Trajectory prediction method, device, vehicle-mounted terminal, vehicle and storage medium
By acquiring and integrating vehicle environment and driving information, determining multiple trajectory points and fitting, the problem of inaccurate vehicle trajectory prediction in the prior art is solved, and stable and efficient trajectory prediction in complex scenarios is achieved.
Patent Information
- Application Number
- CN202411111992.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-14
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-08-14
AI Technical Summary
The existing trajectory prediction methods cannot predict the vehicle's driving trajectory stably and accurately in complex vehicle driving scenarios, resulting in low accuracy of trajectory prediction.
By obtaining the environmental information of the target vehicle and the vehicle driving information, multiple trajectory points are determined, and fitting them based on these trajectory points, the predicted trajectory of the target vehicle is obtained. The specific steps include calculating trajectory prediction points, historical trajectory points and environmental information based on dynamic formulas, fusion and weight determination of trajectory points based on lane line information, and finally trajectory fitting.
It realizes the stable and accurate prediction of the vehicle's driving trajectory in complex vehicle driving scenarios, and improves the efficiency and accuracy of trajectory prediction.
Smart Images

Figure CN118701110B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of computer technology, and in particular to a trajectory prediction method, device, vehicle-mounted terminal, vehicle and storage medium. Background Art
[0002] In intelligent driving systems, the trajectory prediction of autonomous vehicles is crucial to ensure vehicle safety and more intelligent driving and avoid collisions with other vehicles. In existing trajectory prediction methods, dynamic formulas are often used to calculate the vehicle's trajectory over a period of time, or lane line information is introduced to constrain the vehicle's trajectory, thereby predicting the vehicle's trajectory.
[0003] However, due to the complexity of vehicle driving scenarios, existing trajectory prediction methods are unable to stably and accurately predict the vehicle's driving trajectory, resulting in low accuracy of trajectory prediction. Summary of the invention
[0004] The embodiments of the present application provide a trajectory prediction method, device, vehicle-mounted terminal, vehicle and storage medium, which can stably and accurately predict the vehicle's driving trajectory, thereby improving the trajectory prediction efficiency.
[0005] In order to achieve the above object, according to a first aspect of the present application, a trajectory prediction method is provided, the method comprising:
[0006] Obtaining environmental information and vehicle driving information of the target vehicle;
[0007] Determining multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information;
[0008] Trajectory fitting is performed based on the trajectory points to obtain a target predicted trajectory of the target vehicle.
[0009] In some embodiments, the trajectory points include at least two of a first trajectory prediction point, a second trajectory prediction point, a third trajectory prediction point and a historical trajectory point, the first trajectory prediction point is calculated based on a dynamic formula, the historical trajectory point and the second trajectory prediction point are calculated based on the historical trajectory of the target vehicle, and the third trajectory prediction point is calculated based on environmental information of the target vehicle.
[0010] In some embodiments, the trajectory point includes a first trajectory prediction point, and the determining of multiple trajectory points of the target vehicle based on the vehicle driving information and the environmental information includes: determining whether the target vehicle stops within a preset time period based on the vehicle driving information; if so, calculating the driving trajectory points of the target vehicle at multiple time points within the preset time period based on the vehicle driving information to obtain the first trajectory prediction point of the target vehicle; if not, calculating the driving trajectory points of the target vehicle at multiple time points within the preset time period based on the vehicle driving information and the curvature information of the target vehicle to obtain the first trajectory prediction point of the target vehicle.
[0011] In some embodiments, the method of calculating the driving trajectory points of the target vehicle at multiple time points within the preset time length based on the vehicle driving information and the curvature information of the target vehicle to obtain a first trajectory prediction point of the target vehicle includes: for multiple first time points within the preset time length, calculating the driving trajectory points at the first time points based on the speed information in the vehicle driving information to obtain the first trajectory prediction point; for at least one second time point within the preset time length, calculating the total driving distance of the target vehicle within the preset time length according to the speed information; calculating the driving trajectory points at the second time points according to the total driving distance and the curvature information of the target vehicle to obtain the first trajectory prediction point of the target vehicle; wherein the second time point is located after the first time point.
[0012] In some embodiments, the calculating the driving trajectory point at the second time point according to the total driving distance and the curvature information of the target vehicle to obtain the first trajectory prediction point of the target vehicle includes: calculating the first driving trajectory point and the second driving trajectory point of the target vehicle in a spatial rectangular coordinate system according to the total driving distance, the curvature in the curvature information, and the curvature change rate; performing coordinate transformation on the first driving trajectory point and the second driving trajectory point to obtain the transformed first driving trajectory point and the transformed second driving trajectory point in the vehicle coordinate system of the target vehicle; performing trajectory point fusion based on the transformed first driving trajectory point and the transformed second driving trajectory point to obtain the first trajectory prediction point of the target vehicle.
[0013] In some embodiments, the trajectory points include second trajectory prediction points and historical trajectory points. The determining of multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information includes: calculating the historical trajectory points of the target vehicle based on the vehicle driving information; performing curve fitting based on the historical trajectory points to obtain a historical trajectory curve; and sampling in the historical trajectory curve according to the ordinate of the first trajectory prediction point to obtain a second trajectory prediction point of the target vehicle.
[0014] In some embodiments, the trajectory points include a third trajectory prediction point, and determining multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information includes: obtaining a lane line equation of a target lane where the target vehicle is located according to the environmental information; and calculating a third trajectory prediction point on the centerline of the target lane based on the lane line equation.
[0015] In some embodiments, after calculating the third trajectory prediction point on the lane centerline of the target lane based on the lane line equation, it also includes: determining a trajectory point weight corresponding to the third trajectory prediction point based on the third trajectory prediction point and the lane centerline.
[0016] In some embodiments, determining the trajectory point weight corresponding to the third trajectory prediction point according to the third trajectory prediction point and the lane centerline includes: obtaining a target angle between the target vehicle and the target lane, a lateral distance between the target vehicle and the lane centerline, and a longitudinal distance between the target vehicle and the third trajectory prediction point based on a lane line equation for the lane centerline; determining the trajectory point weight corresponding to the third trajectory prediction point based on the target angle, the lateral distance, and the longitudinal distance.
[0017] In some embodiments, performing trajectory fitting based on the trajectory points to obtain the target predicted trajectory of the target vehicle includes: identifying the driving scene of the target vehicle based on the vehicle driving information of the target vehicle, the environmental information and the trajectory points; performing trajectory point fitting based on multiple credible trajectory points in the driving scene to obtain the target predicted trajectory of the target vehicle.
[0018] In some embodiments, performing trajectory point fitting based on multiple credible trajectory points in the driving scenario to obtain the target predicted trajectory of the target vehicle includes: when the driving scenario is a normal driving scenario, obtaining the trajectory point weight of the third trajectory prediction point, and based on the trajectory point weight, fusing the first trajectory prediction point and the third trajectory prediction point to obtain a first fused trajectory point; based on the trajectory point weight, fusing the second trajectory prediction point and the third trajectory prediction point to obtain a second fused trajectory point; performing trajectory fitting on the first fused trajectory point, the second fused trajectory point and the historical trajectory point to obtain the target predicted trajectory of the target vehicle.
[0019] In some embodiments, the trajectory point fitting is performed based on multiple credible trajectory points in the driving scenario to obtain the target predicted trajectory of the target vehicle, including: when the driving scenario is an abnormal driving scenario, determining the credibility weight corresponding to each of the trajectory points, wherein for uncredible trajectory points, the credibility weight is 0; and trajectory point fitting is performed on the trajectory points based on the credibility weights to obtain the target predicted trajectory of the target vehicle.
[0020] According to a second aspect of the present application, a trajectory prediction device is provided, the device comprising:
[0021] A data acquisition module is used to acquire the environmental information and vehicle driving information of the target vehicle;
[0022] A trajectory point determination module, used to determine multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information;
[0023] The trajectory fitting module is used to perform trajectory fitting based on the trajectory points to obtain the target predicted trajectory of the target vehicle.
[0024] According to a third aspect of the present application, a vehicle-mounted terminal is provided, comprising a processor and a memory, wherein the memory stores an application program, and the processor is used to run the application program in the memory to implement the model transformation method provided in an embodiment of the present application.
[0025] According to a fourth aspect of the present application, a vehicle is provided, comprising the vehicle-mounted terminal provided by the third aspect of the present application.
[0026] According to a fifth aspect of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and the computer program is suitable for loading by a processor to execute the steps in any trajectory prediction method provided in the embodiments of the present application.
[0027] According to the sixth aspect of the present application, a computer program product is provided, which includes a computer program, and the computer program is stored in a computer-readable storage medium; when the processor of a vehicle-mounted terminal reads the computer program from the computer-readable storage medium, the processor executes the computer program, so that the vehicle-mounted terminal performs the steps in the trajectory prediction method provided in an embodiment of the present application.
[0028] In the trajectory prediction method, device, vehicle-mounted terminal, vehicle and storage medium of the embodiment of the present application, the environmental information and vehicle driving information of the target vehicle are obtained; based on the vehicle driving information and environmental information, multiple trajectory points of the target vehicle are determined; based on the trajectory points, trajectory fitting is performed to obtain the target predicted trajectory of the target vehicle. In this way, multiple trajectory points of the target vehicle can be determined according to the environmental information and vehicle driving information of the target vehicle, so that the predicted trajectory of the target vehicle can be obtained by trajectory fitting of multiple trajectory points, and the driving trajectory of the vehicle can be predicted stably and accurately, thereby improving the efficiency of trajectory prediction. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.
[0030] Figure 1 It is a schematic diagram of an implementation scenario of a trajectory prediction method provided in an embodiment of the present application;
[0031] Figure 2 It is a flow chart of a trajectory prediction method provided in an embodiment of the present application;
[0032] Figure 3a It is a schematic diagram of a vehicle coordinate system of a trajectory prediction method provided in an embodiment of the present application;
[0033] Figure 3b is a schematic diagram of a trajectory coordinate system of a trajectory prediction method provided in an embodiment of the present application;
[0034] Figure 4a is a trajectory prediction schematic diagram of a trajectory prediction method provided in an embodiment of the present application;
[0035] Figure 4b is another trajectory prediction schematic diagram of a trajectory prediction method provided by an embodiment of the present application;
[0036] Figure 4c is another trajectory prediction schematic diagram of a trajectory prediction method provided in an embodiment of the present application;
[0037] Figure 4d is another trajectory prediction schematic diagram of a trajectory prediction method provided in an embodiment of the present application;
[0038] Figure 5 is a schematic diagram of the structure of a trajectory prediction device provided in an embodiment of the present application;
[0039] Figure 6It is a structural diagram of the vehicle-mounted terminal provided in an embodiment of the present application. DETAILED DESCRIPTION
[0040] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.
