the second translation distance based on the product of the lateral deviation distance and a first coefficient, the vehicle curve driving control method further comprises: determining the first coefficient based on at least one of a lane width of the lane where the target vehicle is located, a road surface adhesion coefficient, a traffic flow density, and the radius of curvature, wherein the first coefficient is inversely proportional to the radius of curvature, the lane width, and the traffic flow density, and the first coefficient is proportional to the road surface adhesion coefficient.

[0012] In some embodiments, determining the target planning trajectory based on the target translation distance and the lateral deviation distance comprises: determining a target planning starting point based on the target translation distance and the lateral deviation distance; and generating a target planning trajectory of a current planning period converging to the lane centerline based on the target planning starting point.

[0013] In some embodiments, determining the target planning starting point based on the target translation distance and the lateral deviation distance comprises: determining a translation ratio based on a ratio of the target translation distance to the lateral deviation distance; and determining the target planning starting point on a line connecting the first projection point and the second projection point based on the translation ratio.

[0014] In some embodiments, generating the target planning trajectory of the current planning period converging to the lane centerline based on the target planning starting point comprises: performing trajectory solving based on the target planning starting point and the lane centerline by an iterative linear quadratic regulator optimization algorithm to generate the target planning trajectory.

[0015] In some embodiments, controlling the target vehicle to drive on the curve based on the target planning trajectory comprises: determining a desired yaw rate based on the target planning trajectory; generating a first steering control instruction based on a first deviation of the desired yaw rate from an actual yaw rate of the target vehicle; sending the first steering control instruction to a steering actuator of the target vehicle to drive the target vehicle to travel along the target planning trajectory; or determining a desired front wheel steering angle based on the target planning trajectory; generating a second steering control instruction based on a second deviation of the desired front wheel steering angle from an actual front wheel steering angle of the target vehicle; and sending the second steering control instruction to the steering actuator of the target vehicle to drive the target vehicle to travel along the target planning trajectory.

[0016] In some embodiments, the determining the lane centerline of the lane where the target vehicle is located based on the radius of curvature comprises: in a case where the radius of curvature is less than a preset radius threshold, performing lane line recognition on the lane where the target vehicle is located to obtain lane line information corresponding to the lane; and determining the lane centerline of the lane where the target vehicle is located based on the lane line information.

[0017] In some embodiments, the lane line information comprises geometric feature information and confidence evaluation information of the lane line; and the determining the lane centerline of the lane where the target vehicle is located based on the lane line information comprises: extracting a first feature point sequence and first curvature information sequence from the geometric feature information; performing interpolation processing on the first feature point sequence and the first curvature information sequence respectively based on the confidence evaluation information to obtain a second feature point sequence and a second curvature information sequence; and determining the lane centerline based on the second feature point sequence and the second curvature information sequence.

[0018] In a second aspect, the present application further provides a vehicle curve driving control device applied to a target vehicle, comprising: an information acquisition unit configured to acquire vehicle speed information of the target vehicle, a vehicle positioning point, and a radius of curvature of a lane where the target vehicle is located; a centerline determination unit configured to determine a lane centerline of the lane where the target vehicle is located based on the radius of curvature; a deviation determination unit configured to determine a lateral deviation distance between a historical planning trajectory of a previous planning period and the lane centerline based on the vehicle positioning point and the lane centerline; a trajectory determination unit configured to determine a target planning trajectory of a current planning period based on the lateral deviation distance and the vehicle speed information; and a vehicle control unit configured to control the target vehicle to drive on a curve based on the target planning trajectory.

[0019] In a third aspect, the present application further provides an electronic device, comprising a memory and a processor, wherein the processor is configured to implement the steps of the vehicle curve driving control method of the first aspect when executing a computer program stored in the memory.

[0020] In a fourth aspect, the present application further provides a computer readable storage medium storing a computer program, wherein the computer program is configured to implement the steps of the vehicle curve driving control method of the first aspect when executed by a processor.

[0021] In a fifth aspect, the present application further provides a computer program product comprising a computer program or computer executable instructions, wherein the computer program or computer executable instructions are configured to implement the steps of the vehicle curve driving control method provided in the embodiments of the present application when executed by a processor.

[0022] In summary, the vehicle curve driving control method provided by the application can accurately identify the geometric characteristics of the curve road section where the vehicle is located by obtaining the vehicle speed information, vehicle positioning point and curvature radius of the lane, and can build subsequent control basis based on this, so that the curve driving control has more accurate environmental understanding ability; the lane center line of the lane where the vehicle is located is determined based on the curvature radius, so that the vehicle has a clear, reliable and consistent reference path with the curve structure during curve driving, which helps to plan a trajectory closely following the actual road trend and improves the accuracy of the vehicle driving along the lane center line; by calculating the lateral deviation distance between the historical planning trajectory of the last planning period and the lane center line, the deviation degree of the vehicle in the curve can be reflected in real time, which provides a quantitative basis for trajectory correction, thereby effectively reducing the lateral error generated by the vehicle in the curve; when determining the target planning trajectory of the current planning period, the lateral deviation distance and the vehicle speed are considered comprehensively, so that the trajectory planning not only corrects the deviation, but also adapts to the current dynamic state of the vehicle, improving the rationality of the planning result and the smoothness of the control; finally, the vehicle is controlled based on the target planning trajectory, so that the vehicle can pass through the curve according to the corrected and optimized trajectory, realizing more stable, continuous and dynamically adaptive curve driving effect, thereby improving the curve safety and the overall control performance of the vehicle. In summary, the vehicle curve driving control method provided by the application determines the lane center line by combining the vehicle speed, vehicle positioning point and lane curvature, and dynamically optimizes the planning trajectory based on the historical lateral deviation, so that the vehicle can drive more accurately, smoothly and safely along the corrected target trajectory in the curve. BRIEF DESCRIPTION OF DRAWINGS

[0023] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not intended to limit the scope of the present description. Moreover, like reference numerals designate like parts throughout the several views in the drawings. In the drawings: Figure 1 A flowchart of a vehicle curve driving control method provided by an embodiment of the application is shown in FIG. 1; Figure 2 A component structure diagram of a vehicle curve driving control device provided by an embodiment of the application is shown in FIG. 2; Figure 3 A component structure diagram of an electronic device provided by an embodiment of the application is shown in FIG. 3. DETAILED DESCRIPTION

[0024] The terms in the specification, claims and drawings of the present application, such as "first", "second", "third", "fourth" and the like (if any), are used to distinguish similar objects, not to describe a particular order or sequence. Therefore, it is understood that these terms can be used interchangeably, as appropriate, to achieve the described embodiments, unless the drawings or description specifically require otherwise. In addition, the terms "is" and "has" and any variants thereof in the present application are intended to cover non-exclusive inclusion of all possible constituent elements. For example, a process, method, system, product or device including several steps or units does not necessarily limit to only the steps or units explicitly listed, but can also include other steps or units not explicitly listed, or steps or units inherent to the process, method, product or device.

[0025] In the present application, "module" or "unit" refers to a computer program or a part of a computer program with a specific function, and works with other related parts to achieve a predetermined target. These modules or units can be implemented by software, hardware (such as processing circuitry or memory) or a combination of both. One or more processors or memories can implement one or more modules or units. At the same time, each module or unit can also be part of a larger module or unit.

[0026] The technical solutions in the present application will be described in detail below in conjunction with the drawings in the embodiments. It should be noted that the described embodiments are only a part of the present application, not all embodiments. In the following description, "some embodiments" mentioned is only a subset of all possible embodiments, which can be the same or different subset, and different embodiments can be combined with each other without conflict.

[0027] Figure 1 is a flowchart of a vehicle curve driving control method provided by an embodiment of the present application. For example, referring to Figure 1 The vehicle curve driving control method provided by the embodiments of the present application applied to a target vehicle can include the following steps 101 to 105. Step 101, obtaining the speed information of the target vehicle, the vehicle positioning point and the curvature radius of the lane where the target vehicle is located.

[0028] In some examples, the target vehicle is a specific vehicle to which the vehicle curve driving control method of the application is applied, and is the main body of trajectory planning and driving control. Its types include but are not limited to family passenger cars, commercial vehicles, new energy vehicles and various types of motor vehicles with auxiliary driving functions. The vehicle speed information is the real-time driving speed of the target vehicle at the current driving moment, which is a key parameter reflecting the dynamic state of the vehicle. The speed data during the driving process of the target vehicle can be collected in real time through the wheel speed sensor, the power control system or the vehicle intelligent driving domain controller, and the accurate vehicle speed information can be obtained after data filtering processing. For example, when the target vehicle drives on the high-speed ramp, the collected vehicle speed information may be 12.5 meters per second; when driving on the curve of the urban expressway, the collected vehicle speed information may be 16 meters per second. The vehicle positioning point is the real-time spatial position coordinate of the target vehicle in the current driving environment, which is used to accurately represent the specific position of the vehicle in the lane. The two-dimensional spatial coordinates (longitude, latitude) based on the geodetic coordinate system can be output as the vehicle positioning point by combining the global positioning system (Global Positioning System, GPS), the Beidou satellite navigation system (Beidou Navigation Satellite System, BDS) and the vehicle inertial measurement unit (Inertial Measurement Unit, IMU) on the target vehicle, and using high-precision map matching technology. For example, when the target vehicle drives on a curve of a certain expressway, the obtained vehicle positioning point may be east longitude 116.39748 degrees and north latitude 39.90882 degrees, and the positioning accuracy can reach centimeter level, which ensures that the position relationship of the vehicle in the lane can be accurately reflected. The curvature radius is a radius parameter corresponding to the bending degree of the curve of the current driving lane of the target vehicle, which is the core basis for judging whether the lane is a large-curvature curve. The front lane line image can be collected in real time by the vehicle-mounted camera on the target vehicle, and the geometric features of the lane line can be extracted by the image recognition algorithm. Then, the least square method or the Bezier curve fitting algorithm is used to fit the curve of the lane line, and the curvature of the lane is calculated according to the fitted curve equation, and then the curvature radius is derived. For example, when the target vehicle drives on the ramp of the high-speed connection section, the curvature radius calculated by fitting may be 180 meters; when driving on the curve of the ordinary urban road, the obtained curvature radius may be 300 meters; and when driving on the straight road, the curvature radius is infinite, so that the straight road and the curve road scenes can be distinguished.

