Prediction trajectory evaluation method and device, electronic equipment and computer program product

Abstract

The invention discloses a predicted trajectory evaluation method and device, electronic equipment and a computer program product. The method comprises the following steps: acquiring an obstacle predicted trajectory output by an autonomous vehicle; performing multi-dimensional evaluation on the obstacle prediction trajectory by using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, the multi-dimensional evaluation index including a physical evaluation dimension, a road traffic evaluation dimension and a precision evaluation dimension; and determining a comprehensive evaluation result of the obstacle prediction trajectory according to the multi-dimensional evaluation result of the obstacle prediction trajectory. According to the prediction trajectory evaluation method provided by the embodiment of the invention, the obstacle trajectory predicted by the automatic driving vehicle is comprehensively evaluated from multiple dimensions of a physical evaluation dimension, a road traffic evaluation dimension and a precision evaluation dimension; the influence of factors such as physical rules of obstacle trajectory motion and road geometric constraints is further considered, and the accuracy and comprehensiveness of obstacle prediction trajectory evaluation are improved.

Description

Technical Field

[0001] The present application relates to the field of autonomous driving technology, and in particular to a prediction trajectory evaluation method, device, electronic equipment, and computer program product. Background Art

[0002] In autonomous driving scenarios, accurately predicting the future movement trajectory of surrounding obstacles is crucial to achieving safe and efficient autonomous driving. The quality of this prediction capability is directly related to whether the autonomous vehicle can make reasonable decisions in a complex and changing traffic environment, thereby avoiding potential collisions and ensuring driving safety.

[0003] In order to quantitatively evaluate the quality of the predicted trajectory, the industry usually uses some evaluation indicators, including average displacement error (ADE), final displacement error (FDE) and miss rate (MR). These indicators provide an intuitive data basis for evaluation by calculating the spatial distance difference between the predicted trajectory and the actual movement trajectory of the obstacle.

[0004] However, the existing evaluation system ignores the constraints of factors such as the physical and kinematic characteristics of the predicted trajectory and road traffic rules. These factors are not only important bases for measuring the rationality of the predicted trajectory, but also key elements for evaluating its drivability in the actual road environment. Therefore, the current predicted trajectory evaluation system has obvious deficiencies in comprehensiveness and accuracy. Summary of the invention

[0005] The embodiments of the present application provide a prediction trajectory evaluation method, device, electronic device, and computer program product to improve the accuracy and comprehensiveness of obstacle prediction trajectory evaluation.

[0006] The present application embodiment adopts the following technical solutions:

[0007] In a first aspect, an embodiment of the present application provides a prediction trajectory evaluation method, wherein the prediction trajectory evaluation method includes:

[0008] Obtain obstacle prediction trajectory output by the autonomous vehicle;

[0009] Performing a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, wherein the multi-dimensional evaluation index includes a physical evaluation dimension, a road traffic evaluation dimension, and an accuracy evaluation dimension;

[0010] A comprehensive evaluation result of the obstacle prediction trajectory is determined according to the multi-dimensional evaluation result of the obstacle prediction trajectory.

[0011] Optionally, the obstacle prediction trajectory includes a plurality of prediction trajectory points, the physical evaluation dimension includes at least one physical evaluation sub-dimension of speed, acceleration and angular velocity, and the multi-dimensional evaluation result of the obstacle prediction trajectory obtained by performing a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index includes:

[0012] Calculate the index values ​​of multiple predicted trajectory points in each physical evaluation sub-dimension;

[0013] Determine the maximum index value of the multiple predicted trajectory points in each physical evaluation sub-dimension according to the index values ​​of the multiple predicted trajectory points in each physical evaluation sub-dimension;

[0014] Compare the maximum index values ​​of multiple predicted trajectory points in each physical evaluation sub-dimension with the index threshold corresponding to each physical evaluation sub-dimension to obtain the evaluation score of each physical evaluation sub-dimension;

[0015] The physical evaluation score of the obstacle prediction trajectory is calculated according to the evaluation scores of each physical evaluation sub-dimension and the corresponding weight coefficient.

[0016] Optionally, the road traffic assessment dimension includes at least one road traffic assessment sub-dimension of a road edge line, a lane solid line, a speed limit, and a road topology, and the multi-dimensional assessment of the obstacle prediction trajectory using a multi-dimensional assessment index to obtain a multi-dimensional assessment result of the obstacle prediction trajectory includes:

[0017] Obtaining map data of the location of the autonomous driving vehicle, wherein the map data includes road edge line data, lane solid line data, speed limit data, and road connection relationship data;

[0018] Calculating the evaluation scores of each road traffic evaluation sub-dimension according to the map data and the obstacle prediction trajectory;

[0019] The road traffic evaluation score of the obstacle prediction trajectory is calculated based on the evaluation scores of each road traffic evaluation sub-dimension and the corresponding weight coefficient.

[0020] Optionally, the calculating the evaluation scores of each road traffic evaluation sub-dimension according to the map data and the obstacle prediction trajectory includes:

[0021] Calculating the distances between a plurality of predicted trajectory points and the road edge line and the lane solid line according to the road edge line data and the lane solid line data, wherein the distances include the distance between the first predicted trajectory point and the road edge line and the lane solid line;

[0022] Determine the minimum distance among the distances between the plurality of predicted trajectory points and the road edge line and the lane solid line;

[0023] A road edge line evaluation score and a lane solid line evaluation score are calculated based on the minimum distance, the distance between the first predicted trajectory point and the road edge line and the lane solid line, and a distance threshold.

