Slope risk monitoring method and system based on three-dimensional laser and two-dimensional phase scanning radar

Abstract

The application discloses a kind of based on three-dimensional laser and two-dimensional phase scanning radar's side slope risk monitoring method and system, method includes steps: S1, the digital elevation model of side slope is obtained by three-dimensional laser radar, and the radar imaging result of side slope is obtained by two-dimensional phase scanning radar;S2, the digital elevation model is with radar imaging result Coordinate system is unified and projection conversion, and conversion relationship lookup table is generated;S3, based on digital elevation model, radar imaging result and conversion relationship lookup table, obtain target real deformation value;S4, to target real deformation value carries out time series analysis, judges side slope state, and when identifying acceleration deformation, calculates tangent angle to determine early warning level;S5, according to early warning level automatically switches the working mode of two-dimensional phase scanning radar;S6, in moving target monitoring mode, moving target is tracked and predicted, carries out collision judgment and target re-identification.The application has the advantages of high precision and intelligentization.

Description

Technical Field

[0001] This invention mainly relates to the field of slope monitoring technology, specifically to a slope risk monitoring method and system based on three-dimensional laser and two-dimensional phase-scanning radar. Background Technology

[0002] A slope refers to a geological body composed of natural soil or rock with an inclined surface, as well as a sloping surface formed by artificial excavation and filling during engineering construction. Unstable slopes pose a significant threat to human life and property safety, and seriously affect the stability and development of society and the economy. Therefore, automated monitoring of potentially hazardous slopes has important theoretical significance and practical value.

[0003] Early slope deformation monitoring relied primarily on leveling and global navigation satellite systems, which were not only inaccurate but also limited to single-point deformation monitoring, consuming significant manpower and resources. With increasing monitoring demands and technological advancements, radar technology has been gradually applied to slope deformation monitoring, making substantial contributions to successful landslide early warning on numerous occasions. Radar differential interferometry extracts deformation phase information from the target point phase, obtaining deformation values ​​with millimeter-level or even sub-millimeter-level accuracy. Spaceborne and airborne radars, due to their long revisit cycles and meter-level resolution, are suitable for early, large-scale surveys, while ground-based radars, with decimeter-level resolution and refresh rates down to the second level, are more suitable for long-term monitoring of potentially hazardous slopes. Lidar, with its high-density three-dimensional point cloud data, can effectively capture subtle cracks and ridges caused by early landslides. Through multi-period data registration and comparison, surface displacement with millimeter-level accuracy can be obtained.

[0004] While differential interferometry can obtain deformation measurements with sub-millimeter precision, it is generally limited to stable, strongly reflective objects such as buildings, trihedral, and dihedral objects. Furthermore, when radar resolution does not satisfy the assumptions of the Nyquist sampling theorem, the calculated deformation values ​​still exhibit entanglement deformation, affecting subsequent deformation-based analysis and early warning prediction. Although lidar offers high precision and high resolution, its resolution decreases with increasing distance and it is susceptible to extreme weather conditions, severely limiting its widespread application.

[0005] With the deepening research on landslide monitoring technology, some experts and scholars have proposed a combined landslide monitoring scheme using lidar and differential interferometric radar. For example, in spaceborne synthetic aperture radar, removing the terrain phase and using the digital elevation model provided by lidar can estimate the terrain phase more accurately than SRTM (a typical external digital elevation model library). Furthermore, some scholars have pointed out in their papers that directly using coordinate transformation to fuse the results of ground-based radar and 3D lidar can lead to deviations in slant range projection calculations. It is necessary to consider the radar's attitude and tilt angle during actual data acquisition, thereby improving the accuracy of coordinate transformation. Other patents indicate the use of 3D lidar and ground-based radar to invert 3D deformation, thus achieving rapid acquisition of high-precision 3D displacement fields.

[0006] Ground-based radar differential interferometry includes steps such as interferometry, filtering, unwrapping, atmospheric correction, and deformation inversion. Unwrapping requires retrieving the wrapped phase (which differs from the true value). The sampling rate (integer multiples of the actual phase) is converted to the true phase, which assumes that the sampling rate satisfies the Nyquist sampling theorem, thus avoiding multiple solutions and gradient direction errors, and obtaining the correct surface deformation. However, when the landslide undergoes severe deformation, the theorem assumption is broken, and the unwrapping results in some strong reflection areas also become unreliable.