[0029] Exemplarily, when the target vehicle starts the auxiliary driving function and is about to enter or is driving in a curve, the data acquisition process can be continuously triggered according to a preset acquisition frequency. First, the real-time driving state of the vehicle is synchronously acquired by the vehicle-mounted wheel speed sensor and the power control system, and after filtering and denoising processing, accurate vehicle speed information is generated. At the same time, the fusion positioning scheme of the global positioning system, the Beidou satellite navigation system and the inertial measurement unit is used, combined with high-precision map matching correction, to output the two-dimensional spatial coordinates of the vehicle in the current driving environment as the vehicle positioning point. In addition, the front lane line image is captured by the vehicle-mounted camera in real time, and the curvature parameter of the current driving lane is continuously calculated through image recognition and curve fitting algorithm, and then the corresponding curvature radius is obtained.

[0030] Through the implementation of step 101, the vehicle speed information, vehicle positioning point and curvature radius of the lane of the target vehicle are acquired, which can provide complete and key dynamic parameters for subsequent curve driving control, accurately master the current motion state of the vehicle and the geometric characteristics of the curve, and thus lay a reliable data foundation for trajectory planning and vehicle control.

[0031] In step 102, based on the curvature radius, the lane center line of the lane where the target vehicle is located is determined.

[0032] In some examples, the lane center line is the geometric center axis of the current driving lane of the target vehicle, which is a reference line for guiding the vehicle to keep driving in the center of the lane in trajectory planning. Its form is consistent with the bending characteristics of the lane, which is a straight line in a straight lane scene and a smooth curve matching the lane curvature in a curve scene. The function of the lane center line is to provide a clear target reference for vehicle trajectory planning and ensure that the vehicle maintains an equal safety distance from both sides of the lane during driving. The lane center line can be generated by algorithm fitting based on the fusion calculation of the geometric feature data (such as lane line position and curvature parameter) and the lane width parameter of the lane. The lane center line can be generated by geometric fitting and coordinate calculation based on the lane curvature radius obtained in step 101 as the core constraint, combined with the lane line position information and the lane width parameter. For example, when the curvature radius obtained in step 101 is 200 meters (large curvature curve), and the recognized lane width is 3.75 meters, a circular arc-shaped lane center line with the same curvature as the lane line and 1.875 meters away from the lane line on both sides is generated. When the curvature radius is infinite (straight line), a straight line-shaped lane center line parallel to the lane line on both sides is generated.

[0033] For example, first, the real-time lane line image data collected by the vehicle-mounted camera is called to extract the edge coordinate point sequence of the two side lane lines through image recognition algorithm, and the preset lane width standard parameter or the real-time recognized current lane width data is obtained; then, the lane type is determined according to the curvature radius, if the curvature radius is infinite, it is determined as a straight lane, the center position point sequence of the two side lane lines is fitted as a straight line through linear fitting algorithm, as the lane center line of the straight lane scene; if the curvature radius is a finite value, it is determined as a curve, and a smooth curve with the same curvature as the two side lane lines and located at the geometric center is generated as the lane center line of the curve scene by using curve fitting algorithm (such as least square method) with the curvature radius as the constraint condition; the process is updated synchronously with the data collection of step 101, and the lane center line is corrected according to the latest curvature radius and lane line information in each planning period to ensure that it is always accurately matched with the actual lane shape, providing a reliable reference for subsequent calculation of lateral deviation based on vehicle positioning point and planning of target trajectory.

[0034] Through the implementation of step 102, the lane center line is determined based on the curvature radius, so that the vehicle has a clear reference path in the curve scene and conforms to the actual curve shape, which helps to improve the consistency of trajectory planning and actual road geometry, thereby improving the accuracy of the vehicle driving along the lane center line.

[0035] Step 103, based on the vehicle positioning point and the lane center line, the lateral deviation distance between the historical planning trajectory of the last planning period and the lane center line is determined.

[0036] In some examples, the previous planning period is a previous fixed time interval period of the current trajectory planning period, is a time unit of trajectory planning in the auxiliary driving system, is used to distinguish trajectory data generated at different times, and ensures continuity and timing of trajectory planning; the start and end times of the planning period can be recorded by a built-in timing module, and a previous complete period of the current period is automatically identified as the previous planning period according to a preset planning period length (such as 100 milliseconds). For example, if the current planning period is the nth frame (the start time is t), and the preset planning period length is 100 milliseconds, then the previous planning period is the (n-1)th frame (the start time is t-100 milliseconds), and the trajectory planning and data storage of the last time in this period are completed. The historical planning trajectory is a planning path generated based on the vehicle state and lane environment data at that time in the previous planning period, is used to guide the vehicle to travel, and is continuous path data reflecting the trajectory planning result at the last time; the generated target trajectory data (including two-dimensional coordinates of points on the trajectory, curvature and other parameters) can be automatically stored in the data cache module after each planning period ends, and the historical planning trajectory can be obtained by calling the cache data of the previous planning period in the current step; for example, the historical planning trajectory generated in the previous planning period is a smooth curve with a curvature radius of 200 meters and an adaptive speed of 14 meters per second at that time, and the data form is a sequence containing 100 continuous coordinate points (such as longitude 116.39748-116.39850 degrees, latitude 39.90882-39.90920 degrees). The lateral deviation distance is a straight line distance between the historical planning trajectory of the previous planning period and the lane center line in the direction perpendicular to the extension direction of the lane center line, and is a core parameter representing the degree of deviation of the historical planning trajectory from the lane center; the spatial relationship between the historical planning trajectory and the lane center line can be established by the vehicle positioning point, and the actual distance in the lateral direction (perpendicular to the lane direction) can be calculated; for example, when the historical planning trajectory deviates to the outside of the lane, the lateral deviation distance can be 0.4 meters, and the larger the value, the more serious the deviation of the historical trajectory from the lane center line, and the higher the risk of vehicle cutting. If the historical planning trajectory fits the lane center line, the lateral deviation distance can be 0.05 meters. Based on the vehicle positioning point and the lane center line, the process of determining the lateral deviation distance between the historical planning trajectory of the previous planning period and the lane center line is to establish the position relationship between the historical planning trajectory and the lane center line with the current vehicle positioning point as the spatial reference, and to obtain the lateral perpendicular distance between the two by geometric calculation.

[0037] For example, the historical planning trajectory complete data stored in the last planning period can be first called, and the lane center line coordinate sequence determined in step 102 and the current vehicle positioning point data obtained in step 101 are synchronously acquired; then, by means of a spatial coordinate matching algorithm, the current vehicle positioning point is taken as a reference to accurately locate the associated point of the historical planning trajectory, so as to ensure that the associated point can truly reflect the corresponding relationship between the historical trajectory and the current vehicle position; subsequently, a geometric distance calculation method is adopted to solve the perpendicular distance from the associated point to the lane center line, and the distance is the required lateral deviation distance.

[0038] By implementing step 103, the lateral deviation distance between the historical planning trajectory and the lane center line is calculated, which can reflect the deviation degree of the vehicle from the lane center line in the last period in real time, and provide a quantitative basis for subsequent trajectory adjustment, thereby effectively reducing the lateral deviation of the vehicle in the curve.

[0039] In step 104, the target planning trajectory of the current planning period is determined based on the lateral deviation distance and the vehicle speed information.

[0040] In some examples, the current planning period is a fixed time interval period in which trajectory planning is being performed, and is used to generate a driving path that adapts to the current vehicle state and environment, so as to ensure the real-time and dynamic adaptability of trajectory planning; the complete time interval in which trajectory calculation is being performed can be identified as the current planning period by means of a built-in timing module (such as a timer in a vehicle-mounted microcontroller) according to a preset planning period length (consistent with the last planning period, such as 100 milliseconds); for example, if the last planning period is the n-1 frame (starting time t-100 milliseconds and ending time t), then the current planning period is the n frame (starting time t and ending time t+100 milliseconds), and the calculation and output of the target planning trajectory will be completed in this period. The target planning trajectory is the optimal path generated in the current planning period based on the real-time vehicle state (vehicle speed information) and deviation data (lateral deviation distance), and is used to guide the target vehicle to stably travel along the lane center line, and is a direct reference for subsequent vehicle driving. The form of the target planning trajectory needs to take into account the smoothness, safety and lane centering requirements; the target planning trajectory can be generated by taking the lateral deviation distance as the core correction basis (reflecting the degree of deviation of the historical trajectory from the lane center), taking the vehicle speed information as the dynamic constraint condition (reflecting the current driving dynamics of the vehicle), and combining the lane center line reference, so as to calculate the optimal driving path in the current period as the target planning trajectory; for example, the target planning trajectory can be a smooth circular arc curve with a curvature radius of 200 meters and an adaptation to the current vehicle speed of 13 meters per second, and the coordinate sequence is east longitude 116.39760-116.39860 degrees and north latitude 39.90890-39.90930 degrees, and the overall converges to the lane center line, so as to ensure that the vehicle gradually returns to the central position.