[0024] Optionally, the calculating the evaluation scores of each road traffic evaluation sub-dimension according to the map data and the obstacle prediction trajectory includes:

[0025] Calculating a topological structure of a road where an obstacle is located according to the road connection relationship data, wherein the topological structure of the road where the obstacle is located includes a lane centerline;

[0026] Performing interpolation processing on the lane centerline to obtain an interpolated lane centerline;

[0027] Calculating a minimum average displacement error between a plurality of predicted trajectory points and the interpolated lane centerline;

[0028] A road topology evaluation score is calculated based on the minimum average displacement error and the average displacement error threshold.

[0029] Optionally, the accuracy assessment dimension includes at least one accuracy assessment sub-dimension of average displacement error and endpoint displacement error, and the multi-dimensional assessment of the obstacle prediction trajectory using the multi-dimensional assessment index to obtain the multi-dimensional assessment result of the obstacle prediction trajectory includes:

[0030] Obtain the actual trajectory points of the obstacle, and transform the predicted trajectory points and actual trajectory points of the obstacle into the obstacle coordinate system;

[0031] Calculate the evaluation scores of each accuracy evaluation sub-dimension according to the predicted trajectory points and the actual trajectory points in the obstacle coordinate system;

[0032] The accuracy evaluation score of the obstacle prediction trajectory is calculated based on the evaluation scores of each accuracy evaluation sub-dimension and the corresponding weight coefficients as well as the current driving scenario.

[0033] Optionally, the average displacement error includes an average lateral displacement error and an average longitudinal displacement error, the endpoint displacement error includes an endpoint lateral displacement error and an endpoint longitudinal displacement error, and the accuracy evaluation score of the obstacle prediction trajectory calculated according to the evaluation scores of each accuracy evaluation sub-dimension and the corresponding weight coefficient and the current driving scene includes:

[0034] When the current driving scene is an obstacle lane change scene or a turning scene, setting a magnification factor corresponding to the average lateral displacement error and the end point lateral displacement error;

[0035] The accuracy evaluation score of the obstacle prediction trajectory is calculated according to the evaluation scores and corresponding weight coefficients of each accuracy evaluation sub-dimension, and the magnification coefficients corresponding to the average lateral displacement error and the end point lateral displacement error.

[0036] In a second aspect, an embodiment of the present application further provides a predicted trajectory evaluation device, wherein the predicted trajectory evaluation device comprises:

[0037] An acquisition unit, used for acquiring an obstacle prediction trajectory output by the autonomous driving vehicle;

[0038] An evaluation unit, configured to perform a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, wherein the multi-dimensional evaluation index includes a physical evaluation dimension, a road traffic evaluation dimension, and an accuracy evaluation dimension;

[0039] A determination unit is used to determine a comprehensive evaluation result of the obstacle prediction trajectory according to the multi-dimensional evaluation result of the obstacle prediction trajectory.

[0040] In a third aspect, an embodiment of the present application further provides an electronic device, including:

[0041] A processor; and a memory arranged to store computer executable instructions, wherein when the executable instructions are executed, the processor performs any of the aforementioned prediction trajectory evaluation methods.

[0042] In a fourth aspect, an embodiment of the present application further provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements any of the aforementioned prediction trajectory evaluation methods.

[0043] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects: the predicted trajectory evaluation method of the embodiments of the present application first obtains the obstacle prediction trajectory output by the autonomous driving vehicle; then uses multi-dimensional evaluation indicators to perform multi-dimensional evaluation on the obstacle prediction trajectory to obtain the multi-dimensional evaluation result of the obstacle prediction trajectory, and the multi-dimensional evaluation indicators include physical evaluation dimensions, road traffic evaluation dimensions, and accuracy evaluation dimensions; finally, the comprehensive evaluation result of the obstacle prediction trajectory is determined according to the multi-dimensional evaluation results of the obstacle prediction trajectory. The predicted trajectory evaluation method of the embodiments of the present application comprehensively evaluates the obstacle trajectory predicted by the autonomous driving vehicle from multiple dimensions such as physical evaluation dimensions, road traffic evaluation dimensions, and accuracy evaluation dimensions. On the basis of spatial distance, it further considers the physical laws of obstacle trajectory movement and the influence of factors such as road geometric constraints, thereby improving the accuracy and comprehensiveness of obstacle prediction trajectory evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0045] Figure 1 A schematic diagram of a flow chart of a prediction trajectory evaluation method in an embodiment of the present application;

[0046] Figure 2 A schematic diagram of an obstacle prediction trajectory evaluation system in an embodiment of the present application;

[0047] Figure 3 This is a schematic diagram of the structure of a prediction trajectory evaluation device in an embodiment of the present application;

[0048] Figure 4 This is a schematic diagram of the structure of an electronic device in an embodiment of the present application. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0050] The technical solutions provided by various embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0051] The present application embodiment provides a prediction trajectory evaluation method, such as Figure 1 As shown, a flowchart of a prediction trajectory evaluation method in an embodiment of the present application is provided, and the prediction trajectory evaluation method at least includes the following steps S110 to S130:

[0052] Step S110, obtaining the obstacle prediction trajectory output by the autonomous driving vehicle.

