Method and device for predicting driving trajectory, vehicle and storage medium

Abstract

The application discloses a driving track prediction method and device, a vehicle and a storage medium. The driving track prediction method comprises the following steps: acquiring a first historical track of a self-vehicle in a driving process, a lateral prediction track, currently collected road information and a second historical track of a target vehicle followed; determining a first weight corresponding to the first historical track, a second weight corresponding to the road information, a third weight corresponding to the second historical track and a fourth weight corresponding to the lateral prediction track according to a current driving road scene; and predicting a driving track of the self-vehicle. The method combines the historical track of the self-vehicle, the road information, the lateral prediction track and the historical track of the target vehicle followed to jointly predict the driving track of the self-vehicle, overcomes the lateral driving defects caused by the ACC technology longitudinal driving and avoids the driving track prediction errors caused by unclear road information.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, and more specifically, to a method, apparatus, vehicle, and storage medium for predicting driving trajectories. Background Technology

[0002] In recent years, with the continuous development of the automotive industry, the popularity of driver assistance systems has been increasing, and more and more models are equipped with them. Adaptive Cruise Control (ACC) technology, as a member of driver assistance systems, has received widespread attention. ACC can reduce driver fatigue in longitudinal control and is of great significance for improving driving safety.

[0003] With the development of the times, ACC has gradually evolved from high-speed scenarios to be applicable to urban roads. The traffic conditions on urban roads are complex, and simply using ACC for vehicle control cannot meet the driving needs on urban roads. Summary of the Invention

[0004] In view of the above problems, this application proposes a method, device, vehicle, and storage medium for predicting driving trajectory.

[0005] In a first aspect, embodiments of this application provide a method for predicting driving trajectories. The method includes: acquiring a first historical trajectory of a vehicle during its driving process, a lateral predicted trajectory based on the wheel angles of the vehicle, road information currently collected by the vehicle, and a second historical trajectory of a target vehicle being followed by the vehicle; determining a first weight corresponding to the first historical trajectory, a second weight corresponding to the road information, a third weight corresponding to the second historical trajectory, and a fourth weight corresponding to the lateral predicted trajectory based on the road scene currently being driven by the vehicle, wherein the road scene is determined by the road information collected by the vehicle in real time; and predicting the driving trajectory of the vehicle based on the first historical trajectory, the first weight, the road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight.

[0006] In an optional embodiment, predicting the vehicle's driving trajectory based on the first historical trajectory, first weight, road information, second weight, second historical trajectory, third weight, lateral predicted trajectory, and fourth weight includes: optimizing the parameters in the target trajectory prediction algorithm based on the first historical trajectory, first weight, road information, second weight, second historical trajectory, third weight, lateral predicted trajectory, and fourth weight to obtain target parameters corresponding to the target trajectory prediction algorithm; and determining the vehicle's driving trajectory based on the target parameters and the target trajectory prediction algorithm.

[0007] In an optional embodiment, optimizing the parameters in the target trajectory prediction algorithm based on the first historical trajectory, first weight, road information, second weight, second historical trajectory, third weight, lateral predicted trajectory, and fourth weight to obtain the target parameters corresponding to the target trajectory prediction algorithm includes: optimizing the parameters in the target trajectory prediction algorithm based on the first historical trajectory to obtain first optimized parameters; optimizing the parameters in the target trajectory prediction algorithm based on the road information to obtain second optimized parameters; optimizing the parameters in the target trajectory prediction algorithm based on the second historical trajectory to obtain third optimized parameters; optimizing the parameters in the target trajectory prediction algorithm based on the lateral predicted trajectory to obtain fourth optimized parameters; and determining the target parameters corresponding to the target trajectory prediction algorithm based on the first weight, first optimized parameters, second weight, second optimized parameters, third weight, third optimized parameters, fourth weight, and fourth optimized parameters.

[0008] In one optional embodiment, the lateral predicted trajectory is obtained by: obtaining the wheel angle and yaw rate of the vehicle; determining the radius of curvature of the vehicle at the current moment based on the wheel angle and the yaw rate; and obtaining the lateral predicted trajectory of the vehicle based on the radius of curvature and the wheel angle.

[0009] In an optional embodiment, before acquiring the wheel angle and yaw rate of the vehicle, the method further includes: acquiring vehicle data of the vehicle at the current moment, the vehicle data including the front wheel sideslip stiffness, rear wheel sideslip stiffness, distance from the front axle to the center of gravity of the vehicle, distance from the rear axle to the center of gravity of the vehicle, mass of the vehicle, oblique velocity at the center of gravity, moment of inertia, and sideslip angle at the center of gravity; determining the state data of the vehicle at the current moment based on the vehicle data; and determining the yaw rate of the vehicle at the current moment based on the state data and a target data processing algorithm.

[0010] In an optional embodiment, the first historical trajectory is obtained by: obtaining multiple historical trajectory points of the vehicle, each historical trajectory point being determined based on the vehicle coordinate system at each historical moment; converting the multiple historical trajectory points into historical trajectory points in the current vehicle coordinate system based on the current vehicle coordinate system to obtain a target historical trajectory point; and determining the first historical trajectory based on the target historical trajectory point.

[0011] In an optional embodiment, after predicting the vehicle's trajectory based on the first historical trajectory, the road information, the second historical trajectory, and the lateral predicted trajectory, the method further includes:

[0012] A target range is obtained, which is used for the subsequent driving plan of the vehicle; based on the target range, the driving trajectories of other vehicles within the target range of the driving trajectory are obtained as target driving trajectories; the subsequent driving plan of the vehicle is determined according to the driving trajectory of the vehicle and the target driving trajectory.

[0013] Secondly, embodiments of this application provide a driving trajectory prediction device, the device comprising: a data acquisition module, configured to acquire a first historical trajectory of a vehicle during driving, a lateral predicted trajectory based on the wheel angle of the vehicle, road information currently collected by the vehicle, and a second historical trajectory of a target vehicle followed by the vehicle; a weight determination module, configured to determine a first weight corresponding to the first historical trajectory, a second weight corresponding to the road information, a third weight corresponding to the second historical trajectory, and a fourth weight corresponding to the lateral predicted trajectory based on the road scene currently being driven by the vehicle, wherein the road scene is determined by the road information collected by the vehicle in real time; and a driving trajectory prediction module, configured to predict the driving trajectory of the vehicle based on the first historical trajectory, the first weight, the road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight.