[0041] Exemplarily, firstly, the lateral deviation distance data calculated in step 103 is called, and the current vehicle speed information collected in step 101 and the lane center line coordinate sequence determined in step 102 are synchronously acquired; then, the correction direction and the demand intensity are judged according to the lateral deviation distance, and the correction strategy is dynamically adjusted in combination with the vehicle speed information, the single correction amplitude is reduced in the medium and high speed scene to guarantee the stability, and the correction efficiency can be moderately improved in the low speed scene to quickly adhere to the center line; subsequently, the target planning trajectory that is continuous, smooth and converges to the lane center line is calculated through the preset trajectory generation algorithm by fusing the above data, and the trajectory data is stored to the cache module.

[0042] Through the implementation of step 104, when determining the target planning trajectory of the current planning period, the lateral deviation distance and the vehicle speed information are considered at the same time, so that the trajectory planning can be more targeted and adaptive in the curve, not only the deviation can be corrected, but also the current speed state of the vehicle can be matched, thereby improving the rationality and smoothness of the planning.

[0043] In step 105, based on the target planning trajectory, the target vehicle is controlled to drive in the curve.

[0044] In some examples, the target planning trajectory generated in step 104 can be converted into a control instruction executable by the vehicle, and the target vehicle is driven along the trajectory to stably complete the curve driving through the driving of the vehicle steering system and other execution mechanisms, so as to realize the precise linkage of trajectory planning and vehicle execution, and ensure that the vehicle adheres to the lane center in the curve driving.

[0045] Exemplarily, firstly, the complete data of the target planning trajectory generated in step 104 can be called, including key information such as two-dimensional coordinates of each point, curvature change and steering trend; then, the core control demand is extracted through the trajectory analysis module, the vehicle dynamics model is used for adaptive calculation in combination with the real-time updated vehicle speed information in step 101, so as to ensure that the generated control parameters meet the requirements of trajectory following and comply with the stability constraints of vehicle driving (such as avoiding shaking of the vehicle body caused by too fast steering rate); subsequently, the calculated steering control parameters are packaged into standardized control instructions and sent to the steering actuator of the vehicle to drive the steering system to accurately adjust the steering state of the front wheel according to the instructions, so that the vehicle drives along the target planning trajectory; the whole control process is updated synchronously with the trajectory planning, and the control instruction is corrected according to the latest target planning trajectory in each planning period, so as to ensure that the vehicle always adheres to the trajectory and stably maintains the lane center in the curve driving, thereby effectively improving the safety and stability of the curve driving.

[0046] Through the implementation of step 105, the vehicle is controlled based on the target planning trajectory, so that the vehicle can pass through the curve along the ideal trajectory that is optimized and corrected, thereby realizing more stable, more accurate and safer curve driving effect, and improving the overall curve passing performance and driving stability.

[0047] In summary, the embodiment of the present application can accurately identify the geometric characteristics of the curved road section where the vehicle is located by obtaining the vehicle speed information, the vehicle positioning point and the curvature radius of the lane, and construct the subsequent control basis based on this, so that the curved driving control has more accurate environmental understanding ability; determine the lane center line of the lane where the vehicle is located based on the curvature radius, so that the vehicle has a clear, reliable and consistent with the structure of the curved reference path during curved driving, which helps to plan the trajectory closely along the real road direction and improve the accuracy of the vehicle driving along the lane center line; by calculating the lateral deviation distance between the historical planning trajectory of the last planning period and the lane center line, the deviation degree of the vehicle in the curve can be reflected in real time, providing a quantitative basis for trajectory correction, thereby effectively reducing the lateral error generated by the vehicle in the curve; when determining the target planning trajectory of the current planning period, the lateral deviation distance and the vehicle speed are considered comprehensively, so that the trajectory planning not only corrects the deviation, but also adapts to the current dynamic state of the vehicle, improves the rationality of the planning result and the smoothness of the control; finally, the vehicle is controlled based on the target planning trajectory, so that the vehicle can pass through the curve according to the corrected and optimized trajectory, realize more stable, continuous and dynamically adaptive curved driving effect, and improve the curved safety and the overall control performance of the vehicle. In summary, the vehicle curved driving control method provided by the embodiment of the present application determines the lane center line by combining the vehicle speed, the vehicle positioning point and the lane curvature, and dynamically optimizes the planning trajectory based on the historical lateral deviation, so that the vehicle can drive more accurately, smoothly and safely along the corrected target trajectory in the curve.

[0048] In some embodiments, the foregoing step 103 can include: vertically projecting the vehicle positioning point onto the historical planning trajectory to obtain a first projection point; vertically projecting the first projection point onto the lane center line to obtain a second projection point; determining the lateral deviation distance based on the distance between the first projection point and the second projection point.

[0049] In some examples, the vertical projection is a geometric operation in which the vehicle positioning point is projected onto the target path (historical planned trajectory or lane centerline) along a direction perpendicular to the extension direction of the target path, forming a unique projection point. The core is to ensure that the projection direction is perpendicular to the tangent direction of the target path, so as to ensure the uniqueness and position relevance of the projection point. The spatial coordinate sequence of the target path can be analyzed to calculate the tangent direction vector of the target path at the corresponding position, and then the projection point coordinates can be obtained based on the spatial point coordinates and the tangent direction vector through geometric perpendicular relationship. For example, for a historical planned trajectory in the form of an arc, the tangent direction at a potential projection position on the trajectory is calculated first, and then the vehicle positioning point is projected onto the trajectory along a direction perpendicular to the tangent, thereby completing the vertical projection operation. The first projection point is a spatial point formed by projecting the vehicle positioning point onto the historical planned trajectory through vertical projection, and is a key intermediate point for associating the current vehicle position with the historical planned trajectory, and is used to accurately position the corresponding position of the historical trajectory and the current vehicle state. For example, the vehicle positioning point is east longitude 116.39760 degrees and north latitude 39.90890 degrees, and the historical planned trajectory is a circular arc curve with a radius of curvature of 200 meters. The first projection point coordinates calculated by vertical projection are east longitude 116.39758 degrees and north latitude 39.90889 degrees. The second projection point is a spatial point formed by projecting the first projection point onto the lane centerline through vertical projection, and is a key intermediate point for associating the historical planned trajectory with the lane centerline, and is used to establish the lateral position association between the two. For example, based on the first projection point coordinates described above, the lane centerline is a circular arc line with the same radius of curvature, and the second projection point coordinates calculated by vertical projection are east longitude 116.39752 degrees and north latitude 39.90886 degrees. The lateral deviation distance is the straight line distance between the first projection point and the second projection point. Since the generation of the two projection points is based on the projection logic perpendicular to the corresponding target path, this distance exactly reflects the deviation degree of the historical planned trajectory relative to the lane centerline in the lateral direction (perpendicular to the lane extension direction), and is the core quantitative basis for trajectory correction. The straight line distance can be directly solved by substituting the two-dimensional coordinates of the first projection point and the second projection point into the distance calculation formula between two points in space; for example, the two-dimensional coordinates of the first projection point and the second projection point are substituted into the formula, and the calculated lateral deviation distance is 0.32 meters, indicating that the historical planned trajectory deviates outward by 0.32 meters relative to the lane centerline.

[0050] For example, the two-dimensional coordinates of the vehicle positioning point obtained in step 101, the complete coordinate sequence of the historical planning trajectory in the last planning period, and the coordinate sequence of the lane center line determined in step 102 can be first invoked; then, the tangent direction vector of the historical planning trajectory is analyzed by the spatial geometry operation module, the vehicle positioning point is projected along the direction perpendicular to the tangent to the historical planning trajectory to accurately generate a first projection point; subsequently, the same perpendicular projection logic is used to analyze the tangent direction vector of the lane center line, and the first projection point is projected to the lane center line to obtain a second projection point; finally, the straight-line distance between the first projection point and the second projection point is solved by a two-point distance calculation formula, and the distance is the lateral deviation distance.

[0051] Through the implementation of the above embodiments, the vehicle positioning point is sequentially projected onto the historical planning trajectory and the lane center line to calculate the lateral deviation, an accurate and geometrically meaningful deviation distance can be obtained, the deviation measurement is free from the influence of noise interference and path curvature change, thereby providing a more reliable deviation signal for subsequent trajectory correction, and further improving the deviation identification accuracy of the vehicle in a curve.

[0052] In some embodiments, the aforementioned step 104 can include: determining a first translation distance corresponding to the current planning period based on the vehicle speed information; determining a second translation distance based on the product of the lateral deviation distance and a first coefficient; determining a target translation distance as the smaller one of the first translation distance and the second translation distance; and determining a target planning trajectory based on the target translation distance and the lateral deviation distance.