[0053] When performing predicted trajectory evaluation, it is necessary to first obtain the obstacle prediction trajectory output by the autonomous driving vehicle. The obstacles here mainly refer to the surrounding motor vehicles on the road where the autonomous driving vehicle is located. The predicted trajectory evaluation method of the embodiment of the present application mainly focuses on the predicted trajectory evaluation problem of motor vehicles traveling on the road. Non-motor vehicles and pedestrians are obviously different from motor vehicles in terms of their movement range and movement laws, and therefore are not discussed in this application.

[0054] The obstacle prediction trajectory can be obtained based on the prediction module of the autonomous vehicle or related sensor data. These prediction trajectories are usually generated based on the vehicle's sensor input (such as lidar, camera, etc.) and the vehicle's current motion state (such as speed, acceleration, direction, etc.). The prediction trajectory usually includes information such as the future position, speed, acceleration, etc. of the obstacle, which is usually expressed in the form of a time series, that is, predicting the possible position and state of the obstacle at different time points in the future.

[0055] Step S120, performing a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, wherein the multi-dimensional evaluation index includes a physical evaluation dimension, a road traffic evaluation dimension, and an accuracy evaluation dimension.

[0056] The embodiment of the present application designs a multi-dimensional evaluation index for the evaluation of the predicted trajectory of obstacles, and comprehensively and accurately evaluates the predicted trajectory of obstacles from multiple different aspects, specifically including the following aspects:

[0057] 1) Physical evaluation dimension: This dimension focuses on the rationality of the physical motion of the predicted trajectory, such as whether the parameters such as speed, acceleration, angular velocity, etc. are within the physically achievable range. For example, if the acceleration in the predicted trajectory suddenly becomes very large or very small, it may indicate that there is something unreasonable about the prediction.

[0058] 2) Road traffic evaluation dimension: This dimension evaluates whether the predicted trajectory complies with road traffic rules and customs. For example, whether the predicted trajectory follows the correct lane and whether it complies with the constraints of road traffic signs.

[0059] 3) Accuracy evaluation dimension: This dimension mainly evaluates the error between the predicted trajectory and the actual trajectory. This usually includes traditional evaluation indicators such as average displacement error (ADE), terminal displacement error (FDE), and miss rate (MR). However, the embodiments of the present application can also be expanded on this basis to consider the error sensitivity in different directions (such as horizontal and vertical directions), as well as the error distribution of the predicted trajectory at different time points.

[0060] Step S130, determining a comprehensive evaluation result of the obstacle prediction trajectory according to the multi-dimensional evaluation results of the obstacle prediction trajectory.

[0061] Based on the evaluation results of multiple dimensions obtained in the above steps, the final evaluation result of the obstacle prediction trajectory is comprehensively calculated by combining the evaluation results of multiple aspects.

[0062] The predicted trajectory evaluation method of the embodiment of the present application comprehensively evaluates the obstacle trajectory predicted by the autonomous driving vehicle from multiple dimensions, including physical evaluation dimension, road traffic evaluation dimension, and accuracy evaluation dimension. On the basis of spatial distance, it further considers the influence of factors such as the physical laws of obstacle trajectory movement and road geometric constraints, thereby improving the accuracy and comprehensiveness of obstacle prediction trajectory evaluation.

[0063] In some embodiments of the present application, the obstacle prediction trajectory includes multiple predicted trajectory points, the physical evaluation dimension includes at least one physical evaluation sub-dimension of speed, acceleration and angular velocity, and the use of multi-dimensional evaluation indicators to perform multi-dimensional evaluation on the obstacle prediction trajectory to obtain the multi-dimensional evaluation result of the obstacle prediction trajectory includes: calculating the index values ​​of the multiple predicted trajectory points in each physical evaluation sub-dimension; determining the maximum index value of the multiple predicted trajectory points in each physical evaluation sub-dimension according to the index values ​​of the multiple predicted trajectory points in each physical evaluation sub-dimension; comparing the maximum index values ​​of the multiple predicted trajectory points in each physical evaluation sub-dimension with the index threshold corresponding to each physical evaluation sub-dimension to obtain the evaluation score of each physical evaluation sub-dimension; calculating the physical evaluation score of the obstacle prediction trajectory according to the evaluation score of each physical evaluation sub-dimension and the corresponding weight coefficient.

[0064] The obstacle prediction trajectory of the embodiment of the present application is composed of multiple prediction trajectory points. The physical evaluation dimension can be further composed of multiple physical evaluation sub-dimensions such as speed, acceleration and angular velocity. The obstacle prediction trajectory is physically evaluated from multiple physical evaluation sub-dimensions, and finally the physical evaluation score S1 of the obstacle prediction trajectory is comprehensively calculated.

[0065] Specifically, the speed threshold T1 (representing the maximum speed at which the obstacle can actually travel), the acceleration threshold T2 (representing the maximum acceleration that the obstacle can actually generate), and the angular speed threshold T3 (representing the maximum angular speed at which the obstacle can actually travel) can be set in advance according to traffic regulations and the physical laws of obstacle movement.

[0066] Then, the speed, acceleration, and angular velocity corresponding to each predicted trajectory point are calculated based on the position of the predicted trajectory point and the time interval between adjacent trajectory points, and then the maximum speed Vmax, maximum acceleration Amax, and maximum angular velocity Wmax among these trajectory points are determined respectively.

