Data evaluation method and device and electronic equipment

Abstract

The invention discloses a data evaluation method and device and electronic equipment, and the method comprises the steps: obtaining a first-class data point set and a second-class data point set, and when the initial recording time of each second-class data point in the second-class data point set is adjusted to the respective corresponding second recording time, determining the first-class data point set and the second-class data point set; establishing association relationships between the first type of data points and the second type of data points according to a preset alignment rule, determining position errors and attitude errors corresponding to the first type of data points and the second type of data points in each association relationship, and processing the position errors and the attitude errors corresponding to each association relationship according to a preset statistical rule, and generating an evaluation result corresponding to the to-be-evaluated data. According to the method, the second type of data point set output by the to-be-measured positioning equipment and the first type of data point set output by the truth value equipment are aligned and the association relationship is established, so that both the position error and the attitude error are absolute errors, and the accuracy and rapidity of obtaining an evaluation result are improved.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent vehicle positioning, and in particular to a data evaluation method, device and electronic equipment. Background Art

[0002] High-precision positioning systems are increasingly used in smart vehicles. High-precision positioning systems can determine the location information and posture information of smart vehicles. The positioning information of high-precision positioning systems is closely related to the safety of smart vehicle driving. When the positioning information of high-precision positioning systems is inaccurate, the safety of smart vehicles during driving will face greater risks. Therefore, an evaluation method for the positioning information of high-precision positioning systems is introduced. The specific evaluation process is as follows:

[0003] The true value data corresponding to the true value device and the positioning data to be evaluated corresponding to the device to be measured are obtained, where the true value device is a device that can provide accurate positioning data, and the true value device and the device to be measured are installed on the same carrier platform; the true value data and the positioning data to be evaluated are calculated and processed by a positioning accuracy evaluation device to obtain the accuracy of the device under test, and the accuracy of the device to be measured is evaluated by an evaluation system.

[0004] The accuracy of the above-mentioned positioning equipment to be measured includes horizontal accuracy and elevation accuracy. The horizontal accuracy represents the accuracy of the ground points in the generated digital elevation model on the horizontal plane, and the elevation accuracy represents the accuracy of the ground points in the generated digital elevation model in the vertical direction, indicating that the positioning data to be evaluated is calculated through longitude and latitude. Therefore, the error between the positioning data to be evaluated and the true data calculated through longitude and latitude can only represent the error in longitude / latitude / altitude, and cannot show the longitudinal and lateral errors of the intelligent vehicle.

[0005] In addition, the true value data collected by the true value device and the positioning data to be evaluated collected by the positioning device to be measured are obtained by multiple reciprocating collections on the same road section. After obtaining the positioning data collected multiple times, the positioning data will be filtered to filter out part of the positioning data, so that the filtered data is non-continuous positioning data. The error determined based on the non-continuous positioning data can only represent the overall error of some sections. There is a difference between the true error corresponding to a specific positioning data in some sections and the determined overall error, which leads to low accuracy of the evaluation system in evaluating the accuracy of the positioning device to be measured.

[0006] In summary, how to accurately and quickly evaluate the positioning information of high-precision systems has become a problem that needs to be solved at present. Summary of the invention

[0007] The present application provides a data evaluation method, device and electronic device for improving the accuracy of evaluation results determined by an evaluation system.

[0008] In a first aspect, the present application provides a data evaluation method, the method comprising:

[0009] Acquire a first type of data point set and a second type of data point set, wherein the first type of data point set includes the position, posture and first recording time of each first type of data point, and the second type of data point set includes the position, posture and initial recording time of each second type of data point;

[0010] When the initial recording time of each second-category data point in the second-category data point set is adjusted to the respective corresponding second recording time, an association relationship between the first-category data point and the second-category data point is established according to a preset alignment rule;

[0011] Determine the position error and attitude error corresponding to the first type of data point and the second type of data point in each association relationship;

[0012] The position error and the posture error corresponding to each association relationship are processed according to preset statistical rules to generate an evaluation result corresponding to the data to be evaluated.

[0013] Through the above method, an association relationship between the first type of data points and the second type of data points is established, so that the first type of data points and the second type of data points are time-aligned, so that the evaluation result corresponding to each association relationship can be determined, ensuring the accuracy of the determined evaluation result.

[0014] In a possible design, before obtaining the first type of data point set and the second type of data point set, the following is further included:

[0015] Parsing out initial positioning coordinates corresponding to each second type of data point in the second type of data point set, wherein the initial positioning coordinates are coordinates of the second carrier in a preset coordinate system;

[0016] Parse the first positioning coordinates corresponding to each first-category data point in the first-category data point set, and determine a preset coordinate conversion matrix between all first positioning coordinates corresponding to the first-category data point set and all initial positioning coordinates corresponding to the second-category data point set, wherein the first positioning coordinates are the coordinates of the first carrier in the preset coordinate system, and the preset coordinate conversion matrix is ​​a rigid body transformation matrix from the second carrier coordinate system to the first carrier coordinate system;

[0017] The second carrier coordinate system is aligned to the first carrier coordinate system based on the preset coordinate transformation matrix, and each initial positioning coordinate after the alignment is referred to as the second positioning coordinate of the corresponding second type data point.

[0018] Through the above method, the first type of data points and the second type of data points are converted into representations under the same coordinate system, and the second positioning coordinates of the second type of data points are obtained, so that the first type of data points and the second type of data points have the same evaluation criteria, thereby being able to compare the first type of data points with the second type of data points.