[0041] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.
[0042] The embodiment of the present application provides a trajectory prediction method, device, vehicle-mounted terminal, vehicle and storage medium. The trajectory prediction device can be integrated in an electronic device, which can be a server or a terminal.
[0043] Among them, the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, network acceleration services (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms. Terminals may include but are not limited to mobile phones, computers, intelligent voice interaction devices, smart home appliances, vehicle-mounted terminals, aircraft, etc. Terminals and servers can be directly or indirectly connected via wired or wireless communications, and this application does not limit this.
[0044] Among them, the vehicle-mounted terminal can be integrated in a vehicle, and the vehicle can be a fuel vehicle, a plug-in hybrid vehicle or a new energy vehicle, etc. This application does not make specific limitations on this.
[0045] See also Figure 1 , taking the trajectory prediction device integrated into the vehicle terminal as an example, Figure 1A schematic diagram of an implementation scenario of the trajectory prediction method provided in an embodiment of the present application, wherein the vehicle-mounted terminal can be integrated on a vehicle, and the vehicle-mounted terminal can obtain environmental information and vehicle driving information of the target vehicle; based on the vehicle driving information and environmental information, multiple trajectory points of the target vehicle are determined; and trajectory fitting is performed based on the trajectory points to obtain a target predicted trajectory of the target vehicle.
[0046] It should be noted that Figure 1 The implementation environment scenario diagram of the trajectory prediction method shown is only an example. The implementation environment scenario of the trajectory prediction method described in the embodiment of the present application is to more clearly illustrate the technical solution of the embodiment of the present application, and does not constitute a limitation on the technical solution provided by the embodiment of the present application. It is known to those skilled in the art that with the evolution of data processing and the emergence of new business scenarios, the technical solution provided by this application is also applicable to similar technical problems.
[0047] The solutions provided by the embodiments of the present application are specifically described by the following embodiments. It should be noted that the description order of the following embodiments is not intended to limit the preferred order of the embodiments.
[0048] This embodiment will be described from the perspective of a trajectory prediction device, which may be specifically integrated in a vehicle-mounted terminal.
[0049] See also Figure 2 , Figure 2 : is a flow chart of a trajectory prediction method provided in an embodiment of the present application. The trajectory prediction method comprises:
[0050] Step S101, obtaining the environment information and vehicle driving information of the target vehicle.
[0051] The target vehicle may be a vehicle for which trajectory prediction is to be performed, and the environmental information may be information indicating the environment in which the target vehicle is currently traveling, for example, it may include lane line information of the lane where the target vehicle is located, obstacle positions and other related information, and the lane line information may describe the lane line or curb information of the lane where the target vehicle is located, and visual perception and other technologies may be used to obtain the lane line information of the lane where the target vehicle is located. The vehicle driving information may be information collected during the driving process of the target vehicle, and may be used to describe the driving state of the target vehicle, for example, it may include the driving speed, acceleration, heading angle and other information of the target vehicle.
[0052] Step S102, determining various trajectory points of the target vehicle based on the vehicle driving information and the environment information.
[0053] The trajectory point may be a possible trajectory point of the target vehicle determined based on vehicle driving information and environmental information.
[0054] Among them, the trajectory points of the target vehicle can be of multiple types. For example, the trajectory points of the target vehicle can include at least two of a first trajectory prediction point, a second trajectory prediction point, a third trajectory prediction point and a historical trajectory point. Among them, the first trajectory prediction point can be calculated based on a dynamic formula, the historical trajectory point and the second trajectory prediction point are calculated based on the historical trajectory of the target vehicle, the historical trajectory point can be a historical driving trajectory point of the target vehicle, the second trajectory prediction point can be a trajectory prediction point obtained by fitting based on the historical trajectory point, and the third trajectory prediction point can be calculated based on the environmental information of the target vehicle.
[0055] Among them, there can be multiple ways to determine the multiple trajectory points of the target vehicle based on the vehicle driving information and the environmental information. For example, when the trajectory points include the first trajectory prediction point, it can be determined whether the target vehicle stops within a preset time period based on the vehicle driving information. If so, the driving trajectory points of the target vehicle at multiple time points within the preset time period are calculated based on the vehicle driving information to obtain the first trajectory prediction point of the target vehicle. If not, the driving trajectory points of the target vehicle at multiple time points within the preset time period are calculated based on the vehicle driving information and the curvature information of the target vehicle to obtain the first trajectory prediction point of the target vehicle.
[0056] Among them, the preset duration can be a pre-set duration, which can be used to further determine the motion state of the target vehicle, so as to determine the calculation method of the target vehicle's driving trajectory point. For example, the preset duration can be 1 second, 2 seconds, etc. The specific duration value can be set according to the actual situation, and the embodiment of the present application is not limited here. The time point can be a time node in the preset duration. For example, assuming that the preset duration is 1 second, the time point can be the 0th second, 1 / 3 second, 2 / 3 second, 1 second, etc. The specific time point can be set according to actual needs. The driving trajectory point can be a trajectory point of the vehicle during driving. The curvature information can be curvature-related information of the driving trajectory of the target vehicle predicted based on the vehicle driving information, and can include information such as curvature and curvature change rate.
[0057] Optionally, the target vehicle's driving trajectory points at multiple time points within multiple preset time lengths may be repeatedly calculated to obtain the target vehicle's first trajectory prediction point. For example, assuming that the preset time length is 1 second, four time points may be set within each preset time length, namely, the 0th second, 1 / 3 second, 2 / 3 second, and 1 second time points, so that the target vehicle's driving trajectory points at multiple time points within each preset time length within four seconds may be calculated, so that 13 driving trajectory points of the target vehicle within 4 seconds may be obtained, and the 2nd to 12th points may be taken as the first trajectory prediction point (P1).
[0058] Among them, based on the vehicle driving information, there are many ways to determine whether the target vehicle stops within the preset time. For example, based on the dynamic formula, according to the speed and acceleration of the target vehicle, it can be calculated whether the target vehicle will stop within the preset time.
[0059] If the target vehicle stops within the preset time, the target vehicle's driving trajectory points at multiple time points within the preset time can be calculated based on the vehicle's driving information to obtain the target vehicle's first trajectory prediction point. There are multiple ways to calculate the target vehicle's driving trajectory points at multiple time points within the preset time based on the vehicle's driving information. For example, the distance the target vehicle travels within the preset time can be calculated based on the target vehicle's speed and acceleration, which can include the lateral distance y and the longitudinal distance x in the vehicle coordinate system. Based on the lateral distance y, the longitudinal distance x and the initial coordinates (x0, y0) of the target vehicle, the first predicted trajectory points of the target vehicle at multiple time points within the preset time can be obtained as (x0+x, y0+y).
[0060] The vehicle coordinate system of the target vehicle can be the direction pointed by the front of the target vehicle (the direction of the vehicle's travel) as the X-axis, and the direction pointing to the left side of the vehicle and perpendicular to the X-axis as the Y-axis. In the vehicle coordinate system, the coordinates on the X-axis are called the ordinate, and the coordinates on the Y-axis are called the abscissa. For example, please refer to Figure 3a , Figure 3a It is a schematic diagram of a vehicle coordinate system of a trajectory prediction method provided in an embodiment of the present application, wherein the direction pointed by the front of the target vehicle is the X-axis, and the direction pointing to the left side of the vehicle and perpendicular to the X-axis is the Y-axis.
[0061] Among them, based on the vehicle driving information and the curvature information of the target vehicle, the driving trajectory points of the target vehicle at multiple time points within a preset time length are calculated, and there can be multiple ways to obtain the first trajectory prediction point of the target vehicle. For example, for multiple first time points within the preset time length, the driving trajectory point at the first time point can be calculated based on the speed information in the vehicle driving information to obtain the first trajectory prediction point. For at least one second time point within the preset time length, the total driving distance of the target vehicle within the preset time length is calculated according to the speed information, and the driving trajectory point at the second time point is calculated according to the total driving distance and the curvature information of the target vehicle to obtain the first trajectory prediction point of the target vehicle, wherein the second time point is located after the first time point.
[0062] Among them, the total travel distance can be the straight-line distance traveled by the target vehicle within the preset time length, the second time point can be a time point after the first time point, and the time span corresponding to the second time point is greater than the first time point. For example, assuming that the preset time length is 1 second, the first time point is the 0th second, the 1 / 3th second and the 2 / 3th second within the preset time length, and the second time point can be the 1st second within the preset time length, that is, the time span between the first time points is 1 / 3 second, and the time span of the second time point is 1 second. When the preset time length is 4 seconds, the first time point can be the 0th second, the 1 / 3th second and the 2 / 3th second within each second, and the second time point can be the first second within each second, etc. Therefore, when calculating the trajectory point corresponding to the second time point, the curvature information of the target vehicle can be introduced according to the specific driving conditions of the target vehicle, so as to more accurately calculate the driving trajectory point corresponding to the second time point, and improve the accuracy of the calculated first trajectory prediction point.