[0007] Even though some radars on the market have integrated differential interferometry and moving target detection technologies, they cannot achieve intelligent automatic switching. Installing two radars will not only cause co-channel interference, but also waste resources. Switching between the two modes at a fixed frequency using time-division multiplexing technology will result in some resource waste. Summary of the Invention

[0008] To address the technical problems existing in the prior art, this invention provides a high-precision, intelligent slope risk monitoring method and system based on three-dimensional laser and two-dimensional phase-scanning radar.

[0009] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows:

[0010] A slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar includes the following steps:

[0011] S1. Obtain the digital elevation model of the slope using a three-dimensional lidar and obtain the radar imaging results of the slope using a two-dimensional phase-scanning radar.

[0012] S2. Perform coordinate system-1 and projection transformation on the digital elevation model and the radar imaging results to generate a transformation relationship lookup table;

[0013] S3. Based on the digital elevation model, the radar imaging results, and the transformation relationship lookup table, obtain the true deformation value of the target;

[0014] S4. Perform time-series analysis on the target's actual deformation value to determine the slope condition, and calculate the tangent angle when accelerated deformation is identified to determine the warning level.

[0015] S5. Automatically switch the working mode of the two-dimensional phase-scanning radar according to the warning level to realize time-division multiplexing of deformation monitoring and moving target monitoring;

[0016] S6. In the moving target monitoring mode, the moving target is tracked and predicted, and collision judgment and target re-identification are performed in combination with the high-precision digital elevation model.

[0017] Preferably, the specific process of step S3 is as follows:

[0018] S3.1 Extracting the displacement values ​​of 3D LiDAR point cloud through multi-period point cloud data registration;

[0019] S3.2 Based on radar imaging results, multiple differential interferograms are obtained using differential interferometry.

[0020] S3.3 Using the three-dimensional displacement value as prior information, the three-dimensional phase is unwrapped using the minimum cost flow method to recover the true phase;

[0021] S3.4 Perform atmospheric phase correction and deformation inversion to obtain the true deformation value of the slope and integrate it into the digital elevation model.

[0022] Preferably, the specific process of step S3.3 is as follows:

[0023] S3.3.1. Perform time-dimension unwrapping on the permanent scattering points in the differential interferogram;

[0024] S3.3.2 Based on the transformation relationship lookup table, the displacement values ​​of the three-dimensional lidar point cloud are projected onto the radar line of sight, and then spatial filtering is used to operate on the point cloud data in the two-dimensional phase-scanning radar pixel unit to obtain the prior deformation value and prior confidence.

[0025] S3.3.3. Convert the prior deformation value into phase using the radar wavelength, and then obtain the prior entanglement value;

[0026] S3.3.4. Set virtual source and virtual sink points for pixels with known prior entanglement values, and set cost weights based on prior confidence levels;

[0027] S3.3.5. Use the minimum cost flow method to solve for the global entanglement value and complete the phase untangling.

[0028] Preferably, the formula for obtaining the prior winding value in step S3.3.3 is:

[0029]

[0030] in, This is the prior entanglement value; The phase is the phase after the a priori deformation value is converted.

[0031] Preferably, the specific process of step S4 is as follows:

[0032] S4.1 Analyze the actual deformation value of the target, identify pixels whose deformation exceeds the preset threshold and cluster them to form potential risk areas;

[0033] S4.2 Filter the deformation and velocity values ​​of the identified risk areas; in the spatial dimension, obtain the displacement curve of the regional deformation; in the time dimension, obtain the filtered displacement and velocity curves.

[0034] S4.3 Identify the time node that begins acceleration in the filtered displacement curve; identify the time node that begins acceleration in the velocity curve; and take the average of the time nodes that begin acceleration in the displacement curve and the identified time nodes that begin acceleration in the velocity curve to obtain the average value of the time nodes that begin acceleration.

[0035] S4.4. Based on the displacement and time corresponding to the average time node when acceleration begins, and the initial displacement and time when the slope begins to deform, calculate the velocity value of the constant velocity deformation stage.

[0036] S4.5. Based on the velocity value during the constant velocity deformation stage and the current velocity value of the risk area, calculate the tangent angle value at the current moment, and compare the tangent angle value with the preset warning rule threshold to determine the current warning level of the risk area.

