Road surface state data acquisition method based on array radar sensor

By adjusting the acquisition interval and analysis matrix of array radar sensors, the problem of difficulty in identifying noise data at uphill or downhill is solved, and the accurate collection and noise removal of road surface status data is achieved, reducing road surface risk.

CN118962623AActive Publication Date: 2024-11-15HAINAN VOCATIONAL COLLEGE OF SCI & TECH
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202411065470.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-05
Publication Date
2024-11-15
Estimated Expiration
2044-08-05

AI Technical Summary

Technical Problem

When the existing array radar sensors drive the vehicle to the uphill or downhill, the collected road surface status data is inconsistent with the normal data, and the noise data is difficult to identify, which increases the potential risks on the road.

Method used

By obtaining the speed of the acquisition vehicle and the flatness of the road state matrix, adjusting the acquisition interval, and constructing a pavement state analysis matrix based on the data consistency and change amplitude of the road state matrix in different directions, identifying and denoising data to ensure the data accuracy.

Benefits of technology

Effectively identify and remove noise data, promptly detect abnormal road conditions, and reduce potential risks on the road.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118962623A_ABST
    Figure CN118962623A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of data processing, in particular to a road surface state data acquisition method based on an array radar sensor, which comprises the following steps: acquiring the speed of an acquisition vehicle, and determining a road surface state matrix acquired next time according to the flatness of the road surface state matrix; obtaining the data consistency of the road surface state matrix in each direction and obtaining the minimum direction of the road surface state matrix; obtaining a partial matrix of the pavement state matrix and a corresponding matrix of the partial matrix, and further obtaining a pavement state analysis matrix; when the flatness of the pavement state matrix is smaller than or equal to a preset pavement flatness threshold value, the corresponding pavement state analysis matrix serves as an uneven pavement matrix, de-noised pavement state data are obtained according to the uneven pavement matrix and the pavement state matrix, and pavement state data collection based on the array radar sensor is completed. The invention aims to accurately identify noise data in the collected data, timely discover an abnormal road surface state and reduce potential risks on the road surface.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method for collecting road surface status data based on an array radar sensor. Background Art

[0002] Array radar sensors are radar systems that utilize multiple receiving and transmitting antennas. These antennas are integrated to form multiple beams, enabling more accurate determination of target location and other characteristics. By using array radar sensors to monitor road conditions and denoising and cleaning the collected data, the accuracy of road condition data is ensured, helping to promptly detect defects such as cracks and potholes, thereby mitigating potential road hazards.

[0003] Current methods for denoising collected road surface condition data typically identify noise data based on the differences between each data point and the rest of the data, and then perform denoising. When a vehicle encounters an uphill or downhill curve, the angle between the vehicle and the road surface can cause the collected data to be inconsistent with normal road surface data. The noise data may also fall within the normal road surface data range, leading to inaccurate identification of noise data within the collected data. This can prevent timely detection of abnormal road conditions and increase potential road safety risks. Summary of the Invention

[0004] The present invention provides a road surface status data acquisition method based on an array radar sensor to solve the existing problems.

[0005] The present invention provides a method for collecting road surface condition data based on an array radar sensor using the following technical solutions:

[0006] An embodiment of the present invention provides a method for collecting road surface condition data based on an array radar sensor, the method comprising the following steps:

[0007] Obtain the speed of the collection vehicle, and obtain the road surface state matrix for the next collection based on the flatness of the road surface state matrix collected each time;

[0008] According to the data variation amplitude of the road surface state matrix in different directions, the data consistency of the road surface state matrix collected each time in each direction is obtained; and the direction corresponding to the minimum consistency of the road surface state matrix collected each time is recorded as the minimum direction of the road surface state matrix collected each time;

[0009] Obtaining a partial matrix of the road surface state matrix collected each time and a corresponding matrix of the partial matrix according to the speed of the collection vehicle; obtaining a road surface state analysis matrix collected each time according to the data difference between the partial matrix collected each time and the corresponding matrix of the partial matrix in the minimum direction;

[0010] When the flatness of the road surface state matrix is ​​less than or equal to the preset road surface flatness threshold, the road surface state analysis matrix corresponding to the number of acquisitions is used as the uneven road surface matrix. Denoising is performed based on the uneven road surface matrix and the road surface state matrix to obtain the denoised road surface state data, completing the road surface state data acquisition based on the array radar sensor.

[0011] Furthermore, the specific method of obtaining the road surface state matrix collected next time is:

[0012] The acquisition interval of the sth collection is obtained as follows:

[0013]

[0014] t s =t×ρ

[0015] Among them, t s represents the acquisition interval of the sth acquisition; σ s-1 represents the standard deviation of the road surface state matrix collected at the s-1th time; t represents the initial collection interval; δ represents the preset hyperparameter; ρ s Indicates the flatness of the road surface collected for the sth time;

[0016] The road surface state matrix of the s-th collection is obtained using the collection interval of the s-th collection.