[0053] In some examples, the first translation distance is a maximum translation amplitude of the planning starting point allowed in the current planning period calculated based on the target vehicle current speed information, used to constrain the single translation distance according to the vehicle dynamic state, to avoid excessive adjustment amplitude leading to vehicle instability due to excessively high speed; the preset speed correlation algorithm can be used for calculation in combination with the current real-time speed information, and the algorithm logic needs to ensure the adaptability of the speed and the translation distance. The first coefficient is a preset fixed proportional coefficient, used to control the proportion of the lateral deviation distance to the second translation distance, to avoid excessive adjustment amplitude based on the deviation, and to ensure the smoothness of the trajectory adjustment; the fixed optimal coefficient value can be calibrated through a large number of simulation experiments and real vehicle debugging before the vehicle is shipped, and stored in the parameter configuration module of the vehicle control system, which can be directly called when used; the value of the first coefficient needs to consider the adjustment efficiency and stability, and can be set to about 0.1 (i.e. 10%), for example, the first coefficient is 0.1, which means that the second translation distance is calculated based on only 10% of the lateral deviation distance. The second translation distance is a theoretical translation amplitude based on the current deviation degree obtained by multiplying the lateral deviation distance by the first coefficient, which is used to quantify the deviation proportion to be corrected and ensure that the translation operation is always carried out around the goal of reducing the lateral deviation; for example, when the lateral deviation distance is 0.4 meters and the first coefficient is 0.1, the second translation distance is 0.4*0.1=0.04 meters; when the lateral deviation distance is 0.2 meters and the first coefficient is 0.1, the second translation distance is 0.2*0.1=0.02 meters. By comparing the first translation distance (upper limit of speed constraint) and the second translation distance (deviation proportion amplitude), the smaller one of the two is selected as the actual translation amplitude of the current planning period, to balance the efficiency and safety of trajectory adjustment, to avoid excessive translation due to high speed, and to avoid excessive adjustment due to large deviation; for example, when the first translation distance is 0.03 meters and the second translation distance is 0.04 meters, 0.03 meters is selected as the target translation distance; when the first translation distance is 0.02 meters and the second translation distance is 0.015 meters, 0.015 meters is selected as the target translation distance. Based on the target translation distance and the lateral deviation distance, the process of determining the target planning trajectory is to take the lateral deviation distance as the adjustment direction guide (the direction of converging to the center line of the lane), take the target translation distance as the actual adjustment amplitude, adjust the planning starting point, and then generate the optimal path of the current planning period through the trajectory generation algorithm; for example, the lateral deviation distance is 0.35 meters (the trajectory deviates to the outside of the lane), and the target translation distance is 0.03 meters, the planning starting point of the last period is translated 0.03 meters to the center line of the lane to obtain a new starting point, and the target planning trajectory with gentle curvature and gradually converging to the center line is generated based on the new starting point.

[0054] For example, the real-time vehicle speed information of step 101 and the lateral deviation distance data of step 103 can be first called, and a preset first coefficient is obtained from the parameter configuration module; then, the first translation distance is calculated by a vehicle speed correlation algorithm, the second translation distance is obtained by multiplying the lateral deviation distance by the first coefficient, and the smaller of the two is selected as the target translation distance through numerical comparison. Subsequently, the system determines the translation direction according to the lateral deviation distance, translates the starting point of the trajectory of the last planning period by the target translation distance in the direction to obtain a new starting point, combines the lane center line data determined in step 102, and generates a smooth and safe target planning trajectory through a trajectory generation algorithm.

[0055] Through the implementation of the above embodiments, the first translation distance is determined based on the vehicle speed, the second translation distance is determined based on the lateral deviation, and the smaller of the two is taken as the final compensation amount, so that the trajectory correction can reflect the actual deviation and will not affect the driving smoothness due to excessive correction, and more reasonable, gradual and vehicle dynamics characteristic trajectory adjustment can be achieved, effectively improving the stability and precision of the curve trajectory planning.

[0056] In some embodiments, the aforementioned determination of the first translation distance corresponding to the current planning period based on the vehicle speed information can include: determining a third translation distance based on the product of the vehicle speed information and a second coefficient; if the third translation distance is within a preset distance range, the third translation distance is determined as the first translation distance, otherwise, a boundary value in the preset distance range that matches the third translation distance is determined as the first translation distance.

[0057] In some examples, the second coefficient is a preset vehicle speed and translation distance correlation coefficient for converting vehicle speed information into a quantitative translation amplitude reference value, which establishes a linear correspondence between vehicle speed and translation distance and ensures that the translation amplitude adapts to the vehicle dynamic state. The optimal fixed value can be obtained through a large number of simulation experiments and real vehicle road tests before the vehicle is shipped and stored in the parameter configuration module of the vehicle control system (VCS) for direct retrieval when used. The value of the second coefficient needs to balance the adjustment efficiency and driving stability, and can be set to 0.003, that is, the third translation distance increases by 0.003 meters for every 1 meter per second increase in vehicle speed. The third translation distance is an initial translation amplitude calculation value obtained by multiplying the current vehicle speed information by the second coefficient, and is the basic reference data of the first translation distance, which serves to preliminarily quantify the reasonable translation range corresponding to the vehicle speed. For example, if the current vehicle speed information is 14 meters per second and the second coefficient is 0.003, the third translation distance is 14 x 0.003 = 0.042 meters. When the current vehicle speed is 8 meters per second, the third translation distance is 8 x 0.003 = 0.024 meters. The preset distance range is a preset translation distance safety constraint interval for limiting the extreme value of the third translation distance to avoid the translation amplitude exceeding the vehicle dynamics bearing range due to excessively high or low vehicle speed. The fixed interval can be calibrated before shipment in combination with the vehicle steering system performance and dynamics characteristics and stored in the parameter configuration module. The interval range needs to consider the adjustment effectiveness and control stability, and can be set to 0 meters to 0.03 meters, which ensures that the translation distance is not negative (no invalid adjustment) at low speed and avoids excessive translation distance (no excessive adjustment) at high speed. When the value of the third translation distance is between the upper and lower boundaries of the preset distance range, the calculation value is directly used as the first translation distance of the current planning period. The core logic is to retain the original adjustment amplitude corresponding to the vehicle speed within the safety constraint to ensure that the translation distance adapts to the vehicle speed and does not exceed the safety range. That is, if the third translation distance and the boundary value of the preset distance range satisfy the condition preset lower limit ≤ third translation distance ≤ preset upper limit, the third translation distance is directly assigned to the first translation distance. When the third translation distance exceeds the preset distance range, the original calculation value is abandoned and the corresponding interval boundary value is used as the first translation distance, which serves to avoid out-of-control translation amplitude through boundary constraint and ensure vehicle driving stability.The boundary value in the preset distance range matching the third translation distance is an interval boundary value corresponding to the direction deviating from the third translation distance when the third translation distance exceeds the preset distance range. If the third translation distance is greater than the upper limit of the preset distance range, the upper limit value of the distance range is matched. If the third translation distance is less than the lower limit of the preset distance range, the lower limit value of the distance range is matched. The position relationship of the third translation distance relative to the preset range can be identified through the deviating direction judgment algorithm, and then the corresponding boundary value is called. For example, the preset distance range is 0 meters to 0.03 meters, the third translation distance is 0.04 meters (exceeding the upper limit), and the matched boundary value is 0.03 meters. The third translation distance is -0.005 meters (exceeding the lower limit), and the matched boundary value is 0 meters.

[0058] Through the implementation of the above embodiments, the third translation distance is calculated according to the product of the vehicle speed and the second coefficient, and the boundary value is replaced when it exceeds the safe range, which can avoid excessive translation correction caused by high speed, keep the trajectory planning within a controllable range, and ensure the controllability and safety of vehicle deviation correction in high-speed curves, thereby further improving the stability of curve driving.

[0059] In some embodiments, before determining the third translation distance based on the product of the aforementioned vehicle speed information and the second coefficient, the vehicle curve driving control method can further include: obtaining a calibration coefficient value corresponding to the driving mode of the target vehicle from a preset parameter mapping table as the second coefficient, wherein the calibration mapping table stores calibration coefficient values corresponding to multiple driving modes, and the calibration coefficient value is used to calculate the third translation distance.

[0060] In some examples, the preset parameter mapping table is a data association table pre-stored in the target vehicle control system, which establishes a one-to-one correspondence between the driving mode and the calibration coefficient value for quick query and matching. The calibration coefficient value is a core parameter pre-calibrated for different driving modes for converting vehicle speed information into the third translation distance, and its value directly determines the calculation result of the third translation distance, thereby affecting the smoothness or response speed of trajectory adjustment; for example, the calibration coefficient value corresponding to the comfort mode is 0.002, which ensures smoother trajectory adjustment; the calibration coefficient value corresponding to the sports mode is 0.004, which makes the trajectory adjustment response more rapid; the calibration coefficient value corresponding to the standard mode is 0.003, which balances the smoothness and responsiveness. The driving mode is a driving style mode preset by the target vehicle to adapt to different driving needs and scenarios, which can match the user's driving preferences by adjusting the vehicle control parameters (such as trajectory adjustment amplitude and response speed); the driving mode setting state of the current vehicle can be read in real time by the vehicle control module, which can be manually selected by the user through the in-vehicle physical buttons or the center touch screen, or automatically switched by the vehicle according to the driving scenario; for example, common driving modes include comfort mode (emphasizing smooth driving, adjusting smoothly), standard mode (balancing smoothness and response speed), and sports mode (emphasizing dynamic response, adjusting more sensitively), and the user may select the comfort mode when driving on a high-speed ramp and the standard mode when driving on a city expressway curve. The driving mode of the current vehicle can be identified first, and then the calibration coefficient value corresponding to the mode can be queried from the preset parameter mapping table to determine the second coefficient required for calculating the third translation distance, and the core logic is to realize the dynamic adaptation of the second coefficient to the driving mode.

[0061] Through the implementation of the above embodiments, the second coefficient adapted to the current driving mode is obtained from the calibration mapping table, so that the trajectory translation correction amount can adapt to the control style under different driving modes (such as comfort, sports, and economy), improve the adaptability of trajectory planning in multiple scenarios, and enable the vehicle to maintain more reasonable and stable curve centering performance in multiple driving modes.

