Water level monitoring method

Abstract

The invention belongs to the technical field of water level monitoring, and particularly relates to a water level monitoring method which comprises the following steps: acquiring point cloud data of a single-line laser radar for scanning a target area; determining a water surface height range according to the lowest point in the point cloud data and the maximum water penetration distance of the single-line laser radar in clear water; clustering the point cloud data, selecting data clusters within the water surface height range, and screening out the data cluster with the minimum distance from the lowest point to the single-line laser radar from the data clusters as a target data cluster; calculating the relative echo slope of each point in the target data cluster, and determining the point with the maximum value as a water bank junction point according to the relative echo slope of each point; and determining a water level value according to the water-bank junction point. According to the method, data errors caused by laser water penetration can be eliminated, meanwhile, the data errors caused by radar image inclination are eliminated by calibrating the horizontal angle of the single-line laser radar, and finally the accuracy of water level monitoring is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of water level monitoring, and particularly relates to a water level monitoring method. Background Art

[0002] Water level monitoring is an important means for flood forecasting, flood control command, safe operation of water conservancy projects, water resources management and protection. Traditional water level monitoring mainly relies on hydrological stations, shipborne sonars, single-beam or multi-beam bathymeters, airborne bathymeters, etc. to obtain bathymetric maps. Such methods often consume a large amount of manpower and material resources and cannot achieve real-time and accurate monitoring.

[0003] LiDAR has the advantages of high measurement accuracy, fast response speed and strong anti-interference ability. The point cloud data obtained by LiDAR contains the position (distance, azimuth angle) of the target. Its spatial resolution can reach the centimeter level, and the time resolution can reach the millisecond level. It can work continuously day and night. The resolution of the data is usually higher than that of millimeter-wave radar or camera, and it can accurately image and form a high-precision three-dimensional point cloud. Compared with hydrological stations, LiDAR can be flexibly installed and has a lower cost; compared with shipborne sonars, single-beam or multi-beam bathymeters and airborne bathymeters, LiDAR does not need to consider corrosion or adhesion caused by contact with the water surface like contact type monitors, which may lead to inaccurate measurement. At present, LiDAR has been widely used in water level monitoring.

[0004] The Chinese patent application document with the publication number of CN113124959A discloses a method for measuring water level based on laser scanning of the embankment. Its main steps include: setting LiDAR to scan along the scanning line on the embankment slope; calculating the distance between the scanning point and LiDAR; calculating the elevation of the scanning point relative to the Yellow Sea as the elevation of LiDAR minus the elevation difference from the scanning point to LiDAR; calculating the water level height corresponding to each water level scanning point. The above method has the problem of laser penetrating water in actual use. When the water quality is clear, the laser has the characteristics of penetrating water and no echo on the water surface. When there is no echo on the water surface, the data points obtained by the laser are lower than the water-land interface due to laser penetration, and the actual water level value cannot be accurately obtained; and when the water quality changes, the penetration distance of the laser will also change accordingly, affecting the accuracy of water level measurement.

[0005] The Chinese invention patent application document with the publication number CN114353905A discloses a water level monitoring device and a water level monitoring method. Among them, the water level monitoring device includes a measuring device, a gyroscope and a processing device. The measuring device uses a lidar to scan the intersection point of the water level liquid surface and a target object and obtain a point cloud map of the initial scanning surface, and transmits the point cloud map of the initial scanning surface to the processing device; the gyroscope is used to detect the current inclination information of the measuring device and transmit it to the processing device; the processing device is used to obtain the point cloud map of the real scanning surface corresponding to the point cloud map of the initial scanning surface according to the current inclination information, and obtain the height of the water level liquid surface according to the point cloud map of the real scanning surface. The above gyroscope is a dynamic monitoring device. In the static state, there is no angular velocity input to the gyroscope, and there is a zero bias error in practice. Therefore, in static use, it is impossible to accurately judge whether the lidar is perpendicular to the horizontal plane, resulting in a certain error between the coordinate system obtained by laser scanning and the actually installed coordinate system, causing the image scanned by the lidar to be tilted, and the measured water level height deviating from the actual value, affecting the water level monitoring accuracy.

[0006] In addition, when the lidar is far from the detection target, if the device is slightly disturbed by external forces, the interference with the acquisition of the position of the far-distance target will be amplified. The installation rod of the lidar is also easily affected by external force fluctuations such as common strong winds, which will cause the coordinate system to shift and affect the monitoring accuracy. Summary of the Invention

[0007] In view of this, the purpose of the present invention is to provide a water level monitoring method to solve the problem of low accuracy of water level monitoring in the prior art.