[0017] Furthermore, obtaining the data consistency of the road surface state matrix collected each time in each direction includes:

[0018] The cumulative sum of the standard deviations of all data in the u-th direction in the pavement state matrix collected for the s-th time is recorded as the data consistency of the pavement state matrix collected for the s-th time in the u-th direction.

[0019] Furthermore, the specific method of obtaining the direction is:

[0020] The four directions of the road surface state matrix collected each time are the horizontal direction, the vertical direction and the two diagonal directions of the road surface state matrix.

[0021] Furthermore, the obtaining of a partial matrix of the road surface state matrix collected each time and a corresponding matrix of the partial matrix includes:

[0022] The last row of the road surface state matrix collected at the s-th acquisition is used as the standard row of the road surface state matrix collected at the s-th acquisition. For the v-th acquisition after the s-th acquisition, the standard row of the road surface state matrix collected at the s-th acquisition is the ratio of the change distance in the v-th acquisition to the distance between the road surface positions collected by two adjacent sensors on the array radar sensor. The difference between the total number of rows of the road surface state matrix and the ratio of the distances is recorded as the number of rows of the standard row of the road surface state matrix collected at the s-th acquisition in the v-th acquisition.

[0023] The number of rows in the road surface state matrix is ​​recorded as L, and the corresponding row in the road surface state matrix collected at the vth time is recorded as the lth row. v,s Row, the Llth row of the road surface state matrix collected for the sth time v,s The row corresponds to the first row in the road surface state matrix collected for the vth time, and the Llth row in the road surface state matrix collected for the sth time. v,s The partial matrix of the sth acquisition is composed of the rows to the standard rows, and the 1st row to the lth row of the road surface state matrix collected in the vth time v,s The rows form the correspondence matrix of the vth acquisition to the sth acquisition.

[0024] Furthermore, the specific method of obtaining the change distance is:

[0025] The product of the speed of the collection vehicle and the collection interval from the sth collection to the vth collection is recorded as the change distance of the standard row of the road state matrix of the sth collection in the vth collection.

[0026] Furthermore, obtaining the road surface condition analysis matrix collected each time includes:

[0027] All the road surface condition data in the minimum direction of the partial matrix collected at the sth time and all the road surface condition data in the same direction of the corresponding matrix collected at the vth time are constructed with the minimum direction as the horizontal axis of the matrix and the vertical axis perpendicular to the minimum direction as the matrix. The matrices are recorded as the co-location matrix of the sth time collection and the corresponding co-location matrix of the vth time collection for the sth time collection;

[0028] The difference between the average value of the co-location matrix corresponding to the s-th acquisition and the average value of the co-location matrix of the s-th acquisition is recorded as the compensation value of the co-location matrix corresponding to the s-th acquisition for the v-th acquisition;

[0029] The difference between the mth road surface condition data in the collocation matrix collected at the sth time and the compensation value of the corresponding collocation matrix collected at the vth time is recorded as the mth adjusted road surface condition data in the collocation matrix collected at the sth time;

[0030] All adjusted road surface condition data in the co-location matrix collected for the sth time constitute the road surface condition analysis matrix collected for the sth time.

[0031] Furthermore, performing denoising based on the uneven road surface matrix and the road surface state matrix to obtain denoised road surface state data includes:

[0032] When the flatness of the sth acquisition is greater than the road surface flatness threshold When , the road surface state matrix collected for the sth time is recorded as a flat road surface matrix;

[0033] Obtain the abnormality degree of the wth data in the flat road matrix collected for the sth time;

[0034] When the abnormality degree of the wth data in the flat road matrix collected for the sth time is greater than or equal to the preset data abnormality threshold, the wth data in the flat road matrix collected for the sth time is recorded as abnormal data;

[0035] When the number of times the wth abnormal data is continuous with other abnormal data in the minimum direction in the flat road matrix collected for the sth time is greater than or equal to the abnormal data number threshold, the wth abnormal data is recorded as noise data;

[0036] When the flatness of the road surface is less than or equal to the preset road surface flatness threshold, the road surface condition analysis matrix collected for the sth time is recorded as an uneven road surface matrix;

[0037] Normalize the accumulated absolute value of the difference between the a-th road surface condition data in the uneven road surface matrix collected at the s-th time and the remaining road surface condition data in the uneven road surface matrix collected at the s-th time, excluding the a-th road surface condition data, and record the normalized result as the abnormality degree of the a-th road surface condition data in the uneven road surface matrix collected at the s-th time;

[0038] Normalize the sum of the absolute values ​​of the differences between the abnormality degree of the ath data in the uneven road matrix collected at the sth time and the abnormality degree of each data in the analysis sequence of the ath data in the uneven road matrix collected at the sth time, and record the normalized result as the possibility that the ath data in the uneven road matrix collected at the sth time is noise data;

[0039] When the possibility that the ath data in the uneven road surface matrix collected for the sth time is noise data is greater than the noise threshold, the ath data in the uneven road surface matrix collected for the sth time is recorded as noise data.