[0019] In a possible design, the initial recording time of each second-category data point in the second-category data point set is adjusted to the second recording time corresponding to the second category, including:

[0020] Parsing out the first recording time corresponding to each first-category data point, and parsing out the initial recording time corresponding to each second-category data point of a preset number;

[0021] Determine a first preset time search range corresponding to the initial recording time of each second-category data point, and calculate a first interval distance and a first heading angle difference between each second-category data point and each first-category data point within the corresponding first preset time search range;

[0022] Filter out first-category data points corresponding to the minimum first interval distance in the first preset distance threshold and the first heading angle difference value being lower than the first preset heading angle threshold from all the first interval distances, and establish a pairing relationship between the first-category data points and the second-category data points;

[0023] Substitute the first recording time corresponding to each first-category data point and the initial recording time corresponding to each second-category data point in all paired relationships into a preset time compensation formula to calculate the time compensation value corresponding to the second-category data point set;

[0024] The initial recording time corresponding to each second type of data point is adjusted according to the time compensation value to obtain the second recording time corresponding to each second type of data point.

[0025] By the above method, the time compensation value of the second type of data point is calculated, and the initial recording time of the second type of data point is adjusted based on the time compensation value, so that the first type of data point is time-aligned with the second data point.

[0026] In a possible design, calculating the first interval distance and the first heading angle difference between each second-category data point and all first-category data points within the corresponding first preset time search range includes:

[0027] Parse out the second heading angle and the second coordinate corresponding to each second-category data point, and parse out the first heading angle and the first coordinate corresponding to each first-category data point within each first preset time search range;

[0028] For each second-category data point, perform the following operations:

[0029] Calculate the first interval distance between the second coordinate of a single second-category data point and each first coordinate within the corresponding first preset time search range; and

[0030] The first heading angle differences between the second heading angle of a single second-category data point and each first heading angle within the corresponding first preset time search range are calculated.

[0031] Through the above method, different ways are used to calculate the first interval distance and the first heading angle difference between the first type of data points and the second type of data points, which is conducive to calculating the time compensation value between the first type of data points and the second type of data points, and is conducive to faster calculation of the position error and attitude error between the first type of data points and the second type of data points.

[0032] In a possible design, establishing an association relationship between the first type of data points and the second type of data points according to a preset alignment rule includes:

[0033] For each second-category data point, perform the following operations:

[0034] Parsing out the first coordinates and the first heading angle corresponding to each first-category data point within the second preset time search range, and parsing out the second coordinates and the second heading angle corresponding to a single second-category data point;

[0035] Calculating second interval distances between the second coordinates and each first coordinate; and

[0036] Calculating first heading angle differences between the second heading angle and each first heading angle;

[0037] In the second preset time search range, first-category data points having a second interval distance lower than the minimum second interval distance in the second preset distance threshold and a second heading angle difference lower than the second preset heading angle threshold are screened out, and an association relationship between the first-category data points and the second-category data points is established.

[0038] Through the above method, the association relationship between the first category data points and the second category data points is determined, so that each second category data point has an associated first category data point, so that the difference between the first category data points and the second category data points can be compared based on the first category data points as a benchmark.

[0039] In a possible design, determining the position error and posture error corresponding to the first type of data point and the second type of data point in each association relationship includes:

[0040] Parsing out a first position vector and a first rotation matrix of a first type of data point in each association relationship, and parsing out a second position vector and a second rotation matrix of a second type of data point in each association relationship;

[0041] The method for determining the position error and attitude error corresponding to each association relationship is as follows:

[0042] Substitute the first rotation matrix, the first position vector and the second position vector into a preset position error formula to calculate the position error corresponding to a single second-category data point; and calculate the difference matrix between the first rotation matrix and the second rotation matrix, and use the difference matrix as the posture error corresponding to a single second-category data point.

[0043] By using the above method, the position error and posture error corresponding to each association relationship are calculated, which is conducive to evaluating the difference between the first type of data points and the second type of data points.

[0044] In a possible design, after processing the position error and the posture error corresponding to each association relationship according to a preset statistical rule, the method further includes:

[0045] Determining a first target score corresponding to the position error according to a correspondence between a preset position error range and a first preset score; and

[0046] A second target score corresponding to the posture error is determined according to a correspondence between a preset posture error range and a second preset score.

[0047] By using the above method, a first target score for position error is determined and a second target score for posture error is determined, so that the evaluation results determined by the evaluation system are more diversified.

[0048] In a second aspect, the present application provides a data evaluation device, the device comprising:

[0049] An acquisition module, used for acquiring a first type of data point set and a second type of data point set;

[0050] an association module, configured to establish an association relationship between the first-category data points and the second-category data points according to a preset alignment rule when the initial recording time of each second-category data point in the second-category data point set is adjusted to the respective corresponding second recording time;

[0051] An error module, used to determine the position error and attitude error corresponding to the first type of data point and the second type of data point in each association relationship;

[0052] The evaluation module is used to process the position error and the posture error corresponding to each association relationship according to preset statistical rules to generate an evaluation result corresponding to the data to be evaluated.

[0053] In one possible design, the acquisition module is specifically used to parse out the initial positioning coordinates corresponding to each second-category data point in the second-category data point set, parse out the first positioning coordinates corresponding to each first-category data point in the first-category data point set, determine a preset coordinate transformation matrix between all first positioning coordinates corresponding to the first-category data point set and all initial positioning coordinates corresponding to the second-category data point set, align the second carrier coordinate system to the first carrier coordinate system based on the preset coordinate transformation matrix, and call each aligned initial positioning coordinate the second positioning coordinate of the corresponding second-category data point.

[0054] In a possible design, the association module is specifically used to parse out the first recording time corresponding to each first-category data point, and parse out the initial recording time corresponding to each preset number of second-category data points, determine the first preset time search range corresponding to the initial recording time of each second-category data point, calculate the first interval distance and the first heading angle difference between each second-category data point and each first-category data point in the corresponding first preset time search range, filter out the first-category data points corresponding to the minimum first interval distance in the first preset distance threshold and the first heading angle difference below the first preset heading angle threshold from all the first interval distances, establish a pairing relationship between the first-category data points and the second-category data points, respectively bring the first recording time corresponding to each first-category data point and the initial recording time corresponding to each second-category data point in all the pairing relationships into the preset time compensation formula, calculate the time compensation value corresponding to the second-category data point set, adjust the initial recording time corresponding to each second-category data point according to the time compensation value, and obtain the second recording time corresponding to each second-category data point.