[0014] Thirdly, embodiments of this application provide a vehicle, including: one or more processors; a memory; and one or more application programs, wherein the one or more application programs are stored in the memory and configured to be executed by the one or more processors, and the one or more application programs are configured to perform the driving trajectory prediction method provided in the first aspect above.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium storing program code, which can be called by a processor to execute the driving trajectory prediction method provided in the first aspect above.

[0016] The solution provided in this application combines the vehicle's historical trajectory, the first weight corresponding to the vehicle's historical trajectory, road information, the second weight corresponding to the road information, the lateral predicted trajectory, the third weight corresponding to the lateral predicted trajectory, the historical trajectory of the target vehicle being followed, and the fourth weight of the historical trajectory of the target vehicle being followed to jointly predict the vehicle's driving trajectory. This not only overcomes the lateral driving defects caused by the longitudinal driving of ACC technology, but also incorporates the historical trajectory of the target vehicle being followed and road information into the predicted driving trajectory. When the historical trajectory of the target vehicle and road information are combined, it can avoid driving trajectory prediction errors caused by unclear road information, and it can also avoid the phenomenon of inaccurate driving trajectory prediction when there is no target vehicle following on the driving road, thereby improving the accuracy of driving trajectory prediction and improving the driver's driving experience. Moreover, different weights are determined according to the current road scenario in which the vehicle is driving, so that the predicted vehicle trajectory can be more closely related to different road scenarios, thus making the predicted vehicle trajectory more accurate. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating a method for predicting driving trajectories provided in an embodiment of this application is shown.

[0019] Figure 2 A flowchart illustrating a method for predicting driving trajectories provided in another embodiment of this application is shown.

[0020] Figure 3 A flowchart illustrating a method for predicting driving trajectories according to another embodiment of this application is shown.

[0021] Figure 4 A structural block diagram of a driving trajectory prediction device provided in an embodiment of this application is shown.

[0022] Figure 5 A structural block diagram of a vehicle for performing a driving trajectory prediction method according to an embodiment of this application is shown. Detailed Implementation

[0023] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0024] In related technologies, when using ACC technology to predict driving trajectories on urban roads, the vehicle's historical trajectory and road information are taken into account. However, this technology cannot meet the needs of situations where lane lines are unclear or missing, which can easily lead to a large difference between the predicted driving trajectory and the actual trajectory.

[0025] To address the background technology and the aforementioned technical problems, the inventors have proposed a method, device, vehicle, and storage medium for predicting driving trajectories. This method combines the vehicle's historical trajectory, a first weight corresponding to the historical trajectory, road information, a second weight corresponding to the road information, a lateral predicted trajectory, a third weight corresponding to the lateral predicted trajectory, the historical trajectory of the target vehicle being followed, and a fourth weight corresponding to the historical trajectory of the target vehicle being followed, to jointly predict the vehicle's driving trajectory. This not only overcomes the lateral travel defects caused by the longitudinal movement of ACC technology, but also incorporates the historical trajectory of the target vehicle being followed and road information into the predicted driving trajectory. When the target vehicle's historical trajectory and road information are combined, it avoids driving trajectory prediction errors caused by unclear road information and also avoids inaccurate driving trajectory prediction when there is no target vehicle following on the road. This improves the accuracy of driving trajectory prediction and enhances the driver's driving experience. Furthermore, by determining different weights for different trajectories or information based on the current road scenario, the predicted vehicle trajectory can be more closely approximated to different road scenarios, resulting in a more accurate predicted vehicle trajectory.

[0026] Please see Figure 1 , Figure 1 A flowchart illustrating a driving trajectory prediction method according to an embodiment of this application is shown. In a specific embodiment, the driving trajectory prediction method is applied to, for example... Figure 4 The driving trajectory prediction device 200 shown and the vehicle equipped with the driving trajectory prediction device 200.

[0027] The following will address... Figure 1 The process shown will be described in detail. The method for predicting the driving trajectory may specifically include the following steps:

[0028] Step S110: Obtain the first historical trajectory of the vehicle during its driving process, the lateral predicted trajectory based on the wheel angle of the vehicle, the road information currently collected by the vehicle, and the second historical trajectory of the target vehicle being followed by the vehicle.

[0029] The first historical trajectory can be the trajectory traveled from the moment the vehicle starts to the present moment, or it can be the historical driving trajectory within a preset time period. The length of the historical driving trajectory is not specifically limited here.

[0030] Wheel angle refers to the yaw angle of the front wheels of a vehicle. Wheel angle can be obtained by wheel angle sensors, or by first obtaining the steering wheel angle using a steering wheel angle sensor, and then determining the wheel angle based on the correspondence between the steering wheel angle and the wheel angle. No specific limitation is made here on the method of obtaining wheel angle.

[0031] During car driving, the horizontal direction of the vehicle's movement, or lateral direction, is determined by the wheel angles. While Adaptive Cruise Control (ACC) technology only controls the vehicle's longitudinal movement, the vehicle's trajectory involves both longitudinal and lateral travel. If the prediction of the vehicle's trajectory only considers longitudinal movement, the predicted trajectory may not adapt to the driver's driving habits, reducing the driving experience. Therefore, a better approach is to first predict the vehicle's lateral trajectory based on the wheel angles, and then reference the driver-controlled lateral trajectory during the overall trajectory prediction. This ensures that the final predicted trajectory better matches the driver's driving habits, improving the overall driving experience.

[0032] The method for predicting a vehicle's lateral trajectory based on its wheel angles can be as follows: First, input the wheel angles from historical trajectories into a lateral trajectory prediction model pre-trained using the wheel angles of sample vehicles. Second, input the wheel angles from historical trajectories into the lateral trajectory prediction equation. Third, obtain the current driving radius based on the vehicle's current wheel angle and determine the lateral trajectory accordingly. No specific limitations are imposed on the method of obtaining the lateral trajectory.

[0033] Road information includes, but is not limited to, road boundary information, lane center lines, and lane boundary lines. Road information can be acquired through image acquisition devices (cameras or radar devices) on the vehicle, or through specific lane information displayed in navigation software. No limitation is made on the method of acquiring road information here.

[0034] The acquisition of the second historical trajectory of the target vehicle being followed is achieved through vehicle-to-vehicle communication. This communication method can be Vehicle-to-Everything (V2X), which enables communication between vehicles, between vehicles and roads, and between vehicles and pedestrians. Another method is the Vehicle Internet of Things (VIoT). VIoT connects vehicles to surrounding smart devices (other vehicles or infrastructure with network communication capabilities) to achieve interactive communication.