[0040] Furthermore, the abnormality degree of the wth data in the flat road matrix collected for the sth time includes:

[0041] Normalize the accumulated absolute value of the difference between the wth data in the flat road matrix collected for the sth time and the remaining data in the flat road matrix collected for the sth time, excluding the wth data. The normalized result is recorded as the abnormality degree of the wth data in the flat road matrix collected for the sth time.

[0042] Furthermore, the analysis sequence includes:

[0043] The data corresponding to the same road surface position of the a-th data is obtained in the corresponding co-location matrix of the s-th acquisition at the s+1-th acquisition. When the data corresponding to the same road surface position of the a-th data can be obtained in the corresponding co-location matrix of the s-th acquisition at the s+1-th acquisition, the data corresponding to the same road surface position of the a-th data is obtained in the corresponding co-location matrix of the s+2-th acquisition. And so on, until there is a constant c such that the data corresponding to the same road surface position of the a-th data cannot be obtained in the corresponding co-location matrix of the s+c-1-th acquisition at the s+c-th acquisition. Then, all the data corresponding to the same road surface position obtained by the a-th data constitute the analysis sequence of the a-th data in the uneven road surface matrix of the s-th acquisition.

[0044] The beneficial effects of the technical solution of the present invention are: obtaining the speed of the collection vehicle, obtaining the road surface state matrix for the next collection based on the flatness of the road surface state matrix collected each time, and at uneven road surface locations, in order to better judge whether it is noise or interference caused by factors such as road surface slope, it is necessary to shorten the collection interval so that multiple collections can be performed on the same road surface location, abnormal conditions are analyzed from the road surface state data collected multiple times, and the road surface state matrix is ​​re-acquired based on the adjusted collection interval; according to the data change amplitude of the road surface state matrix in different directions, the data consistency of the road surface state matrix collected each time in each direction is obtained, and the direction corresponding to the minimum data consistency is recorded as the minimum direction of the road surface state matrix collected each time. When the road surface state matrix changes in a certain direction, the noise data therein may be close to the normal road surface state data, and thus cannot be directly identified, but in the case of high consistency. In terms of direction, it will not be affected in this way. Therefore, the data consistency of each direction in the collected road surface state matrix is ​​determined, and the direction corresponding to the minimum data consistency can better reflect the abnormal data in the road surface state matrix. According to the speed of the collection vehicle, the partial matrix of the road surface state matrix collected each time and the corresponding matrix of the partial matrix are obtained. According to the data difference between the partial matrix collected each time and the corresponding matrix of the partial matrix in the minimum direction, the road surface state analysis matrix collected each time is obtained. When the flatness of the road surface state matrix is ​​less than or equal to the preset road surface flatness threshold, the road surface state analysis matrix corresponding to the number of collections is used as an uneven road surface matrix. Denoising is performed based on the uneven road surface matrix and the road surface state matrix to obtain the denoised road surface state data, and the road surface state data collection based on the array radar sensor is completed. The noise data in the collected data is accurately identified, the abnormal road surface state is discovered in time, and the potential risks on the road surface are reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0046] Figure 1 The present invention is a flowchart of the steps of a method for collecting road condition data based on an array radar sensor. DETAILED DESCRIPTION

[0047] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of a road surface condition data acquisition method based on an array radar sensor. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0048] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0049] The following describes in detail a specific solution of a road surface status data collection method based on an array radar sensor provided by the present invention with reference to the accompanying drawings.

[0050] See also Figure 1 , which shows a flowchart of a method for collecting road surface condition data based on an array radar sensor according to an embodiment of the present invention, the method comprising the following steps:

[0051] Step S001: Obtain the speed of the data collection vehicle, and obtain the next collected road surface state matrix based on the flatness of the road surface state matrix collected each time.

[0052] An array radar sensor is fixed under a data collection vehicle, which is then driven at a constant speed on the road surface, and its speed is recorded. The array radar sensor is used to collect the distance between the road surface and the vehicle chassis at an initial collection interval of t seconds, obtaining a road surface state matrix collected at each time. Because the array radar sensor's sensor distribution resembles a matrix, the road surface state matrix collected at each moment is arranged into a matrix according to the sensor positions. In this embodiment, the preset initial collection interval t = 10 seconds is used as an example for description. Other values ​​may be set in other implementations and are not limited in this embodiment.