[0008] To solve the above problems, a water level monitoring method provided by the present invention adopts the following technical solutions: A water level monitoring method includes: S1. Scanning a target area with a single-line lidar and obtaining point cloud data; S2. Determining the water surface height range according to the lowest point in the point cloud data and the maximum water penetration distance of the single-line lidar in clear water; S3. Clustering the point cloud data, selecting the data clusters located within the water surface height range, and screening out the data cluster with the smallest distance from the lowest point to the single-line lidar as the target data cluster from these data clusters; S4. Calculating the relative echo slope of each point in the target data cluster, determining the point with the largest value as the water bank intersection point according to the relative echo slope of each point; determining the water level value according to the water bank intersection point.

[0009] Further, it also includes the step of performing multi-level sliding filtering on the water bank intersection point data.

[0010] Furthermore, the method of multi-stage sliding filtering is: the y value of the output water-shore junction point is input into a 1-minute window, and the average value in the window is input into a 5-minute window. Similarly, the average value outputted by the 5-minute window is the water level value.

[0011] Furthermore, in S2, the calibration method of the maximum water penetration distance is: In the laboratory, the single-line laser radar used was fixed at a set distance from the water tank without water, and a point at a certain angle was determined as a reference point; Gradually add water to the water tank, and stop adding water when the above reference point disappears; Place a floating object on the water surface, obtain the reference point at that moment, calculate the distance difference between the two reference points and the single-line laser radar, and obtain the maximum water penetration distance of the single-line laser radar.

[0012] Furthermore, the method for determining the water surface height range is: find the lowest point in the point cloud data, record the coordinates of the lowest point as (x1, y1), the distance from the lowest point to the origin as d, the angle as θ, and the maximum water penetration distance as C; the origin is the location of the single-line laser radar, calculate the Cartesian coordinate system (x2, y2) corresponding to the polar coordinate system (dC, θ), then y∈[y1, y2] is the height range of the water surface.

[0013] Furthermore, it is determined whether the y value of the data point in each data cluster is located in [y1, y2]. If not, the data cluster is eliminated. The distance from the lowest point to the origin in the remaining data clusters is compared, and the data cluster with the smallest distance value is the target data cluster containing the water-shore junction point.

[0014] Furthermore, in S4, the data points in the target data cluster are sorted according to the index point number. The sequence is the sequence number of each laser sent by the single-line laser radar. According to the formula

[0015] Calculate the relative echo slope of each data point. The data point with the maximum relative echo slope is the water-shore junction point. is the echo rate of the data point, is the relative echo slope of the data point.

[0016] Furthermore, the point cloud data is clustered using a DBSCAN clustering algorithm.

[0017] Furthermore, before acquiring the point cloud data, the horizontal angle of the single-line laser radar is calibrated.

[0018] Further, the method for horizontal angle calibration is as follows: On the water surface of the scanning section of the single-line lidar, lay two suspended objects separated by a set distance. Obtain the data points of the two suspended objects in the lidar image, calculate the included angle between the two data points, and this included angle is the angle offset of the single-line lidar. Substitute this included angle into the coordinate system of the single-line lidar for correction.

[0019] The beneficial effects of a water level monitoring method of the present invention are as follows: 1. Since the laser has the characteristics of penetrating water and having no echo on the water surface when the water quality is clear, the present invention first determines the height range of the water surface according to the lowest point of the point cloud data scanned by the single-line lidar and the maximum water penetration distance of the laser in clear water. This height range is not affected by water quality changes, and target data clusters can be accurately selected within the above height range. By vertically scanning the cross-section of the water surface with a single-line lidar, an image of the embankment can be obtained. Using the principle that the echo of the embankment is strong and the echo of the embankment submerged in water is weak or even has no echo, by calculating the relative echo slope of the data points in the target data cluster and finding the point where the first numerical mutation occurs in the order of numbers, which is the water-bank intersection point. The water level height can be calculated based on the water-bank intersection point. The present invention can solve the measurement error problem caused by laser water penetration in the prior art, and the accuracy of water level measurement is higher.

[0020] 2. After the single-line lidar is installed, the present invention performs horizontal angle calibration on the single-line lidar, can calculate the angle offset of the single-line lidar at the installation site of this time, and substitute this angle offset into the coordinate system of the single-line lidar, so that the required horizontal plane of the single-line lidar is the real horizontal plane, thereby correcting the image tilt and data error problems caused by the installation angle deviation and improving the accuracy of water level monitoring.

[0021] 3. The present invention can filter out unnecessary data points such as floating debris on the water surface and the opposite bank through a clustering algorithm, accurately obtain the data cluster near the water-bank intersection, and obtain the actual position of the water-bank intersection through the change of the echo rate of each data point and the echo rate between adjacent points.