[0035] Since the vehicle is following more than one target vehicle in the first historical trajectory, the length of the target vehicle's second historical trajectory can be different from the length of the first historical trajectory. Here, we do not limit the length of the second historical trajectory or the method of obtaining it.

[0036] Step S120: Based on the road scene where the vehicle is currently traveling, determine the first weight corresponding to the first historical trajectory, the second weight corresponding to the road information, the third weight corresponding to the second historical trajectory, and the fourth weight corresponding to the lateral predicted trajectory. The road scene is determined by the road information collected by the vehicle in real time.

[0037] In this scenario, the first, second, third, and fourth weights differ depending on the road conditions. In the straight-ahead scenario, when the lane centerline obtained from the road information is unclear, the road information has limited reference value for predicting the vehicle's trajectory; therefore, the second weight corresponding to the road information is reduced. In this case, the prediction of the vehicle's trajectory focuses more on the target vehicle's trajectory and the vehicle's historical trajectory; therefore, the first and third weights are increased. In the turning scenario, road information is scarce or nonexistent, and the target vehicle only provides longitudinal following; the turn is controlled by the driver. Therefore, the second and third weights are reduced, and the first and fourth weights are increased. In the intersection scenario, the road information is similar to that in the turning scenario; therefore, the second weight is reduced. However, in the intersection scenario, the vehicle's movement is significantly affected by the target vehicle's trajectory; therefore, the third weight can be increased. It should be noted that in this embodiment, initial values ​​are set for the first, second, third, and fourth weights. These initial values ​​can be specifically set by the driver or are default settings. The initial values ​​of the first, second, third, and fourth weights are not specifically defined here.

[0038] Furthermore, road scenarios include, but are not limited to, straight-ahead scenarios, turning scenarios, and intersection scenarios. When multiple consecutive lane centerlines in the road information are straight, the current road scenario is determined to be a straight-ahead scenario. When multiple consecutive lane centerlines in the road information are curved, the current road scenario is determined to be a turning scenario. When no lane line data is collected in the road information, and there are multiple intersections, the current road scenario is determined to be an intersection scenario. Road scenarios can also be determined directly based on the data collection device on the vehicle; therefore, the specific method for determining road scenarios is not limited here.

[0039] Step S130: Based on the first historical trajectory, the first weight, the road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight, predict the driving trajectory of the vehicle.

[0040] In this embodiment of the application, predicting the driving trajectory of a vehicle can be achieved by inputting a first historical trajectory, a first weight, road information, a second weight, a second historical trajectory, a third weight, a lateral predicted trajectory, and a fourth weight into a pre-trained trajectory prediction model to predict the driving trajectory. The trajectory prediction model is trained based on the first historical trajectory, the first weight, road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight of the sample vehicle.

[0041] Furthermore, predicting the vehicle's trajectory can also involve first using the first historical trajectory, road information, the second historical trajectory, and the lateral prediction trajectory to predict the vehicle's trajectory, resulting in four vehicle trajectories. Then, based on the weight corresponding to each trajectory, the final trajectory can be predicted.

[0042] In some implementations, the prediction method for the driving trajectory may involve inputting multiple coordinate points in the vehicle's current coordinate system from the first historical driving trajectory, road information, second historical trajectory, and lateral prediction trajectory into the trajectory equation, and then generating the corresponding driving trajectory based on the corresponding weights, which serves as the vehicle's predicted driving trajectory. No specific limitations are imposed on the method of driving trajectory prediction here.

[0043] The driving trajectory prediction method provided in this application combines the vehicle's historical trajectory, the first weight corresponding to the vehicle's historical trajectory, road information, the second weight corresponding to the road information, the lateral predicted trajectory, the third weight corresponding to the lateral predicted trajectory, the historical trajectory of the target vehicle being followed, and the fourth weight of the historical trajectory of the target vehicle being followed to jointly predict the vehicle's driving trajectory. This not only overcomes the lateral driving defects caused by the longitudinal driving of ACC technology, but also incorporates the historical trajectory of the target vehicle being followed and road information into the predicted driving trajectory. When the historical trajectory of the target vehicle and road information are combined, it can avoid driving trajectory prediction errors caused by unclear road information, and can also avoid inaccurate driving trajectory prediction when there is no target vehicle following on the driving road, thereby improving the accuracy of driving trajectory prediction and improving the driver's driving experience. Moreover, different weights corresponding to different trajectories or information are determined according to the current road scenario in which the vehicle is driving, so that the predicted vehicle trajectory can be closer to different road scenarios, thereby making the predicted vehicle trajectory more accurate.

[0044] Please see Figure 2 , Figure 2 A flowchart illustrating a method for predicting driving trajectories according to another embodiment of this application is shown. The following will focus on... Figure 2 The process shown will be described in detail. The method for predicting the driving trajectory may specifically include the following steps:

[0045] Step S210: Obtain the wheel rotation angle and yaw rate of the vehicle.

[0046] Yaw rate refers to the speed at which a vehicle rotates around its vertical axis during cornering, and it can be used to assess vehicle stability and handling performance. Yaw rate is related to the vehicle's speed and lateral acceleration during cornering. Lateral acceleration is related to the forces acting on the vehicle during cornering and the vehicle's mass. Therefore, the yaw rate of a vehicle can be obtained by acquiring the lateral acceleration during cornering using an accelerometer, and then calculating the yaw rate based on the vehicle's current speed. Alternatively, it can be obtained by first acquiring the lateral acceleration based on the forces acting on the vehicle during cornering and the vehicle's mass, and then calculating the yaw rate based on the vehicle's current speed. No specific limitations are placed on the method of obtaining the yaw rate here.

[0047] In some implementations, determining the yaw rate includes: acquiring vehicle data for the vehicle at the current moment, the vehicle data including the front wheel sideslip stiffness, rear wheel sideslip stiffness, distance from the front axle to the vehicle's center of gravity, distance from the rear axle to the center of gravity, the vehicle's mass, oblique velocity at the center of gravity, moment of inertia, and sideslip angle at the center of gravity. Based on the vehicle data, determining the vehicle's state data for the current moment. Based on the state data and a target data processing algorithm, determining the vehicle's yaw rate for the current moment.