[0067] The calculated maximum speed Vmax, maximum acceleration Amax, and maximum angular velocity Wmax are compared with the corresponding speed threshold T1, acceleration threshold T2, and angular velocity threshold T3, respectively, to determine whether the maximum values ​​of speed, acceleration, and angular velocity exceed the corresponding thresholds. If so, the corresponding scores are assigned to -1, otherwise to 0. For example, if only the maximum speed exceeds the speed threshold, only the speed score is assigned to -1, and the rest are 0. If all three maximum values ​​exceed the corresponding thresholds, the speed score, acceleration score, and angular velocity score are assigned to -1, respectively. Of course, the specific assignment scores and methods can be flexibly adjusted according to actual needs, and are not specifically limited here.

[0068] Finally, the physical evaluation score S1 of the obstacle prediction trajectory is calculated comprehensively, which can be calculated by weighted summation, for example, as follows:

[0069] S1=w1*S_v+w2*S_a+w3*S_w,

[0070] Among them, S_v is the speed score, S_a is the acceleration score, S_w is the angular velocity score, and w1, w2, and w3 are the corresponding weight coefficients.

[0071] It should be noted that in the above embodiment, it is only necessary to compare the maximum index values ​​of multiple predicted trajectory points in each physical evaluation sub-dimension with the corresponding threshold. As long as the maximum value exceeds the threshold requirement, it will be considered that the trajectory quality is affected and the score will be deducted accordingly. It is not necessary to compare the index values ​​of each predicted trajectory point, thereby improving the efficiency of the evaluation calculation.

[0072] In some embodiments of the present application, the predicted obstacle trajectory may be sampled at equal intervals first, and a plurality of predicted trajectory points after sampling may be obtained for subsequent evaluation and calculation. Compared with directly processing all trajectory points, the processing efficiency can be improved.

[0073] In some embodiments of the present application, the road traffic evaluation dimension includes at least one road traffic evaluation sub-dimension of road edge line, lane solid line, speed limit, and road topology, and the use of multi-dimensional evaluation indicators to perform multi-dimensional evaluation on the obstacle prediction trajectory to obtain the multi-dimensional evaluation result of the obstacle prediction trajectory includes: obtaining map data of the location of the autonomous driving vehicle, the map data including road edge line data, lane solid line data, speed limit data, and road connection relationship data; calculating the evaluation score of each road traffic evaluation sub-dimension based on the map data and the obstacle prediction trajectory; calculating the road traffic evaluation score of the obstacle prediction trajectory based on the evaluation score of each road traffic evaluation sub-dimension and the corresponding weight coefficient.

[0074] The road traffic assessment dimension designed in the embodiment of the present application can be specifically composed of multiple road traffic assessment sub-dimensions such as road edge lines, lane solid lines, speed limits, road topology, etc., and the road traffic assessment is performed on the obstacle prediction trajectory from multiple road traffic assessment sub-dimensions, and finally the road traffic assessment score S2 of the obstacle prediction trajectory is comprehensively calculated.

[0075] Specifically, the distance threshold T4, speed limit threshold T5 and average displacement error threshold T6 can be set based on the actual road conditions. Data such as road edge lines, lane solid lines, speed limits and road connection relationships are extracted from the map data of the current location of the autonomous driving vehicle. The road edge lines are located on both sides of the road, and are mainly used to determine whether the obstacle prediction trajectory presses the edge lines based on the distance threshold. The vehicle solid lines are located on both sides of the lane, and are mainly used to determine whether the obstacle prediction trajectory presses the lane solid line based on the distance threshold. Different roads and different sections may have different speed limit requirements. The speed limit threshold is mainly used to determine whether the speed of the obstacle prediction trajectory point exceeds the speed limit requirement of the section. The road topology reflects the road structure and connection relationship information in front of the obstacle, and is mainly used to determine whether the obstacle prediction trajectory is traveling according to the road topology structure based on the average displacement error threshold.

[0076] Therefore, based on the above data such as road edge line, lane solid line, speed limit and road connection relationship, the evaluation scores of the obstacle prediction trajectory in the road edge line, lane solid line, speed limit, road topology and other road traffic evaluation sub-dimensions can be calculated respectively, and finally the road traffic evaluation score S2 of the obstacle prediction trajectory is comprehensively calculated, which can be calculated by weighted summation, for example, it can be expressed as follows:

[0077] S2=w4*S_e+w5*S_l+w6*S_s+w7*S_r,

[0078] Among them, S_e is the road edge score, S_l is the lane solid line score, S_s is the speed limit score, S_r is the driving score along the road topology, and w4, w5, w6, and w7 are the corresponding weight coefficients.

[0079] In some embodiments of the present application, the calculation of the evaluation scores of each road traffic evaluation sub-dimension based on the map data and the obstacle prediction trajectory includes: calculating the distances between multiple predicted trajectory points and the road edge lines and lane solid lines based on the road edge line data and the lane solid line data, the distances including the distances between the first predicted trajectory point and the road edge line and the lane solid line; determining the minimum distance among the distances between multiple predicted trajectory points and the road edge line and the lane solid line; calculating the road edge line evaluation score and the lane solid line evaluation score based on the minimum distance, the distance between the first predicted trajectory point and the road edge line and the lane solid line, and the distance threshold.

[0080] For the calculation of the road edge score and lane solid line score in the above-mentioned embodiment, the distance between each predicted trajectory point and the road edge line and the lane solid line can be calculated first, and the Euclidean distance can be used for calculation. Then, the minimum distance is determined among the distances between multiple predicted trajectory points and the road edge line and the lane solid line, and it is determined whether the minimum distance to the road edge line and the minimum distance to the lane solid line are less than the distance threshold T4. If the minimum distance to the road edge line is less than the distance threshold, it is considered that the curb has been pressed. If the minimum distance to the lane solid line is less than the distance threshold T4, it is considered that the lane solid line has been pressed, and the corresponding road edge line score and lane solid line score are assigned to -1, otherwise they are assigned to 0.

