An automated evaluation method and system for the relative accuracy of an autonomous driving positioning system

Abstract

The present invention discloses an automatic evaluation method and system for the relative accuracy of an autonomous driving positioning system. The method includes: taking the current moment as the center, obtaining the curves of the positioning trajectory and its corresponding GPS trajectory in a preset time interval before and after; performing a similarity transformation on the curve of the preset time interval of the obtained positioning trajectory towards the curve of the preset time interval of the obtained GPS trajectory, so that the two ends of the two curves coincide; taking the current moment as the center, intercepting the curves of the preset time interval of the positioning trajectory and the curves of the preset time interval of the GPS trajectory in a preset time interval before and after the two ends coincide; in the intercepted curves of the preset time interval, calculating the distance from the positioning curve to the GPS curve at each moment to obtain the maximum value M. If M is less than the preset distance value, the positioning quality at the current moment is qualified. This method fully automatically evaluates the quality of the positioning result, saves manpower, and the evaluation result is more reliable.

Description

Technical Field

[0001] The present invention belongs to the technical field of automotive autonomous driving, and more specifically, relates to an automated evaluation method and system for the relative accuracy of an autonomous driving positioning system. Background Art

[0002] In autonomous driving, obtaining the position of a vehicle in a lane is one of the very important requirements. To meet this requirement, a common approach is to model the lane lines. This model is also called a map, which stores the positions of the lane lines. During autonomous driving, after calculating the coordinates of the vehicle, by comparing with the positions of the lane lines in the map, it is possible to determine the current position of the vehicle in which lane and at what location.

[0003] One method to achieve this requirement is to use visual positioning. The principle of visual positioning is to match the image at the current moment with the feature points in the map, and then obtain the current position relative to the map. After visual positioning, it is necessary to manually determine the positioning quality, which consumes a large amount of manpower, has strong subjectivity, cannot be quantified, and the evaluation results are not reliable enough.

[0004] The invention with publication number CN108121346A discloses an autonomous driving direction control system and method based on a high-precision navigation positioning system. It includes: a high-precision navigation positioning system for obtaining the rough positioning coordinates of the vehicle from its own positioning data; an autonomous driving control system for receiving the information of the high-precision navigation positioning system, and combining the data collected by the vehicle detection system and the high-precision map of the vehicle to obtain the precise positioning coordinates of the vehicle, the curvature of the curve in the vehicle driving direction, traffic sign information, and the distance data between the vehicle and the curve, and determining the speed and driving direction that the vehicle needs to control according to the current vehicle speed and the obtained data; a vehicle control system for controlling the reduction of vehicle speed and the change of driving direction. This invention can position the vehicle, obtain the control information of the vehicle in advance, control the vehicle control system to take measures for vehicle speed adjustment and driving direction change, but this invention uses GPS positioning and does not evaluate the quality of the positioning, and cannot guarantee the accuracy of the positioning result.

[0005] Compared with the prior art, the present invention segments the positioning trajectory and makes a similarity transformation with the GPS trajectory to obtain a positioning trajectory aligned with the GPS trajectory, and judges the quality of the positioning based on the maximum distance between the two trajectories, fully automatically evaluating the quality of the positioning result, saving manpower, and the evaluation results are more reliable. Summary of the Invention

[0006] In view of the above defects and improvement requirements of the prior art, the present invention provides an automated evaluation method and system for the relative accuracy of an autonomous driving positioning system, aiming to provide an automated evaluation method and system for the relative accuracy of an autonomous driving positioning system with high accuracy, thereby solving the technical problems of consuming a large amount of manpower and unreliable evaluation results existing in the prior art.

[0007] To achieve the above object, the present invention provides an automated evaluation method for the relative accuracy of an autonomous driving positioning system, including the steps of:

[0008] S1. Centered on the current moment, obtain the curves of the positioning trajectory and its corresponding GPS trajectory in the preset time intervals before and after.