[0037] Preferably, the formula for calculating the velocity value during the constant velocity deformation stage in step S4.4 is as follows:

[0038]

[0039] in, The speed value during the constant speed change phase; These represent the displacement and time corresponding to the average time point at which acceleration begins; This represents the initial displacement and time at which the slope begins to deform.

[0040] Preferably, the specific formula for calculating the tangent angle value at the current moment in step S4.5 is as follows:

[0041]

[0042] in, The tangent angle value. This represents the current speed value for the risk area.

[0043] Preferably, the preset early warning rules are as follows:

[0044] when When the angle is ≤45°, the warning level is alert level and the alarm type is blue.

[0045] When 45° < When the temperature is below 80°, the warning level is alert, and the alarm type is yellow.

[0046] When 80°≤ When the temperature is below 85°, the warning level is alert level and the alarm type is orange.

[0047] When 85°≤ At that time, the warning level is alarm level, and the alarm type is red.

[0048] Preferably, the specific process of step S5 is as follows:

[0049] When the risk area is under orange alert and there is no red alert area, insert a moving target monitoring between two deformation monitoring;

[0050] When the risk area is under red alert, moving target monitoring is continuously performed between two deformation monitoring sessions, and when the speed of the moving target is detected to exceed the warning value, the process switches to step S6.

[0051] Preferably, step S6 specifically includes:

[0052] S6.1 Identify moving targets and filter false alarm targets by combining slope information;

[0053] S6.2. Based on the current speed, position information, and slope and aspect information of the moving target, predict the target's position at the next moment, and perform a collision judgment to obtain the collision judgment result;

[0054] S6.3. Based on the collision judgment results, the scanning mode of the two-dimensional phase-scanning radar is adaptively adjusted.

[0055] Preferably, the specific process of step S6.1 is as follows:

[0056] The two-dimensional phase-scanning radar performs full-scene imaging in moving target monitoring mode, identifies moving targets in the image, and obtains distance and angle information;

[0057] Subsequently, the moving target is located in the real three-dimensional spatial coordinate system centered on the radar according to the transformation relationship lookup table generated in step S2; and the confidence level is judged by using the slope information pre-acquired by the three-dimensional lidar; if the slope of the area where the target is located is lower than the preset slope threshold, it is regarded as a false alarm and filtered out; if no moving target is detected in the scene for a preset number of consecutive times, it is determined that the target movement has stopped.

[0058] Preferably, the specific process of step S6.2 is as follows:

[0059] Based on the current velocity and position information of the moving target, and combined with the slope and aspect information provided by the high-precision digital elevation model, the distance and azimuth of the target at the next phase-scan radar acquisition time are estimated using the integral method and motion model; simultaneously, collision detection is performed on the predicted trajectory.

[0060] If the slope change in the predicted trajectory is constant or monotonically increasing, a parabolic model is used to calculate the landing point and determine whether the target reaches the landing point. If it does, it is a collision; otherwise, it is not a collision.

[0061] If the slope change in the predicted trajectory reaches a minimum or decreases monotonically, it is determined that a collision will occur.

[0062] Preferably, the parabolic model is:

[0063]

[0064] Where m is the longitudinal displacement value and n is the horizontal displacement value. Let g represent the horizontal and vertical velocities, and g represent the acceleration due to gravity.

[0065] Preferably, the coefficient of restitution is used to quantify the change in velocity and energy loss before and after the collision, and the calculation formula is as follows:

[0066]

[0067] in, These are the normal and tangential velocity components before the collision. These are the normal and tangential velocity components after the collision. These are the normal and tangential restitution coefficients.

[0068] Preferably, the specific process of step S6.3 is as follows:

[0069] If it is determined that no collision has occurred, the radar is switched to narrow beam scanning mode to enhance the radar scanning energy, and the beam is pointed to the predicted position and azimuth for precise tracking; if the target can continue to be detected, the process jumps to step S6.2 to predict the next moment.

[0070] If a collision or target loss is detected, the radar switches back to full-scene scanning mode and returns to step S6.1 to re-perform moving target detection and identification.

[0071] The present invention also discloses a slope risk monitoring system based on three-dimensional laser and two-dimensional phase-scanning radar, including a memory and a processor connected to each other. The memory stores a computer program, which executes the steps of the method described above when run by the processor.