[0055] In one possible design, the association module is also used to parse out the second heading angle and the second coordinate corresponding to each second-category data point, and to parse out the first heading angle and the first coordinate corresponding to each first-category data point within each first preset time search range. For each second-category data point, the following operations are performed respectively: calculating the first interval distance between the second coordinate of a single second-category data point and each first coordinate within the corresponding first preset time search range, and calculating the first heading angle difference between the second heading angle of a single second-category data point and each first heading angle within the corresponding first preset time search range.

[0056] In one possible design, the association module is also used to perform the following operations for each second-category data point: parse out the first coordinates and the first heading angle corresponding to each first-category data point within the second preset time search range corresponding to the second recording time of each second-category data point, and parse out the second coordinates and the second heading angle corresponding to a single second-category data point, calculate the second interval distance between the second coordinates and each first coordinate, and calculate the second heading angle difference between the second heading angle and each first heading angle, screen out the first-category data points whose minimum second interval distance is lower than the second preset distance threshold and whose second heading angle difference is lower than the second preset heading angle threshold within the second preset time search range, and establish an association relationship between the first-category data points and the second-category data points.

[0057] In one possible design, the error module is specifically used to parse the first position vector and the first rotation matrix of the first type of data point in each association relationship, and to parse the second position vector and the second rotation matrix of the second type of data point in each association relationship. The method for determining the position error and posture error corresponding to each association relationship is as follows: Substitute the first rotation matrix, the first position vector and the second position vector into a preset position error formula, calculate the position error corresponding to a single second type of data point, and calculate the difference matrix between the first rotation matrix and the second rotation matrix, and use the difference matrix as the posture error corresponding to a single second type of data point.

[0058] In one possible design, the evaluation module is also used to determine the first target score corresponding to the position error according to the correspondence between the preset position error range and the first preset score, and to determine the second target score corresponding to the posture error according to the correspondence between the preset posture error range and the second preset score.

[0059] In a third aspect, the present application provides an electronic device, including:

[0060] Memory, used to store computer programs;

[0061] The processor is used to implement the above-mentioned data evaluation method steps when executing the computer program stored in the memory.

[0062] In a fourth aspect, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned data evaluation method are implemented.

[0063] For each aspect from the first to the fourth aspect and the technical effects that may be achieved by each aspect, please refer to the above description of the technical effects that can be achieved by the first aspect or various possible schemes in the first aspect, and no further details will be given here. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A flowchart of the steps of a data evaluation method provided for this application;

[0065] Figure 2 A schematic diagram of the position error provided by the present application plotted on a target shooting diagram;

[0066] Figure 3 A schematic diagram of the structure of a data evaluation device provided in this application;

[0067] Figure 4 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0068] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The specific operating methods in the method embodiments can also be applied to device embodiments or system embodiments. It should be noted that in the description of the present application, "multiple" is understood as "at least two". "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A is connected to B, which can represent: A is directly connected to B and A is connected to B through C. In addition, in the description of the present application, words such as "first" and "second" are only used to distinguish the purpose of description, and cannot be understood as indicating or implying relative importance, nor can they be understood as indicating or implying order.

[0069] In previous technologies, when evaluating the accuracy of the device under test, the accuracy of the device under test includes horizontal accuracy and elevation accuracy, indicating that the positioning data to be evaluated is calculated through longitude and latitude. Therefore, the error between the positioning data to be evaluated and the true data calculated through longitude and latitude can only represent the error in longitude / latitude / altitude, and cannot show the longitudinal and lateral errors of the intelligent vehicle. In addition, the determined true data and the positioning data to be evaluated are non-overall positioning data, which will cause the error determined by the evaluation system to be the overall error of some sections. There is a difference between the real error corresponding to a specific positioning data in some sections and the determined overall error, which leads to low accuracy of the evaluation system in evaluating the accuracy of the positioning device to be measured. Therefore, how to accurately and quickly evaluate the positioning information of a high-precision system has become a current problem to be solved.

[0070] In order to solve the above-described problems, the embodiments of the present application provide a data evaluation method for realizing accurate evaluation of positioning information of a high-precision system. The method and device described in the embodiments of the present application are based on the same technical concept. Since the principles of the problems solved by the method and the device are similar, the embodiments of the device and the method can refer to each other, and the repeated parts will not be repeated.

[0071] The embodiments of the present application are described in detail below in conjunction with the accompanying drawings.

[0072] Reference Figure 1 , the present application provides a data evaluation method, which can improve the evaluation accuracy of positioning information of a high-precision system. The implementation process of the method is as follows:

[0073] Step S1: Obtain a first type of data point set and a second type of data point set.

[0074] In order to accurately evaluate the positioning information of the high-precision system, the evaluation system needs to obtain all the data output by the true value device and the device to be measured, so that the true value device and the device to be measured cover all continuous data of the entire road section for analysis. The first type of data point set and the second type of data point set are not partial sections or discrete sampling points, so that the evaluation system inputs complete and actual road sampling data, ensuring the reliability of the evaluation results.

[0075] Furthermore, all data output by the true value device are taken as a first-category data point set, which includes the position, posture and first recording time of each first-category data point; all data output by the device to be measured are taken as a second-category data point set, which includes the position, posture and initial recording time of each second-category data point; the true value device and the device to be measured are both fixed on the intelligent vehicle; the frequency of the true value device outputting the first-category data points is higher than the frequency of the device to be measured outputting the second-category data points; and the rigid body change between the true value device and the device to be measured has been calibrated.