[0009] S2. Make a similarity transformation of the curve of the preset time interval of the obtained positioning trajectory to the curve of the preset time interval of the obtained GPS trajectory, so that the two ends of the two curves coincide.

[0010] S3. Centered on the current moment, intercept the curves of the preset time intervals of the positioning trajectory and the GPS trajectory with the two ends coinciding in the preset time intervals before and after.

[0011] S4. In the intercepted curves of the preset time intervals, calculate the distance from the positioning curve to the GPS curve at each moment to obtain the maximum value M. If M is less than the preset distance value, the positioning quality at the current moment is qualified.

[0012] Further, before step S1, it further includes the step of:

[0013] Obtain the positioning trajectory of autonomous driving and its corresponding GPS trajectory within a period of time.

[0014] Further, after step S4, it further includes the step of:

[0015] Judge whether the positioning quality corresponding to all moments in the positioning trajectory is qualified, and calculate the accumulated time of the moments with qualified positioning quality.

[0016] Further, it further includes the step of:

[0017] Calculate the percentage of the accumulated time of the moments with qualified positioning quality in the total time of the positioning trajectory, and this percentage is the quality score of the positioning trajectory.

[0018] Further, it further includes the step of:

[0019] Feed back the quality score of the positioning trajectory.

[0020] Correspondingly, it also provides an automated evaluation system for the relative accuracy of an autonomous driving positioning system, including:

[0021] An acquisition module, configured to acquire, centered on the current moment, a positioning trajectory and curves in a preset time interval before and after the corresponding GPS trajectory;

[0022] A variation module, configured to perform a similarity variation on the curve of the preset time interval of the acquired positioning trajectory towards the curve of the preset time interval of the acquired GPS trajectory;

[0023] An interception module, configured to intercept, centered on the current moment, curves in a preset time interval before and after the curves of the preset time interval of the positioning trajectory and the GPS trajectory with both ends coinciding;

[0024] A calculation module, configured to calculate the distance from the positioning curve to the GPS curve at each moment in the intercepted curves of the preset time interval and obtain the maximum value M;

[0025] An evaluation module, configured to evaluate that the positioning quality at the current moment is qualified when M is less than a preset distance value.

[0026] Further, it further includes:

[0027] An acquisition module, configured to acquire the positioning trajectory of autonomous driving and its corresponding GPS trajectory within a period of time.

[0028] Further, it further includes:

[0029] A judgment module, configured to judge whether the positioning quality corresponding to all moments in the positioning trajectory is qualified;

[0030] A calculation module, configured to calculate the accumulated time of the moments when the positioning quality is qualified.

[0031] Further, it further includes:

[0032] A calculation module, configured to calculate the percentage of the accumulated time of the moments when the positioning quality is qualified in the total time of the positioning trajectory.

[0033] Further, it further includes:

[0034] A feedback module, configured to feedback the quality score of the positioning trajectory.

[0035] Compared with the prior art, the present invention has the following advantages:

[0036] By obtaining the positioning trajectory and the corresponding GPS trajectory within a preset time interval before and after a specific moment, segmenting the positioning trajectory and making a similarity transformation with the GPS trajectory, obtaining the positioning trajectory aligned with the GPS trajectory, judging whether the positioning quality at the specific moment is qualified based on the maximum distance between the two trajectories, and finally scoring the quality of the entire positioning trajectory according to the proportion of the accumulated qualified time to the total time, the quality of the positioning result is evaluated completely automatically, saving manpower and making the evaluation result more reliable. Description of the Drawings

[0037] Figure 1 is a flowchart of an automatic evaluation method for the relative accuracy of an autonomous driving positioning system provided in Embodiment 1;

[0038] Figure 2 is a schematic diagram of an automatic evaluation system for the relative accuracy of an autonomous driving positioning system provided in Embodiment 1;

[0039] Figure 3 is a flowchart of an automatic evaluation method for the relative accuracy of an autonomous driving positioning system provided in Embodiment 2;

[0040] Figure 4 is a schematic diagram of an automatic evaluation system for the relative accuracy of an autonomous driving positioning system provided in Embodiment 2. Detailed Embodiments

[0041] The following are specific embodiments of the present invention in combination with the accompanying drawings, and the technical solutions of the present invention are further described, but the present invention is not limited to these embodiments.