[0072] Compared with the prior art, the advantages of the present invention are as follows:

[0073] This invention fully considers the possibility of drastic displacement during slope deformation monitoring. By using a high-precision coordinate fusion and transformation relationship lookup table, the three-dimensional deformation prior information acquired by the lidar is introduced into the radar phase unwrapping process, which effectively solves the unwrapping error when the Nyquist sampling theorem is not satisfied, and significantly improves the accuracy and reliability of deformation inversion.

[0074] Secondly, the tangent angle warning level obtained from deformation time sequence analysis is innovatively linked with the radar working mechanism, realizing the autonomous switching and time-division multiplexing of the two-dimensional phase-scanning radar between deformation monitoring and moving target monitoring modes based on risk level. This solves the problems of single function and rigid mode switching of traditional equipment, and realizes unmanned intelligent monitoring of the entire time and process of "before-middle-after landslide".

[0075] Finally, during the landslide movement, the digital elevation model and prediction algorithm were used to predict the three-dimensional trajectory and determine the collision of the moving target, and the radar was guided to switch to a narrow beam for precise tracking. This greatly improved the ability to capture, continuously track and warn of fast-moving targets, and provided critical decision-making time for disaster prevention and mitigation. Attached Figure Description

[0076] Figure 1 This is a flowchart of an embodiment of the slope risk monitoring method of the present invention.

[0077] Figure 2 This is a schematic diagram of the tangent angle warning rule in this invention. Detailed Implementation

[0078] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0079] like Figure 1 As shown in the figure, the slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar provided by the embodiments of the present invention includes the following steps:

[0080] S1. Obtain a digital elevation model using a 3D lidar and acquire radar imaging results of the scene using a 2D phase-scanning radar.

[0081] The raw data acquired by the 3D LiDAR is used to generate a high-precision digital elevation model through steps such as position calculation, point cloud denoising and registration, automatic classification, interpolation and resampling;

[0082] Two-dimensional phase-scanning radar generates two-dimensional radar images through steps such as pulse compression, channel correction, and angle estimation. The range resolution of the radar image is [missing information]. Where C is the speed of light and B represents the pulse width of the transmitted signal. The azimuth resolution of the radar image is related to the number of equivalent antennas and can be adjusted according to the set angle difference during actual imaging. After imaging, the entire image consists of multiple pixel cells whose range resolution is multiplied by the set angle difference. Each pixel can be represented by the following formula:

[0083]

[0084] Where Signal represents the imaging result of each pixel; A is the target amplitude, which reflects the target backscattering coefficient; It is the imaginary unit; The target phase is the phase corresponding to the distance, and its range is... .

[0085] S2. Fuse the three-dimensional lidar digital elevation model with the two-dimensional phase-scanning radar imaging results to generate a conversion relationship lookup table;

[0086] S2.1 Transform the coordinates of the 3D laser point cloud to the radar's 3D spatial coordinate system with the radar center as the origin, unifying the coordinate origins of the 2D phase-scanning radar and the 3D laser radar. In the radar's 3D spatial coordinate system, the X-axis represents the azimuth direction of the 2D phase-scanning radar, the Z-axis represents the vertical direction, and the Y-axis is perpendicular to the XOZ plane.

[0087] S2.2, Considering the elevation angle of a two-dimensional phase-scanned radar The projection transformation relationship under the given conditions projects the 3D laser point cloud data onto the 2D phase-scanned radar. Specifically, the point cloud coordinate transformation is performed using the following transformation relationship:

[0088]

[0089] in, The coordinates of the point cloud after unifying the origin. The point cloud is located in the three-dimensional body coordinate system of the two-dimensional phased radar. The X-axis of the three-dimensional body coordinate system is the azimuth direction of the two-dimensional phased radar, the Y-axis is the beam pointing direction of the two-dimensional phased radar, the Z-axis is perpendicular to the XOY plane, and D is the distance of the point cloud coordinates from the origin of the three-dimensional spatial coordinate system. The plane in which the radar beam is pointed is the radar elevation angle, and the angle between this plane and the ground plane is the radar elevation angle.

[0090] Then, use the projection transformation formula to transform the point cloud coordinates to radar coordinates. The transformation formula is as follows:

[0091]

[0092] in, The coordinates of the point cloud are in the two-dimensional body coordinate system of the two-dimensional phased radar. Specifically, the X-axis of the radar's two-dimensional body coordinate system is the azimuth direction of the two-dimensional phased radar, and the Y-axis is the beam pointing direction of the two-dimensional phased radar. The radar's two-dimensional body coordinate system blurs the height information.