[0062] In some embodiments, before determining the second translation distance based on the product of the aforementioned lateral deviation distance and the first coefficient, the vehicle curve driving control method can further include: determining the first coefficient based on at least one of the lane width, the road adhesion coefficient, the traffic flow density, and the radius of curvature of the lane where the target vehicle is located, wherein the first coefficient is inversely proportional to the radius of curvature, the lane width, and the traffic flow density, and the first coefficient is directly proportional to the road adhesion coefficient.

[0063] In some examples, the lane width is the horizontal distance between the edges of the lane lines on both sides of the current lane of the target vehicle, is a core parameter reflecting the lane space constraint, and directly affects the safety redundancy space of vehicle trajectory adjustment; the lane line image and two-dimensional environment data in front of the target vehicle can be collected in real time by a vehicle camera (VC) or a light detection and ranging (LiDAR) carried by the target vehicle, edge coordinates of the lane lines on both sides are extracted through image recognition and spatial ranging algorithm, and the straight-line distance between the two is calculated; for example, the lane width of a city ordinary road can be 3.5 meters, the lane width of a highway can be 3.75 meters, and the lane width of a ramp can be 3.0 meters. The road adhesion coefficient is the friction coefficient between the road and the vehicle tire, which is used to represent the road grip performance and is a key dynamic parameter affecting the stability of vehicle trajectory adjustment; the road adhesion coefficient can be calculated in real time or matched by combining the road type recognition algorithm (such as recognizing dry asphalt, wet asphalt in rainy days, snow-covered road surface, etc.) through the wheel speed sensor (WSS) and inertial measurement unit of the target vehicle to collect dynamic data (such as wheel speed difference, lateral acceleration) during vehicle driving; or a preset common road adhesion coefficient database can be directly called for adaptation; for example, the adhesion coefficient of a dry asphalt road can be 0.85, the adhesion coefficient of a wet and slippery asphalt road in rainy days can be 0.45, and the adhesion coefficient of a snow-covered road can be 0.2. The traffic flow density is the number of vehicles driving in a unit length of road section, which is used to represent the congestion degree of the current road and is a scene-based parameter that restricts the adjustment range of vehicle trajectory; the vehicle distribution in front of and around the lane can be detected in real time by the vehicle camera and the LiDAR, and the number of vehicles in a preset road length (such as 100 meters) is counted; or the regional traffic flow data issued by the road side unit (RSU) can be received through the vehicle-to-everything (V2X) technology, and the traffic flow density of the current lane is converted; for example, the traffic flow density under the free-flow scenario of the highway can be 20 vehicles per kilometer, and the traffic flow density under the peak congestion scenario of the urban road can be 100 vehicles per kilometer. At least one parameter of the lane width, the road adhesion coefficient, the traffic flow density, and the curvature radius can be input into a preset multi-parameter fusion algorithm, and a dynamically adapted first coefficient value can be calculated in combination with the proportional relationship that is inversely proportional to the curvature radius, the lane width, and the traffic flow density, and is proportional to the road adhesion coefficient; for example, in the scenario of a curvature radius of 200 meters, a lane width of 3.75 meters, a traffic flow density of 30 vehicles per kilometer, and a road adhesion coefficient of 0.8, the first coefficient can be 0.12; in the scenario of a curvature radius of 150 meters, a lane width of 3.0 meters, a traffic flow density of 80 vehicles per kilometer, and a road adhesion coefficient of 0.5, the first coefficient can be 0.05.

[0064] Through implementation of the foregoing embodiments, the first coefficient is determined based on road and environmental factors such as lane width, road adhesion coefficient, traffic flow density, and radius of curvature, so that the deviation correction range can be dynamically adjusted according to road conditions and the environment, thereby improving the environmental adaptability of trajectory planning, and in particular, in a narrow lane, a low adhesion road, or a large-curvature curve, the deviation can be more accurately and stably corrected, and the ability of the vehicle to keep the center of the lane can be improved.

[0065] In some embodiments, the foregoing determining the target planning trajectory based on the target translation distance and the lateral deviation distance can include: determining a target planning starting point based on the target translation distance and the lateral deviation distance; and generating the target planning trajectory of the current planning period that converges to the lane center line based on the target planning starting point.

[0066] In some examples, the target planning starting point is a starting spatial point of the target planning trajectory in the current planning period, and is a core reference point obtained by adjusting the target translation distance and the lateral deviation distance. The target planning starting point functions to guide the trajectory to converge to the lane center line through starting point offset, and avoids trajectory planning lagging behind the deviation correction demand. For example, the coordinates of the planning starting point associated with the last period are 116.39758 degrees east and 39.90889 degrees north, the lateral deviation distance is 0.32 meters (the trajectory deviates to the outside of the lane), and the target translation distance is 0.03 meters. After translation in the direction toward the lane center line, the coordinates of the target planning starting point are 116.39755 degrees east and 39.90887 degrees north. Based on the target planning starting point, the process of generating the target planning trajectory of the current planning period that converges to the lane center line is to generate a smooth and gradually close-to-lane-center-line path by taking the target planning starting point as the starting end of the trajectory, combining the geometric shape of the lane center line, the current dynamic state (such as the vehicle speed) of the vehicle, and the dynamic constraints, and the core is to achieve smooth convergence of the trajectory from the starting point to the reference line to solve the problem of deviation of the historical trajectory. For example, the target planning starting point deviates from the lane center line by 0.29 meters, the generated target planning trajectory is a cubic polynomial curve, the coordinate range is 116.39755-116.39855 degrees east and 39.90887-39.90927 degrees north, the trajectory curvature is 0.005 meters -1 and gradually transitions to be consistent with the curvature of the lane center line, and the lateral distance of the end to the lane center line is reduced to 0.26 meters, thereby achieving smooth convergence.

[0067] Through implementation of the foregoing embodiments, the planning starting point is determined based on the target translation distance and the lateral deviation distance, and the planning trajectory that converges to the lane center line is generated from the starting point, so that the trajectory planning process can gradually and continuously approach the lane center line, and the vehicle is prevented from being unstable due to abrupt correction, thereby providing a more smooth, continuous, and stable path planning for a curve scenario.

[0068] In some embodiments, the determining the target planning origin based on the target translation distance and the lateral deviation distance can include: determining a translation ratio based on a ratio of the target translation distance to the lateral deviation distance; and determining the target planning origin on a line connecting the first projection point and the second projection point based on the translation ratio.

[0069] In some examples, the translation ratio is a ratio of the target translation distance to the lateral deviation distance, is a core parameter for quantifying the translation amplitude of the target planning origin on the line connecting the two points, can clearly determine the proportion of the translation from the first projection point to the second projection point, and ensure that the translation amplitude and the deviation correction requirement are accurately matched; the ratio value (the value range is 0 to 1) can be directly solved by performing a basic division operation of dividing the target translation distance determined in the current planning period by the lateral deviation distance obtained in step 103; for example, when the target translation distance is 0.03 meters and the lateral deviation distance is 0.3 meters, the translation ratio is 0.03 ÷ 0.3 = 0.1 (i.e., 10%); when the target translation distance is 0.015 meters and the lateral deviation distance is 0.2 meters, the translation ratio is 0.015 ÷ 0.2 = 0.075 (i.e., 7.5%). The line connecting the first projection point and the second projection point is a spatial straight line segment connecting the first projection point and the second projection point, the extension direction of the line segment is exactly the deviation correction direction from the historical planning trajectory to the center line of the lane, and is a translation path reference of the target planning origin. The first projection point can be taken as a starting end, and translation is performed along the line connecting the two points to the second projection point (in the direction of the center line of the lane), the translation distance is the length of the line connecting the two points × the translation ratio, and the spatial point finally obtained is the target planning origin. The core logic is to achieve accurate positioning of the origin translation through proportional control, and to ensure that the translation amplitude of each frame is controllable and meets the deviation correction requirement; for example, when the length of the line connecting the two points is 0.3 meters and the translation ratio is 0.1, the actual translation distance is 0.3 × 0.1 = 0.03 meters, and after moving 0.03 meters from the first projection point to the second projection point, the target planning origin coordinates are 116.39757 degrees east longitude and 39.90888 degrees north latitude.

[0070] Through implementation of the above embodiments, the ratio of the target translation distance to the lateral deviation distance is calculated as the translation ratio, and the target planning origin is determined by interpolation between the historical trajectory point and the center line of the lane, so that the adjustment of the planning origin has a clear geometric basis, and the deviation can be continuously and progressively converged according to the size, thereby improving the smoothness and control accuracy of trajectory convergence.

[0071] In some embodiments, the generating the target planning trajectory of the current planning period that converges to the center line of the lane based on the target planning origin can include: performing trajectory solving based on the target planning origin and the center line of the lane through an iterative linear quadratic regulator optimization algorithm to generate the target planning trajectory.

[0072] In some examples, the Iterative Linear Quadratic Regulator (ILQR) is an optimal trajectory optimization algorithm for nonlinear dynamic systems, which linearizes the nonlinear system model step by step through multiple iterations, solves a quadratic programming problem, and outputs an optimal trajectory that meets system constraints and optimization objectives. The algorithm can be pre-installed in the trajectory planning module of the vehicle control system before the vehicle is shipped, combined with the dynamic characteristics of the target vehicle (such as the steering angle range and lateral acceleration limit), to calibrate and verify the algorithm parameters, and ensure that the algorithm adapts to the actual operating state of the vehicle. For example, the algorithm has been integrated into the trajectory planning unit of the auxiliary driving system, and through a large number of simulation and real vehicle testing of curved road scenes, the core parameters such as the number of iterations and the convergence threshold that adapt to passenger cars have been optimized.