[0081] A special case that needs to be emphasized here is that if the distance between the first trajectory point of the obstacle prediction trajectory (i.e., the current position) and the edge of the road or the solid line of the lane is less than the distance threshold T4, it is considered that the obstacle is currently on the edge of the road or the solid line of the lane. At this time, the predicted trajectory will inevitably press the curb or the solid line, and the value is directly assigned to 0, that is, this situation is not considered to be an abnormality in the predicted trajectory.

[0082] Similarly, for the calculation of the speed limit score, it can be determined whether the maximum speed in the predicted trajectory point exceeds the speed limit threshold T5. If it exceeds, the corresponding speed limit score is assigned to -1, otherwise it is 0.

[0083] In some embodiments of the present application, the calculation of the evaluation scores of each road traffic evaluation sub-dimension based on the map data and the obstacle prediction trajectory includes: calculating the topological structure of the road where the obstacle is located based on the road connection relationship data, the topological structure of the road where the obstacle is located includes the lane centerline; interpolating the lane centerline to obtain the interpolated lane centerline; calculating the minimum average displacement error between multiple predicted trajectory points and the interpolated lane centerline; and calculating the road topology evaluation score based on the minimum average displacement error and the average displacement error threshold.

[0084] When calculating the driving score along the road topology, the embodiment of the present application can first calculate the topological structure of the road in front of the obstacle based on the road connection relationship data, specifically including the lane centerline and the connection relationship between the lane centerlines. Since the lane centerline information provided in the map data may only include the starting position and the end position of the lane centerline, the embodiment of the present application can interpolate the lane centerline so that the interval between the map lane centerline points is a fixed distance such as 1.0 meter.

[0085] Then, the average displacement error ADE is calculated based on each predicted trajectory point and the nearest lane centerline point in the road topology, and the minimum average displacement error is determined and compared with the average displacement error threshold T6. If it exceeds, it is considered that the predicted trajectory obviously does not follow the road topology, and the corresponding driving score along the road topology is assigned to -1, otherwise it is assigned to 0.

[0086] In some embodiments of the present application, the accuracy assessment dimension includes at least one accuracy assessment sub-dimension of average displacement error and endpoint displacement error, and the use of multi-dimensional evaluation indicators to perform multi-dimensional evaluation on the obstacle prediction trajectory to obtain the multi-dimensional evaluation result of the obstacle prediction trajectory includes: obtaining the real trajectory points of the obstacle, and converting the predicted trajectory points and the real trajectory points of the obstacle to the obstacle coordinate system; calculating the evaluation scores of each accuracy assessment sub-dimension based on the predicted trajectory points and the real trajectory points in the obstacle coordinate system; calculating the accuracy assessment score of the obstacle prediction trajectory based on the evaluation scores of each accuracy assessment sub-dimension and the corresponding weight coefficients and the current driving scene.

[0087] Considering that in some critical scenarios, such as when an obstacle suddenly cuts into the lane where the autonomous vehicle is located, it is difficult to meet the evaluation requirements of specific scenarios by relying solely on existing spatial distance indicators. Because in this scenario, the importance of lateral distance error far exceeds that of longitudinal distance error, and even a small lateral deviation may lead to misjudgment of the obstacle's movement intention, thereby endangering driving safety. The evaluation needs in this scenario highlight the shortcomings of traditional indicators in capturing the sensitivity of errors in specific directions.

[0088] Based on this, the embodiment of the present application, based on the design of accuracy assessment sub-dimensions such as average displacement error and endpoint displacement error, further combines the specific driving scenario to perform multi-dimensional accuracy assessment on the obstacle prediction trajectory, and finally comprehensively calculates the accuracy assessment score S3 of the obstacle prediction trajectory.

[0089] Specifically, when calculating the accuracy score, you can first obtain the actual trajectory points of the obstacle, and transform both the predicted trajectory points and the actual trajectory points into the obstacle coordinate system, that is, the direction of the obstacle front is the X-axis, the left side of the front is the Y-axis, and the location of the obstacle is the coordinate origin.

[0090] Then, the average displacement error ADE and the end point displacement error FDE are calculated respectively according to the predicted trajectory points and the corresponding real trajectory points in the obstacle coordinate system. The corresponding weight coefficients are adjusted according to the influence of the current driving scene. Finally, the accuracy evaluation score of the obstacle prediction trajectory is calculated to improve the accuracy of the accuracy score calculation in specific scenarios.

[0091] In some embodiments of the present application, the average displacement error includes an average lateral displacement error and an average longitudinal displacement error, the endpoint displacement error includes an endpoint lateral displacement error and an endpoint longitudinal displacement error, and the accuracy evaluation score of the obstacle prediction trajectory calculated according to the evaluation scores and corresponding weight coefficients of each accuracy evaluation sub-dimension and the current driving scene includes: when the current driving scene is an obstacle lane change scene or a turning scene, setting a magnification coefficient corresponding to the average lateral displacement error and the endpoint lateral displacement error; calculating the accuracy evaluation score of the obstacle prediction trajectory according to the evaluation scores and corresponding weight coefficients of each accuracy evaluation sub-dimension, and the magnification coefficient corresponding to the average lateral displacement error and the endpoint lateral displacement error.