[0022] 4. The present invention performs multi-level filtering on the determined data of the water-bank intersection point, reduces the data jitter caused by the system error of the single-line lidar itself and external interference, can output smooth and stable water level data, and the output water level data is more accurate. Description of the Drawings

[0023] Figure 1 It is a step block diagram of a water level monitoring method of the present invention. Detailed Embodiments

[0024] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Those skilled in the art should know that the embodiments described below are part of the embodiments of the present disclosure, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present invention.

[0025] In addition, it should be understood that the magnitudes of the sequence numbers of the steps in the following embodiments do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0026] An embodiment of a water level monitoring method provided by the present invention: As Figure 1 shown, a water level monitoring method mainly includes the following steps: S1. Use a single-line lidar to scan the target area and obtain point cloud data; S2. Determine the water surface height range according to the lowest point in the point cloud data and the maximum water penetration distance of the single-line lidar in clear water; S3. Cluster the point cloud data, select the data clusters located within the water surface height range, and screen out the data cluster with the smallest distance from the lowest point to the single-line lidar from these data clusters as the target data cluster; S4. Calculate the slope of the echo rate of each point in the target data cluster, and determine the point with the largest value as the water-land intersection point according to the slope of the echo rate of each point; determine the water level value according to the water-land intersection point.

[0027] In S2, for different single-line lidars, the maximum water penetration distance in clear water is calibrated in the laboratory to obtain the water surface value range and reduce the subsequent calculation amount. Specifically, the calibration method for the maximum water penetration distance of the single-line lidar in clear water is as follows: S21. In the laboratory, fix the used single-line lidar at a position 5 meters away from the water tank without water, and determine a point at a certain angle as the reference point; S22. Gradually fill water into the water tank until the water stops flowing when the above reference point disappears; S23. Place a floating object on the current water surface to obtain the reference point at this moment; S24. Compare the distances between the two reference points and the single-line lidar, and calculate the distance difference between the two reference points and the single-line lidar, which is the maximum water penetration distance of the single-line lidar.

[0028] It should be noted that since there are many single-line lidar brands, with different performances, laser types and powers, the water penetration distance of each single-line lidar in clear water is different. Therefore, it is necessary to measure and calibrate the maximum water penetration distance of the used single-line lidar. This calibration only needs to be carried out once in the laboratory for the same single-line lidar, and it has nothing to do with the subsequent use scenarios and does not need to be calibrated again.

[0029] In S2, the method for determining the water surface height range is as follows: find the lowest point in the point cloud data, record the coordinates of the lowest point as (x1, y1), the distance from the lowest point to the origin (single-line lidar) is d, the angle is θ, and the maximum water penetration distance of the single-line lidar in clear water is C. Calculate the Cartesian coordinates (x2, y2) corresponding to the polar coordinate system (d - C, θ), x 2= (d - C)×cosθ, y 2= (d - C)×sinθ, then y ∈ [y1, y2] is the height range where the water surface is located.

[0030] In S3, the DBSCAN clustering algorithm is used to classify the point cloud data to obtain multiple different data clusters. As a density clustering algorithm, the DBSCAN clustering algorithm can automatically identify clustering clusters according to the spatial distribution density of data points by defining the concepts of "core points" and "density reachable", and regard points in low-density areas (such as interference points) as noise points and eliminate them, so as to retain high-density effective data clusters. The DBSCAN clustering algorithm does not need to specify the number of clusters in advance and can automatically determine the number of clusters according to the data distribution, adapting to the point cloud data classification requirements in different environments and improving the versatility and robustness of the algorithm. In this application, the data clusters obtained through clustering include embankment clusters, floating object clusters, opposite bank clusters, etc. Among them, the embankment cluster contains the point cloud data of the embankment; the floating object cluster contains the point cloud data of floating objects on the water surface; the opposite bank cluster includes the point cloud data of the distant river bank or obstacles, which are far from the position of the single-line lidar.

[0031] The method for screening the target data cluster is as follows: judge whether the y value of the data points in each data cluster is located in [y1, y2], if not, then eliminate the data cluster; compare the distances from the lowest points of the remaining data clusters to the origin, and the data cluster with the smallest distance value is the target data cluster containing the water-edge intersection point.