[0076] In order to use the first type of data point set collected by the true value device as an evaluation benchmark, it is necessary to obtain the first type of data point set output by the true value device and the second type of data point set output by the positioning device to be measured, and parse out the initial positioning coordinates corresponding to each second type of data point in the second type of data point set. The first type of data point set represents the position and posture of the center of mass of the true value device in the initial coordinate system, and the second type of data point set represents the position and posture of the center of mass of the device to be evaluated in the initial coordinate system. The initial coordinate system can be the World Geodetic System (WGS84). In order to realize the coordinate transformation of the first type of data points and the second type of data points, it is necessary to perform coordinate transformation on the first type of data points and the second type of data points respectively, so that the first type of data points are transformed from the initial coordinate system to the preset coordinate system, and the second type of data points are transformed from the initial coordinate system to the preset coordinate system. The preset coordinate system can be the station center coordinate system (local Cartesian coordinates coordinate system, ENU).

[0077] After the coordinate transformation of the first type of data points and the second type of data points is completed, due to the deviation between the center of mass of the true value device and the center of mass of the device to be positioned, in order to align the coordinates of the first type of data points with the second type of data points, it is necessary to determine the preset coordinate transformation matrix between the first type of data points and the second type of data points. The preset coordinate transformation matrix is ​​the rigid body transformation matrix from the second carrier coordinate system to the first carrier coordinate system. The first carrier can be the true value device, and the second carrier can be the device to be positioned. Since the first carrier and the second carrier are fixedly installed on the intelligent vehicle, the preset coordinate transformation matrix between the first carrier and the second carrier is constant.

[0078] In addition, it is also necessary to determine the initial positioning coordinates of the second type of data points, which are the coordinates of the second carrier in the preset coordinate system, and determine the first positioning coordinates of the first type of data points, which are the coordinates of the first carrier in the preset coordinate system. In order to enable the first type of data points and the second type of data points to be converted into representations in the same coordinate system, it is necessary to align the second carrier coordinate system to the first carrier coordinate system according to the preset coordinate conversion matrix, convert the initial positioning coordinate representation of the second carrier of the second type of data points in the preset coordinate system into the first positioning coordinate representation of the first carrier in the preset coordinate system, and call each aligned initial positioning coordinate the second positioning coordinate of the corresponding second type of data point, so that the first type of data points and the second type of data points can be evaluated using the same evaluation criteria. The coordinate alignment formula for aligning the second carrier coordinate system to the first carrier coordinate system is as follows:

[0079] T 1 =T 2 ·T C

[0080] In the above formula 1, T 1 Represents the pose matrix of the second positioning coordinates of the second type of data points after transformation, T 2 The pose matrix representing the initial positioning coordinates of the second type of data points, T C Represents a preset coordinate transformation matrix between the first carrier coordinate system and the second carrier coordinate system.

[0081] Through the above method, the positioning poses of the first type of data points and the second type of data points have the same physical meaning, which is convenient for direct comparison.

[0082] Step S2: when the initial recording time of each second type data point in the second type data point set is adjusted to the corresponding second recording time, an association relationship between the first type data point and the second type data point is established according to a preset alignment rule.

[0083] After obtaining the first type of data point set and the second type of data point set, due to the different clock sources of the true value device and the device to be measured, the recording time of the true value device and the device to be measured will be inconsistent when they pass the same position. When the recording time of the true value device is significantly different from that of the device to be measured, the accuracy of the evaluation result will be reduced. Therefore, it is necessary to align the first recording time of the first type of data point with the initial recording time of the second type of data point. The specific process of time alignment is as follows:

[0084] The first recording time corresponding to each first-category data point and the initial recording time corresponding to each second-category data point are parsed to determine a preset number of second-category data points among all second-category data points. Each second-category data point in the preset number needs to determine a matching first-category data point. Since the process of matching each second-category data point in the preset number with the corresponding first-category data point is the same, only one second-category data point is matched with the corresponding first-category data point for explanation here.

[0085] Determine the initial recording time corresponding to the second type of data point, and determine the first preset time search range corresponding to the initial recording time. For example, the initial recording time of the second type of data point is 13:46:12 on September 3, 2023, and the first preset time search range is within 10 seconds of the initial recording time, that is, the first preset time search range is [13:46:02 on September 3, 2023, 13:46:22 on September 3, 2023]. Then determine each first type of data point within the first preset time search range, and parse the first preset time search range. The first heading angle and the first coordinate of each first-category data point within the time search range, and the second heading angle and the second coordinate corresponding to the second-category data point are determined, the first interval distance and the first heading angle difference between each first-category data point and the second-category data point within the first preset time search range are calculated, and the first-category data points with the minimum first interval distance below the first preset distance threshold and the first heading angle difference below the first preset heading angle threshold are screened out from all the first interval distances, and a pairing relationship between the screened first-category data points and the second-category data points is established.

[0086] According to the method described above, the first-category data points that match each second-category data point in the preset number are determined to obtain a preset number of pairing relationships, and then the first recording time of the first-category data point and the initial recording time of the second-category data point in each pairing relationship are respectively brought into the preset time compensation formula. The preset time compensation formula is as follows:

[0087]

[0088] In the above preset time compensation formula, N is the number of matching pairs, t 1i represents the i-th first record time corresponding to the first type of data point, t 2i It represents the i-th initial recording time corresponding to the second type of data point, and △T is the time compensation value.

[0089] The time compensation value corresponding to the second type of data point set can be calculated through the above preset time compensation formula, and then the initial recording time corresponding to each second type of data point is adjusted according to the time compensation value to obtain the second recording time corresponding to each second type of data point. The second recording time is obtained by the following formula:

[0090] t' 2i =t 2i +△T

[0091] In the above formula, t' 2i Represents the second recording time of the second type of data point after the time compensation value is compensated, t 2i represents the initial recording time of the second type of data points, and △T represents the time compensation value.