[0092] The average displacement error specifically calculated in the embodiment of the present application includes an average lateral displacement error ADE_Y and an average longitudinal displacement error ADE_X, and the endpoint displacement error includes an endpoint lateral displacement error FDE_Y and an endpoint longitudinal displacement error FDE_X.

[0093] In scenarios such as when an obstacle suddenly cuts into the lane where the autonomous vehicle is located, the importance of the lateral distance error far exceeds the longitudinal distance error. If the original weight coefficient is still used for weighted calculation, the impact of the lateral error will be weakened. Therefore, when the embodiment of the present application identifies that the obstacle is currently in a lane change or turn, an amplification factor α (α>1.0) can be set for the lateral displacement error, and the importance of the lateral displacement error can be increased by multiplying the amplification factor α. The adjusted average displacement error ADE and end point displacement error FDE can be expressed as follows:

[0094] ADE=(ADE_X^2+(α*ADE_Y)^2)^(1 / 2),

[0095] FDE=(FDE_X^2+(α*FDE_Y)^2)^(1 / 2),

[0096] Finally, the weighted score S3 of the accuracy score is calculated, which can be expressed as follows:

[0097] S3=w8*S_ADE+w9*S_FDE,

[0098] Among them, S_ADE is the average displacement error score, which is the inverse of the average displacement error ADE, S_FDE is the end point displacement error score, which is the inverse of the end point displacement error FDE, and w8 and w9 are the corresponding weighting coefficients.

[0099] Finally, the physical evaluation score S1, road traffic evaluation score S2 and accuracy evaluation score S3 calculated in the above embodiment are combined to calculate the comprehensive score S=S1+S2+S3 of the obstacle prediction trajectory. Since the scoring calculation in the above embodiment adopts the deduction system, S, S1, S2 and S3 are all non-positive numbers. Of course, the addition system can also be used, which will not be elaborated here.

[0100] The higher the predicted trajectory score is, the better the predicted trajectory quality is. If the predicted trajectory does not exceed the speed, acceleration, corner speed threshold, does not exceed the lane, does not cross the line, does not exceed the road speed limit, and travels along the road topology, the predicted trajectory score is only related to the accuracy score. Otherwise, the corresponding item score will be subtracted from the accuracy score, and the trajectory quality will be reduced accordingly.

[0101] In order to facilitate intuitive understanding of the above embodiments, Figure 2 As shown, a schematic diagram of an evaluation system for obstacle prediction trajectory in an embodiment of the present application is provided.

[0102] In summary, the prediction trajectory evaluation method of the present application has achieved at least the following technical effects:

[0103] 1) The predicted trajectory evaluation method proposed in this application fully considers factors such as physical properties and road traffic rules, takes into account the kinematic requirements of speed, acceleration, and angular velocity of the predicted trajectory, as well as traffic rules such as not pressing the curb, not pressing the solid line, complying with the road speed limit, and driving along the road topology. The physical score, road traffic score, and accuracy score are combined to calculate the final score of the predicted trajectory. The factors considered are more comprehensive, the scoring is more scientific and reasonable, and a weight coefficient is introduced to flexibly adjust the weights of each score to meet the evaluation needs of different scenarios.

[0104] 2) This application introduces a magnification factor for specific scenarios, which increases the importance of the lateral position error and overcomes the shortcoming of only using the Euclidean distance to weaken the impact of the lateral position error.

[0105] The present application embodiment also provides a prediction trajectory evaluation device 300, such as Figure 3 As shown, a schematic diagram of the structure of a predicted trajectory evaluation device in an embodiment of the present application is provided, wherein the predicted trajectory evaluation device 300 includes: an acquisition unit 310, an evaluation unit 320, and a determination unit 330, wherein:

[0106] An acquisition unit 310 is used to acquire the obstacle prediction trajectory output by the autonomous driving vehicle;

[0107] An evaluation unit 320 is used to perform a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, wherein the multi-dimensional evaluation index includes a physical evaluation dimension, a road traffic evaluation dimension, and an accuracy evaluation dimension;

[0108] The determination unit 330 is used to determine the comprehensive evaluation result of the obstacle prediction trajectory according to the multi-dimensional evaluation result of the obstacle prediction trajectory.

[0109] In some embodiments of the present application, the obstacle prediction trajectory includes multiple predicted trajectory points, the physical evaluation dimension includes at least one physical evaluation sub-dimension of speed, acceleration and angular velocity, and the evaluation unit 320 is specifically used to: calculate the index values ​​of the multiple predicted trajectory points in each physical evaluation sub-dimension; determine the maximum index value of the multiple predicted trajectory points in each physical evaluation sub-dimension according to the index values ​​of the multiple predicted trajectory points in each physical evaluation sub-dimension; compare the maximum index value of the multiple predicted trajectory points in each physical evaluation sub-dimension with the index threshold corresponding to each physical evaluation sub-dimension, respectively, to obtain the evaluation score of each physical evaluation sub-dimension; calculate the physical evaluation score of the obstacle prediction trajectory according to the evaluation score of each physical evaluation sub-dimension and the corresponding weight coefficient.