[0032] In practical applications, the single-line lidar is fixedly installed on the shore base and vertically scans the cross-section of the water surface and the embankment. As a solid object, the embankment appears as a structure higher than the water surface in the scanned cross-section, and the junction of its bottom and the water surface is the water-shore junction. In the scan, the water surface may have no echo due to the laser passing through the water, or only produce a weak echo when there are impurities on the water surface. The embankment cluster is located in the near-shore area in front of the single-line lidar. For the same scanned cross-section, the distance between the single-line lidar and the embankment is necessarily less than the distance between the single-line lidar and the opposite shore or distant obstacles. Floating objects are located on the water surface. Although their height is within the height range of the water surface, the horizontal distance between the floating objects and the single-line lidar is farther. Therefore, the distance from the lowest point of the floating object cluster to the single-line lidar is also greater than the distance from the lowest point in the embankment cluster to the single-line lidar. The logic of the above target data cluster screening essentially utilizes the geometric uniqueness of the embankment as the near-shore solid boundary to quickly locate the data cluster where the water-shore junction point is located through distance comparison.

[0033] To ensure the accuracy of the measurement by the single-line lidar, after the single-line lidar is fixedly installed on-site, its horizontal angle is calibrated. The specific calibration method is as follows: On the water surface in the scanned cross-section of the lidar, lay two floating objects at a certain distance apart, and obtain the coordinate data points of the two floating objects in the radar image. Calculate the included angle between the two data points. The included angle is calculated through the arctangent function, and the obtained angle result is the angle offset of the single-line lidar at the current installation site. Substitute this angle into the coordinate system of the single-line lidar for correction. If the single-line lidar undergoes displacement, the horizontal angle needs to be calibrated again.

[0034] In S4, the format of each data point in the target data cluster is as follows: class Point{ int index; / / Point number double angle; / / Angle in the polar coordinate system of the point double dis; / / Distance in the polar coordinate system of the point double x; / / x in the Cartesian coordinate system of the point double y; / / y in the Cartesian coordinate system of the point double rssi; / / Echo rate of the point } Extract the data points with y values within the water surface height range in the data cluster, and sort them according to the index point number. This sequence is the sequence number of each laser emission by the single-line lidar. The calculation formula for the relative echo slope of each data point is:

[0035] The first half of the formula is the ratio of the maximum echo rate to the minimum echo rate between the calculated data point and several adjacent data points, which reflects the relative strength comparison of the signal. The second half of the formula is the absolute value of the index difference between the data points corresponding to the maximum echo rate and the data points corresponding to the minimum echo rate among the calculated data point and several adjacent data points, which normalizes the proportional change to the unit index interval, and finally obtains the relative change rate within the unit index interval.

[0036] The relative echo slope formula adopts the form of "maximum echo rate / minimum echo rate÷index interval", breaking through the traditional "difference - slope" calculation mode, and instead using the relative proportional characteristics of the signal, solving the problem of unstable absolute value of the lidar echo rate, and realizing device - independent and environment - robust water - shore boundary detection.

[0037] Obtain the first occurrence according to the index order When the maximum value is reached, it is the water - shore intersection point. Based on the water - shore intersection point, the water level height can be determined. Its core principle is that: the difference in the reflection characteristics of the laser at the junction of the solid bank and the liquid water surface will cause a sudden change in the echo rate, and the slope calculation can quantify this sudden change and locate it to specific data points. When the laser hits the bank, effective reflection occurs, generating a relatively high echo rate. The bank surface is relatively rough or has a stable structure, so the echo rates of the bank data points within the same cluster are usually stable and relatively high. In a clear - water scenario, the laser may penetrate the water surface (water - penetration phenomenon), and there is no reflector or weak reflection underwater, resulting in a significant decrease in the echo rate (the rssi value approaches 0 or is close to the noise level); in a water surface scenario with impurities, the turbid water may produce a small amount of reflection, but the echo rate is still lower than that of the bank and is unstable with changes in water quality. In the transition area from the bank to the water surface, the laser reflection characteristics change suddenly from "high reflection" to "low reflection", resulting in a step - like decrease in the echo rate, forming a point with the largest echo rate gradient, that is, the water - shore intersection point.

[0038] It should be noted that the above - mentioned calibration of the maximum water - penetration distance in clear water, calibration of the horizontal angle at the installation site, classification of each data point in the radar map to obtain data clusters, confirmation of the height range where the water surface is located, and calculation of the cluster where the water - shore intersection point is located can be combined or used separately.