[0063] Among them, for multiple first time points within the preset time length, based on the speed information in the vehicle driving information, the driving trajectory point at the first time point is calculated, and there are multiple ways to obtain the first trajectory prediction point. For example, assuming that the preset time length is 1s, the first time point is the 0th second, 1 / 3 second and 2 / 3 second within the preset time length, and the second time point is the 1st second within the preset time length, then the dynamic formula can be used to calculate the coordinates of the driving trajectory point of the target vehicle at the first time points 0s, 1 / 3s and 2 / 3s. For example, for the calculation of the longitudinal coordinate of the driving trajectory point, at the first time point 0s, the longitudinal coordinate of the driving trajectory point can be x0=x last_1 , where x last_1 It can be expressed as the ordinate of the driving trajectory point at the 1st second of the previous preset duration. At the first time point 1 / 3s, the ordinate of the driving trajectory point can be expressed as x 1 / 3 =x0+vt=x0+v(1 / 3)=x0+1 / 3×v, where v can be expressed as the current speed of the target vehicle, which can be the speed at the 0th second. The speed of the target vehicle at the first time point 1 / 3s can be expressed as v1=v+at=v+a / 3, a is the acceleration of the target vehicle. At the first time point 2 / 3s, the ordinate of the driving trajectory point can be expressed as x 2 / 3 =x 1 / 3 +v1t+1 / 2×at 2 =x0+v(1 / 3)=x0+2 / 3×v+1 / 6×a. Correspondingly, the calculation method of the horizontal coordinate of the driving trajectory point can be the same as the vertical coordinate, so that the coordinate of the first trajectory prediction point corresponding to the first time point can be calculated.
[0064] Among them, for at least one second time point within the preset time length, the total driving distance of the target vehicle within the preset time length is calculated according to the speed information, and the driving trajectory point at the second time point is calculated according to the total driving distance and the curvature information of the target vehicle. There can be multiple ways to obtain the first trajectory prediction point of the target vehicle. For example, for calculating the driving trajectory point at the second time point according to the total driving distance and the curvature information of the target vehicle to obtain the first trajectory prediction point of the target vehicle, the first driving trajectory point and the second driving trajectory point of the target vehicle in the spatial rectangular coordinate system can be calculated according to the total driving distance, the curvature in the curvature information, and the curvature change rate; the first driving trajectory point and the second driving trajectory point are converted to obtain the converted first driving trajectory point and the converted second driving trajectory point in the vehicle coordinate system of the target vehicle; the trajectory points are fused based on the converted first driving trajectory point and the converted second driving trajectory point to obtain the first trajectory prediction point of the target vehicle.
[0065] Among them, the first driving trajectory point can be the position of the target vehicle at the second time point when it is assumed that the curvature of the target vehicle remains unchanged during the driving process, and the second driving trajectory point can be the position of the target vehicle at the second time point when it is assumed that the curvature of the target vehicle changes based on the curvature change rate during the driving process.
[0066] For example, assuming that the preset time length is 1 second, the first time point is the 0th second, the 1 / 3th second, and the 2 / 3th second within the preset time length, and the second time point is the 1st second within the preset time length, the total travel distance ds of the target vehicle within the preset time length of 1 second can be calculated according to the speed information, and the calculation process of the total travel distance ds can be expressed as:
[0067]
[0068] Among them, v can be represented as the current speed of the target vehicle, a represents the acceleration, and t can be represented as the preset duration.
[0069] Then, the total travel distance ds, the curvature c in the curvature information, and the curvature change rate can be used to calculate the curvature change rate. , calculate the first driving trajectory point and the second driving trajectory point of the target vehicle in the spatial rectangular coordinate system. For example, the curvature radius of the target vehicle's driving trajectory can be calculated according to the curvature and the curvature change rate in the curvature information:
[0070]
[0071]
[0072] Among them, r1 represents the radius of curvature when the curvature remains unchanged, and r2 represents the curvature c based on the curvature change rate. Then, the angle of rotation corresponding to the different curvature radii at the second time point 1s can be calculated:
[0073]
[0074]
[0075] Among them, β1 represents the turning angle of the target vehicle driving trajectory based on r1 at the second time point (i.e. within 1 second), and β2 represents the turning angle of the target vehicle driving trajectory based on r2 at the second time point. Then, the first driving trajectory point and the second driving trajectory point of the target vehicle at the second time point can be calculated in the spatial rectangular coordinate system:
[0076]
[0077]
[0078]
[0079]
[0080] Among them, P 1x , P 1y They are respectively represented as the ordinate and abscissa of the first driving trajectory point, P 2x , P 2y Then, the first driving trajectory point and the second driving trajectory point can be transformed to obtain the transformed first driving trajectory point and the transformed second driving trajectory point in the vehicle coordinate system of the target vehicle:
[0081]
[0082]
[0083] in, , They are respectively represented as the ordinate and abscissa of the first driving trajectory point after conversion, , They are respectively represented as the ordinate and abscissa of the second driving trajectory point after conversion.
[0084] Then, trajectory points can be fused based on the converted first driving trajectory point and the converted second driving trajectory point to obtain the first trajectory prediction point of the target vehicle. For example, a weight can be set for the converted first driving trajectory point and the converted second driving trajectory point, for example, the weights can be 0.7, 0.3, etc., so that the converted first driving trajectory point and the converted second driving trajectory point can be fused according to the weights corresponding to the converted first driving trajectory point and the converted second driving trajectory point to obtain the first trajectory prediction point of the target vehicle:
[0085]
[0086] Among them, x last_1 It can be expressed as the initial ordinate of the target vehicle, or as the ordinate of the driving trajectory point at the second time point 1s in the previous preset time period, Y last_1 It can be expressed as the initial horizontal coordinate of the target vehicle, or as the horizontal coordinate of the driving trajectory point at the second time point 1s in the previous preset time period.
[0087] Optionally, there are multiple ways to determine the multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information. For example, when the trajectory points include the second trajectory prediction points and the historical trajectory points, the historical trajectory points of the target vehicle can be calculated based on the vehicle driving information; curve fitting can be performed based on the historical trajectory points to obtain the historical trajectory curve; and sampling can be performed in the historical trajectory curve according to the ordinate of the first trajectory prediction point to obtain the second trajectory prediction point of the target vehicle.
[0088] Among them, there are many ways to calculate the historical trajectory points of the target vehicle based on the vehicle driving information. For example, the historical trajectory points of the target vehicle at historical moments can be calculated based on the target vehicle's speed, vehicle acceleration and other data. The number of calculated historical trajectory points can be set according to actual needs, for example, it can be 100. Sampling can be performed on the calculated historical trajectory points to obtain the historical trajectory points that ultimately participate in trajectory fitting. For example, the number of historical trajectory points can be 10, 11, 15, etc.
[0089] In one embodiment, an array of size 100 may be set to store the calculated historical trajectory points. Specifically, the current index (CurrentId=1) and the oldest index (OldestId=1) of the array may be initialized. Then, each time a frame of driving-related data of the target vehicle is received, the turning angle dYaw, lateral displacement dx, and longitudinal displacement dy of the target vehicle in two adjacent frames are calculated:
[0090]
[0091]
[0092]
[0093] Among them, α can be expressed as the heading angle of the target vehicle, t can be expressed as the time difference between two frames, v can be expressed as the current speed of the target vehicle, and w can be expressed as the angular frequency of the target vehicle.
[0094] Then, the current heading can be calculated, with a value range of [0, 360] / [-360, 0]. The specific calculation formula can be expressed as:
[0095]
[0096] Among them, α on the left side of the formula can be expressed as the heading angle of the target vehicle in the current frame, and α on the right side of the formula can be expressed as the heading angle of the target vehicle in the previous frame.
[0097] Each time a new frame of data is received, the index value of the array can be updated. Each time a new frame of data is received, CurrentId can be increased by 1. If CurrentId is greater than 100, CurrentId can be set to 1. If CurrentId is equal to OldestId, OldestId can be increased by 1. If OldestId is greater than 100, OldestId can be set to 1. Thus, all historical track points can be updated. Then, the historical track points can be converted to the current vehicle coordinate system of the target vehicle. For example, please refer to Figure 3b , Figure 3b This is a schematic diagram of a trajectory coordinate system of a trajectory prediction method provided by an embodiment of the present application. The initial coordinate system X0O0Y0 of the target vehicle can be determined according to the orientation of the vehicle at the start time. The historical trajectory of the vehicle provides points in the X1O1Y1 coordinate system. Since the trajectory points of the vehicle calculated according to the vehicle dynamics model are all coordinates in the current vehicle coordinate system, it is necessary to convert the calculated historical trajectory points to the current vehicle coordinate system X2O1Y2 through coordinate conversion. The specific conversion process can be expressed as:
[0098]
[0099] Among them, (X i , Y i ) can be expressed as coordinates in the X1O1Y1 coordinate system, It can be expressed as the coordinates in the X2O1Y2 coordinate system, and α can be expressed as the heading angle of the current historical trajectory point.
[0100] Optionally, sampling may be performed among the calculated historical trajectory points according to the number of historical trajectory points required for trajectory fitting of the target vehicle, thereby obtaining the historical trajectory points (P2) for final trajectory fitting.
[0101] For example, assuming that the number of calculated historical trajectory points is 100, a certain screening logic can be used to determine the 10 historical trajectory points (P2) involved in the final trajectory fitting, so that a historical trajectory curve can be obtained by fitting the screened historical trajectory points through the least squares method, and 11 second trajectory prediction points (P3) can be determined based on the fitted historical trajectory curve. Specifically, the current historical trajectory point ID (currentPointer) and the oldest historical trajectory point ID (oldestPointer) can be determined. If currentPointer is less than 10 and currentPointer is greater than oldestPointer, all historical trajectory points can be used as the historical trajectory points that finally participate in trajectory fitting; if currentPointer is greater than 10, currentPointer is greater than oldestPointer and currentPointer is not equal to 100, the maximum sampling interval MaxDiff+floor(currentPointer / 10) can be calculated, so that points can be taken in sequence from the historical trajectory points stored in the array according to the sampling interval to obtain the historical trajectory points that finally participate in trajectory fitting; if currentPointer is less than oldestPointer or currentPointer is equal to 100, points can be sampled in sequence from the stored historical trajectory points according to the pre-set sampling interval to obtain the historical trajectory points that finally participate in trajectory fitting.
[0102] Then, a cubic curve, i.e., the historical trajectory curve, can be obtained by least square fitting based on the historical sampling points that finally participate in the trajectory fitting after the sampling process. Thus, a point can be taken on the historical trajectory curve based on the ordinate (X coordinate) of the first trajectory prediction point P1 to obtain the second trajectory prediction point (P3).
[0103] Optionally, there may be multiple ways to determine the multiple trajectory points of the target vehicle based on the vehicle driving information and the environmental information. For example, when the trajectory points include a third trajectory prediction point, the lane line equation of the target lane where the target vehicle is located may be obtained based on the environmental information; based on the lane line equation, the third trajectory prediction point on the centerline of the target lane may be calculated.