[0042] Embodiment 1

[0043] This embodiment provides an automatic evaluation method for the relative accuracy of an autonomous driving positioning system, as Figure 1 shown, including the steps:

[0044] S11. Taking the current moment as the center, obtain the curves of the positioning trajectory and its corresponding GPS trajectory within a preset time interval before and after;

[0045] S12. Make a similarity transformation of the curve of the preset time interval of the obtained positioning trajectory to the curve of the preset time interval of the obtained GPS trajectory, so that the two ends of the two curves coincide;

[0046] S13. Taking the current moment as the center, intercept the curves of the preset time interval of the positioning trajectory and the GPS trajectory with the two ends coinciding within a preset time interval before and after;

[0047] S14. In the curve of the intercepted preset time interval, calculate the distance from the positioning curve to the GPS curve at each moment to obtain the maximum value M. If M is less than the preset distance value, the positioning quality at the current moment is qualified.

[0048] Further, before step S11, the following step is also included:

[0049] Obtain the positioning trajectory of autonomous driving and its corresponding GPS trajectory within a period of time.

[0050] First, obtain the positioning trajectory within a period of time and its corresponding GPS trajectory, where the coordinate value of the positioning result corresponds to each moment. Then, with the current moment as the center, take out the curves of the preset time interval before and after the positioning trajectory and the GPS trajectory. For example, the curves of the 10s interval before and after the positioning trajectory can be taken out, and at the same time, the curves of the 10s interval before and after the GPS trajectory are taken out. Make a similarity transformation of the taken-out curve of the positioning trajectory to the GPS curve. The similarity transformation includes changes such as translation, rotation, and scaling. The purpose is to make the two ends of the positioning trajectory curve coincide with the GPS curve. After the similarity transformation is completed, still with the current moment as the center, take out the positioning trajectory curve and the GPS curve of the preset time interval before and after. For example, the positioning trajectory curve and the GPS curve of the first and last 2s can be removed, and then calculate the distance from the positioning trajectory curve to the GPS curve at each moment within this preset time interval to obtain the maximum value M. If M is less than the preset distance value, it is determined that the positioning quality at this moment is qualified. The preset distance value should be set according to the ratio of the coordinates of the positioning trajectory to the coordinates of the GPS trajectory, and should not be determined by the visible distance of the trajectory in the figure. For example, if the preset distance value is 2 meters, it should be the difference between the coordinates of the position in the positioning trajectory curve corresponding to the maximum value M and the coordinates of the position in the GPS trajectory curve.

[0051] The reason for first taking out the curves of the 10s interval before and after the positioning trajectory and the curves of the 10s interval before and after the GPS trajectory, and then intercepting the positioning trajectory curve and the GPS curve of the 2s interval after the similarity transformation is completed is that if the curves of the 2s interval are directly intercepted, it is possible that the curves within the 2s interval are very close, which is not convenient for calculation and determination. By first intercepting the curves within a longer time interval to more completely display the trajectory changes, and then intercepting the curves of the 2s interval from the curves of the 10s interval, the curves of the 2s interval will be clearer.

[0052] Further, after step S14, the following step is also included:

[0053] Judge whether the positioning quality corresponding to all moments in the positioning trajectory is qualified, and calculate the accumulated time of the moments when the positioning quality is qualified.

[0054] Further, the following step is also included:

[0055] Calculate the percentage of the accumulated time at the moments when the positioning quality is qualified in the total time of the positioning trajectory, and this percentage is the quality score of the positioning trajectory.