[0048] Front wheel lateral stiffness refers to the change in the slip angle of the front tire under lateral force. Rear wheel lateral stiffness is the ratio of the change in the slip angle of the rear tire under lateral force to the lateral force itself. The oblique velocity at the center of mass refers to the velocity component in a plane perpendicular to a given axis when the vehicle rotates around its center of mass on a certain axis. Moment of inertia refers to the vehicle's ability to resist angular acceleration when rotating around its center of mass on a certain axis. The slip angle at the center of mass is a measure of the degree of deflection of the vehicle in a certain direction.

[0049] The vehicle's current state data can be directly used as its current state data, or the current vehicle data can be transformed to obtain the corresponding state data. The yaw rate of the vehicle is then determined based on the state data. State data better reflects the dynamic characteristics of the vehicle at the current moment.

[0050] In one feasible implementation, the purpose of obtaining the yaw rate is to predict the lateral trajectory of the vehicle. This can be achieved by transforming the current vehicle data into state data, which is then defined as a two-degree-of-freedom equation for the vehicle. This equation considers not only the vehicle's lateral motion but also its yaw motion.

[0051] The vehicle's two-degree-of-freedom equations are as follows:

[0052]

[0053]

[0054] Where k1 is the front wheel sideslip stiffness, k2 is the rear wheel sideslip stiffness, a is the distance from the front axle to the vehicle's center of gravity, b is the distance from the rear axle to the vehicle's center of gravity, m is the vehicle's mass, u is the tangential velocity at the vehicle's center of gravity, Iz is the moment of inertia, β is the sideslip angle at the center of gravity, and ω... r δ represents the yaw rate, and δ represents the wheel rotation angle.

[0055] The above formulas (1) and (2) can be used to determine the yaw rate of the vehicle.

[0056] Furthermore, during vehicle operation, noise is generated due to vibrations, electromagnetic interference between various controllers, and manual adjustments to the steering wheel. To improve the accuracy of the yaw rate, a target data processing algorithm is needed to further process the vehicle's two-degree-of-freedom equations. This target data processing algorithm can be a noise filtering algorithm, such as a Kalman filter, median filter, or mean filter. In this embodiment, the target data processing algorithm is preferably a Kalman filter.

[0057] The Kalman filter algorithm includes the following equations:

[0058] State prediction equation:

[0059]

[0060] Predictive covariance matrix equation:

[0061] P k =AP k―1 A T +Q (4)

[0062] Current optimal estimation equation update equation:

[0063]

[0064] Kalman gain matrix equation:

[0065] K = P k H T HP k H T +R) ―1 (6)

[0066] Error covariance matrix update equation:

[0067] P′ k =P k ―KHPk (7)

[0068] Where A is the vehicle state matrix, B is the coefficient matrix of the vehicle state, Q is the noise covariance matrix, zk is the observation value, H is the observation matrix, and R is the observation covariance matrix of the actual vehicle state.

[0069] By inputting the above two-degree-of-freedom equations of the vehicle as the vehicle's state data into formulas (3)(4)(5)(6)(7), a more accurate yaw rate can be obtained.

[0070] Step S220: Determine the radius of curvature of the vehicle at the current moment based on the wheel rotation angle and the yaw rate.

[0071] Step S230: Obtain the lateral predicted trajectory of the vehicle based on the radius of curvature and the wheel angle.

[0072] In this embodiment, the road curvature corresponding to the vehicle at the current moment can be obtained based on the wheel rotation angle and yaw rate. Then, based on the correspondence between road curvature and radius of curvature, the radius of curvature of the vehicle at the current moment is determined. Finally, the lateral predicted trajectory of the vehicle is determined based on the radius of curvature of the vehicle at the current moment and the direction of the wheel rotation angle.

[0073] In one feasible implementation, the formula for obtaining road curvature can be:

[0074]

[0075] In another feasible implementation, the formula for obtaining road curvature can also be:

[0076]

[0077] Where L is the wheelbase of the vehicle and v is the vehicle speed.

[0078] During low-speed driving, noise has little impact on wheel angle, so the road curvature can be obtained according to formula (8). During high-speed driving, noise has a greater impact on wheel angle, and using formula (8) to obtain road curvature is prone to large errors. Since the yaw rate in this embodiment is obtained after processing the noise, using the yaw rate to calculate the road curvature at high speed is more accurate.

[0079] Based on the obtained correspondence between road curvature and radius of curvature, the radius of curvature of the vehicle at the current moment is determined. The correspondence is as follows:

[0080]

[0081] According to formula (10), the radius of curvature of the vehicle at the current moment is determined.

[0082] The lateral predicted trajectory of the vehicle is determined based on the radius of curvature and the vehicle's coordinate system at the current moment. The coordinate system can be established with the vehicle's center at the moment of startup as the origin, or it can be established with the vehicle's center at each moment as the center. No specific limitation is made on the coordinate system here.

[0083] The predicted lateral trajectory is as follows:

[0084]

[0085] In formula (11), the sign of y can be determined based on the magnitude or direction of the wheel angle. The magnitude or direction of the wheel angle determines whether the vehicle is turning left or right.

[0086] Step S240: Obtain the first historical trajectory of the vehicle during its driving process, the lateral predicted trajectory based on the wheel angle of the vehicle, the road information currently collected by the vehicle, and the second historical trajectory of the target vehicle being followed by the vehicle.

[0087] For a detailed explanation of step S240, please refer to step S110 in the foregoing embodiments, which will not be repeated here.

[0088] In some implementations, obtaining the first historical trajectory may include: acquiring multiple historical trajectory points of the vehicle, each historical trajectory point being determined based on the vehicle's coordinate system at each historical moment; converting the multiple historical trajectory points into historical trajectory points in the current vehicle coordinate system based on the current moment's vehicle coordinate system to obtain a target historical trajectory point; and determining the first historical trajectory based on the target historical trajectory point.

[0089] In this embodiment, the coordinate system is preferably established with the center of the vehicle's rear axle as the origin. Since the position of the origin changes during vehicle movement, multiple historical trajectory points are converted to the current coordinate system to improve trajectory prediction efficiency.

[0090] Transformation of a spatial coordinate system consists of two parts: translational coordinate transformation and rotational coordinate transformation. Transformation in a two-dimensional coordinate system only considers translational coordinate transformation.