[0104] Among them, the lane line equation can be a cubic curve equation. The lane line equation can be constructed by using the lane perception data of the target vehicle obtained through visual perception and other technologies, so as to obtain the lane line equations of the left and right lane lines of the target lane where the target vehicle is located.
[0105] Among them, there can be multiple ways to calculate the third trajectory prediction point on the lane centerline of the target lane based on the lane line equation. For example, the coordinates of the centerline point of the target lane can be calculated according to the lane line equation of the left lane line to obtain the first lane centerline point, and the coordinates of the centerline point of the target lane can be calculated according to the lane line equation of the right lane line to obtain the second lane centerline point, so that the weights of the left and right lane lines can be calculated according to the effective lengths of the left and right lane lines, so that the first lane centerline point and the second lane centerline point can be fused according to the weights of the left and right lane lines to obtain the sampling point coordinates of the lane centerline of the target lane, and then curve fitting can be performed according to the sampling point coordinates of the lane centerline of the target lane to obtain the lane line equation of the lane centerline of the target lane, and then, sampling can be performed on the curve of the lane centerline according to the longitudinal coordinate of the first trajectory prediction point to obtain the third trajectory prediction point on the lane centerline of the target lane.
[0106] The effective length of the lane line may be the length of the lane line perceived by visual perception technology. The longer the length of the lane line is, that is, the larger the effective length is, the greater the corresponding weight is.
[0107] Optionally, a weight corresponding to the third trajectory prediction point may also be calculated for subsequent trajectory point fusion. For example, the trajectory point weight corresponding to the third trajectory prediction point may be determined based on the third trajectory prediction point and the lane centerline.
[0108] Among them, there are many ways to determine the trajectory point weight corresponding to the third trajectory prediction point according to the third trajectory prediction point and the lane centerline. For example, based on the lane line equation of the lane centerline, the target angle between the target vehicle and the target lane, the lateral distance between the target vehicle and the lane centerline, and the longitudinal distance between the target vehicle and the third trajectory prediction point can be obtained; based on the target angle, lateral distance and longitudinal distance, the trajectory point weight corresponding to the third trajectory prediction point is determined.
[0109] The target angle may be the angle between the driving direction of the target vehicle and the center line of the target lane, and the trajectory point weight may be the weight corresponding to the third trajectory prediction point, which is used to indicate the credibility of the third trajectory prediction point.
[0110] There are multiple ways to obtain the target angle between the target vehicle and the target lane, the lateral distance between the target vehicle and the lane centerline, and the longitudinal distance between the target vehicle and the third trajectory prediction point based on the lane line equation of the lane centerline. For example, the lane line equation of the lane centerline of the target lane can be calculated based on the perceived left and right lane line equations. Since the lane line equation of the lane centerline is a cubic curve, the cubic curve can contain a constant term coefficient, a linear term coefficient, a quadratic term coefficient, and a cubic term coefficient. The constant term coefficient of the lane line equation of the lane centerline can be expressed as the lateral distance from the target vehicle to the lane centerline, the linear term coefficient is tan(θ), where θ is the angle between the X-axis of the vehicle coordinate system of the ego vehicle and the lane centerline, that is, the target angle, the quadratic term coefficient can be expressed as one-half of the curvature of the lane centerline, and the cubic term coefficient can be one-sixth of the curvature change rate of the lane centerline. In this way, the target angle between the target vehicle and the target lane can be obtained according to the linear term coefficient in the lane line equation of the center line of the lane, the lateral distance between the target vehicle and the center line of the lane can be obtained according to the constant term coefficient, and then the longitudinal distance between the target vehicle and the third trajectory prediction point can be obtained according to the longitudinal coordinate of the third trajectory prediction point.
[0111] Among them, based on the target angle, lateral distance and longitudinal distance, there are many ways to determine the trajectory point weight corresponding to the third trajectory prediction point. For example, when the target vehicle has a certain angle with the target lane, it is considered that the target vehicle is not completely driving along the lane direction. The larger the absolute value of the target angle, the greater the tendency of the target vehicle to deviate from the lane direction, and the smaller the trajectory point weight of the third trajectory prediction point, and vice versa. In addition, since it is impossible to accurately predict the driving behavior of the driver of the target vehicle, it cannot be guaranteed that it will definitely drive along the center line of the lane. Therefore, the greater the lateral distance of the self-vehicle from the center line of the lane of the target lane, the smaller the trajectory point weight of the third trajectory prediction point, and vice versa. In addition, since both the first trajectory prediction point calculated according to the vehicle dynamics model, the historical trajectory point calculated according to the historical trajectory of the self-vehicle, and the second trajectory prediction point are relatively accurate at close distances from the vehicle, but the trajectory credibility at long distances is low. Therefore, the farther the distance between the third trajectory prediction point and the target vehicle, the greater the trajectory point weight of the third trajectory prediction point, and vice versa.
[0112] In this way, when calculating the trajectory point weight corresponding to the third trajectory prediction point, the embodiment of the present application takes into account the relative relationship between the target vehicle, the third trajectory prediction point and the lane, thereby accurately calculating the trajectory point weight corresponding to the third trajectory prediction point, thereby being able to more accurately predict the actual movement trend and driving trajectory of the target vehicle.
[0113] Step S103, performing trajectory fitting based on the trajectory points to obtain a target predicted trajectory of the target vehicle.
[0114] The target predicted trajectory may be information describing the predicted driving trajectory of the target vehicle, or may be a trajectory curve obtained by fitting at least one trajectory point.
[0115] There are many ways to obtain the target predicted trajectory of the target vehicle by performing trajectory fitting based on trajectory points. For example, the driving scene of the target vehicle can be identified based on the vehicle driving information, environmental information and trajectory points of the target vehicle; trajectory point fitting can be performed based on multiple reliable trajectory points in the driving scene to obtain the target predicted trajectory of the target vehicle.
[0116] Among them, the driving scene can be the scene in which the target vehicle is in the process of driving, and the driving scene can include normal driving scenes and abnormal driving scenes. Different driving scenes can be divided according to the target vehicle's own vehicle motion state, the quality of the sampled trajectory points and the effectiveness of the lane lines.
[0117] For example, the normal driving scenario is a standard driving scenario, which may refer to a scenario where the vehicle is driving straight or changing lanes when the vehicle speed meets normal driving conditions, the historical trajectory is available, and at least one lane line is valid. In a normal scenario, the error between the historical trajectory point and the second trajectory prediction point is within an acceptable range and can reflect the movement trend of the vehicle to a certain extent. At the same time, the lane line exists and is valid. Therefore, trajectory fitting can be performed based on the first trajectory prediction point, the historical trajectory point, the second trajectory prediction point, and the third trajectory prediction point to obtain the target predicted trajectory of the target vehicle.
[0118] Abnormal driving scenarios are special driving scenarios, which may include large curve scenarios, lane-crossing scenarios, vehicles going straight and turning without lane lines, vehicles going straight and turning with lane lines but invalid lane line data (scenarios such as lane width not meeting requirements and chaotic lane lines), continuous lane change scenarios, slow vehicle driving, large steering wheel angle rate when turning, and invalid historical trajectory prediction points. In different abnormal driving scenarios, different trajectory points can be selected for trajectory fusion to obtain the target predicted trajectory of the target vehicle.
[0119] There are many ways to obtain the target predicted trajectory of the target vehicle by performing trajectory point fitting based on multiple reliable trajectory points in the driving scene. For example, when the driving scene is a normal driving scene, the trajectory point weight of the third trajectory prediction point can be obtained, and based on the trajectory point weight, the first trajectory prediction point and the third trajectory prediction point are fused to obtain a first fused trajectory point; based on the trajectory point weight, the second trajectory prediction point and the third trajectory prediction point are fused to obtain a second fused trajectory point; the first fused trajectory point, the second fused trajectory point and the historical trajectory point are trajectory fitted to obtain the target predicted trajectory of the target vehicle.
[0120] There are multiple ways to obtain a first fused trajectory point by fusing the first trajectory prediction point and the third trajectory prediction point based on the trajectory point weight, and obtain a second fused trajectory point by fusing the second trajectory prediction point and the third trajectory prediction point based on the trajectory point weight. For example, the first fused trajectory point can be obtained by fusing the first trajectory prediction point and the third trajectory prediction point based on the trajectory point weight, which can be specifically expressed as:
[0121]
[0122] in, , is the coordinate of the predicted point of the ith third trajectory, w i is the trajectory point weight of the i-th third trajectory prediction point, X1 i 、Y1 i Point_x can be the coordinates of the first trajectory prediction point calculated based on the vehicle dynamics model, i 、Point_y i It can be the first fusion trajectory point.
[0123] Furthermore, based on the trajectory point weight, the second trajectory prediction point and the third trajectory prediction point can be fused to obtain a second fused trajectory point, which can be specifically expressed as:
[0124]
[0125] in, is the trajectory point weight of the ith third trajectory prediction point, X2 i 、Y2 i The coordinates of the predicted point for the second trajectory, Point_x i 、Point_y i It can be the second fusion trajectory point.
[0126] Optionally, the first fused trajectory point and the second fused trajectory point may be adjusted to optimize abnormal trajectory points that exceed the curb or lane line such as the lane, so as to improve the accuracy of trajectory prediction.
[0127] Specifically, for the fused trajectory points beyond the curb, since the area outside the curb is an unreachable area, for the fused trajectory points beyond the curb (x i ,y i ) (i.e. the trajectory point after being fused with the third trajectory prediction point), can be optimized in the following way:
[0128] If the target vehicle exceeds the left curb and the distance from the target vehicle to the left curb is greater than 1m, the fused trajectory point (x i ,y i) is optimized to obtain the optimized fusion trajectory point ( , ):
[0129]
[0130] Among them, C0-C3 are the four coefficients of the lane line equation of the left / right side of the roadside. The lane line equation of the lane line and the roadside is a cubic curve, which can contain a constant term coefficient (C0), a first-order term coefficient (C1), a second-order term coefficient (C2) and a third-order term coefficient (C3). Among them, C0 can be expressed as the lateral distance from the center of the rear axle of the vehicle to the lane line or the roadside, C1 is tan(θ), where θ is the angle between the X-axis of the vehicle coordinate system of the vehicle and the lane line or the roadside, C2 is expressed as half of the curvature of the lane line / roadside, and C3 can be one-sixth of the curvature change rate of the lane line or the roadside.