[0039] To ensure the accuracy of the output water-edge intersection point data, multi-level sliding filtering is performed on the water-edge intersection point data. The method of multi-level sliding filtering is as follows: the y value of the output water-edge intersection point is input into a 1-minute window, which only stores the data within 1 minute, and the average value in this window is input into a 5-minute window. Similarly, the average value output by the 5-minute window is the water level value. Through filtering processing, the fluctuations in water level data caused by the system errors of the single-line lidar itself (such as environmental interference and equipment fluctuations) are reduced, and a smooth and stable water level value is output, improving the data reliability. The filtering method in this embodiment is a simple window average calculation, which does not require complex algorithms or dedicated hardware, has a small calculation amount, adapts to the low-power consumption characteristics of the single-line lidar, and avoids increasing the equipment energy consumption or cost due to high computing power requirements.

[0040] Through the analysis of the echo rate intensity of the data points near the water-edge intersection, the present invention can remove the incorrect data points obtained by the laser passing through water, further improving the accuracy of water level detection. The present invention obtains the angle between the 0-degree angle of laser scanning and the horizontal plane through pre-calibration to solve the problem of inaccurate data caused by the inclination of the lidar image. Through the clustering algorithm, the present invention clusters the data points obtained near the water-edge intersection, and at the same time can filter out the unnecessary data points such as floating debris on the water surface and the opposite bank. The present invention analyzes the data points near the water-edge intersection, and obtains the actual position of the water-edge intersection through the change in the echo rate between each data point and the echo rate of the adjacent points. Further, the present invention performs multiple filtering on the obtained point data to reduce the data jitter caused by the system errors of the laser scanning lidar itself and output smooth water level data.

[0041] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A water level monitoring method, characterized in that, include: S1, use a single-line laser radar to scan the target area and obtain point cloud data; S2. Determine the water surface height range according to the lowest point in the point cloud data and the maximum water penetration distance of the single-line laser radar in clear water; S3, clustering the point cloud data, and selecting data clusters within the water surface height range, and selecting from these data clusters the data cluster with the shortest distance from the lowest point to the single-line laser radar as the target data cluster; S4. Calculate the relative echo slope of each point in the target data cluster, and determine the point with the largest value as the water-shore junction point according to the relative echo slope of each point; and determine the water level value according to the water-shore junction point.

2. The water level monitoring method according to claim 1, characterized in that, The method also includes the step of performing multi-level sliding filtering on the water-bank junction point data.

3. The water level monitoring method according to claim 2, characterized in that, The method of multi-stage sliding filtering is: input the y value of the output water-shore junction into a 1-minute window, and input the average value in the window into a 5-minute window. Similarly, the average value output by the 5-minute window is the water level value.

4. A water level monitoring method according to claim 1, characterized in that, In S2, the calibration method of the maximum water penetration distance is: In the laboratory, the single-line laser radar used was fixed at a set distance from the water tank without water, and a point at a certain angle was determined as a reference point; Gradually add water to the water tank, and stop adding water when the above reference point disappears; Place a floating object on the water surface to obtain a reference point at the moment; The distance difference between the two reference points and the single-line laser radar is calculated to obtain the maximum water penetration distance of the single-line laser radar.

5. A water level monitoring method according to claim 4, characterized in that The method for determining the water surface height range is: find the lowest point in the point cloud data, record the coordinates of the lowest point as (x1, y1), the distance from the lowest point to the origin as d, the angle as θ, and the maximum water penetration distance as C; the origin is the location of the single-line laser radar, calculate the Cartesian coordinate system (x2, y2) corresponding to the polar coordinate system (dC, θ), then y∈[y1, y2] is the height range of the water surface.

6. The water level monitoring method according to claim 5, characterized in that, Determine whether the y value of the data point in each data cluster is located in [y1, y2]. If not, the data cluster is eliminated; compare the distance from the lowest point to the origin in the remaining data clusters, and the data cluster with the smallest distance value is the target data cluster containing the water-shore junction point.

7. A water level monitoring method according to claim 6, characterized in that, In S4, the data points in the target data cluster are sorted according to the index point number. The sequence is the order number of each laser sent by the single-line laser radar. According to the formula Calculate the relative echo slope of each data point, and the data point with the maximum relative echo slope is the water-edge intersection point; in the formula, is the echo rate of the data point, is the relative echo slope of the data point.

8. A water level monitoring method according to claim 1, characterized in that The point cloud data is clustered using the DBSCAN clustering algorithm.

9. A water level monitoring method according to claim 1, characterized in that, Before acquiring point cloud data, the single-line laser radar is calibrated at the horizontal angle.

10. A water level monitoring method according to claim 8, characterized in that, The method for horizontal angle calibration is: on the water surface of the section scanned by the single-line laser radar, two suspended objects separated by a set distance are laid, the data points of the two suspended objects are obtained in the radar image, and the angle between the two data points is calculated. This angle is the angle offset of the single-line laser radar, and this angle is brought into the coordinate system of the single-line laser radar for correction.

Images