[0073] The nonlinear optimization capability of the ILQR algorithm can be used to obtain an optimal driving path by integrating vehicle dynamics constraints and trajectory smoothness requirements through multiple iterations, with the target planning starting point as the trajectory starting reference and the lane centerline as the final convergence target. This can achieve smooth convergence of the trajectory from the starting point to the lane centerline while meeting the dynamics feasibility of vehicle driving. For example, the target planning starting point coordinates are 116.39755 degrees east and 39.90887 degrees north, and the lane centerline is a circular arc with a radius of 180 meters. After 20 iterations of the ILQR algorithm, a smooth trajectory with continuous curvature and no abrupt changes is generated. The starting end of the trajectory coincides with the target planning starting point, and the lateral deviation of the end from the lane centerline is reduced from 0.29 meters to 0.26 meters, and the maximum lateral acceleration constraint is met throughout the journey.

[0074] Through the implementation of the above embodiments, the ILQR algorithm is used to solve the trajectory that converges to the lane centerline based on the target planning starting point. This makes the generated target trajectory meet both the vehicle dynamics constraints and the convergence requirements, improves the smoothness, stability, and executability of the trajectory, and enhances the path tracking effect of the vehicle in the curve.

[0075] In some embodiments, the aforementioned step 105 can include determining a desired yaw rate based on the target planning trajectory, generating a first steering control instruction based on a first deviation of the desired yaw rate from an actual yaw rate of the target vehicle, sending the first steering control instruction to a steering actuator of the target vehicle to drive the target vehicle to travel along the target planning trajectory, or determining a desired front wheel steering angle based on the target planning trajectory, generating a second steering control instruction based on a second deviation of the desired front wheel steering angle from an actual front wheel steering angle of the target vehicle, and sending the second steering control instruction to the steering actuator of the target vehicle to drive the target vehicle to travel along the target planning trajectory.

[0076] In some examples, the expected yaw rate is a rotational speed around the vertical axis of the vehicle body that the target vehicle should theoretically reach when driving along the target planning trajectory, and is a core dynamic parameter representing the adaptability of the vehicle steering response to the trajectory; for example, when the current vehicle speed is 12.5 meters per second and the curvature radius of the target planning trajectory corresponding point is 180 meters, the expected yaw rate is 12.5 ÷ 180 ≈ 0.069 rad / s; if the trajectory curvature increases (the radius decreases to 150 meters), the expected yaw rate is 12.5 ÷ 150 ≈ 0.083 rad / s. The actual yaw rate is the actual rotational speed around the vertical axis of the vehicle body during the current driving process of the target vehicle, and is a measured parameter reflecting the real-time steering state of the vehicle; for example, when the vehicle is driving along the curve, the actual yaw rate collected and output by the inertial measurement unit may be 0.062 rad / s, which has a slight deviation from the expected yaw rate. The first deviation is the numerical difference between the expected yaw rate and the actual yaw rate, and is a parameter for quantifying the deviation of the vehicle steering state from the target trajectory adaptability, which provides a basis for generating a steering control instruction; for example, when the expected yaw rate is 0.069 rad / s and the actual yaw rate is 0.062 rad / s, the first deviation is 0.069-0.062=0.007 rad / s; if the actual yaw rate is 0.075 rad / s, the first deviation is -0.006 rad / s. The first steering control instruction is a standardized execution instruction generated based on the first deviation for adjusting the steering state of the vehicle, and the role is to correct the yaw rate deviation and make the vehicle steering response fit the target planning trajectory; the first deviation can be input into a preset control algorithm, such as a proportional-integral-derivative controller (PID Controller), combined with the characteristics of the vehicle steering system (such as response sensitivity, delay time) to output a standardized instruction (such as a CAN bus compatible instruction) containing the steering adjustment direction and adjustment amplitude; for example, when the first deviation is 0.007 rad / s, the first steering control instruction generated by the PID algorithm is “increase the steering angle of the front wheels to the inside of the curve by 0.15 degrees”; if the first deviation is -0.006 rad / s, the instruction is “correct the steering angle of the front wheels to the outside of the curve by 0.12 degrees”. The first steering control instruction can be transmitted to the vehicle steering actuator through the vehicle-mounted communication bus, and the actuator executes the steering action.

[0077] The expected front wheel steering angle is a steering angle that the front wheel theoretically should have when the target vehicle travels along the target planning trajectory, is a mechanical control parameter directly adapted to the curvature of the trajectory, and can be calculated based on the real-time curvature of the target planning trajectory and the wheelbase parameter of the vehicle through a geometric relationship formula (expected front wheel steering angle ≈ wheelbase ÷ curvature radius of the corresponding point of the target trajectory, small angle approximation). For example, when the wheelbase of the target vehicle is 2.8 meters and the curvature radius of the corresponding point of the target planning trajectory is 180 meters, the expected front wheel steering angle is 2.8 ÷ 180 ≈ 0.0156 radian (converted to 0.89 degrees). If the curvature radius of the trajectory is reduced to 150 meters, the expected front wheel steering angle is 2.8 ÷ 150 ≈ 0.0187 radian (converted to 1.07 degrees). The actual front wheel steering angle is the actual steering angle of the front wheel of the target vehicle at present, is a measured parameter reflecting the execution state of the steering system, and may be slightly different from the expected front wheel steering angle. For example, the actual front wheel steering angle collected and output by the steering angle sensor may be 0.82 degrees. If the vehicle is over-steered, the actual front wheel steering angle may be 1.15 degrees. The second deviation is the numerical difference between the expected front wheel steering angle and the actual front wheel steering angle, is a core parameter for quantifying the steering state of the front wheel and the deviation of the target trajectory, and provides a basis for generating a targeted steering control instruction. For example, when the expected front wheel steering angle is 0.89 degrees and the actual front wheel steering angle is 0.82 degrees, the second deviation is 0.89-0.82=0.07 degrees. If the actual front wheel steering angle is 1.15 degrees, the second deviation is -0.26 degrees. The second steering control instruction is a standardized execution instruction for accurately adjusting the steering angle of the front wheel, is generated based on the second deviation, and directly corrects the steering deviation of the front wheel to make the steering angle of the front wheel conform to the expected value, thereby ensuring trajectory following. The second steering control instruction can be transmitted to the steering actuator through the vehicle-mounted communication bus, the steering angle of the front wheel is directly adjusted by the actuator, and trajectory following is achieved.

[0078] Through implementation of the above embodiments, the expected yaw rate or the expected front wheel steering angle is obtained based on the target planning trajectory, and a steering control instruction is generated based on the deviation of the actual state, which can make the steering execution of the vehicle more accurately conform to the planning trajectory, achieve high-precision tracking of the trajectory, and thus improve the vehicle posture stability and lane keeping ability in curve driving.

[0079] In some embodiments, the foregoing step 102 can include: in the case where the curvature radius is less than a preset radius threshold, performing lane line recognition on the lane where the target vehicle is located to obtain lane line information corresponding to the lane; and determining a lane center line of the lane where the target vehicle is located based on the lane line information.

[0080] In some examples, the preset radius threshold is a preset curvature radius threshold value for distinguishing a large-curvature curve from a normal curve (including a straight road), and is used to trigger targeted lane line recognition and center line determination logic to adapt to the visual detection characteristics in a large-curvature curve scenario. For example, the preset radius threshold can be set to 250 meters. When the lane curvature radius obtained in step 101 is less than this value, it is determined that it is a large-curvature curve, and the exclusive recognition process is started. If it is greater than or equal to this value, it is processed according to the normal straight road / normal curve logic. The lane line information is data obtained by sensor recognition, representing the current lane line shape and reliability, and is the core basis for determining the lane center line, including the geometric features of the lane line and the recognition reliability evaluation data. The lane line data can be obtained by processing the front road image and two-dimensional environment data collected by the vehicle-mounted camera and laser radar of the target vehicle through image recognition algorithms (such as edge detection and feature extraction algorithms). For example, the lane line information can include the edge coordinate point sequence of the lane line, the curvature change trend, the line segment integrity identifier, and the confidence evaluation value, such as 0.9 indicating that the recognition result is reliable and 0.3 indicating that the recognition result is unstable.

[0081] Based on the lane line information, the process of determining the lane center line of the lane where the target vehicle is located is based on the recognized lane line information, and a lane geometric center axis is generated through geometric fitting operation. In the implementation process, first, the edge coordinate point sequence and curvature data of the two-side lane lines can be extracted from the lane line information. Second, the midpoint of each group of corresponding coordinate points is calculated according to the symmetry relationship and geometric distance of the two-side lane lines. Finally, all midpoints are connected and smoothed through a curve fitting algorithm (such as the least squares method) to generate a lane center line consistent with the lane line shape.

[0082] Through the implementation of the above embodiments, lane line recognition is triggered only when the curvature radius is small and the road curvature is large, and the lane center line and subsequent steps are determined based on the lane line information, which can avoid unnecessary calculation and perception resource consumption in a straight road or a gentle curve scenario, while significantly improving the estimation accuracy of the lane center line in a curve scenario, thereby improving the stability and safety of curve trajectory planning and vehicle attitude control.

[0083] In some embodiments, the lane line information can include geometric feature information and confidence evaluation information of the lane line. The lane center line of the lane where the target vehicle is located can be determined based on the lane line information, which can include: extracting a first feature point sequence and a first curvature information sequence from the geometric feature information; performing interpolation processing on the first feature point sequence and the first curvature information sequence based on the confidence evaluation information to obtain a second feature point sequence and a second curvature information sequence; and determining the lane center line based on the second feature point sequence and the second curvature information sequence.