[0053] It's important to note that in relatively flat locations, where data variance is minimal, the presence of noise can be analyzed by analyzing the differences in road condition data collected at each moment. Therefore, to improve collection efficiency, the collection interval can be extended. In locations with uneven roads, however, to better determine whether interference is caused by noise or factors like road slope, the collection interval should be shortened. This allows for multiple collections at the same road location, allowing analysis of anomalies from these multiple acquisitions. Therefore, the initial collection interval should be adjusted based on the measured road surface flatness.

[0054] Specifically, for the s-th collection, the acquisition method of the s-th collection interval is:

[0055]

[0056] t s =t×ρ

[0057] Among them, t s represents the acquisition interval of the sth acquisition; σ s-1 represents the standard deviation of the road surface state matrix collected for the s-1th time; t represents the initial collection interval; δ represents a preset hyperparameter. This embodiment is described using δ = 0.1 as an example. Other embodiments may set other values; ρ s Indicates the flatness of the road surface collected for the sth time.

[0058] When the standard deviation of the pavement state matrix collected for the s-1th time is zero, it means that the data in the pavement state matrix collected for the s-1th time is consistent, the pavement collected for the sth time is flat, and the collection interval of the sth time is consistent with the initial collection interval.

[0059] Furthermore, the road surface state matrix of the sth collection is obtained using the collection interval of the sth collection. Similarly, the road surface state matrices of all collections in one day are obtained according to the collection interval of each collection.

[0060] At this point, the adjusted collection interval is obtained, and the road surface state matrix is ​​recollected according to the adjusted collection interval.

[0061] Step S002: Obtain the data consistency of the road surface state matrix collected each time in each direction based on the data variation amplitude of the road surface state matrix in different directions; record the direction corresponding to the minimum consistency as the minimum direction of the road surface state matrix collected each time.

[0062] It should be noted that when the road surface state matrix varies in a certain direction, the noise data may be similar to the normal road surface state data, making it difficult to directly identify. However, in highly consistent directions, this effect is not affected. Therefore, it is necessary to first determine the data consistency in each direction of the collected road surface state matrix. The direction corresponding to the minimum data consistency is more representative of the abnormal data in the road surface state matrix.

[0063] Specifically, this embodiment uses the four directions of the pavement state matrix collected each time as an example for description. Other implementation methods may set it to other directions, which is not limited in this embodiment. The four directions of the pavement state matrix collected each time are the horizontal, vertical and two diagonal directions of the pavement state matrix. The horizontal direction is from the first column to the last column of the pavement state matrix, and the vertical direction is from the first row to the last row of the pavement state matrix. The two diagonal directions refer to from the lower left corner to the upper right corner of the pavement state matrix and from the lower right corner to the upper left corner of the pavement state matrix, respectively.

[0064] The calculation formula for the data consistency of the road surface state matrix collected for the sth time in the uth direction is:

[0065]

[0066] Among them, r s,u Indicates the data consistency of the u-th direction in the road surface state matrix collected for the s-th time; n s,u Indicates the number of data rows in the u-th direction of the road surface state matrix collected for the s-th time; σ s,u,k It represents the standard deviation of the k-th row of data in the u-th direction in the road surface state matrix collected for the s-th time.

[0067] Similarly, the data consistency of each direction in the road surface state matrix collected for the sth time is obtained, and the direction corresponding to the minimum data consistency is recorded as the minimum direction of the road surface state matrix collected each time.

[0068] Step S003: Obtain a partial matrix of the road surface state matrix collected each time and a corresponding matrix of the partial matrix according to the speed of the collection vehicle; and obtain a road surface state analysis matrix collected each time according to the data difference between the partial matrix collected each time and the corresponding matrix of the partial matrix in the minimum direction.

[0069] It's important to note that the minimum direction of each collected pavement state matrix best reflects whether the matrix contains noise data. Therefore, to analyze the degree of anomaly for a particular data point in each collected pavement state matrix, it's necessary to identify all the data points corresponding to that data point in the minimum direction in the pavement state matrix as reference data to determine the anomaly. Because the amount of data used as reference data for a single row of data in the minimum direction of a single collected pavement state matrix is ​​too small and may be affected by random factors and unreliable, it's necessary to further identify multiple reference data points at the same location. For any given data point in each collected pavement state matrix, due to the adjusted collection interval, the same pavement location may be captured multiple times during multiple collections. Therefore, it's necessary to determine the position of the data corresponding to the same pavement location in the pavement state matrices obtained from different collections.