[0091] For example, taking a spatial coordinate system as an example, the position vector of point P in the historical coordinate system {B} is: B p. Displacement vector A p B The vector representing the position of point P in the current coordinate system {A}, used to describe the position of the historical coordinate system {B} relative to the current coordinate system {A}. Ap can be represented as:

[0092] A p = B p+ A p B (12)

[0093] For coordinate rotation, it can be used Representing rotational transformations between coordinate systems:

[0094]

[0095] Combining translation and rotation transformations can be achieved using a transformation matrix. express:

[0096]

[0097] In this embodiment, the coordinate system at each moment is a local coordinate system. The positioning module in the vehicle can obtain the translation amount in the three directions of x, y, and z axes relative to the starting coordinate system at each moment, as well as the quaternion that can be used for coordinate rotation transformation. The quaternion can be converted into a rotation matrix.

[0098] For the coordinate transformation of a moving vehicle between adjacent moments... i-1 T represents the transformation matrix of the vehicle relative to the world coordinate system at time i-1. i T represents the transformation matrix of the vehicle at time i relative to the world coordinate system. Then, the coordinate transformation at time i-1 relative to time i is:

[0099]

[0100] According to formulas (12)(13)(14)(15), the trajectory coordinates of the vehicle in the historical coordinate system are converted into coordinates in the current coordinate system, and the historical trajectory of the vehicle is obtained based on the converted coordinates and the time corresponding to the converted coordinates.

[0101] The historical trajectory of the target vehicle can be obtained according to the method of obtaining the historical trajectory of the vehicle itself, which will not be elaborated here.

[0102] Furthermore, after transforming the coordinates of adjacent moments, the historical trajectory lengths of the self-vehicle and the target vehicle are recorded according to different scenarios. The self-vehicle's historical trajectory can be recorded at fixed time intervals, such as recording 8 or 12 historical trajectories per cycle (100ms) or 1 second. Alternatively, it can be recorded according to the required trajectory length L (m). Assuming the self-vehicle moves at a constant speed v (m / s) at the current moment, and the trajectory is recorded once per calculation cycle T (e.g., 100ms), then the required number of recording cycles is:

[0103] N = max(3, L / v * 10)

[0104] The number 10 indicates that there are 10 cycles within 1 second (e.g., T = 100ms), and the minimum number of records is set to 3 (this is just an example and can be adjusted according to the actual situation).

[0105] Step S250: Based on the first historical trajectory, the first weight, the road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight, optimize the parameters in the target trajectory prediction algorithm to obtain the target parameters corresponding to the target trajectory prediction algorithm.

[0106] The target trajectory prediction algorithm refers to an algorithm for predicting the driving trajectory of a vehicle. This algorithm can be a road equation, where parameters are determined based on a first historical trajectory, a first weight, road information, a second weight, a second historical trajectory, a third weight, a lateral predicted trajectory, and a fourth weight, thereby determining the road equation and predicting the vehicle's trajectory. Alternatively, the target trajectory prediction algorithm can be a neural network algorithm, where parameters are determined based on the same parameters, and the neural network algorithm then predicts the vehicle's trajectory. No specific limitations are imposed on the target trajectory prediction algorithm here.

[0107] The target parameters can be obtained by comprehensively calculating the parameters corresponding to the first historical trajectory, first weight, road information, second weight, second historical trajectory, third weight, lateral predicted trajectory, and fourth weight, based on the first historical trajectory, first weight, road information, second weight, second historical trajectory, third weight, lateral predicted trajectory, and fourth weight. Alternatively, the target trajectory prediction algorithm can be used to calculate the corresponding parameters separately using the first historical trajectory, road information, second historical trajectory, and lateral predicted trajectory. Then, the target parameters are obtained by combining the first weight, second weight, third weight, and fourth weight. The target parameters can be the sum of the products of each weight and its corresponding parameter, or simply the sum of each weight and its corresponding parameter; no specific limitation is placed on the method for determining the target parameters here.

[0108] Furthermore, the required lengths of the first historical trajectory, the second historical trajectory, and the lateral prediction trajectory differ depending on the road scenario.

[0109] For example, when the road scenario is a straight-ahead scenario, there is little demand for lateral travel. In this scenario, the vehicle's travel depends more on road information and the target vehicle's historical trajectory. Therefore, the length of the lateral predicted trajectory can be reduced, while the lengths of the second historical trajectory and the first historical trajectory can be appropriately increased.

[0110] When the road scene is a turning scene, the demand for lateral predicted trajectory increases, while the demand for the second historical trajectory of the target vehicle decreases. In this case, the length of the lateral predicted trajectory is increased and the length of the second historical trajectory is decreased.

[0111] When the road scenario is an intersection scenario, the demand for lateral travel is not significant compared to the straight-through scenario. Therefore, the length of the lateral predicted trajectory is reduced, and the length of the second historical trajectory is increased.

[0112] In some implementations, the parameters in the target trajectory prediction algorithm are optimized based on the first historical trajectory, the first weight, road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight to obtain the target parameters corresponding to the target trajectory prediction algorithm. Specifically, this includes: optimizing the parameters in the target trajectory prediction algorithm based on the first historical trajectory to obtain first optimized parameters; optimizing the parameters in the target trajectory prediction algorithm based on the road information to obtain second optimized parameters; optimizing the parameters in the target trajectory prediction algorithm based on the second historical trajectory to obtain third optimized parameters; optimizing the parameters in the target trajectory prediction algorithm based on the lateral predicted trajectory to obtain fourth optimized parameters; and determining the target parameters corresponding to the target trajectory prediction algorithm based on the first weight, the first optimized parameters, the second weight, the second optimized parameters, the third weight, the third optimized parameters, the fourth weight, and the fourth optimized parameters.

[0113] In this embodiment, the target trajectory prediction algorithm is preferably a road equation. Since the vehicle's trajectory is a curve, the road equation is determined to be a higher-order equation so that it better matches the actual trajectory.

[0114] The first historical trajectory and its first weight, the road information and its second weight, the second historical trajectory and its third weight, and the lateral predicted trajectory and its fourth weight are respectively substituted into the road equation to obtain the first, second, third, and fourth optimization parameters. The optimization parameters can be the coefficients corresponding to the variables of each order in the road equation, or they can be the residual values ​​corresponding to the road equation. No specific limitations are imposed on the optimization parameters here.