[0092] After determining the second recording time corresponding to the second type of data point, it is still impossible to ensure that the first recording time of the first type of time point is exactly the same as the second recording time of the second type of time point when passing the same position, so as to avoid the problem of introducing large errors when establishing an association relationship based on timestamps by using the interpolation and extrapolation method. In this embodiment of the present application, each association relationship is determined by using the principle of the closest distance. The specific association process is as follows:

[0093] Since the true value device and the device to be measured collect a large amount of actual data, the number of first-category data points and second-category data points will also be large. It takes a lot of time and computing resources to find the nearest neighbor points in a large amount of data. Therefore, in order to improve the efficiency of the association process, it is necessary to determine all first-category data points within the second preset time search range. Due to the time orderliness of the first-category data points, when traversing the first-category data points, after the first-category data point is found, the subsequent first-category data points are traversed in turn, thereby improving the matching efficiency between the first-category data points and the second-category data points.

[0094] Since the process of determining the first-class data point corresponding to each second-class data point is the same, only the matching process of the first-class data point corresponding to one second-class data point is described. The process of matching other second-class data points to first-class data points is described as follows:

[0095] The evaluation system needs to obtain the first coordinates and the first heading angle corresponding to each first-category data point within the second preset time search range of the second-category data point, as well as the second coordinates and the second heading angle of the second-category data point, calculate the second interval distance between each first coordinate and the second coordinate within the second preset time search range, and calculate the difference between each first heading angle and the second heading angle within the second preset time search range, and then screen out the first-category data points within the second preset time search range whose minimum second interval distance is lower than the second preset distance threshold and whose second heading angle difference is lower than the second preset heading angle threshold. The first preset distance threshold and the second preset distance threshold can be set according to actual conditions, and the first preset heading angle threshold and the second preset heading angle threshold can be adjusted according to actual conditions, to establish an association relationship between the first-category data point and the second-category data point, and all association relationships can be expressed as {p 2i , p 1i}, i=1,2,3,......N,p 2i represents the second type of data points, p 1i The first type of data points that represent the associations of the second type of data points.

[0096] Through the above method, after the initial recording time of the second type of data points is adjusted by the time compensation value, they are aligned in the time dimension and the distance dimension, and the association relationship between the first type of data points and the second type of data points is determined, so that the first recording time of the first type of data points and the second recording time of the second type of data points are accurately aligned, ensuring the accuracy of the evaluation results determined by the evaluation system.

[0097] Step S3: Determine the position error and attitude error corresponding to the first type of data point and the second type of data point in each association relationship

[0098] After determining the association relationship between the first type of data points and the second type of data points, in order to determine the error between the first type of data points and the second type of data points, the first position vector and the first rotation matrix of the first type of data points in each association relationship are parsed, and the second position vector and the second rotation matrix of the second type of data points in each association relationship are parsed. The first position vector and the second position vector both represent the position of the intelligent vehicle, the first rotation matrix can be converted into a first attitude angle, and the second rotation matrix can also be converted into a second attitude angle. The first attitude angle and the second attitude angle both describe the direction and attitude of the intelligent vehicle moving in space, and the first attitude angle and the second attitude angle can both represent heading angle, pitch angle, roll angle, etc.

[0099] There are position errors and attitude errors between the first type of data points and the second type of data points. The position error and attitude error can be obtained through the position error formula and the attitude error formula. The position error calculates the position error under the intelligent vehicle carrier. Compared with the position error in the world coordinate system, it can more intuitively reflect the longitudinal and lateral position errors of the vehicle. The position error vector formula is as follows:

[0100]

[0101] In the above position error formula, represents the position error vector, The second position vector representing the second type of data point, Represents the first position vector of the first type of data point, R -1 Represents the inverse of the first rotation matrix.

[0102] Assume that the x-axis of the intelligent vehicle carrier coordinate system points forward, the y-axis points to the left, and the z-axis points upward.

[0103] Position error dx of the intelligent vehicle in the X-axis direction i , that is, the longitudinal position error is directly taken as The x-component of the intelligent vehicle, the position error dy in the Y-axis direction i , that is, the lateral position error is directly taken as The y component of the intelligent vehicle’s position error dz in the Z-axis direction i , that is, the height position error is directly taken as The z component of:

[0104]

[0105] If the position error in all directions is considered, the calculation formula is as follows:

[0106]

[0107] In the above formula for calculating the position error in all directions, dist i Represents the position error in all directions, dx i Represents the longitudinal position error in the X-axis direction, dy i Represents the lateral position error in the Y-axis direction, dz i Represents the height position error in the Z-axis direction.

[0108] If the position error in the plane direction is considered, the calculation formula is as follows:

[0109]

[0110] In the above formula for calculating the position error in the plane direction, dist i Represents the position error in the plane direction, dx i Represents the longitudinal position error in the X-axis direction, dy i Represents the lateral position error in the Y-axis direction.

[0111] If the position error of the intelligent vehicle in the lateral direction is considered, the calculation formula is as follows:

[0112] dist i =|dy i |

[0113] In the above formula for calculating the position error in the lateral direction, dist i Represents the position error in the lateral direction, dy i Represents the position error in the lateral direction.

[0114] When calculating the attitude error, you can select the attitude angle error in the roll angle, pitch angle, and yaw angle directions as needed. The more common method is to consider the yaw angle error. The formula for calculating the heading angle error is as follows:

[0115] △w=|w 2i -w 1i |

[0116] In the above formula for calculating the heading angle error, △w represents the attitude error between the first type of data point and the second type of data point, w2i The second yaw angle representing the second type of data points, w 1i Represents the first yaw angle of the first type of data points.

[0117] When the attitude error takes into account the roll angle and pitch angle, you can refer to the above formula for the yaw angle, which will not be explained in detail here.

[0118] Based on the above method, the position error and attitude error between the first type of data points and the second type of data points are calculated. The position error and attitude error are both absolute errors, thus ensuring the accuracy of the position error and attitude error.

[0119] Step S4: Processing the position error and the posture error corresponding to each association relationship according to preset statistical rules to generate an evaluation result corresponding to the data to be evaluated.

[0120] After determining the position error and posture error corresponding to the first type of data point and the second type of data point in each association relationship, the position error and posture error corresponding to each association relationship are processed according to preset statistical rules to generate an evaluation result of the evaluation system.