[0070] Taking the rear of the vehicle as a reference, i.e., the last row of data collected in the road condition matrix, the position of the same road surface location in the road condition matrix will change with different acquisition times. For the same road surface location, the data corresponding to the road condition matrix and the corresponding position in the road condition data collected at different times are obtained based on the position change of the same road surface location in the road condition matrix with different acquisition times.

[0071] Furthermore, the last row of the pavement state matrix collected at the sth time is used as the standard row of the pavement state matrix collected at the sth time. This embodiment uses the standard row of the pavement state matrix collected at the sth time as a reference for the same pavement position in the pavement state data. Other embodiments use other data marking methods to represent the same pavement position, which is not specifically limited in this embodiment. Then, for the vth acquisition after the sth acquisition, the number of rows of the standard row of the pavement state matrix collected at the sth time in the vth acquisition is:

[0072]

[0073] Among them, l v,s is the number of rows of the standard row of the pavement state matrix collected at the sth time in the vth collection; L is the total number of rows of the pavement state matrix; S v,s The distance changed by the standard row of the road surface state matrix collected at the sth time during the vth collection is obtained by multiplying the speed of the collection vehicle by the collection interval from the sth collection to the vth collection. x represents the distance between the road surface positions collected by two adjacent sensors on the array radar sensor.

[0074] It should be noted that when the collection vehicle moves forward, the column corresponding to the same road surface position in the road surface state matrix during different collection processes will not change. Therefore, in the road surface state matrix obtained by the sth collection, find the row closest to the last row of data in the road surface state matrix collected for the vth time. The data in the same column as the road surface position data collected for the sth time is the position of the data corresponding to the same road surface position in the road surface state matrix obtained by different collections.

[0075] It is particularly important to note that the standard row of the road surface state matrix collected for the sth time corresponds to the lth row of the road surface state matrix collected for the vth time. v,s Row, the Llth row of the road surface state matrix collected for the sth time v,s The row corresponds to the first row in the road surface state matrix collected for the vth time, and the Llth row in the road surface state matrix collected for the sth time. v,s The partial matrix of the sth acquisition is composed of the rows from row 1 to row 1 in the road surface state matrix of the vth acquisition.v,s The sizes of the partial matrices of the vth acquisition formed by the rows are the same, so the partial matrix of the sth acquisition is recorded as the corresponding matrix of the vth acquisition to the sth acquisition.

[0076] At this point, the corresponding matrix of the same data in different acquisition times is obtained.

[0077] It should be noted that since the angle of the vehicle relative to the ground changes when the collection vehicle goes uphill or downhill, the same road surface position may change in different collection processes. When the vehicle moves forward, the columns of the same collection position will not change, but the number of rows will change as the vehicle moves forward. Therefore, when the collection vehicle is going uphill or downhill, the data difference in the same row of the road surface state matrix is ​​smaller than the data difference in the same column. When the tires on one side of the collection vehicle are higher than the tires on the other side, the data difference in the same column of the road surface state matrix is ​​smaller than the data difference in the same row. In order to facilitate finding the noise data, it is necessary to first perform height compensation on the data in different collection processes based on the average value of the data in different collection processes and the minimum direction of the data change difference in the same direction in the road surface state matrix collected each time. Then, all collected road surface state data are adjusted according to the compensation value, and a road surface state analysis matrix is ​​established based on the adjusted road surface state data in the road surface area.

[0078] Specifically, based on the different road conditions encountered by the collection vehicle, all road condition data in the minimum direction of the partial matrix collected for the sth time and all road condition data in the same direction of the corresponding matrix of the vth collection for the sth collection are taken as the horizontal axis of the matrix and the vertical axis perpendicular to the minimum direction. The matrices constructed are recorded as the co-location matrix of the sth collection and the corresponding co-location matrix of the vth collection for the sth collection. The co-location matrix and the corresponding co-location matrix have the same size.

[0079] Furthermore, the difference between the average value of the corresponding co-location matrix of the v-th acquisition for the s-th acquisition and the average value of the co-location matrix of the s-th acquisition is recorded as the compensation value of the corresponding co-location matrix of the v-th acquisition for the s-th acquisition; the difference between the m-th road surface condition data in the co-location matrix of the s-th acquisition and the compensation value of the corresponding co-location matrix of the v-th acquisition for the s-th acquisition is recorded as the m-th adjusted road surface condition data in the co-location matrix of the s-th acquisition; all adjusted road surface condition data in the co-location matrix of the s-th acquisition constitute the road surface condition analysis matrix of the s-th acquisition.

[0080] Step S004: When the flatness of the road surface state matrix is ​​less than or equal to a preset road surface flatness threshold, the road surface state analysis matrix corresponding to the number of acquisitions is used as an uneven road surface matrix. Denoising is performed based on the uneven road surface matrix and the road surface state matrix to obtain denoised road surface state data, thereby completing road surface state data acquisition based on the array radar sensor.