[0115] For example, the road equation is y = a0x 3 +a1x 2 +a2x+a3, optimizes the residual values ​​corresponding to the road equation parameters. The trajectory points corresponding to the first historical trajectory, the road information in the road information, the trajectory points corresponding to the second historical trajectory, and the trajectory points corresponding to the lateral predicted trajectory are input into the road equation to construct the following residual equations:

[0116] (a0x i3 +a1x i 2 +a2x i +a3―y i (16)

[0117] Substituting the first historical trajectory into formula (16) yields the first optimized parameter. Substituting the road information into formula (16) yields the second optimized parameter. Substituting the second historical trajectory into formula (16) yields the third optimized parameter. Substituting the laterally predicted driving trajectory into formula (16) yields the fourth optimized parameter. The first, second, third, and fourth weights are used as the corresponding importance levels of the first, second, third, and fourth optimized parameters. Therefore, the target parameter is the sum of the products of the weights and the optimized parameters.

[0118] In other implementations, to avoid Runge's phenomenon, the coefficients of higher-order terms in the road equation need to be limited according to expected values.

[0119] For example, continuing with the road equation y = a0x 3 +a1x 2 Taking +a2x+a3 as an example, construct the residual equation based on the expected value and the actual value:

[0120] w 11 (a0′―a0) (17)

[0121] w 12 (a1′―a1) (18)

[0122] Where a0′ is the expected value of the coefficient of the cubic term, a1′ is the expected value of the coefficient of the quadratic term, and w 11 w represents the importance of the third-order term coefficients in the determination of the objective parameters in the residual equation. 12 The coefficients of the second-order terms represent the importance of the residual equation in determining the target parameters.

[0123] In determining the target parameters, the target parameters are determined according to formulas (16)(17)(18) and the first weight, second weight, third weight and fourth weight to avoid Runge phenomenon.

[0124] In some other implementations, the determination of the target parameters can also refer to the optimized historical target parameters, thereby making a comprehensive judgment on the target parameters.

[0125] For example, continuing with the road equation y = a0x 3 +a1x 2 Taking +a2x+a3 as an example, construct the residual equation based on the historical target parameters:

[0126]

[0127] in, is the historical target parameter, and w2 is the importance of this residual equation in the process of determining the target parameter.

[0128] In the process of determining the target parameters, the target parameters are determined according to formulas (16)(17)(18)(19) and the first weight, second weight, third weight and fourth weight, so that the road equation corresponding to the finally determined target parameters is closer to the real vehicle driving trajectory.

[0129] Step S260: Determine the driving trajectory of the vehicle based on the target parameters and the target trajectory prediction algorithm.

[0130] The vehicle's trajectory is determined by combining the target trajectory prediction algorithm and the target parameters corresponding to the target trajectory prediction algorithm.

[0131] The driving trajectory prediction method provided in this application, when predicting the driving trajectory of a vehicle, also determines the weights corresponding to the first historical trajectory, the second historical trajectory, road information, and the lateral prediction trajectory according to different road scenarios. This allows for flexible selection of the original parameters and constraints used to generate the vehicle's pre-driving reference line under different scenarios, better representing the road shape and avoiding the driving trajectory predicted under uniform weights from not conforming to the actual driving trajectory, thus improving the driver's driving experience.

[0132] Please see Figure 3 , Figure 3 A flowchart illustrating a method for predicting driving trajectories according to another embodiment of this application is shown. The following will focus on... Figure 3 The process shown will be described in detail. The method for predicting the driving trajectory may specifically include the following steps:

[0133] Step S310: Obtain the first historical trajectory of the vehicle during its driving process, the lateral predicted trajectory based on the wheel angle of the vehicle, the road information currently collected by the vehicle, and the second historical trajectory of the target vehicle being followed by the vehicle.

[0134] Step S320: Based on the road scene where the vehicle is currently traveling, determine the first weight corresponding to the first historical trajectory, the second weight corresponding to the road information, the third weight corresponding to the second historical trajectory, and the fourth weight corresponding to the lateral predicted trajectory. The road scene is determined by the road information collected by the vehicle in real time.

[0135] Step S330: Based on the first historical trajectory, the first weight, the road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight, predict the driving trajectory of the vehicle.

[0136] For a detailed explanation of steps S310 to S330, please refer to steps S110 to S130 in the foregoing embodiments, which will not be repeated here.

[0137] Step S340: Obtain the target range, which is used for the subsequent driving planning of the vehicle.

[0138] The target range can be a circular area with the vehicle's center as the origin and a preset distance as the radius, or a circular area with each trajectory point in the predicted driving trajectory as the origin and a preset distance as the radius, or a rectangular area with each trajectory point in the predicted driving trajectory as the midpoint or vertex and a preset distance as the side length. No specific limitations are placed on the shape and size of the target range.

[0139] Step S350: Based on the target range, obtain the driving trajectories of other vehicles within the target range of the driving trajectory, and use them as the target driving trajectory.

[0140] The driving trajectory can be a historical driving trajectory, a driving trajectory predicted by other vehicles according to the driving trajectory prediction method provided in the embodiments of this application, or a driving trajectory predicted according to other driving trajectory prediction methods. The determination of the driving trajectory is not specifically limited here.

[0141] The acquisition of driving trajectories can be achieved through communication between the two vehicles. This communication can be vehicle-to-everything (V2X) technology, which enables communication between vehicles, between vehicles and roads, and between vehicles and pedestrians. Another communication method is the Vehicle Internet of Things (VIoT). VIoT connects vehicles to surrounding smart devices (other vehicles or infrastructure with network communication capabilities) to achieve interactive communication.

[0142] Step S360: Determine the subsequent driving plan of the vehicle based on the vehicle's driving trajectory and the target driving trajectory.

[0143] In this embodiment, the target trajectory is preferably a predicted driving trajectory of other vehicles. Driving planning includes, but is not limited to, following, accelerating, and decelerating driving operations. By analyzing the target driving trajectories of other vehicles within the target range and the vehicle's own driving trajectory, it can be determined whether there are any intersection points between the target and vehicle's trajectories. If an intersection point exists, it indicates a future collision risk between the two vehicles. In this case, acceleration or deceleration can be determined based on the distance between the intersection point and the vehicle's current trajectory point to ensure driving safety. Furthermore, the number of points of overlap between the target and vehicle's trajectories can be used to determine the next vehicle to follow, thereby enabling the selection of following vehicles using predicted driving trajectories and improving the driver's driving experience.