[0121] In addition, after obtaining the position error and attitude error, in order to reflect the diversity of the evaluation results obtained by the evaluation system, the preset statistical rules include a variety of statistical methods for statistical position errors and / or attitude errors. The specific statistical methods are as follows:

[0122] The first target score corresponding to the position error is determined according to the correspondence between the preset position error range and the first preset score. The correspondence between the preset position error range and the first preset score is shown in Table 1 below:

[0123] Preset position error range First preset score [0，0.1m] 100 (0.1m, 0.2m] 90 (0.2m, 0.3m] 80 (0.3m, 0.4m] 70 (0.4m, 0.5m] 60 (0.5m, 0.6m] 50 (0.6m, 0.7m] 40 (0.7m, 0.8m] 30 (0.8m, 0.9m] 20 (0.9m, 1m] 10 (1m,∞m] 0

[0124] Table 1

[0125] In the above Table 1, the correspondence between the preset position error range and the first preset score is recorded. After the position error is determined through Table 1, the first preset score corresponding to the position error is determined. The scoring rules in the above Table 1 can be adjusted according to actual needs, and no further explanation is given here.

[0126] Through the above Table 1, the first preset score corresponding to the position error can be directly found, and the corresponding first preset score is used as the first target score.

[0127] The correspondence between the preset posture error and the second preset score can also be scored by designing corresponding scoring rules with reference to the distance error. After the posture error is determined, the second target score of the posture error can be determined.

[0128] It should be noted that, in order to display the evaluation results intuitively, the preset rule statistics method can set the presentation method to a scoring table, so that the evaluation results can be displayed in the form of a scoring table, as shown in Table 2: When the evaluation results only involve position errors, scoring rules can be formulated for position errors. The closer the position, the higher the score. According to the specific scoring rules, each position error item will be assigned a score. The number of position error items falling within each score range is counted, and its percentage in the total number of position error items is calculated.

[0129] First target score Position error range Points percentage 100 [0, 10cm] 2133242 79.6％ 90 (10cm, 20cm] 1738465 2.9％ 80 (20cm, 30cm] 78959 1.8％ 70 (30cm, 40cm] 49739 1.64％ …… …… …… ……

[0130] Table 2

[0131] In the above Table 2, the position error range corresponding to the first target score, the number of position error items falling within each score segment, and the percentage of the number of position error items within each score segment to the total number of position error items are recorded. The above Table 2 is a scoring table corresponding to the known mileage and mileage recording time of the intelligent vehicle. The above Table 2 is only used as an example. The scoring table corresponding to other mileage and mileage recording time of the intelligent vehicle refers to the example in Table 2, and no further explanation is given here.

[0132] In addition, the preset rule statistics method can also set the presentation mode to a target shooting diagram, and plot all position errors on the target shooting diagram. The schematic diagram of position errors plotted on the target shooting diagram is as follows: Figure 2 As shown, in Figure 2 In the figure, each black dot represents a position error. The center of the target image represents the origin of the first coordinate system of the true value device. The dx axis represents the longitudinal position error dx of the intelligent vehicle. i The dy axis represents the lateral position error of the intelligent vehicle. A ring on the target chart represents a distance of 0.1m. The closer to the center of the target, the smaller the position error. The target chart can intuitively display the distribution of the position error in the spatial position of the vehicle coordinate system, and show the distribution of the lateral position error and longitudinal position error of the intelligent vehicle.

[0133] The preset rule statistics method can also set the presentation method to a heat map, and plot the position errors on the heat map. The heat map uses color to distinguish the number of position errors in each area. The darker the color, the more position errors there are.

[0134] The embodiment of the present application may also set a score threshold for the evaluation result of the evaluation system. The evaluation system outputs trajectories that are less than the score threshold, and the trajectories that are less than the score threshold are used as regression trajectories, which are used to guide the forward iteration of the algorithm.

[0135] Through the above method, the evaluation results determined by the evaluation system record the position error and posture error of each first-category data point and second-category data point of the intelligent vehicle in the actual scene, avoiding the use of statistical indicators to replace the error of a certain road section, resulting in the problem that the actual error of each data point is different from the statistical indicator. The evaluation results in the embodiment of the present application can be displayed in a variety of ways, and the evaluation results can be comprehensively and intuitively displayed. The determined evaluation results record the actual error of each data point, ensuring the accuracy of the evaluation results determined by the evaluation system.

[0136] Based on the above method, taking the first type of data point set as the benchmark, the position error and attitude error between the first type of data points and the second type of data points are determined, and the position error and attitude error are absolute errors, which avoids the problem of inaccurate evaluation results caused by using partial road sections or discrete sampling points for analysis. In addition, the scoring table, target map and heat map are used to comprehensively display the overall distribution of the position evaluation results, ensuring the accuracy and observability of the evaluation results.

[0137] Based on the same inventive concept, a data evaluation device is also provided in the embodiment of the present application. The data evaluation device is used to implement the function of a data evaluation method. Figure 3 , the device comprises:

[0138] An acquisition module 301 is used to acquire a first type of data point set and a second type of data point set;

[0139] An association module 302, configured to establish an association relationship between the first-category data point and the second-category data point according to a preset alignment rule when the initial recording time of each second-category data point in the second-category data point set is adjusted to the respective corresponding second recording time;

[0140] An error module 303 is used to determine the position error and posture error corresponding to the first type of data point and the second type of data point in each association relationship;

[0141] The evaluation module 304 is used to process the position error and the posture error corresponding to each association relationship according to preset statistical rules to generate an evaluation result corresponding to the data to be evaluated.

[0142] In one possible design, the acquisition module 301 is specifically used to parse out the initial positioning coordinates corresponding to each second-category data point in the second-category data point set, parse out the first positioning coordinates corresponding to each first-category data point in the first-category data point set, determine a preset coordinate transformation matrix between all first positioning coordinates corresponding to the first-category data point set and all initial positioning coordinates corresponding to the second-category data point set, align the second carrier coordinate system to the first carrier coordinate system based on the preset coordinate transformation matrix, and refer to each aligned initial positioning coordinate as the second positioning coordinate of the corresponding second-category data point.