[0131] If the target vehicle exceeds the right curb and the distance from the target vehicle to the right curb is greater than 1m, the fused trajectory point (x i ,y i ) is optimized to obtain the optimized fusion trajectory point ( , ):
[0132]
[0133] If the target vehicle exceeds the left or right curb but the distance from the curb is less than 1m, the fused trajectory point (x i ,y i ) is optimized to obtain the optimized fusion trajectory point ( , ):
[0134]
[0135] Specifically, for the optimization of the fused trajectory point beyond the lane, when the first fused trajectory point or the second fused trajectory point meets the following conditions, the optimization of the fused trajectory point beyond the lane is performed:
[0136] (1) When the target vehicle's state changes from lane change to straight driving (lane change completed), the points that exceed the left and right lane lines are optimized over the next period of time;
[0137] (2) The angle between the target vehicle and the lane line is greater than 0, the absolute value of the angle is greater than 6 degrees, and the steering wheel angle is greater than 60 degrees;
[0138] (3) The angle between the target vehicle and the lane line is less than 0, the absolute value of the angle is greater than 6 degrees, and the steering wheel angle is less than -60 degrees.
[0139] For the fused trajectory point that meets the above conditions, if it exceeds the left lane line and the distance from the target vehicle to the left lane line is greater than 1m, the fused trajectory point (x i ,y i ) is optimized to obtain the optimized fusion trajectory point ( , ):
[0140]
[0141] If the fused trajectory point that meets the above conditions exceeds the right lane line and the distance from the target vehicle to the right lane line is greater than 1m, the fused trajectory point (x i ,y i ) is optimized to obtain the optimized fusion trajectory point ( , ):
[0142]
[0143] Then, the first fused trajectory point, the second fused trajectory point and the historical trajectory point can be subjected to trajectory fitting to obtain the target predicted trajectory of the target vehicle. There are multiple ways to perform trajectory fitting on the first fused trajectory point, the second fused trajectory point and the historical trajectory point to obtain the target predicted trajectory of the target vehicle. For example, assuming that the number of the first trajectory prediction point, the second trajectory prediction point and the third trajectory prediction point is 11, and the number of the historical trajectory points is 10, then in a normal driving scenario, the number of trajectory points finally involved in trajectory fitting can be 32, including 10 historical trajectory sampling points (P2), 11 second fused trajectory points after the second trajectory prediction point (P3) and the third trajectory prediction point (P4) are fused, and 11 first fused trajectory points after the first trajectory prediction point (P1) and the third trajectory prediction point (P4) are fused. In this way, 10 historical trajectory points (P2) and 22 fused trajectory points (the first fused trajectory point and the second fused trajectory point) can be used as trajectory points that finally participate in the target vehicle's driving trajectory fitting. The sampling least squares method can be used to fit the 32 trajectory points, and the obtained cubic curve can be used as the target predicted trajectory of the target vehicle.
[0144] Please refer to Figure 4a , Figure 4aThis is a trajectory prediction diagram of a trajectory prediction method provided by an embodiment of the present application, wherein, in the target prediction trajectory of the target vehicle, the yellow trajectory point is the first fused trajectory point after the third trajectory prediction point is fused with the first trajectory prediction point, the green trajectory point in front of the vehicle is the second fused trajectory point, and the blue trajectory point behind the vehicle is the historical trajectory point of the vehicle. It can be seen that in normal driving scenarios, the trajectory prediction method provided by the embodiment of the present application can output the target prediction trajectory of the target vehicle more accurately and stably.
[0145] Optionally, there are multiple ways to obtain the target predicted trajectory of the target vehicle by performing trajectory point fitting based on multiple credible trajectory points in the driving scenario. For example, when the driving scenario is an abnormal driving scenario, the credibility weight corresponding to each trajectory point is determined, wherein the credibility weight for uncredible trajectory points is 0; and trajectory points are fitted based on the credibility weights to obtain the target predicted trajectory of the target vehicle.
[0146] Among them, the credibility weight can be information indicating the credibility of each trajectory point. For example, the credibility weight can include 0 and 1. When the credibility weight is 0, it can be indicated that the trajectory point is not credible in the current driving scene, and the trajectory point is not used for trajectory fitting when the trajectory of the target vehicle is fitted. When the credibility weight is 1, it can be indicated that the trajectory point is credible in the current driving scene, and the trajectory point is used for trajectory fitting when the trajectory of the target vehicle is fitted. For another example, the credibility weight can include any value from 0 to 1, which is used to indicate the credibility of the trajectory point in the current driving scene, so that the trajectory point can be fitted according to the credibility weight to obtain the target predicted trajectory of the target vehicle.
[0147] Among them, when the driving scene is an abnormal driving scene, there can be multiple ways to determine the credibility weights corresponding to each trajectory point. For example, according to the specific abnormal driving scene, the credibility weights corresponding to various trajectory points in the abnormal driving scene can be determined.
[0148] For example, a slow-moving vehicle scene may refer to a scene where the vehicle speed is low or the vehicle is stationary. When the vehicle is moving slowly or is stationary, the historical trajectory points will gradually concentrate at one point, and the first trajectory prediction point calculated based on the vehicle speed and dynamics model will also be relatively concentrated (slow speed) or completely concentrated (stationary) at one point. At this time, the effect of trajectory point fitting based on the target vehicle is unstable and its accuracy cannot be guaranteed. Therefore, for an abnormal driving scene where the vehicle is moving slowly and there is no lane line or the lane line does not meet the conditions, the credibility weight of the first trajectory prediction point can be determined to be 1, and the credibility weights of the historical trajectory points, the second trajectory prediction points, and the third trajectory prediction points can be determined to be 0, and a default speed can be given to the target vehicle, so that the first trajectory prediction point can be calculated according to the default speed, so that the trajectory fitting based on the first trajectory prediction point (P1) can be performed as the final trajectory point participating in the trajectory fitting to obtain the target predicted trajectory of the target vehicle.
[0149] For abnormal driving scenarios where the vehicle is traveling slowly and the lane lines are valid, the credibility weights of the first trajectory prediction point and the third trajectory prediction point can be determined to be 1, and the credibility weights of the historical trajectory point and the second trajectory prediction point can be determined to be 0, so that the first fused trajectory point can be used as the final trajectory point involved in trajectory fitting to perform trajectory fitting and obtain the target predicted trajectory of the target vehicle.
[0150] For example, see Figure 4b , Figure 4b is another trajectory prediction schematic diagram of a trajectory prediction method provided in an embodiment of the present application, which is a prediction result of the target prediction trajectory of the target vehicle after following and stopping the vehicle in front based on the adaptive cruise control system (ACC). At this time, although the vehicle speed of the target vehicle is 0, a valid target prediction trajectory can still be fitted through the trajectory prediction method provided in an embodiment of the present application.
[0151] For abnormal driving scenarios where the historical trajectory prediction point is invalid, when the second trajectory prediction point (P3) and the first trajectory prediction point (P1) meet any of the following conditions, the second trajectory prediction point can be considered invalid:
[0152] (1) The curvature of the trajectory of the second trajectory prediction point has an opposite sign to the curvature of the trajectory of the first trajectory prediction point, and the difference in the lateral coordinates of the two trajectory end points is greater than 15m;
[0153] (2) The lateral coordinate of the trajectory end point of the first trajectory prediction point is less than 1.5 m, and the lateral coordinate of the trajectory end point of the second trajectory prediction point is greater than 15 m;
[0154] (3) The lateral coordinate of the trajectory end point of the first trajectory prediction point is greater than 30m;
[0155] (4) The difference in lateral distance between the end points of two trajectory points is greater than 20m.
[0156] Therefore, for an abnormal driving scenario where the second trajectory prediction point is invalid and the lane line is invalid, the credibility weights of the first trajectory prediction point and the historical trajectory point can be determined to be 1, and the credibility weights of the third trajectory prediction point and the second trajectory prediction point can be determined to be 0, so that the first trajectory prediction point (P1) and the historical trajectory point (P2) can be used as the final trajectory points involved in trajectory fitting for trajectory fitting, and the target predicted trajectory of the target vehicle can be obtained.
[0157] For an abnormal driving scenario where the second trajectory prediction point is invalid and the lane line is valid, the credibility weights of the first trajectory prediction point, the third trajectory prediction point and the historical trajectory point can be determined to be 1, and the credibility weight of the second trajectory prediction point can be determined to be 0, so that the first fused trajectory point and the historical trajectory point (P2) can be used as the final trajectory points involved in trajectory fitting to perform trajectory fitting and obtain the target predicted trajectory of the target vehicle.
[0158] For abnormal driving scenarios where the steering wheel turning rate is large when turning, the scenario refers to the scenario where the lane line does not exist, the vehicle is turning, and the second trajectory prediction point is valid. During the turning process, the constraints of the third trajectory prediction point are not considered. When the steering wheel turning rate is greater than a certain threshold, it can be considered that the driver is turning the steering wheel quickly. After a period of time (for example, 10 frames), the first trajectory prediction point determined based on the steering wheel can be more trusted. Therefore, the first trajectory prediction point can be given a default credibility weight p i =0.8, so that the coordinates of the trajectory point after the fusion of the first trajectory prediction point (P1) and the second trajectory prediction point (P3) can be calculated:
[0159]
[0160] Among them, X i and Y i is the coordinate of the predicted point of the second trajectory, , is the coordinate of the first trajectory prediction point, P_x i 、P_y i is the coordinate of the trajectory point after the fusion of the first trajectory prediction point (P1) and the second trajectory prediction point (P3), p iis the default weight of the first trajectory prediction point. Therefore, in this scenario, the credibility weight of the first trajectory prediction point can be determined to be 0.8, the credibility weight of the second trajectory prediction point can be determined to be 1-0.8=0.2, the credibility weight of the historical trajectory point can be determined to be 1, and the credibility weight of the third trajectory prediction point can be determined to be 0, so that the trajectory point after the historical trajectory point (P2) and the first trajectory prediction point (P1) and the second trajectory prediction point (P3) are fused as the final trajectory points involved in trajectory fitting to obtain the target predicted trajectory of the target vehicle.