[0144] In some implementations, the target range may include a first range and a second range, wherein the first range is smaller than the second range. The target driving trajectories of other vehicles within the first range can be used to determine the following vehicles, and the target driving trajectories of other vehicles within the second range can be used to determine whether there is a collision risk, thereby enabling acceleration or deceleration driving maneuvers.

[0145] The driving trajectory prediction method provided in this application not only predicts the driving trajectory of the vehicle, but also plans the future driving process of the vehicle according to the preset target range. This not only enables the selection of vehicles to follow during future driving, but also avoids vehicle collisions and improves the user's driving experience.

[0146] Please see Figure 4 This document illustrates a structural block diagram of a driving trajectory prediction device 200 provided in an embodiment of this application. The driving trajectory prediction device 200 is applied to a vehicle and includes: a data acquisition module 210, used to acquire a first historical trajectory of the vehicle during its driving process, a lateral predicted trajectory based on the vehicle's wheel angles, currently collected road information of the vehicle, and a second historical trajectory of a target vehicle being followed by the vehicle; a weight determination module 220, used to determine a first weight corresponding to the first historical trajectory, a second weight corresponding to the road information, a third weight corresponding to the second historical trajectory, and a fourth weight corresponding to the lateral predicted trajectory based on the road scene currently being driven by the vehicle, wherein the road scene is determined by the road information collected by the vehicle in real time; and a driving trajectory prediction module 230, used to predict the driving trajectory of the vehicle based on the first historical trajectory, the first weight, the road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight.

[0147] In some embodiments of this application, the driving trajectory prediction module 230 includes: a target parameter acquisition unit, used to optimize the parameters in the target trajectory prediction algorithm based on the first historical trajectory, the first weight, road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight, to obtain the target parameters corresponding to the target trajectory prediction algorithm; and a driving trajectory determination unit, used to determine the driving trajectory of the vehicle based on the target parameters and the target trajectory prediction algorithm.

[0148] In some embodiments of this application, the target parameter acquisition unit includes: a first optimization parameter acquisition subunit, used to optimize the parameters in the target trajectory prediction algorithm based on the first historical trajectory to obtain a first optimization parameter; a second optimization parameter acquisition subunit, used to optimize the parameters in the target trajectory prediction algorithm based on the road information to obtain a second optimization parameter; a third optimization parameter acquisition subunit, used to optimize the parameters in the target trajectory prediction algorithm based on the second historical trajectory to obtain a third optimization parameter; a fourth optimization parameter acquisition subunit, used to optimize the parameters in the target trajectory prediction algorithm based on the lateral prediction trajectory to obtain a fourth optimization parameter; and a target parameter determination subunit, used to determine the target parameters corresponding to the target trajectory prediction algorithm based on the first weight, the first optimization parameter, the second weight, the second optimization parameter, the third weight, the third optimization parameter, the fourth weight, and the fourth optimization parameter.

[0149] In some embodiments of this application, the driving trajectory prediction device 200 further includes: a data acquisition module for acquiring the wheel angle and yaw rate of the vehicle; a radius of curvature determination module for determining the radius of curvature of the vehicle at the current moment based on the wheel angle and the yaw rate; and a lateral predicted trajectory acquisition module for acquiring the lateral predicted trajectory of the vehicle based on the radius of curvature and the wheel angle.

[0150] In some embodiments of this application, the trajectory prediction device 200 further includes: a vehicle data acquisition module, used to acquire vehicle data of the vehicle at the current moment, the vehicle data including the front wheel sideslip stiffness, rear wheel sideslip stiffness, distance from the front axle to the center of gravity of the vehicle, distance from the rear axle to the center of gravity of the vehicle, mass of the vehicle, oblique velocity at the center of gravity, moment of inertia, and sideslip angle at the center of gravity; a state data determination module, used to determine the state data of the vehicle at the current moment based on the vehicle data; and a yaw rate determination module, used to determine the yaw rate of the vehicle at the current moment based on the state data and a target data processing algorithm.

[0151] In some embodiments of this application, the driving trajectory prediction device 200 further includes: a historical trajectory point acquisition module, used to acquire multiple historical trajectory points of the vehicle, each historical trajectory point being determined based on the vehicle coordinate system at each historical moment; a target historical trajectory point acquisition module, used to convert the multiple historical trajectory points into historical trajectory points in the vehicle coordinate system at the current moment, based on the vehicle coordinate system at the current moment, to obtain a target historical trajectory point; and a first historical trajectory determination module, used to determine the first historical trajectory based on the target historical trajectory point.

[0152] In some embodiments of this application, the driving trajectory prediction device 200 further includes: a target range acquisition module, used to acquire a target range, the target range being used for subsequent driving planning of the vehicle; a target driving trajectory acquisition module, used to acquire, based on the target range, the driving trajectories corresponding to other vehicles within the target range of the driving trajectory, as the target driving trajectory; and a driving planning module, used to determine the subsequent driving plan of the vehicle based on the driving trajectory of the vehicle and the target driving trajectory.

[0153] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described device and module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0154] In the several embodiments provided in this application, the coupling between modules can be electrical, mechanical, or other forms of coupling.

[0155] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0156] Please refer to Figure 5 This illustration shows a structural block diagram of a vehicle according to an embodiment of this application. The vehicle can be a switch, a computer, or a control unit with data transmission capabilities. The vehicle in this application may include one or more components such as a processor, a memory, and one or more application programs, wherein the one or more application programs can be stored in the memory and configured to be executed by one or more processors, and the one or more programs are configured to perform the methods described in the foregoing method embodiments.

[0157] A processor may include one or more processing cores. The processor connects to various parts of the vehicle via various interfaces and lines, executing instructions, programs, code sets, or instruction sets stored in memory, and calling data stored in memory to perform various vehicle functions and process data. Optionally, the processor may be implemented using at least one of the following hardware forms: Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). The processor may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing displayed content; and the modem handles wireless communication. It is understood that the modem may also be implemented separately as a communication chip, without being integrated into the processor.

[0158] The memory may include random access memory (RAM) or read-only memory (ROM). The memory can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), and instructions for implementing the various method embodiments described below. The data storage area may also store data generated during vehicle use (such as phonebook data, audio and video data, chat log data, etc.).

[0159] This application also provides a computer-readable storage medium storing program code, which can be called by a processor to execute the methods described in the above method embodiments.