[0084] In some examples, the geometric feature information is core data representing the spatial form and structure of the lane line, is basic information describing the physical properties of the lane line, and directly determines the geometric form of the lane center line; the front road image and two-dimensional environment data can be collected by the vehicle-mounted camera and laser radar carried by the target vehicle, processed by an image recognition algorithm (such as an edge detection algorithm or a feature extraction algorithm), and the key spatial parameters of the lane line can be extracted; for example, the geometric feature information can include discrete coordinate points of the edge of the lane line, a line segment direction, curvature change data, a lane line width, etc., wherein the curvature change data can directly reflect the bending degree of the lane line. The confidence evaluation information is quantitative evaluation data for the reliability of the recognition result of the geometric feature information, and the function is to judge the confidence degree of each segment of the lane line data, and to provide a basis for subsequent data repair; for example, the image of the lane line in the range of 10-30 meters near the ego vehicle is clear and unobstructed, and the corresponding confidence evaluation value is 0.9; the image beyond 50 meters is blurred and the feature is not clear, and the corresponding confidence evaluation value is 0.3.

[0085] The first feature point sequence is a set of discrete spatial coordinate points representing the edge contour of the lane line, which is extracted directly from the geometric feature information and is basic data for constructing the form of the lane line; the key coordinate points of the lane line edge can be selected from the geometric feature information by a feature extraction algorithm, and a continuous sequence is formed by sorting according to the extension direction of the lane line; for example, the first feature point sequence can be a set of discrete coordinates of the left lane line, including (E 116.39700, N 39.90880), (E 116.39720, N 39.90885), and dozens of continuous coordinate points, which completely cover the edge of the lane line in the current detection range. The first curvature information sequence is a set of curvature values corresponding to each feature point of the lane line, which is core data reflecting the bending characteristics of the lane line; based on the coordinate data of the first feature point sequence, the curvature value corresponding to each feature point can be calculated one by one by a curve curvature calculation formula (such as a three-point curvature calculation), and a sequence is formed according to the sorting order of the feature points; for example, the first curvature information sequence can be [0.0053 meters, 0.0054 meters, 0.0055 meters, …], corresponding to the curvature of each point in the first feature point sequence, and the larger the value, the more obvious the bending of the lane line at the point. -1 -1 -1 ……]

[0086] ​​The low-confidence missing or unreliable coordinate points in the first feature point sequence and the abnormal curvature values in the first curvature information sequence can be repaired according to the confidence evaluation value, to generate continuous and reliable sequence data. In the implementation process, first, low-quality data segments with a confidence evaluation value lower than a preset threshold (such as 0.5) can be identified. Second, linear interpolation or cubic spline interpolation algorithm is used to supplement or correct low-confidence data, with high-confidence (such as ≥0.8) data points before and after the low-quality data segment as the reference. Finally, the repaired complete sequence is output. For example, the confidence of three points in a certain segment of the first feature point sequence is 0.2 (low quality), and the adjacent points with a confidence of 0.9 before and after the segment are used as the reference to calculate the missing coordinate points by linear interpolation, to fill the data gap. The second feature point sequence is a continuous and reliable lane line edge coordinate point set obtained after interpolation processing, and the data integrity and accuracy are significantly improved compared with the first feature point sequence. The second feature point sequence is obtained by removing abnormal points and reordering after interpolation processing of the first feature point sequence, to form a coordinate point sequence covering the complete detection range and without data gaps. For example, after interpolation repair, the second feature point sequence supplements three low-confidence missing points in the original first feature point sequence, to form a complete sequence containing continuous coordinate points such as (116.39700 degrees east, 39.90880 degrees north) … (116.39710 degrees east, 39.90883 degrees north) … without discontinuity or abnormal deviation. The second curvature information sequence is a smooth and continuous curvature value set corresponding to the second feature point sequence, and the abnormal fluctuations are eliminated compared with the first curvature information sequence, which is more consistent with the actual bending characteristics of the lane line. The second curvature information sequence is obtained by interpolation processing of the first curvature information sequence, to correct the abnormal curvature values corresponding to the low-confidence, so that the curvature values smoothly transition with the feature point sequence. For example, the second curvature information sequence can be [0.0053 meters, 0.00535 meters, 0.0054 meters, …], which eliminates the 0.008 meters abnormal value in the original first curvature information sequence caused by low confidence, and the curvature changes continuously without mutation. -1 -1 -1 -1

[0087] ​​​​The process of determining the lane center line based on the second feature point sequence and the second curvature information sequence is to generate a lane geometric center axis through geometric operation based on the repaired second feature point sequence and the second curvature information sequence, which can ensure that the center line is accurately matched with the actual form of the lane line. In the implementation process, first, the second feature point sequence and the second curvature information sequence of the two side lane lines can be obtained respectively. Second, the midpoint coordinates (i.e., the center position of the lateral distance) of the corresponding feature points of the two side lane lines are calculated to form a center line feature point sequence. Finally, the center line feature point sequence is smoothed through a curve fitting algorithm (such as the least square method) in combination with the second curvature information sequence to generate a lane center line with the same curvature and continuous smoothness as the lane line. For example, the midpoint coordinates of the corresponding points of the second feature point sequence of the left lane line and the second feature point sequence of the right lane line are (116.39735 degrees east, 39.90882 degrees north), (116.39755 degrees east, 39.90887 degrees north), and so on. After fitting in combination with the second curvature information sequence, a circular arc-shaped lane center line with a curvature radius of 180 meters is obtained.

[0088] Through the implementation of the above embodiments, the geometric feature points and curvature information of the lane line are interpolated in combination with the confidence, and then the lane center line is determined based on the interpolated results, which can effectively repair the local instability caused by noise, occlusion and other factors in visual detection, improve the smoothness and reliability of the lane center line inference, and thus provide a more stable and accurate reference line for curve trajectory planning.

[0089] Further, as an implementation of the foregoing method embodiment, the application also provides a vehicle curve driving control device applied to a target vehicle for implementing the foregoing method embodiment. The device embodiment corresponds to the foregoing method embodiment, and for the sake of reading, the vehicle curve driving control device embodiment will not repeat the details of the foregoing method embodiment one by one, but it should be clear that the device in the application embodiment can correspondingly implement all the contents in the foregoing method embodiment. For example, Figure 2As shown, the vehicle curve driving control device 20 comprises an information acquisition unit 201, a center line determination unit 202, a deviation determination unit 203, a trajectory determination unit 204 and a vehicle control unit 205, wherein the information acquisition unit 201 is configured to acquire vehicle speed information of a target vehicle, a vehicle positioning point and a radius of curvature of a lane where the target vehicle is located; the center line determination unit 202 is configured to determine a lane center line of the lane where the target vehicle is located based on the radius of curvature; the deviation determination unit 203 is configured to determine a lateral deviation distance between a historical planning trajectory of a last planning period and the lane center line based on the vehicle positioning point and the lane center line; the trajectory determination unit 204 is configured to determine a target planning trajectory of a current planning period based on the lateral deviation distance and the vehicle speed information; and the vehicle control unit 205 is configured to control the target vehicle to drive on a curve based on the target planning trajectory.

[0090] In some embodiments, the deviation determination unit 203 is further configured to vertically project the vehicle positioning point onto the historical planning trajectory to obtain a first projection point; vertically project the first projection point onto the lane center line to obtain a second projection point; and determine the lateral deviation distance based on a distance between the first projection point and the second projection point.

[0091] In some embodiments, the trajectory determination unit 204 is further configured to determine a first translation distance corresponding to the current planning period based on the vehicle speed information; determine a second translation distance based on a product of the lateral deviation distance and a first coefficient; determine a target translation distance as a smaller one of the first translation distance and the second translation distance; and determine the target planning trajectory based on the target translation distance and the lateral deviation distance.

[0092] In some embodiments, the trajectory determination unit 204 is further configured to determine a third translation distance based on a product of the vehicle speed information and a second coefficient; if the third translation distance is within a preset distance range, determine the third translation distance as the first translation distance, otherwise, determine a boundary value in the preset distance range that matches the third translation distance as the first translation distance.

[0093] In some embodiments, the trajectory determination unit 204 is further configured to acquire a calibration coefficient value corresponding to a driving mode of the target vehicle from a preset parameter mapping table as the second coefficient, wherein the calibration mapping table stores calibration coefficient values corresponding to a plurality of driving modes, and the calibration coefficient value is used to calculate the third translation distance.

[0094] In some embodiments, the trajectory determination unit 204 is further configured to determine the first coefficient based on at least one of a lane width of the lane where the target vehicle is located, a road surface adhesion coefficient, a traffic flow density and the radius of curvature, wherein the first coefficient is inversely proportional to the radius of curvature, the lane width and the traffic flow density, and the first coefficient is directly proportional to the road surface adhesion coefficient.

[0095] In some embodiments, the trajectory determination unit 204 is further configured to determine a target planning start point based on the target translation distance and the lateral deviation distance; and generate the target planning trajectory of the current planning period converging to the lane center line based on the target planning start point.

[0096] In some embodiments, the trajectory determination unit 204 is further configured to determine a translation ratio based on a ratio of the target translation distance and the lateral deviation distance; and determine the target planning start point on a line connecting the first projection point and the second projection point based on the translation ratio.