[0143] In a possible design, the association module 302 is specifically used to parse out the first recording time corresponding to each first-category data point, and parse out the initial recording time corresponding to each preset number of second-category data points, determine the first preset time search range corresponding to the initial recording time of each second-category data point, calculate the first interval distance and the first heading angle difference between each second-category data point and each first-category data point in the corresponding first preset time search range, filter out the first-category data points corresponding to the minimum first interval distance in the first preset distance threshold and the first heading angle difference below the first preset heading angle threshold from all the first interval distances, establish a pairing relationship between the first-category data points and the second-category data points, respectively bring the first recording time corresponding to each first-category data point and the initial recording time corresponding to each second-category data point in all the pairing relationships into the preset time compensation formula, calculate the time compensation value corresponding to the second-category data point set, adjust the initial recording time corresponding to each second-category data point according to the time compensation value, and obtain the second recording time corresponding to each second-category data point.

[0144] In one possible design, the association module 302 is also used to parse out the second heading angle and the second coordinate corresponding to each second-category data point, and to parse out the first heading angle and the first coordinate corresponding to each first-category data point within each time search range, and to perform the following operations for each second-category data point: calculate the first interval distance between the second coordinate of a single second-category data point and each first coordinate within the corresponding time search range, and calculate the first heading angle difference between the second heading angle of a single second-category data point and each first heading angle within the corresponding time search range.

[0145] In one possible design, the association module 302 is also used to perform the following operations for each second-category data point: parse out the first coordinates and the first heading angle corresponding to each first-category data point within the second preset time search range, and parse out the second coordinates and the second heading angle corresponding to a single second-category data point, calculate the second interval distance between the second coordinates and each first coordinate, and calculate the first heading angle difference between the second heading angle and each first heading angle, filter out the first-category data points whose minimum second interval distance is lower than the second preset distance threshold and whose second heading angle difference is lower than the second preset heading angle threshold within the second preset time search range, and establish an association relationship between the first-category data points and the second-category data points.

[0146] In one possible design, the error module 303 is specifically used to parse the first position vector and the first rotation matrix of the first type of data point in each association relationship, and to parse the second position vector and the second rotation matrix of the second type of data point in each association relationship. The method for determining the position error and posture error corresponding to each association relationship is as follows: Substitute the first rotation matrix, the first position vector and the second position vector into a preset position error formula, calculate the position error corresponding to a single second type of data point, and calculate the difference matrix between the first rotation matrix and the second rotation matrix, and use the difference matrix as the posture error corresponding to a single second type of data point.

[0147] In one possible design, the evaluation module 304 is also used to determine the first target score corresponding to the position error according to the correspondence between the preset position error range and the first preset score, and to determine the second target score corresponding to the posture error according to the correspondence between the preset posture error range and the second preset score.

[0148] Based on the same inventive concept, an electronic device is also provided in an embodiment of the present application, and the electronic device can realize the functions of the aforementioned data evaluation device, referring to Figure 4 , the electronic device comprises:

[0149] At least one processor 401, and a memory 402 connected to the at least one processor 401. The specific connection medium between the processor 401 and the memory 402 is not limited in the embodiment of the present application. Figure 4 In the example, the processor 401 and the memory 402 are connected via the bus 400. The bus 400 is Figure 4 The connection between other components is shown by bold lines, and is not intended to be limiting. The bus 400 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4Only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. Alternatively, the processor 401 can also be called a controller, and there is no limitation on the name.

[0150] In the embodiment of the present application, the memory 402 stores instructions that can be executed by at least one processor 401. The at least one processor 401 can execute a data evaluation method discussed above by executing the instructions stored in the memory 402. The processor 401 can implement Figure 3 The functions of each module in the device shown.

[0151] Among them, the processor 401 is the control center of the device, and can use various interfaces and lines to connect the various parts of the entire control device. By running or executing instructions stored in the memory 402 and calling the data stored in the memory 402, the various functions of the device and process data, the device can be monitored as a whole.

[0152] In one possible design, the processor 401 may include one or more processing units, and the processor 401 may integrate an application processor and a modem processor, wherein the application processor mainly processes an operating system, a user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the modem processor may not be integrated into the processor 401. In some embodiments, the processor 401 and the memory 402 may be implemented on the same chip, and in some embodiments, they may also be implemented separately on separate chips.

[0153] Processor 401 can be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application-specific integrated circuit, a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, and can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor can be a microprocessor or any conventional processor, etc. The steps of a data evaluation method disclosed in conjunction with the embodiments of the present application can be directly embodied as a hardware processor to be executed, or a combination of hardware and software modules in the processor can be executed.

[0154] The memory 402 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules. The memory 402 may include at least one type of storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (Random Access Memory, RAM), a static random access memory (Static Random Access Memory, SRAM), a programmable read-only memory (Programmable Read Only Memory, PROM), a read-only memory (Read Only Memory, ROM), an electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), a magnetic memory, a disk, an optical disk, etc. The memory 402 is any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory 402 in the embodiment of the present application can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.

[0155] By programming the processor 401, the code corresponding to the data evaluation method described in the above embodiment can be fixed into the chip, so that the chip can execute the code when running. Figure 1 A data evaluation step of the embodiment shown. How to design and program the processor 401 is a technique known to those skilled in the art and will not be described in detail here.

[0156] Based on the same inventive concept, an embodiment of the present application further provides a storage medium, which stores computer instructions. When the computer instructions are executed on a computer, the computer executes a data evaluation method discussed above.

[0157] In some possible implementations, various aspects of a data evaluation method provided by the present application may also be implemented in the form of a program product, which includes program code. When the program product is run on an apparatus, the program code is used to enable the control device to execute the steps of a data evaluation method according to various exemplary implementations of the present application described above in this specification.