[0081] It should be noted that the greater the flatness of each acquired data point, the less affected the road surface is by the road surface condition. The data corresponding to that road surface location in the road surface condition matrix is ​​relatively consistent with normal road surface condition data. For a road surface condition matrix acquired when the road surface is flat, the degree of abnormality of the road surface condition data can be directly calculated based on the difference between each data point and the other data points in the road surface condition matrix. The smaller the flatness of each acquired data point, the more affected the road surface is by the road surface condition. For a road surface condition analysis matrix acquired when the road surface is uneven, the degree of abnormality of each data point in the acquired road surface condition analysis matrix can be calculated based on the difference between each data point and the other data points.

[0082] Specifically, the road surface flatness threshold preset in this embodiment is This is described as an example. In other implementations, other values ​​may be set, and this embodiment does not limit this.

[0083] When the flatness of the sth acquisition is greater than the road surface flatness threshold When , the road surface state matrix collected for the sth time is recorded as the flat road surface matrix. The calculation formula for the abnormality degree of the wth data in the flat road surface matrix collected for the sth time is:

[0084]

[0085] Among them, f s,w Indicates the abnormality of the wth data in the flat road matrix collected for the sth time; d s,j represents the jth data in the flat road matrix collected for the sth time; d s,w represents the wth data in the flat road matrix collected for the sth time; n represents the number of sensors on the array radar sensor; norm() represents the linear normalization function, which normalizes the data value to the range [0,1]; || represents the absolute value function.

[0086] In this embodiment, the data anomaly threshold b is set to 0.5, which is used as an example. Other values ​​may be set in other embodiments and are not limited in this embodiment. When the degree of anomaly of the wth data item in the flat road matrix collected for the sth time is greater than or equal to the preset data anomaly threshold b, the wth data item in the flat road matrix collected for the sth time is recorded as an anomaly.

[0087] It should be further explained that abnormal data may be either noise data or road surface defects. Noise data is usually scattered, while road surface defects are continuous. Therefore, noise data can be obtained based on the number of consecutive abnormal data in the flat road surface matrix.

[0088] In this embodiment, the preset abnormal data quantity threshold h = 3 is used as an example for description. Other values ​​may be set in other embodiments and are not limited in this embodiment. When the number of times the w-th abnormal data is continuous with other abnormal data in the minimum direction in the flat road matrix collected for the s-th time is greater than or equal to the abnormal data quantity threshold h, the w-th abnormal data is recorded as noise data.

[0089] When the flatness of the road surface is less than or equal to the preset road surface flatness threshold When , the road condition analysis matrix collected for the sth time is recorded as the uneven road matrix. The calculation formula for the abnormality degree of the ath data in the uneven road matrix collected for the sth time is:

[0090]

[0091] Among them, f s,a represents the abnormality of the ath data in the uneven road matrix collected for the sth time; Y represents the number of data in the uneven road matrix; d s,a represents the ath road surface state data in the uneven road surface matrix collected for the sth time; d s,y represents the y-th road surface condition data in the uneven road surface matrix collected for the s-th time; norm() represents the linear normalization function, which normalizes the data value to the interval [0,1]; || represents the absolute value function.

[0092] It should be further noted that due to the random nature of noise data, if data with significant anomalies is noise data, multiple data collections at the same road surface location will show inconsistencies. Therefore, the differences between multiple data collections at the same road surface location can be used to determine whether the data collected is noise data.

[0093] Specifically, the data corresponding to the same road surface position of the a-th data is obtained in the corresponding co-location matrix of the s-th acquisition at the s+1-th acquisition. When the data corresponding to the same road surface position of the a-th data can be obtained in the corresponding co-location matrix of the s-th acquisition at the s+1-th acquisition, the data corresponding to the same road surface position of the a-th data is obtained in the corresponding co-location matrix of the s+1-th acquisition at the s+2-th acquisition, and so on, until there is a constant c such that the data corresponding to the same road surface position of the a-th data cannot be obtained in the corresponding co-location matrix of the s+c-1-th acquisition at the s+c-th acquisition. All the data corresponding to the same road surface position obtained by the a-th data constitute the analysis sequence of the a-th data in the uneven road surface matrix of the s-th acquisition.

[0094] The calculation formula for the possibility that the ath data in the uneven road surface matrix collected for the sth time is noise data is:

[0095]

[0096] Among them, p s,a Indicates the possibility that the ath data in the uneven road matrix collected for the sth time is noise data; n s,a represents the sequence length of the analysis sequence of the ath data in the uneven road surface matrix collected for the sth time; f s,a represents the abnormality of the ath data in the uneven road matrix collected for the sth time; f a,s,o It represents the abnormality of the o-th data in the analysis sequence of the a-th data in the uneven road surface matrix collected for the s-th time; norm() represents the linear normalization function, which normalizes the data values ​​to the interval [0,1]; || represents the absolute value function.