[0097] In some embodiments, the trajectory determination unit 204 is further configured to perform trajectory solving based on the target planning start point and the lane center line by an iterative linear quadratic regulator optimization algorithm to generate the target planning trajectory.

[0098] In some embodiments, the vehicle control unit 205 is further configured to determine a desired yaw rate based on the target planning trajectory; generate a first steering control instruction based on a first deviation of the desired yaw rate and an actual yaw rate of the target vehicle; send the first steering control instruction to a steering actuator of the target vehicle to drive the target vehicle to travel along the target planning trajectory; or determine a desired front wheel steering angle based on the target planning trajectory; generate a second steering control instruction based on a second deviation of the desired front wheel steering angle and an actual front wheel steering angle of the target vehicle; and send the second steering control instruction to the steering actuator of the target vehicle to drive the target vehicle to travel along the target planning trajectory.

[0099] In some embodiments, the center line determination unit 202 is further configured to perform lane line recognition on a lane in which the target vehicle is located to obtain lane line information corresponding to the lane when the curvature radius is less than a preset radius threshold; and determine the lane center line of the lane in which the target vehicle is located based on the lane line information.

[0100] In some embodiments, the lane line information includes geometric feature information and confidence assessment information of the lane line; and the center line determination unit 202 is further configured to extract a first feature point sequence and a first curvature information sequence from the geometric feature information; perform interpolation processing on the first feature point sequence and the first curvature information sequence respectively based on the confidence assessment information to obtain a second feature point sequence and a second curvature information sequence; and determine the lane center line based on the second feature point sequence and the second curvature information sequence.

[0101] The present application also provides a computer readable storage medium having stored computer executable instructions or computer programs, which when executed by a processor, cause the processor to perform any step of the vehicle curve driving control method provided by the present application.

[0102] In some embodiments, the computer-readable storage media can be a memory such as a random access memory (RAM), a read-only memory (ROM), a flash memory, a magnetic surface memory, an optical disc, or a Compact Disc Read-Only Memory (CD-ROM), and the like. The computer-readable storage media can also be various devices including one or any combination of the above-mentioned memory.

[0103] In some embodiments, the computer-executable instructions can take the form of programs, software, software modules, scripts, or code, written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages; and can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0104] In some embodiments, the computer-executable instructions can, but need not, correspond to a file in a file system. The computer-executable instructions can be stored in a portion of a file that holds other programs or data, for example, one or more scripts stored in a Hypertext Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files, for example, files that store one or more modules, sub programs, or portions of code.

[0105] In some embodiments, the computer-executable instructions can be deployed to be executed on one electronic device, or on multiple electronic devices that are located at one site, or on multiple electronic devices that are distributed across multiple sites and interconnected through a communication network.

[0106] As shown in Figure 3 The present application also provides an electronic device 30, which includes a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, any step of the vehicle cornering control method described above is implemented.

[0107] The present application also provides a computer program product, which includes a computer program or computer-executable instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer program or computer-executable instructions from the computer-readable storage medium, and executes the computer program or computer-executable instructions, so that the electronic device performs any step of the vehicle cornering control method described above.

[0108] The above examples are only used to illustrate the technical solutions of the present application, but not to limit the same; although the present application has been described in detail with reference to the foregoing examples, it should be understood by those of ordinary skill in the art that: it can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for controlling vehicle cornering, characterized in that, Applied to the target vehicle, the vehicle cornering control method includes: Obtain the vehicle speed information, vehicle location point, and radius of curvature of the lane where the target vehicle is located; Based on the radius of curvature, determine the centerline of the lane where the target vehicle is located; Based on the vehicle positioning point and the lane centerline, determine the lateral deviation distance between the historical planning trajectory of the previous planning cycle and the lane centerline; Based on the lateral deviation distance and the vehicle speed information, the target planning trajectory for the current planning cycle is determined; Based on the target planned trajectory, the target vehicle is controlled to drive on curves.

2. The vehicle cornering control method according to claim 1, characterized in that, The step of determining the lateral deviation distance between the historical planning trajectory of the previous planning period and the lane centerline based on the vehicle positioning point and the lane centerline includes: The vehicle positioning point is vertically projected onto the historical planned trajectory to obtain the first projection point; The first projection point is vertically projected onto the center line of the lane to obtain the second projection point; The lateral deviation distance is determined based on the distance between the first projection point and the second projection point.

3. The vehicle cornering control method according to claim 2, characterized in that, The step of determining the target planning trajectory for the current planning period based on the lateral deviation distance and the vehicle speed information includes: Based on the vehicle speed information, determine the first translation distance corresponding to the current planning cycle; The second translation distance is determined based on the product of the lateral deviation distance and the first coefficient; The smaller of the first translation distance and the second translation distance is determined as the target translation distance; The target's planned trajectory is determined based on the target's translational distance and the lateral deviation distance.

4. The vehicle cornering control method according to claim 3, characterized in that, Determining the first translation distance corresponding to the current planning period based on the vehicle speed information includes: The third translation distance is determined based on the product of the vehicle speed information and the second coefficient; If the third translation distance is within a preset distance range, then the third translation distance is determined as the first translation distance; otherwise, the boundary value in the preset distance range that matches the third translation distance is determined as the first translation distance.

5. The vehicle cornering control method according to claim 4, characterized in that, Before determining the third translation distance based on the product of the vehicle speed information and the second coefficient, the vehicle cornering control method further includes: The calibration coefficient value corresponding to the driving mode of the target vehicle is obtained from the preset parameter mapping table and used as the second coefficient. The calibration mapping table stores calibration coefficient values ​​corresponding to multiple driving modes, and the calibration coefficient value is used to calculate the third translation distance.

6. The vehicle cornering control method according to claim 3, characterized in that, Before determining the second translation distance based on the product of the lateral deviation distance and the first coefficient, the vehicle cornering control method further includes: The first coefficient is determined based on at least one of the lane width, road surface adhesion coefficient, traffic flow density, and radius of curvature of the lane where the target vehicle is located, wherein the first coefficient is inversely proportional to the radius of curvature, the lane width, and the traffic flow density, and is directly proportional to the road surface adhesion coefficient.

7. The vehicle cornering control method according to claim 3, characterized in that, Determining the target planned trajectory based on the target translation distance and the lateral deviation distance includes: The starting point for target planning is determined based on the target translation distance and the lateral deviation distance; Based on the target planning starting point, a target planning trajectory for the current planning period that converges to the lane centerline is generated.

8. The vehicle cornering control method according to claim 7, characterized in that, Determining the target planning starting point based on the target translation distance and the lateral deviation distance includes: The translation ratio is determined based on the ratio of the target translation distance to the lateral deviation distance; Based on the translation ratio, the target planning starting point is determined on the line connecting the first projection point and the second projection point.

9. The vehicle curve control method according to claim 7, characterized in that, The step of generating the target planning trajectory for the current planning period, which converges to the lane centerline, based on the target planning starting point, includes: The target planned trajectory is generated by solving the trajectory based on the target planning starting point and the lane centerline using an iterative linear quadratic regulator optimization algorithm.

10. The vehicle cornering control method according to any one of claims 1 to 9, characterized in that, The step of controlling the target vehicle to drive on curves based on the target planned trajectory includes: Based on the target trajectory, determine the desired yaw rate; A first steering control command is generated based on the first deviation between the desired yaw rate and the actual yaw rate of the target vehicle. The first steering control command is sent to the steering actuator of the target vehicle to drive the target vehicle to travel along the target planned trajectory; or, Based on the target planned trajectory, determine the desired front wheel steering angle; A second steering control command is generated based on the second deviation between the desired front wheel steering angle and the actual front wheel steering angle of the target vehicle. The second steering control command is sent to the steering actuator of the target vehicle to drive the target vehicle along the target planned trajectory.

11. The vehicle cornering control method according to any one of claims 1 to 9, characterized in that, Determining the lane centerline of the lane where the target vehicle is located based on the radius of curvature includes: When the radius of curvature is less than a preset radius threshold, lane line identification is performed on the lane where the target vehicle is located to obtain lane line information corresponding to the lane; Based on the lane line information, the center line of the lane where the target vehicle is located is determined.

12. The vehicle cornering control method according to claim 11, characterized in that, The lane line information includes the geometric feature information and confidence assessment information of the lane lines; Determining the lane centerline of the lane where the target vehicle is located based on the lane line information includes: The first feature point sequence and the first curvature information sequence are extracted from the geometric feature information; Based on the confidence assessment information, the first feature point sequence and the first curvature information sequence are interpolated to obtain the second feature point sequence and the second curvature information sequence. The lane centerline is determined based on the second feature point sequence and the second curvature information sequence.

13. A vehicle curve control device, characterized in that, Applied to the target vehicle, the vehicle cornering control device includes: The information acquisition unit is used to acquire the vehicle speed information, vehicle positioning point, and radius of curvature of the lane where the target vehicle is located; A centerline determination unit is used to determine the lane centerline of the lane where the target vehicle is located based on the radius of curvature. The deviation determination unit is used to determine the lateral deviation distance between the historical planning trajectory of the previous planning cycle and the lane centerline based on the vehicle positioning point and the lane centerline; The trajectory determination unit is used to determine the target planning trajectory for the current planning cycle based on the lateral deviation distance and the vehicle speed information. The vehicle control unit is used to control the target vehicle to drive on curves based on the target planned trajectory.

14. An electronic device comprising: The memory and processor are characterized in that the processor, when executing a computer program stored in the memory, implements the steps of the vehicle cornering control method as described in any one of claims 1 to 12.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the vehicle cornering control method as described in any one of claims 1 to 12.

Images