[0110] In some embodiments of the present application, the road traffic evaluation dimension includes at least one road traffic evaluation sub-dimension of road edge line, lane solid line, speed limit, and road topology, and the evaluation unit 320 is specifically used to: obtain map data of the location of the autonomous driving vehicle, the map data including road edge line data, lane solid line data, speed limit data, and road connection relationship data; calculate the evaluation score of each road traffic evaluation sub-dimension based on the map data and the obstacle prediction trajectory; calculate the road traffic evaluation score of the obstacle prediction trajectory based on the evaluation score of each road traffic evaluation sub-dimension and the corresponding weight coefficient.

[0111] In some embodiments of the present application, the evaluation unit 320 is specifically used to: calculate the distances between multiple predicted trajectory points and the road edge line and the lane solid line based on the road edge line data and the lane solid line data, the distances including the distance between the first predicted trajectory point and the road edge line and the lane solid line; determine the minimum distance among the distances between multiple predicted trajectory points and the road edge line and the lane solid line; calculate the road edge line evaluation score and the lane solid line evaluation score based on the minimum distance, the distance between the first predicted trajectory point and the road edge line and the lane solid line, and the distance threshold.

[0112] In some embodiments of the present application, the evaluation unit 320 is specifically used to: calculate the topological structure of the road where the obstacle is located based on the road connection relationship data, and the topological structure of the road where the obstacle is located includes the lane centerline; interpolate the lane centerline to obtain the interpolated lane centerline; calculate the minimum average displacement error between multiple predicted trajectory points and the interpolated lane centerline; calculate the road topology evaluation score based on the minimum average displacement error and the average displacement error threshold.

[0113] In some embodiments of the present application, the accuracy assessment dimension includes at least one accuracy assessment sub-dimension of average displacement error and endpoint displacement error, and the evaluation unit 320 is specifically used to: obtain the actual trajectory points of the obstacle, and convert the predicted trajectory points and the actual trajectory points of the obstacle into the obstacle coordinate system; calculate the evaluation scores of each accuracy assessment sub-dimension according to the predicted trajectory points and the actual trajectory points in the obstacle coordinate system; calculate the accuracy assessment score of the obstacle predicted trajectory according to the evaluation scores of each accuracy assessment sub-dimension and the corresponding weight coefficients and the current driving scene.

[0114] In some embodiments of the present application, the average displacement error includes an average lateral displacement error and an average longitudinal displacement error, and the terminal displacement error includes an terminal lateral displacement error and a terminal longitudinal displacement error. The evaluation unit 320 is specifically used to: when the current driving scene is an obstacle lane change scene or a turning scene, set the amplification coefficient corresponding to the average lateral displacement error and the terminal lateral displacement error; calculate the accuracy evaluation score of the obstacle prediction trajectory according to the evaluation scores and corresponding weight coefficients of each accuracy evaluation sub-dimension, and the amplification coefficient corresponding to the average lateral displacement error and the terminal lateral displacement error.

[0115] It can be understood that the above-mentioned predicted trajectory evaluation device can implement each step of the predicted trajectory evaluation method provided in the above-mentioned embodiment, and the relevant explanations about the predicted trajectory evaluation method are applicable to the predicted trajectory evaluation device, which will not be repeated here.

[0116] Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 4 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0117] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0118] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.

[0119] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a prediction trajectory evaluation device at the logical level. The processor executes the program stored in the memory and is specifically used to perform the following operations:

[0120] Obtain obstacle prediction trajectory output by the autonomous vehicle;

[0121] Performing a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, wherein the multi-dimensional evaluation index includes a physical evaluation dimension, a road traffic evaluation dimension, and an accuracy evaluation dimension;

[0122] A comprehensive evaluation result of the obstacle prediction trajectory is determined according to the multi-dimensional evaluation result of the obstacle prediction trajectory.

[0123] The above application Figure 1The method performed by the prediction trajectory evaluation device disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in the form of software. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as a hardware decoding processor for execution, or a combination of hardware and software modules in the decoding processor for execution. The software module can be located in a storage medium mature in the art, such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0124] The present application also provides a computer program product, which stores one or more programs, wherein the one or more programs include instructions, which, when executed by an electronic device including multiple application programs, enable the electronic device to execute Figure 1 The method performed by the prediction trajectory evaluation device in the illustrated embodiment is specifically used to perform:

[0125] Obtain obstacle prediction trajectory output by the autonomous vehicle;

[0126] Performing a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, wherein the multi-dimensional evaluation index includes a physical evaluation dimension, a road traffic evaluation dimension, and an accuracy evaluation dimension;

[0127] A comprehensive evaluation result of the obstacle prediction trajectory is determined according to the multi-dimensional evaluation result of the obstacle prediction trajectory.

[0128] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0129] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0130] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0131] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0132] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0133] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0134] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0135] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0136] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0137] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A prediction trajectory evaluation method, wherein: The prediction trajectory evaluation method comprises: Obtain obstacle prediction trajectory output by the autonomous vehicle; Performing a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, wherein the multi-dimensional evaluation index includes a physical evaluation dimension, a road traffic evaluation dimension, and an accuracy evaluation dimension; A comprehensive evaluation result of the obstacle prediction trajectory is determined according to the multi-dimensional evaluation result of the obstacle prediction trajectory.