[0161] For the abnormal driving scenario of the vehicle going straight or turning in the scenario without lane lines or invalid lane lines, the scenario may refer to the scenario of the vehicle going straight or turning in which the second trajectory prediction point is valid but the lane line does not exist or the lane line is invalid. Since the lane line does not exist or is invalid, the constraint of the third trajectory prediction point is not considered. Therefore, the credibility weights of the first trajectory prediction point, the second trajectory prediction point and the historical trajectory point can be determined to be 1, and the credibility weight of the third trajectory prediction point can be determined to be 0, so that the historical trajectory point (P2), the first trajectory prediction point (P1) and the second trajectory prediction point (P3) can be used as the final trajectory points involved in trajectory fitting for trajectory fitting, and the target predicted trajectory of the target vehicle can be obtained.
[0162] For abnormal driving scenarios of continuous lane change, this scenario can refer to the scenario of continuously crossing multiple lanes in a short period of time. In this scenario, the vehicle should not be constrained by lane lines. When the following conditions are met, it can be judged that the target vehicle is in a continuous lane change scenario:
[0163] (1) The angle between the target vehicle and the lane line is greater than 0, the absolute value of the angle is greater than 6 degrees, and the steering wheel angle is less than 60 degrees;
[0164] (2) The angle between the target vehicle and the lane line is less than 0, the absolute value of the angle is greater than 6 degrees, and the steering wheel angle is greater than -60 degrees.
[0165] At this time, the credibility weight of the first trajectory prediction point can be determined to be 1, and the credibility weights of the second trajectory prediction point, the third trajectory prediction point and the historical trajectory point can be determined to be 0, so that the first trajectory prediction point (P1) can be used as the final trajectory point involved in trajectory fitting to obtain the target predicted trajectory of the target vehicle.
[0166] For abnormal driving scenarios in large curve scenarios, the scenario can refer to a scenario where the lane line is valid and the lane line curvature is large. Since in large curve scenarios, the trajectory error calculated only by information such as the steering wheel is large, it will obviously deviate from the lane, and at this time the angle between the vehicle and the lane line is large, resulting in a relatively reduced weight of the lane trajectory point, and the trajectory cannot be constrained within the lane. Therefore, in this scenario, when the distance between the third trajectory prediction point and the target vehicle is greater than a certain threshold, the weight of the third trajectory prediction point can be increased. Therefore, for large curve scenarios where the second trajectory prediction point is valid, the credibility weights of the first trajectory prediction point, the second trajectory prediction point, the third trajectory prediction point, and the historical trajectory point can be determined to be 1, so that the historical trajectory point (P2) and 22 fused sampling points (first fused trajectory point, second fused trajectory point) can be used as the final trajectory points participating in trajectory fitting for trajectory fitting, and the target predicted trajectory of the target vehicle can be obtained. For the large curve scene where the second trajectory prediction point is invalid, the credibility weights of the first trajectory prediction point, the historical trajectory point and the third trajectory prediction point can be determined to be 1, and the credibility weight of the second trajectory prediction point can be determined to be 0, so that the first fused trajectory point and the historical trajectory point (P2) can be used as the final trajectory points involved in trajectory fitting to perform trajectory fitting and obtain the target predicted trajectory of the target vehicle.
[0167] For example, see Figure 4c , Figure 4c This is another trajectory prediction diagram of a trajectory prediction method provided by an embodiment of the present application. It is the prediction result of the target prediction trajectory in the large curve scene when the second trajectory prediction point is valid and invalid. In this scene, the predicted trajectory of the target vehicle can be more accurately limited to the target lane to avoid the predicted trajectory of the target vehicle deviating from the target lane due to the large curve. At this time, the predicted trajectory can be used for the adaptive cruise control system (ACC), through which the target ahead can be stably positioned to avoid the problem of target loss due to the trajectory deviating from the target lane.
[0168] For the abnormal driving scenario of the lane crossing scenario, this scenario can refer to the scenario where the vehicle is driving on the lane and the lane crossing is valid. At this time, a new calculation method is needed to calculate the coordinates of the third trajectory prediction point. The coordinates of the sampling points on the lane crossing can be calculated according to the direction and curvature of the lane crossing to obtain the third trajectory prediction point. The trajectory point weight corresponding to the third trajectory prediction point can be determined according to the longitudinal distance from the third trajectory prediction point to the target vehicle. At this time, the coordinates of the third trajectory prediction point can be calculated according to the following equation:
[0169]
[0170] Among them, C1~C3 are the linear to cubic coefficients of the lane line pressed by the target vehicle.
[0171] Therefore, for an abnormal driving scenario where the vehicle crosses the line and the second trajectory prediction point is valid, the credibility weights of the first trajectory prediction point, the second trajectory prediction point, the third trajectory prediction point and the historical trajectory point can be determined to be 1, so that the historical trajectory point (P2) and the first fused trajectory point and the second fused trajectory point can be used as the final trajectory points participating in trajectory fitting for trajectory fitting to obtain the target predicted trajectory of the target vehicle.
[0172] For abnormal driving scenarios where the vehicle crosses the line and the second trajectory prediction point is invalid, the credibility weights of the first trajectory prediction point, the third trajectory prediction point and the historical trajectory point can be determined to be 1, and the credibility weight of the second trajectory prediction point can be determined to be 0, so that the first fused trajectory point and the historical trajectory point (P2) can be used as the final trajectory points involved in trajectory fitting to perform trajectory fitting and obtain the target predicted trajectory of the target vehicle.
[0173] For example, see Figure 4d , Figure 4d This is another trajectory prediction schematic diagram of a trajectory prediction method provided in an embodiment of the present application. The figure shows the prediction results of the target prediction trajectory of the trajectory prediction method provided in the embodiment of the present application in the lane-crossing scenario when the second trajectory prediction point is valid and invalid. It can be seen that through the trajectory prediction method provided in the embodiment of the present application, the predicted trajectory of the target vehicle can be optimized through the lane line information, so that the predicted trajectory can be accurately and stably output, and the actual movement trend of the vehicle is maintained, which is consistent with the actual vehicle movement situation, greatly improving the trajectory prediction accuracy.
[0174] In this way, the trajectory prediction method provided in the embodiment of the present application, by introducing the historical trajectory of the self-vehicle and the lane line information to impose certain constraints on the first trajectory prediction point calculated by the dynamic formula, is helpful to accurately predict the driving trajectory of the self-vehicle at a long distance or for a long time. At the same time, the relative relationship between the target vehicle, the third trajectory prediction point and the lane line is considered to calculate the trajectory point weight corresponding to the third trajectory prediction point, which can further make the fitted self-vehicle trajectory more consistent with the actual movement situation. Then, the driving scene of the target vehicle is identified according to the vehicle driving information, environmental information and trajectory points of the target vehicle, so as to formulate different trajectory point fitting schemes for different driving scenes, and then the self-vehicle predicted trajectory adapted to the current driving scene of the target vehicle can be obtained in a variety of driving scenarios, so as to achieve stable and accurate prediction of the vehicle's driving trajectory in a variety of driving scenarios, effectively solve the problems of unstable trajectory prediction and large errors in long-distance trajectory sampling points in the prior art, and further improve the trajectory prediction efficiency.
[0175] As can be seen from the above, the embodiment of the present application obtains the environmental information and vehicle driving information of the target vehicle; determines multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information; and performs trajectory fitting based on the trajectory points to obtain the target predicted trajectory of the target vehicle. In this way, multiple trajectory points of the target vehicle can be determined based on the environmental information and vehicle driving information of the target vehicle, so that the predicted trajectory of the target vehicle can be obtained by trajectories fitting of multiple trajectory points, and the driving trajectory of the vehicle can be predicted stably and accurately, thereby improving the efficiency of trajectory prediction.
[0176] In order to facilitate better implementation of the trajectory prediction method provided in the embodiment of the present application, the embodiment of the present application also provides a device based on the above trajectory prediction method. The meanings of the terms are the same as those in the above trajectory prediction method, and the specific implementation details can refer to the description in the method embodiment.
[0177] For example, Figure 5 As shown, it is a schematic diagram of the structure of the trajectory prediction device provided in an embodiment of the present application. The trajectory prediction device may include a data acquisition module 201, a trajectory point determination module 202 and a trajectory fitting module 203, as follows:
[0178] The data acquisition module 201 is used to acquire the environment information and vehicle driving information of the target vehicle;
[0179] A trajectory point determination module 202, for determining a plurality of trajectory points of a target vehicle based on vehicle driving information and environmental information;
[0180] The trajectory fitting module 203 is used to perform trajectory fitting based on the trajectory points to obtain a target predicted trajectory of the target vehicle.
[0181] In one embodiment, the trajectory points include at least two of a first trajectory prediction point, a second trajectory prediction point, a third trajectory prediction point and a historical trajectory point, the first trajectory prediction point is calculated based on a dynamic formula, the historical trajectory point and the second trajectory prediction point are calculated based on the historical trajectory of the target vehicle, and the third trajectory prediction point is calculated based on environmental information of the target vehicle.
[0182] In one embodiment, the trajectory point includes a first trajectory prediction point, and the trajectory point determination module 202 includes:
[0183] A state determination submodule, used to determine whether the target vehicle stops within a preset time period based on the vehicle driving information;
[0184] A first trajectory point calculation submodule is used to calculate the driving trajectory points of the target vehicle at multiple time points within a preset time period based on the vehicle driving information to obtain a first trajectory prediction point of the target vehicle;
[0185] The second trajectory point calculation submodule is used to calculate the driving trajectory points of the target vehicle at multiple time points within a preset time period based on the vehicle driving information and the curvature information of the target vehicle to obtain the first trajectory prediction point of the target vehicle.
[0186] In one embodiment, the second trajectory point calculation submodule includes:
[0187] A first trajectory point calculation unit is used to calculate the driving trajectory point at the first time point based on the speed information in the vehicle driving information for a plurality of first time points within a preset time length, so as to obtain a first trajectory prediction point;
[0188] A second trajectory point calculation unit is used to calculate, for at least one second time point within a preset time length, a total travel distance of the target vehicle within the preset time length according to the speed information, and calculate the travel trajectory point at the second time point according to the total travel distance and the curvature information of the target vehicle to obtain a first trajectory prediction point of the target vehicle;
[0189] The second time point is after the first time point.