[0160] Computer-readable storage media can be electronic storage devices such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Optionally, computer-readable storage media includes non-transitory computer-readable storage medium. The computer-readable storage medium has storage space for program code that performs any of the method steps described above. This program code can be read from or written to one or more computer program products. The program code can be compressed, for example, in a suitable form.

[0161] In summary, the solution provided in this application combines the vehicle's historical trajectory, the first weight corresponding to the vehicle's historical trajectory, road information, the second weight corresponding to the road information, the lateral predicted trajectory, the third weight corresponding to the lateral predicted trajectory, the historical trajectory of the target vehicle being followed, and the fourth weight of the historical trajectory of the target vehicle being followed to jointly predict the vehicle's driving trajectory. This not only overcomes the lateral driving defects caused by the longitudinal driving of ACC technology, but also incorporates the historical trajectory of the target vehicle being followed and road information into the predicted driving trajectory. When the historical trajectory of the target vehicle and road information are combined, it can avoid driving trajectory prediction errors caused by unclear road information, and it can also avoid the phenomenon of inaccurate driving trajectory prediction when there is no target vehicle following on the driving road, thereby improving the accuracy of driving trajectory prediction and improving the driver's driving experience. Moreover, by determining different weights corresponding to different trajectories or information based on the current road scenario of the vehicle's driving, the predicted vehicle trajectory can be more closely approximated to different road scenarios, thus making the predicted vehicle trajectory more accurate.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for predicting driving trajectories, characterized in that, The method includes: The system acquires the first historical trajectory of the vehicle during its driving process, the lateral predicted trajectory based on the wheel angle of the vehicle, the road information currently collected by the vehicle, and the second historical trajectory of the target vehicle being followed by the vehicle. Based on the road scenario currently being driven by the vehicle, a first weight corresponding to the first historical trajectory, a second weight corresponding to the road information, a third weight corresponding to the second historical trajectory, and a fourth weight corresponding to the lateral predicted trajectory are determined. The road scenario is determined by the road information collected by the vehicle in real time. The vehicle's driving trajectory is predicted based on the first historical trajectory, the first weight, road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight.

2. The method as described in claim 1, characterized in that, The step of predicting the vehicle's trajectory based on the first historical trajectory, the first weight, road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight includes: Based on the first historical trajectory, the first weight, road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight, the parameters in the target trajectory prediction algorithm are optimized to obtain the target parameters corresponding to the target trajectory prediction algorithm. The vehicle's trajectory is determined based on the target parameters and the target trajectory prediction algorithm.

3. The method as described in claim 2, characterized in that, The optimization of the parameters in the target trajectory prediction algorithm based on the first historical trajectory, first weight, road information, second weight, second historical trajectory, third weight, lateral predicted trajectory, and fourth weight yields the target parameters corresponding to the target trajectory prediction algorithm, including: The parameters in the target trajectory prediction algorithm are optimized based on the first historical trajectory to obtain the first optimized parameters; The parameters in the target trajectory prediction algorithm are optimized based on the road information to obtain the second optimized parameters; The parameters in the target trajectory prediction algorithm are optimized based on the second historical trajectory to obtain the third optimized parameters; The parameters in the target trajectory prediction algorithm are optimized based on the lateral predicted trajectory to obtain the fourth optimized parameter; The target parameters corresponding to the target trajectory prediction algorithm are determined based on the first weight, the first optimization parameter, the second weight, the second optimization parameter, the third weight, the third optimization parameter, the fourth weight, and the fourth optimization parameter.

4. The method as described in claim 1, characterized in that, The lateral predicted trajectory is obtained through the following method: Obtain the wheel rotation angle and yaw rate of the vehicle; The radius of curvature of the vehicle at the current moment is determined based on the wheel rotation angle and the yaw rate. Based on the radius of curvature and the wheel angle, the predicted lateral trajectory of the vehicle is obtained.

5. The method as described in claim 4, characterized in that, Before obtaining the wheel angle and yaw rate of the vehicle, the method further includes: The vehicle data of the vehicle at the current moment is obtained. The vehicle data includes the front wheel lateral stiffness, rear wheel lateral stiffness, distance from the front axle to the center of gravity of the vehicle, distance from the rear axle to the center of gravity of the vehicle, mass of the vehicle, oblique velocity at the center of gravity, moment of inertia, and sideslip angle at the center of gravity. Based on the vehicle data, determine the current state data of the vehicle; Based on the state data and the target data processing algorithm, the yaw rate of the vehicle at the current moment is determined.

6. The method according to any one of claims 1-5, characterized in that, The first historical trajectory was obtained in the following way: Multiple historical trajectory points of the vehicle are obtained, and each historical trajectory point is determined based on the vehicle coordinate system at each historical moment; Based on the current vehicle coordinate system, the multiple historical trajectory points are converted into historical trajectory points in the current vehicle coordinate system to obtain the target historical trajectory point. The first historical trajectory is determined based on the target historical trajectory points.

7. The method according to any one of claims 1-5, characterized in that, After predicting the vehicle's trajectory based on the first historical trajectory, the road information, the second historical trajectory, and the lateral predicted trajectory, the method further includes: Obtain the target range, which is used for the subsequent driving planning of the vehicle; Based on the target range, the driving trajectories of other vehicles within the target range of the driving trajectory are obtained and used as the target driving trajectory. Based on the vehicle's driving trajectory and the target driving trajectory, the vehicle's subsequent driving plan is determined.

8. A device for predicting driving trajectories, characterized in that, include: The data acquisition module is used to acquire the first historical trajectory of the vehicle during its driving process, the lateral predicted trajectory based on the wheel angle of the vehicle, the road information currently collected by the vehicle, and the second historical trajectory of the target vehicle being followed by the vehicle. The weight determination module is used to determine the first weight corresponding to the first historical trajectory, the second weight corresponding to the road information, the third weight corresponding to the second historical trajectory, and the fourth weight corresponding to the lateral predicted trajectory based on the road scene currently being driven by the vehicle. The road scene is determined by the road information collected by the vehicle in real time. The driving trajectory prediction module is used to predict the driving trajectory of the vehicle based on the first historical trajectory, the first weight, road information, the second weight, the second historical trajectory, the third weight, the lateral predicted trajectory, and the fourth weight.

9. A vehicle, characterized in that, include: One or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, the one or more applications being configured to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code that can be invoked by a processor to execute the method as described in any one of claims 1-7.

Images