[0056] In the same way, determine the positioning quality at each moment, calculate the accumulated time at the qualified moments, and finally calculate the percentage of the qualified time in the total time of the entire positioning trajectory. This percentage is used as the quality score of the positioning trajectory.

[0057] For example, if the accumulated time at the qualified moments is calculated to be 89 minutes, the total time of the entire positioning trajectory is 100 minutes, and the proportion of the qualified time in the total time is 89%, then the quality score of the positioning trajectory is: 89.

[0058] Correspondingly, an automatic evaluation system for the relative accuracy of an autonomous driving positioning system is also provided, as Figure 2 shown, including:

[0059] An acquisition module 11, configured to obtain curves of a positioning trajectory and its corresponding GPS trajectory in a preset time interval before and after the current moment;

[0060] A change module 12, configured to perform a similarity change on the curve of the preset time interval of the acquired positioning trajectory to the curve of the preset time interval of the acquired GPS trajectory;

[0061] An intercepting module 13, configured to intercept curves of the preset time interval of the positioning trajectory and the curve of the preset time interval of the GPS trajectory that coincide at both ends in a preset time interval before and after the current moment;

[0062] A calculation module 14, configured to calculate the distance from the positioning curve to the GPS curve at each moment in the intercepted preset time interval curve and obtain the maximum value M;

[0063] An evaluation module 15, configured to evaluate that the positioning quality at the current moment is qualified when M is less than a preset distance value.

[0064] Furthermore, it further includes:

[0065] An acquisition module 11, configured to obtain a positioning trajectory of an autonomous driving and its corresponding GPS trajectory within a period of time.

[0066] The acquisition module 11 acquires the positioning trajectory of the autonomous driving and its corresponding GPS trajectory within a period of time, and then, with the current time as the center, acquires the curves in the preset time intervals before and after the positioning trajectory and its corresponding GPS trajectory. The variation module 12 performs a similarity transformation on the acquired positioning curve towards the acquired GPS curve, and then the interception module 13 intercepts the curves in the preset time intervals before and after the positioning curve and the GPS curve from the curves after the similarity transformation. Then, the calculation module 14 calculates the distance from each moment of the positioning curve to the GPS curve in the intercepted preset time interval curves, and obtains the maximum value M. Finally, the evaluation module 15 determines whether M is less than the preset distance value. When M is less than the preset distance value, the positioning quality at the current moment is evaluated as qualified.

[0067] Further, it further includes:

[0068] A judgment module 16, which is used to judge whether the positioning quality corresponding to all moments in the positioning trajectory is qualified;

[0069] A calculation module 14, which is used to calculate the cumulative time of the moments when the positioning quality is qualified.

[0070] Further, it further includes:

[0071] A calculation module 14, which is used to calculate the percentage of the cumulative time of the moments when the positioning quality is qualified in the total time of the positioning trajectory.

[0072] After the evaluation module 15 evaluates the positioning quality corresponding to all moments according to the same method, the judgment module 16 judges whether the positioning quality corresponding to all moments in the positioning trajectory is qualified, and then the calculation module 14 calculates the cumulative time of the moments when the positioning quality is qualified, and finally calculates the percentage of the cumulative time of the moments when the positioning quality is qualified in the total time of the positioning trajectory.

[0073] In this embodiment, by acquiring the positioning trajectory and the corresponding GPS trajectory in the preset time intervals before and after a specific moment, segmenting the positioning trajectory and performing a similarity transformation with the GPS trajectory to obtain a positioning trajectory aligned with the GPS trajectory, judging whether the positioning quality at the specific moment is qualified based on the maximum distance between the two trajectories, and finally scoring the quality of the entire positioning trajectory according to the proportion of the cumulative time of the qualified moments in the total time, the quality of the positioning result is automatically evaluated completely, saving manpower and making the evaluation result more reliable.