[0190] In one embodiment, the second trajectory point calculation unit is used to:
[0191] Calculate a first driving trajectory point and a second driving trajectory point of the target vehicle in a spatial rectangular coordinate system according to the total driving distance, the curvature in the curvature information, and the curvature change rate;
[0192] The first driving trajectory point and the second driving trajectory point are subjected to coordinate transformation to obtain the transformed first driving trajectory point and the transformed second driving trajectory point in the vehicle coordinate system of the target vehicle;
[0193] Trajectory point fusion is performed based on the converted first driving trajectory point and the converted second driving trajectory point to obtain a first trajectory prediction point of the target vehicle.
[0194] In one embodiment, the trajectory points include the second trajectory prediction points and the historical trajectory points, and the trajectory point determination module 202 includes a third trajectory point calculation submodule for:
[0195] Based on the vehicle driving information, calculate the historical trajectory points of the target vehicle;
[0196] Perform curve fitting based on historical trajectory points to obtain a historical trajectory curve;
[0197] According to the ordinate of the first trajectory prediction point, sampling is performed in the historical trajectory curve to obtain the second trajectory prediction point of the target vehicle.
[0198] In one embodiment, the data processing object includes an application program and a trajectory prediction unit, which is used to:
[0199] The target data is written into the first memory corresponding to the application program through the target thread, and the process of the application program is triggered to call the corresponding data reading function to obtain the target data.
[0200] In one embodiment, the trajectory point includes a third trajectory prediction point, and the trajectory point determination module 202 includes a third trajectory point calculation submodule for:
[0201] Obtaining the lane line equation of the target lane where the target vehicle is located according to the environmental information;
[0202] Based on the lane line equation, the third trajectory prediction point on the lane centerline of the target lane is calculated.
[0203] In one embodiment, the trajectory prediction device further includes a weight determination module, which is used to:
[0204] According to the third trajectory prediction point and the lane center line, a trajectory point weight corresponding to the third trajectory prediction point is determined.
[0205] In one embodiment, the weight determination module is used to:
[0206] Based on the lane line equation of the lane centerline, the target angle between the target vehicle and the target lane, the lateral distance between the target vehicle and the lane centerline, and the longitudinal distance between the target vehicle and the third trajectory prediction point are obtained;
[0207] Based on the target angle, the lateral distance, and the longitudinal distance, a trajectory point weight corresponding to the third trajectory prediction point is determined.
[0208] In some embodiments, the trajectory fitting module 203 includes:
[0209] A scene recognition submodule is used to identify the driving scene of the target vehicle based on the vehicle driving information, environmental information and trajectory points of the target vehicle;
[0210] The trajectory prediction submodule is used to perform trajectory point fitting based on multiple reliable trajectory points in the driving scenario to obtain the target predicted trajectory of the target vehicle.
[0211] In some embodiments, the trajectory prediction submodule is used to:
[0212] When the driving scene is a normal driving scene, a trajectory point weight of the third trajectory prediction point is obtained, and based on the trajectory point weight, the first trajectory prediction point and the third trajectory prediction point are fused to obtain a first fused trajectory point;
[0213] Based on the trajectory point weights, the second trajectory prediction point and the third trajectory prediction point are fused to obtain a second fused trajectory point;
[0214] The first fused trajectory points, the second fused trajectory points and the historical trajectory points are fitted to obtain the target predicted trajectory of the target vehicle.
[0215] In some embodiments, the trajectory prediction submodule is used to:
[0216] When the driving scene is an abnormal driving scene, the credibility weight corresponding to each trajectory point is determined, wherein the credibility weight of an untrustworthy trajectory point is 0;
[0217] The trajectory points are fitted based on the credibility weights to obtain the target predicted trajectory of the target vehicle.
[0218] As can be seen from the above, the embodiment of the present application obtains the environmental information and vehicle driving information of the target vehicle through the data acquisition module 201; the trajectory point determination module 202 determines the multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information; the trajectory fitting module 203 performs trajectory fitting based on the trajectory points to obtain the target predicted trajectory of the target vehicle. In this way, the multiple trajectory points of the target vehicle can be determined according to the environmental information and vehicle driving information of the target vehicle, so that the predicted trajectory of the target vehicle can be obtained by trajectories fitting of multiple trajectory points, and the driving trajectory of the vehicle can be predicted stably and accurately, thereby improving the efficiency of trajectory prediction.
[0219] Accordingly, the embodiment of the present application also provides a vehicle-mounted terminal, such as Figure 6 As shown, Figure 6 A schematic diagram of the structure of the vehicle-mounted terminal provided in an embodiment of the present application. The vehicle-mounted terminal 300 includes a processor 301 having one or more processing cores, a memory 302 having one or more computer-readable storage media, and a computer program stored in the memory 302 and executable on the processor. The processor 301 is electrically connected to the memory 302. Those skilled in the art will appreciate that the vehicle-mounted terminal structure shown in the figure does not constitute a limitation on the vehicle-mounted terminal, and may include more or fewer components than shown in the figure, or combine certain components, or arrange components differently.
[0220] The processor 301 is the control center of the vehicle terminal 300, and uses various interfaces and lines to connect various parts of the entire vehicle terminal 300. It executes various functions and processes data of the vehicle terminal 300 by running or loading software programs and / or units stored in the memory 302, and calling data stored in the memory 302. The processor 301 can be a processor CPU, a graphics processor GPU, a network processor (Network Processor, NP), etc., and can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application.
[0221] In the embodiment of the present application, the processor 301 in the vehicle terminal 300 will load instructions corresponding to the processes of one or more application programs into the memory 302 according to the following steps, and the processor 301 will run the application programs stored in the memory 302 to implement various functions, such as:
[0222] Acquire environmental information and vehicle driving information of the target vehicle; determine multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information; perform trajectory fitting based on the trajectory points to obtain a target predicted trajectory of the target vehicle.
[0223] Furthermore, various functions implemented by running the application stored in the memory 302 can also be found in the description of the aforementioned embodiments, which will not be repeated here.
[0224] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.
[0225] Optional, such as Figure 6 As shown, the vehicle terminal 300 also includes: a touch screen 303, a radio frequency circuit 304, an audio circuit 305, an input unit 306 and a power supply 307. Among them, the processor 301 is electrically connected to the touch screen 303, the radio frequency circuit 304, the audio circuit 305, the input unit 306 and the power supply 307 respectively. Those skilled in the art can understand that Figure 6 The vehicle-mounted terminal structure shown in the figure does not constitute a limitation on the vehicle-mounted terminal, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.
[0226] The touch display screen 303 can be used to display a graphical user interface and receive operation instructions generated by the user acting on the graphical user interface. The touch display screen 303 may include a display panel and a touch panel. Among them, the display panel can be used to display information input by the user or information provided to the user and various graphical user interfaces of the vehicle terminal, which can be composed of graphics, text, icons, videos and any combination thereof. Optionally, the display panel can be configured in the form of a liquid crystal display (LCD, Liquid Crystal Display), an organic light-emitting diode (OLED, Organic Light-Emitting Diode), etc. The touch panel can be used to collect the user's touch operations on or near it (such as the user's operation on the touch panel or near the touch panel using any suitable object or accessory such as a finger, stylus, etc.), and generate corresponding operation instructions, and the operation instructions execute the corresponding program. Optionally, the touch panel may include two parts: a touch detection device and a touch controller. Among them, the touch detection device detects the user's touch direction, detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into the touch point coordinates, and then sends it to the processor 301, and can receive the command sent by the processor 301 and execute it. The touch panel can cover the display panel. When the touch panel detects a touch operation on or near it, it is transmitted to the processor 301 to determine the type of touch event, and then the processor 301 provides a corresponding visual output on the display panel according to the type of touch event. In an embodiment of the present application, the touch panel and the display panel can be integrated into the touch display screen 303 to realize the input and output functions. However, in some embodiments, the touch panel and the touch panel can be used as two independent components to realize the input and output functions. That is, the touch display screen 303 can also be used as a part of the input unit 306 to realize the input function.
[0227] The radio frequency circuit 304 may be used to send and receive radio frequency signals, so as to establish wireless communication with a network device or other vehicle-mounted terminals through wireless communication, and to send and receive signals with the network device or other vehicle-mounted terminals.
[0228] The audio circuit 305 can be used to provide an audio interface between the user and the vehicle-mounted terminal through a speaker and a microphone. The audio circuit 305 can transmit the electrical signal converted from the received audio data to the speaker, which is converted into a sound signal for output; on the other hand, the microphone converts the collected sound signal into an electrical signal, which is received by the audio circuit 305 and converted into audio data, and then the audio data is output to the processor 301 for processing, and then sent to another vehicle-mounted terminal through the radio frequency circuit 304, or the audio data is output to the memory 302 for further processing. The audio circuit 305 may also include an earplug jack to provide communication between an external headset and the vehicle-mounted terminal.
[0229] The input unit 306 may be used to receive input target video and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control.
[0230] The power supply 307 is used to supply power to various components of the vehicle-mounted terminal 300. Optionally, the power supply 307 can be logically connected to the processor 301 through a power management system, so as to manage charging, discharging, and power consumption through the power management system. The power supply 307 can also include one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components.
[0231] although Figure 6 Not shown, the vehicle terminal 300 may also include a camera, a sensor, a wireless fidelity module, a Bluetooth module, etc., which will not be described in detail here.
[0232] In the above embodiments, the description of each embodiment has its own emphasis. For the parts not described in detail in a certain embodiment, please refer to the relevant description of other embodiments. It should be noted that the vehicle-mounted terminal provided in the embodiment of the present application belongs to the same concept as the trajectory prediction method in the above embodiment. The specific implementation process is detailed in the above method embodiment and will not be repeated here.