[0097] In this embodiment, the preset noise threshold g = 0.5 is used as an example for description. Other values ​​may be set in other embodiments and are not limited in this embodiment. When the probability that the a-th data item in the uneven road surface matrix collected for the s-th time is noise data is greater than the noise threshold g, the a-th data item in the uneven road surface matrix collected for the s-th time is recorded as noise data.

[0098] For all noise data, median filtering is performed on the noise data according to the adjacent non-noise data in the road surface state matrix. The result of median filtering is used as the road surface state data after denoising each noise data, and the denoised road surface state data is obtained to complete the collection of road surface state data.

[0099] This concludes the present invention.

[0100] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for collecting road surface status data based on an array radar sensor, characterized in that: The method comprises the following steps: The speed of the collection vehicle is obtained, and the road surface state matrix for the next collection is obtained according to the flatness of the road surface state matrix collected each time; According to the data variation range of the road surface state matrix in different directions, the data consistency of the road surface state matrix collected each time in each direction is obtained; and the direction corresponding to the minimum data consistency is recorded as the minimum direction of the road surface state matrix collected each time; According to the speed of the data collection vehicle, a partial matrix of the road surface state matrix collected each time and a corresponding matrix of the partial matrix are obtained; according to the data difference between the partial matrix collected each time and the corresponding matrix of the partial matrix in the minimum direction, an analysis matrix of the road surface state collected each time is obtained; When the flatness of the road surface state matrix is ​​less than or equal to the preset road surface flatness threshold, the road surface state analysis matrix corresponding to the number of acquisitions is used as the uneven road surface matrix. Denoising is performed based on the uneven road surface matrix and the road surface state matrix to obtain the denoised road surface state data, thereby completing the road surface state data acquisition based on the array radar sensor.

2. The method for collecting road surface status data based on an array radar sensor according to claim 1, characterized in that: The specific method of obtaining the road surface state matrix collected next time is: The acquisition interval of the sth collection is obtained as follows: t s =t×ρ Among them, t s represents the collection interval of the sth collection; σ s-1 represents the standard deviation of the road state matrix collected for the s-1th time; t represents the initial collection interval; δ represents the preset hyperparameter; ρ s Indicates the flatness of the road surface collected for the sth time; The road surface state matrix of the s-th collection is obtained by using the collection interval of the s-th collection.

3. The method for collecting road surface status data based on an array radar sensor according to claim 1, characterized in that: The obtaining of data consistency of the road surface state matrix collected each time in each direction includes: The cumulative sum of the standard deviations of all data in the u-th direction in the road surface state matrix collected for the s-th time is recorded as the data consistency of the road surface state matrix collected for the s-th time in the u-th direction.

4. The method for collecting road surface status data based on an array radar sensor according to claim 3, characterized in that: The specific method of obtaining the direction is: The four directions of the road surface state matrix collected each time are the horizontal direction, the vertical direction and two diagonal directions of the road surface state matrix.

5. The method for collecting road surface status data based on array radar sensors according to claim 1, characterized in that: The method of obtaining a partial matrix of the road surface state matrix collected each time and a corresponding matrix of the partial matrix includes: The last row of the road state matrix collected at the sth time is used as the standard row of the road state matrix collected at the sth time. For the vth collection after the sth collection, the standard row of the road state matrix collected at the sth time is the ratio of the change distance in the vth collection to the distance between the road surface positions collected by two adjacent sensors on the array radar sensor. The difference between the total number of rows of the road state matrix and the ratio of the distance is recorded as the number of rows of the standard row of the road state matrix collected at the sth time in the vth collection. The number of rows in the road surface state matrix is ​​recorded as L, and the corresponding row in the road surface state matrix collected at the vth time is recorded as the lth row. v,s Row, the Llth row of the road surface state matrix collected for the sth time v,s The row corresponds to the first row in the road surface state matrix collected for the vth time, and the Llth row in the road surface state matrix collected for the sth time. v,s The partial matrix of the sth acquisition is composed of rows from row to standard row, and the pavement state matrix of the vth acquisition is composed of rows from row 1 to row l v,s The rows form the correspondence matrix of the vth acquisition to the sth acquisition.

6. The method for collecting road surface status data based on array radar sensors according to claim 5, characterized in that: The specific method of obtaining the change distance is: The product of the speed of the collection vehicle and the collection interval from the sth collection to the vth collection is recorded as the change distance of the standard row of the road state matrix of the sth collection in the vth collection.