2. The prediction trajectory evaluation method according to claim 1, wherein: The obstacle prediction trajectory includes a plurality of prediction trajectory points, the physical evaluation dimension includes at least one physical evaluation sub-dimension of speed, acceleration and angular velocity, and the multi-dimensional evaluation index is used to perform a multi-dimensional evaluation on the obstacle prediction trajectory to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, including: Calculate the index values ​​of multiple predicted trajectory points in each physical evaluation sub-dimension; Determine the maximum index value of the multiple predicted trajectory points in each physical evaluation sub-dimension according to the index values ​​of the multiple predicted trajectory points in each physical evaluation sub-dimension; Compare the maximum index values ​​of multiple predicted trajectory points in each physical evaluation sub-dimension with the index threshold corresponding to each physical evaluation sub-dimension to obtain the evaluation score of each physical evaluation sub-dimension; The physical evaluation score of the obstacle prediction trajectory is calculated according to the evaluation scores of each physical evaluation sub-dimension and the corresponding weight coefficient.

3. The prediction trajectory evaluation method according to claim 1, wherein: The road traffic assessment dimension includes at least one road traffic assessment sub-dimension of a road edge line, a lane solid line, a speed limit, and a road topology. The multi-dimensional assessment of the obstacle prediction trajectory using a multi-dimensional assessment index to obtain a multi-dimensional assessment result of the obstacle prediction trajectory includes: Obtaining map data of the location of the autonomous driving vehicle, wherein the map data includes road edge line data, lane solid line data, speed limit data, and road connection relationship data; Calculating the evaluation scores of each road traffic evaluation sub-dimension according to the map data and the obstacle prediction trajectory; The road traffic evaluation score of the obstacle prediction trajectory is calculated based on the evaluation scores of each road traffic evaluation sub-dimension and the corresponding weight coefficient.

4. The prediction trajectory evaluation method according to claim 3, wherein: The step of calculating the evaluation scores of each road traffic evaluation sub-dimension according to the map data and the obstacle prediction trajectory includes: Calculating the distances between a plurality of predicted trajectory points and the road edge line and the lane solid line according to the road edge line data and the lane solid line data, wherein the distances include the distance between the first predicted trajectory point and the road edge line and the lane solid line; Determine the minimum distance among the distances between the plurality of predicted trajectory points and the road edge line and the lane solid line; A road edge line evaluation score and a lane solid line evaluation score are calculated based on the minimum distance, the distance between the first predicted trajectory point and the road edge line and the lane solid line, and a distance threshold.

5. The prediction trajectory evaluation method according to claim 3, wherein: The step of calculating the evaluation scores of each road traffic evaluation sub-dimension according to the map data and the obstacle prediction trajectory includes: Calculating a topological structure of a road where an obstacle is located according to the road connection relationship data, wherein the topological structure of the road where the obstacle is located includes a lane centerline; Performing interpolation processing on the lane centerline to obtain an interpolated lane centerline; Calculating a minimum average displacement error between a plurality of predicted trajectory points and the interpolated lane centerline; A road topology evaluation score is calculated based on the minimum average displacement error and the average displacement error threshold.

6. The prediction trajectory evaluation method according to claim 1, wherein: The accuracy assessment dimension includes at least one accuracy assessment sub-dimension of average displacement error and endpoint displacement error. The multi-dimensional assessment of the obstacle prediction trajectory using the multi-dimensional assessment index to obtain the multi-dimensional assessment result of the obstacle prediction trajectory includes: Obtain the actual trajectory points of the obstacle, and transform the predicted trajectory points and actual trajectory points of the obstacle into the obstacle coordinate system; Calculate the evaluation scores of each accuracy evaluation sub-dimension according to the predicted trajectory points and the actual trajectory points in the obstacle coordinate system; The accuracy evaluation score of the obstacle prediction trajectory is calculated based on the evaluation scores of each accuracy evaluation sub-dimension and the corresponding weight coefficients as well as the current driving scenario.

7. The prediction trajectory evaluation method according to claim 6, wherein: The average displacement error includes an average lateral displacement error and an average longitudinal displacement error, the end point displacement error includes an end point lateral displacement error and an end point longitudinal displacement error, and the accuracy evaluation score of the obstacle prediction trajectory calculated according to the evaluation scores of each accuracy evaluation sub-dimension and the corresponding weight coefficient and the current driving scene includes: When the current driving scene is an obstacle lane change scene or a turning scene, setting a magnification factor corresponding to the average lateral displacement error and the end point lateral displacement error; The accuracy evaluation score of the obstacle prediction trajectory is calculated according to the evaluation scores and corresponding weight coefficients of each accuracy evaluation sub-dimension, and the magnification coefficients corresponding to the average lateral displacement error and the end point lateral displacement error.

8. A prediction trajectory evaluation device, wherein: The predicted trajectory evaluation device comprises: An acquisition unit, used for acquiring an obstacle prediction trajectory output by the autonomous driving vehicle; An evaluation unit, configured to perform a multi-dimensional evaluation on the obstacle prediction trajectory using a multi-dimensional evaluation index to obtain a multi-dimensional evaluation result of the obstacle prediction trajectory, wherein the multi-dimensional evaluation index includes a physical evaluation dimension, a road traffic evaluation dimension, and an accuracy evaluation dimension; A determination unit is used to determine a comprehensive evaluation result of the obstacle prediction trajectory according to the multi-dimensional evaluation result of the obstacle prediction trajectory.

9. An electronic device, comprising: processor; and a memory arranged to store computer executable instructions, wherein when the executable instructions are executed, the processor executes the prediction trajectory evaluation method according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program or an instruction, wherein when the computer program or the instruction is executed by a processor, the predicted trajectory evaluation method according to any one of claims 1 to 7 is implemented.

Images