[0233] As can be seen from the above, the vehicle-mounted terminal provided in the embodiment of the present application can obtain the environmental information and vehicle driving information of the target vehicle; determine multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information; perform trajectory fitting based on the trajectory points to obtain the target predicted trajectory of the target vehicle. In this way, multiple trajectory points of the target vehicle can be determined based on the environmental information and vehicle driving information of the target vehicle, so that the predicted trajectory of the target vehicle can be obtained by trajectories fitting of multiple trajectory points, and the driving trajectory of the vehicle can be predicted stably and accurately, thereby improving the efficiency of trajectory prediction.
[0234] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.
[0235] To this end, an embodiment of the present application provides a computer-readable storage medium, which includes a computer program. When the computer program is run on a vehicle-mounted terminal, the computer program is used to enable the vehicle-mounted terminal to perform any trajectory prediction method provided in the embodiment of the present application. For example, the computer program can perform the following steps of the trajectory prediction method:
[0236] Acquire environmental information and vehicle driving information of the target vehicle; determine multiple trajectory points of the target vehicle based on the vehicle driving information and environmental information; perform trajectory fitting based on the trajectory points to obtain a target predicted trajectory of the target vehicle.
[0237] Furthermore, for the detailed steps of the above method steps, reference can be made to the description in the above embodiments, which will not be repeated here.
[0238] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.
[0239] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.
[0240] Since the computer program stored in the computer-readable storage medium can execute any trajectory prediction method provided in the embodiments of the present application, the beneficial effects that can be achieved by any trajectory prediction method provided in the embodiments of the present application can be achieved. Please refer to the previous embodiments for details and will not be repeated here.
[0241] According to one aspect of the present application, a computer program product is also provided, including a computer program, which is stored in a computer-readable storage medium; when the processor of a vehicle-mounted terminal reads the computer program from the computer-readable storage medium, the processor executes the computer program, so that the vehicle-mounted terminal executes the methods provided in the various optional implementation methods in the above-mentioned embodiments.
[0242] In the above trajectory prediction device, computer-readable storage medium, vehicle-mounted terminal, and computer program product embodiments, the description of each embodiment has its own emphasis. For parts not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process and beneficial effects of the trajectory prediction device, computer-readable storage medium, computer program product, vehicle-mounted terminal and its corresponding units described above can refer to the description of the trajectory prediction method in the above embodiment, and will not be repeated here.
[0243] The above is a detailed introduction to a trajectory prediction method, device, vehicle-mounted terminal, vehicle, computer-readable storage medium and computer program product provided in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present application.
Claims
1. A trajectory prediction method, characterized in that: include: Obtaining environmental information and vehicle driving information of the target vehicle; Based on the vehicle driving information and environmental information, determining multiple trajectory points of the target vehicle, the trajectory points including at least two of a first trajectory prediction point, a second trajectory prediction point, a third trajectory prediction point and a historical trajectory point, the first trajectory prediction point being calculated based on a dynamics formula, the historical trajectory point and the second trajectory prediction point being calculated based on a historical trajectory of the target vehicle, and the third trajectory prediction point being calculated based on environmental information of the target vehicle; Based on the vehicle driving information of the target vehicle, the environmental information and the trajectory points, identifying the driving scene in which the target vehicle is located; Performing trajectory point fitting based on multiple reliable trajectory points in the driving scenario to obtain a target predicted trajectory of the target vehicle; The performing trajectory point fitting based on the multiple reliable trajectory points in the driving scenario to obtain the target predicted trajectory of the target vehicle includes at least one of the following two methods: Method 1: When the driving scene is a normal driving scene, obtaining a trajectory point weight of the third trajectory prediction point, and based on the trajectory point weight, fusing the first trajectory prediction point and the third trajectory prediction point to obtain a first fused trajectory point; Based on the trajectory point weight, the second trajectory prediction point and the third trajectory prediction point are fused to obtain a second fused trajectory point; Performing trajectory fitting on the first fused trajectory point, the second fused trajectory point, and the historical trajectory point to obtain a target predicted trajectory of the target vehicle; Method 2: When the driving scene is an abnormal driving scene, determining the credibility weight corresponding to each of the trajectory points, wherein for an untrustworthy trajectory point, the credibility weight is 0; The trajectory points are fitted based on the credibility weights to obtain a target predicted trajectory of the target vehicle.
2. The trajectory prediction method according to claim 1, characterized in that: The trajectory point includes a first trajectory prediction point, and the multiple trajectory points of the target vehicle are determined based on the vehicle driving information and the environmental information, including: Based on the vehicle driving information, determining whether the target vehicle stops within a preset time period; If yes, based on the vehicle driving information, the driving trajectory points of the target vehicle at multiple time points within the preset time period are calculated to obtain a first trajectory prediction point of the target vehicle; If not, based on the vehicle driving information and the curvature information of the target vehicle, the driving trajectory points of the target vehicle at multiple time points within the preset time period are calculated to obtain a first trajectory prediction point of the target vehicle.
3. The trajectory prediction method according to claim 2, characterized in that: The step of calculating the driving trajectory points of the target vehicle at multiple time points within the preset time period based on the vehicle driving information and the curvature information of the target vehicle to obtain a first trajectory prediction point of the target vehicle includes: For a plurality of first time points within the preset time period, based on the speed information in the vehicle driving information, the driving trajectory point at the first time point is calculated to obtain a first trajectory prediction point; For at least one second time point within the preset time period, calculating the total travel distance of the target vehicle within the preset time period according to the speed information; Calculating the driving trajectory point at the second time point according to the total driving distance and the curvature information of the target vehicle to obtain a first trajectory prediction point of the target vehicle; The second time point is after the first time point.
4. The trajectory prediction method according to claim 3, characterized in that: The step of calculating the driving trajectory point at the second time point according to the total driving distance and the curvature information of the target vehicle to obtain the first trajectory prediction point of the target vehicle includes: Calculate a first driving trajectory point and a second driving trajectory point of the target vehicle in a spatial rectangular coordinate system according to the total driving distance, the curvature in the curvature information, and the curvature change rate; Performing coordinate transformation on the first driving trajectory point and the second driving trajectory point to obtain a transformed first driving trajectory point and a transformed second driving trajectory point in a vehicle coordinate system of the target vehicle; Trajectory point fusion is performed based on the converted first driving trajectory point and the converted second driving trajectory point to obtain a first trajectory prediction point of the target vehicle.
5. The trajectory prediction method according to claim 1, characterized in that: The trajectory points include second trajectory prediction points and historical trajectory points, and the multiple trajectory points of the target vehicle are determined based on the vehicle driving information and the environment information, including: Based on the vehicle driving information, calculating the historical trajectory points of the target vehicle; Performing curve fitting based on the historical trajectory points to obtain a historical trajectory curve; According to the ordinate of the first trajectory prediction point, sampling is performed in the historical trajectory curve to obtain a second trajectory prediction point of the target vehicle.
6. The trajectory prediction method according to claim 1, characterized in that: The trajectory point includes a third trajectory prediction point, and the multiple trajectory points of the target vehicle are determined based on the vehicle driving information and the environmental information, including: Acquire a lane line equation of a target lane where the target vehicle is located according to the environmental information; Based on the lane line equation, a third trajectory prediction point on the lane centerline of the target lane is calculated.
7. The trajectory prediction method according to claim 6, characterized in that: After calculating the third trajectory prediction point on the lane centerline of the target lane based on the lane line equation, the method further includes: A trajectory point weight corresponding to the third trajectory prediction point is determined according to the third trajectory prediction point and the lane center line.
8. The trajectory prediction method according to claim 7, characterized in that: The determining, according to the third trajectory prediction point and the lane center line, a trajectory point weight corresponding to the third trajectory prediction point includes: Based on the lane line equation of the lane centerline, obtaining a target angle between the target vehicle and the target lane, a lateral distance between the target vehicle and the lane centerline, and a longitudinal distance between the target vehicle and the third trajectory prediction point; A trajectory point weight corresponding to the third trajectory prediction point is determined based on the target angle, the lateral distance, and the longitudinal distance.
9. A trajectory prediction device, characterized in that: include: A data acquisition module is used to acquire the environmental information and vehicle driving information of the target vehicle; a trajectory point determination module, for determining multiple trajectory points of the target vehicle based on the vehicle driving information and the environmental information, the trajectory points comprising at least two of a first trajectory prediction point, a second trajectory prediction point, a third trajectory prediction point and a historical trajectory point, the first trajectory prediction point being calculated based on a dynamics formula, the historical trajectory point and the second trajectory prediction point being calculated based on the historical trajectory of the target vehicle, and the third trajectory prediction point being calculated based on the environmental information of the target vehicle; A trajectory fitting module is used to identify the driving scene of the target vehicle based on the vehicle driving information of the target vehicle, the environmental information and the trajectory points, and perform trajectory point fitting based on multiple reliable trajectory points in the driving scene to obtain a target predicted trajectory of the target vehicle; The performing trajectory point fitting based on the multiple reliable trajectory points in the driving scenario to obtain the target predicted trajectory of the target vehicle includes at least one of the following two methods: Method 1: When the driving scene is a normal driving scene, obtaining a trajectory point weight of the third trajectory prediction point, and based on the trajectory point weight, fusing the first trajectory prediction point and the third trajectory prediction point to obtain a first fused trajectory point; Based on the trajectory point weight, the second trajectory prediction point and the third trajectory prediction point are fused to obtain a second fused trajectory point; Performing trajectory fitting on the first fused trajectory point, the second fused trajectory point, and the historical trajectory point to obtain a target predicted trajectory of the target vehicle; Method 2: When the driving scene is an abnormal driving scene, determining the credibility weight corresponding to each of the trajectory points, wherein for an untrustworthy trajectory point, the credibility weight is 0; The trajectory points are fitted based on the credibility weights to obtain a target predicted trajectory of the target vehicle.
10. A vehicle-mounted terminal, characterized in that: It comprises a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of any one of the methods of claims 1 to 8.
11. A vehicle, characterized in that: The vehicle includes the vehicle-mounted terminal as claimed in claim 10.
12. A computer-readable storage medium, characterized in that: It includes a computer program. When the computer program is run on a vehicle-mounted terminal, the computer program is used to enable the vehicle-mounted terminal to execute the steps of any method described in claims 1 to 8.
Citation Information
Cited By
Method for predicting goods stacking deviation of unmanned clamping and holding vehicle based on kinematics and Kalman filtering
CN119958552A