7. The method for collecting road surface status data based on array radar sensors according to claim 5, characterized in that: The obtaining of the road surface condition analysis matrix collected each time includes: All the road surface condition data in the minimum direction of the partial matrix collected at the sth time and all the road surface condition data in the same direction of the corresponding matrix collected at the vth time are constructed with the minimum direction as the horizontal axis of the matrix and the vertical axis perpendicular to the minimum direction as the matrix vertical axis, and the obtained matrix is ​​recorded as the co-location matrix of the sth collection and the corresponding co-location matrix of the vth collection for the sth collection; The difference between the average value of the co-location matrix corresponding to the s-th acquisition for the v-th acquisition and the average value of the co-location matrix for the s-th acquisition is recorded as the compensation value of the co-location matrix corresponding to the s-th acquisition for the v-th acquisition; The difference between the mth road surface state data in the co-location matrix collected at the sth time and the compensation value of the corresponding co-location matrix collected at the vth time is recorded as the mth adjusted road surface state data in the co-location matrix collected at the sth time; All adjusted road surface condition data in the co-location matrix collected for the sth time constitute the road surface condition analysis matrix collected for the sth time.

8. The method for collecting road surface status data based on array radar sensors according to claim 2, characterized in that: The denoising is performed according to the uneven road surface matrix and the road surface state matrix to obtain the denoised road surface state data, which includes: When the flatness of the sth acquisition is greater than the road surface flatness threshold When , the road surface state matrix collected for the sth time is recorded as a flat road surface matrix; Obtain the abnormality degree of the wth data in the flat road matrix collected for the sth time; When the abnormality degree of the wth data in the flat road matrix collected for the sth time is greater than or equal to the preset data abnormality threshold, the wth data in the flat road matrix collected for the sth time is recorded as abnormal data; When the number of the wth abnormal data that is continuous with other abnormal data in the minimum direction in the flat road matrix collected for the sth time is greater than or equal to the abnormal data quantity threshold, the wth abnormal data is recorded as noise data; When the flatness of the road surface is less than or equal to the preset road surface flatness threshold, the road surface state analysis matrix collected for the sth time is recorded as an uneven road surface matrix; Normalize the accumulated sum of the absolute values ​​of the differences between the ath road surface state data in the uneven road surface matrix collected for the sth time and the remaining road surface state data except the ath road surface state data in the uneven road surface matrix collected for the sth time, and record the normalized result as the abnormality degree of the ath road surface state data in the uneven road surface matrix collected for the sth time; Normalize the accumulated sum of the absolute values ​​of the difference between the abnormality degree of the ath data in the uneven road matrix collected at the sth time and the abnormality degree of each data in the analysis sequence of the ath data in the uneven road matrix collected at the sth time, and record the normalized result as the possibility that the ath data in the uneven road matrix collected at the sth time is noise data; When the possibility that the ath data in the uneven road surface matrix collected for the sth time is noise data is greater than the noise threshold, the ath data in the uneven road surface matrix collected for the sth time is recorded as noise data.

9. The method for collecting road surface status data based on array radar sensors according to claim 8, characterized in that: The abnormality degree of the wth data in the flat road matrix collected for the sth time includes: The accumulated absolute value of the difference between the wth data in the flat road matrix collected for the sth time and the remaining data except the wth data in the flat road matrix collected for the sth time is normalized, and the normalized result is recorded as the abnormality degree of the wth data in the flat road matrix collected for the sth time.

10. The method for collecting road surface status data based on array radar sensors according to claim 8, characterized in that: The analysis sequence includes: In the corresponding co-location matrix of the s+1th acquisition, data corresponding to the same road surface position of the a-th data is obtained. When the data corresponding to the same road surface position of the a-th data can be obtained in the corresponding co-location matrix of the s+1th acquisition, data corresponding to the same road surface position of the a-th data is obtained in the corresponding co-location matrix of the s+2th acquisition, and so on, until there is a constant c such that when the data corresponding to the same road surface position of the a-th data cannot be obtained in the corresponding co-location matrix of the s+cth acquisition for the s+c-1th acquisition, all the data corresponding to the same road surface position obtained by the a-th data constitute the analysis sequence of the a-th data in the uneven road surface matrix of the s-th acquisition.

Citation Information

Patent Citations

  • Traffic data acquisition method and system of matrix radar based on Doppler technology

    CN108010334A

  • Pavement disease area judgment method based on ground penetrating radar original data

    CN115542278A

  • Pavement flatness evaluation method based on laser radar three-dimensional data

    CN116879866A

  • Target detecting apparatus using electronically agile radar

    US20080088498A1

  • Slope estimation using ground radar

    WO2024088506A1