[0074] Embodiment 2

[0075] This embodiment provides an automatic evaluation method for the relative accuracy of an autonomous driving positioning system, as Figure 3 shown, including the steps:

[0076] S21. Centering on the current moment, obtain the positioning trajectory and the curves of the preset time intervals before and after the corresponding GPS trajectory;

[0077] S22. Make a similarity transformation of the curve of the preset time interval of the obtained positioning trajectory to the curve of the preset time interval of the obtained GPS trajectory, so that the two ends of the two curves coincide;

[0078] S23. Centering on the current moment, intercept the curves of the preset time intervals before and after the curves of the preset time intervals of the positioning trajectory and the GPS trajectory where the two ends coincide;

[0079] S24. In the intercepted curves of the preset time intervals, calculate the distance from the positioning curve to the GPS curve at each moment to obtain the maximum value M. If M is less than the preset distance value, the positioning quality at the current moment is qualified.

[0080] Further, before step S21, there is also a step:

[0081] Obtain the positioning trajectory of autonomous driving and its corresponding GPS trajectory within a period of time.

[0082] Further, after step S24, there is also a step:

[0083] Judge whether the positioning quality corresponding to all moments in the positioning trajectory is qualified, and calculate the accumulated time of the moments with qualified positioning quality.

[0084] Further, there is also a step:

[0085] Calculate the percentage of the accumulated time of the moments with qualified positioning quality in the total time of the positioning trajectory, and this percentage is the quality score of the positioning trajectory.

[0086] Further, there is also a step:

[0087] Feed back the quality score of the positioning trajectory.

[0088] Different from Embodiment 1, there is also a step:

[0089] Feed back the quality score of the positioning trajectory.

[0090] When the quality score of the positioning trajectory is calculated, the quality score of the positioning trajectory should be fed back so that the user can know.

[0091] The advantage of this embodiment compared with Embodiment 1 is that:

[0092] Feed back the calculated quality score of the positioning trajectory, let the user know, and improve the user experience.

[0093] Correspondingly, an automatic evaluation system for the relative accuracy of an automatic driving positioning system is also provided, such as Figure 4 shown, including:

[0094] An acquisition module 21, configured to obtain curves of a positioning trajectory and its corresponding GPS trajectory within a preset time interval before and after the current moment as the center;

[0095] A change module 22, configured to perform a similarity change on the curve of the preset time interval of the acquired positioning trajectory towards the curve of the preset time interval of the acquired GPS trajectory;

[0096] An interception module 23, configured to intercept curves of a preset time interval of the positioning trajectory with both ends coinciding and curves of a preset time interval of the GPS trajectory within a preset time interval before and after the current moment as the center;

[0097] A calculation module 24, configured to calculate the distance from the positioning curve to the GPS curve at each moment in the intercepted curves of the preset time interval and obtain the maximum value M;

[0098] An evaluation module 25, configured to evaluate that the positioning quality at the current moment is qualified when M is less than a preset distance value.

[0099] Furthermore, it further includes:

[0100] An acquisition module 21, configured to obtain a positioning trajectory of an automatic driving and its corresponding GPS trajectory within a period of time.

[0101] Furthermore, it further includes:

[0102] A judgment module 26, configured to judge whether the positioning quality corresponding to all moments in the positioning trajectory is qualified;

[0103] A calculation module 24, configured to calculate the accumulated time of the moments when the positioning quality is qualified.

[0104] Furthermore, it further includes:

[0105] A calculation module 24, configured to calculate the percentage of the accumulated time of the moments when the positioning quality is qualified in the total time of the positioning trajectory.

[0106] Furthermore, it further includes:

[0107] A feedback module 27, configured to feedback the quality score of the positioning trajectory.

[0108] Different from Embodiment 1, it further includes a feedback module 27.

[0109] The feedback module 27 is configured to feedback the quality score of the positioning trajectory.