[0158] 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 in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0159] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes 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.

[0160] 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.

[0161] 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.

[0162] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A data evaluation method, It is characterized in that include: Acquire a first type of data point set and a second type of data point set, wherein the first type of data point set includes the position, posture and first recording time of each first type of data point, and the second type of data point set includes the position, posture and initial recording time of each second type of data point; When the initial recording time of each second-category data point in the second-category data point set is adjusted to the respective corresponding second recording time, an association relationship between the first-category data point and the second-category data point is established according to a preset alignment rule; Determine the position error and attitude error corresponding to the first type of data point and the second type of data point in each association relationship; The position error and the posture error corresponding to each association relationship are processed according to preset statistical rules to generate an evaluation result corresponding to the data to be evaluated.

2. The method according to claim 1, It is characterized in that Before obtaining the first type of data point set and the second type of data point set, it also includes: Parsing out initial positioning coordinates corresponding to each second type of data point in the second type of data point set, wherein the initial positioning coordinates are coordinates of the second carrier in a preset coordinate system; Parse the first positioning coordinates corresponding to each first-category data point in the first-category data point set, and determine a preset coordinate conversion matrix between all first positioning coordinates corresponding to the first-category data point set and all initial positioning coordinates corresponding to the second-category data point set, wherein the first positioning coordinates are the coordinates of the first carrier in the preset coordinate system, and the preset coordinate conversion matrix is ​​a rigid body transformation matrix from the second carrier coordinate system to the first carrier coordinate system; The second carrier coordinate system is aligned to the first carrier coordinate system based on the preset coordinate transformation matrix, and each initial positioning coordinate after the alignment is referred to as the second positioning coordinate of the corresponding second type data point.

3. The method according to claim 1, It is characterized in that The initial recording time of each second-category data point in the second-category data point set is adjusted to the respective corresponding second recording time, including: Parsing out the first recording time corresponding to each first-category data point, and parsing out the initial recording time corresponding to each second-category data point of a preset number; Determine a first preset time search range corresponding to the initial recording time of each second-category data point, and calculate a first interval distance and a first heading angle difference between each second-category data point and each first-category data point within the corresponding first preset time search range; Filter out first-category data points corresponding to the minimum first interval distance in the first preset distance threshold and the first heading angle difference value being lower than the first preset heading angle threshold from all the first interval distances, and establish a pairing relationship between the first-category data points and the second-category data points; Substitute the first recording time corresponding to each first-category data point and the initial recording time corresponding to each second-category data point in all paired relationships into a preset time compensation formula to calculate the time compensation value corresponding to the second-category data point set; The initial recording time corresponding to each second type of data point is adjusted according to the time compensation value to obtain the second recording time corresponding to each second type of data point.

4. The method according to claim 3, It is characterized in that The step of calculating the first interval distance and the first heading angle difference between each second-category data point and all first-category data points within the corresponding first preset time search range comprises: Parse out the second heading angle and the second coordinate corresponding to each second-category data point, and parse out the first heading angle and the first coordinate corresponding to each first-category data point within each first preset time search range; For each second-category data point, perform the following operations: Calculate the first interval distance between the second coordinate of a single second-category data point and each first coordinate within the corresponding first preset time search range; and The first heading angle differences between the second heading angle of a single second-category data point and each first heading angle within the corresponding first preset time search range are calculated.

5. The method according to claim 1, It is characterized in that The establishing of the association relationship between the first type of data points and the second type of data points according to the preset alignment rule includes: For each second-category data point, perform the following operations: Parsing out the first coordinates and the first heading angle corresponding to each first-category data point within the second preset time search range corresponding to the second recording time of each second-category data point, and parsing out the second coordinates and the second heading angle corresponding to a single second-category data point; Calculating second interval distances between the second coordinates and each first coordinate; and Calculating second heading angle differences between the second heading angle and each first heading angle; In the second preset time search range, first-category data points having a second interval distance lower than the minimum second interval distance in the second preset distance threshold and a second heading angle difference lower than the second preset heading angle threshold are screened out, and an association relationship between the first-category data points and the second-category data points is established.

6. The method according to claim 1, It is characterized in that The determining of the position error and the posture error corresponding to the first type of data point and the second type of data point in each association relationship includes: Parsing out a first position vector and a first rotation matrix of a first type of data point in each association relationship, and parsing out a second position vector and a second rotation matrix of a second type of data point in each association relationship; The method for determining the position error and attitude error corresponding to each association relationship is as follows: Substituting the first rotation matrix, the first position vector, and the second position vector into a preset position error formula to calculate the position error corresponding to a single second-category data point; and A difference matrix between the first rotation matrix and the second rotation matrix is ​​calculated, and the difference matrix is ​​used as a posture error corresponding to a single second-category data point.

7. The method according to claim 1, It is characterized in that After processing the position error and the posture error corresponding to each association relationship according to the preset statistical rules respectively, the method further includes: Determining a first target score corresponding to the position error according to a correspondence between a preset position error range and a first preset score; and A second target score corresponding to the posture error is determined according to a correspondence between a preset posture error range and a second preset score.

8. A data evaluation device, It is characterized in that include: An acquisition module, used for acquiring a first type of data point set and a second type of data point set; an association module, configured to establish an association relationship between the first-category data points and the second-category data points according to a preset alignment rule when the initial recording time of each second-category data point in the second-category data point set is adjusted to the respective corresponding second recording time; An error module, used to determine the position error and attitude error corresponding to the first type of data point and the second type of data point in each association relationship; The evaluation module is used to process the position error and the posture error corresponding to each association relationship according to preset statistical rules to generate corresponding evaluation results.

9. An electronic device, It is characterized in that include: Memory, used to store computer programs; A processor, for implementing the method steps described in any one of claims 1 to 7 when executing the computer program stored in the memory.

10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method steps described in any one of claims 1 to 7 are implemented.

Images