[0110] After the calculation module 24 calculates the quality score of the positioning trajectory, the feedback module 27 feeds back the quality score so that the user can know.

[0111] Feeding back the quality score of the calculated positioning trajectory for the user to know improves the user experience.

[0112] The specific embodiments described herein are merely illustrative of the spirit of the present invention. Those skilled in the art to which the present invention pertains can make various modifications or supplements to the described specific embodiments or use similar ways for substitution, but will not deviate from the spirit of the present invention or exceed the scope defined by the appended claims.

Claims

1. An automated evaluation method for the relative accuracy of an autonomous driving positioning system, characterized in that Including the steps: S1. Taking the current moment as the center, obtain the curves of the positioning trajectory and its corresponding GPS trajectory within a preset time interval before and after; S2. Make a similarity transformation of the curve of the preset time interval of the obtained positioning trajectory to the curve of the preset time interval of the obtained GPS trajectory, so that the two ends of the two curves coincide; S3. Taking the current moment as the center, intercept the curves of the preset time interval of the positioning trajectory and the preset time interval of the GPS trajectory with the two ends coinciding within a preset time interval before and after; S4. In the intercepted curves of the preset time interval, calculate the distance from the positioning curve to the GPS curve at each moment to obtain the maximum value M. If M is less than the preset distance value, the positioning quality at the current moment is qualified; Before step S1, it also includes the step: Obtain the positioning trajectory of autonomous driving and its corresponding GPS trajectory within a period of time.

2. The automated evaluation method for the relative accuracy of an autonomous driving positioning system according to claim 1, characterized in that After step S4, it also includes the step: Judge whether the positioning quality corresponding to all moments in the positioning trajectory is qualified, and calculate the cumulative time of the moments with qualified positioning quality.

3. The automated evaluation method for the relative accuracy of an autonomous driving positioning system according to claim 2, wherein It also includes the step: Calculate the percentage of the cumulative time of the moments with qualified positioning quality in the total time of the positioning trajectory, and this percentage is the quality score of the positioning trajectory.

4. The automated evaluation method for the relative accuracy of an autonomous driving positioning system according to claim 3, characterized in that It also includes the step: Feedback the quality score of the positioning trajectory.

5. An automated evaluation system for the relative accuracy of an autonomous driving positioning system, characterized in that, Including: An acquisition module, used to take the current moment as the center and obtain the curves of the positioning trajectory and its corresponding GPS trajectory within a preset time interval before and after; A transformation module, used to make a similarity transformation of the curve of the preset time interval of the obtained positioning trajectory to the curve of the preset time interval of the obtained GPS trajectory, so that the two ends of the two curves coincide; An interception module, used to take the current moment as the center and intercept the curves of the preset time interval of the positioning trajectory and the preset time interval of the GPS trajectory with the two ends coinciding within a preset time interval before and after; A calculation module, used to calculate the distance from the positioning curve to the GPS curve at each moment in the intercepted curves of the preset time interval and obtain the maximum value M; An evaluation module, used to evaluate that the positioning quality at the current moment is qualified when M is less than the preset distance value; An acquisition module, used to obtain the positioning trajectory of autonomous driving and its corresponding GPS trajectory within a period of time.

6. An automated evaluation system for the relative accuracy of an autonomous driving positioning system according to claim 5, characterized in that, It also includes: A judgment module, used to judge whether the positioning quality corresponding to all moments in the positioning trajectory is qualified; A calculation module, used to calculate the cumulative time of the moments with qualified positioning quality.

7. An automated evaluation system for the relative accuracy of an autonomous driving positioning system according to claim 6, characterized in that, It also includes: A calculation module, used to calculate the percentage of the cumulative time of the moments with qualified positioning quality in the total time of the positioning trajectory.

8. An automated evaluation system for the relative accuracy of an autonomous driving positioning system according to claim 7, characterized in that, It also includes: A feedback module, used to feedback the quality score of the positioning trajectory.

Images