[0093] S2.3 Generate a transformation relationship lookup table based on the transformation relationship for easy use in subsequent operations;

[0094] Based on the above transformation, a mapping relationship is established between points in the laser point cloud data and pixel units in the two-dimensional radar image, generating a transformation relationship lookup table. This lookup table stores the radar image pixel position corresponding to each three-dimensional point, which is used to subsequently associate and fuse the three-dimensional laser data and the two-dimensional radar data.

[0095] S3. Based on the digital elevation model, radar imaging results, and transformation relationship lookup table, obtain the true deformation value of the target;

[0096] S3.1. Register multi-period point cloud data using the iterative nearest point algorithm, and extract the displacement values ​​of the 3D LiDAR point cloud using the multi-scale model-to-model algorithm;

[0097] S3.2, Two-dimensional phase-scanning radar obtains multiple differential interferograms based on radar images through permanent scatterer (PS) point selection, differential interferometry, and phase filtering steps;

[0098] S3.3 Using the three-dimensional displacement value as prior information, the three-dimensional phase is unwrapped using the minimum cost flow method to recover the true phase of the target;

[0099] S3.3.1. Use a one-dimensional unwrapping algorithm to unwrap the PS points in the differential interferogram in the time dimension;

[0100] S3.3.2 Based on the transformation relationship lookup table, the point cloud displacement value is projected onto the radar line of sight direction by the angle between the point cloud displacement and the radar line of sight direction. Then, spatial filtering is used to operate on the point cloud data in the two-dimensional phase-scanned radar pixel unit to obtain the prior deformation value and prior confidence.

[0101] S3.3.3. The prior deformation value is converted into phase using the radar wavelength, and then the prior winding value is obtained. The calculation formula is as follows:

[0102]

[0103] in, The phase after the transformation of the a priori deformation value. This is the prior entanglement value;

[0104] S3.3.4. Add virtual source and virtual sink points to pixels with known prior entanglement values, and set cost weights based on prior confidence. In a practical processing case, pixels are represented as nodes, edges are connections between adjacent pixels, positive and negative residuals are set as source and sink points, and virtual source point Sp and virtual sink point Tp are set for pixels p with known prior entanglement values ​​in the image. The prior entanglement value is the flow rate, and the cost of Sp flowing to p or the cost of p flowing to Tp is set as the prior confidence.

[0105] S3.3.5. Use the minimum cost flow method to solve for the global entanglement value and complete the phase untangling.

[0106] S3.4. By using atmospheric phase correction and deformation inversion, the true deformation value of the slope monitored by radar is obtained. Then, based on the coordinate transformation relationship and deformation projection relationship, the true deformation value is integrated into the digital elevation model, making the results simpler and easier to understand.

[0107] S4. Perform time-series analysis on the actual deformation value of the target, use the early warning algorithm to judge the slope status, and calculate the tangent angle when accelerated deformation is detected to determine the early warning level.

[0108] S4.1 Analyze the actual deformation value of the target, identify pixels whose deformation exceeds the preset threshold and cluster them to form potential risk areas;

[0109] S4.2 Filter the deformation and velocity values ​​of the risk area; In the spatial dimension, use the pixel with the maximum deformation value and its surrounding pixels to represent the deformation of the area, and obtain the regional deformation displacement curve through the spatial averaging algorithm; In the time dimension, perform interpolation, moving average method and staggered subtraction method on the displacement curve to obtain the filtered displacement curve and velocity curve.

[0110] S4.3 Identify the acceleration time point of the landslide. Use wavelet transform to identify the acceleration time point in the filtered displacement curve; identify the acceleration time point of the velocity curve based on the normal distribution characteristics of the constant velocity stage; if both have values, take the average value; if neither has a result, it indicates that the slope has not started to accelerate and is in a safe stage, at which point it is necessary to jump to step S1.

[0111] S4.4 Calculate the velocity during the constant velocity deformation stage based on the displacement value corresponding to the acceleration start time node and the time and displacement value at the start of deformation. The calculation formula is as follows:

[0112]

[0113] in, The speed during the constant-speed transition phase, These represent the displacement and time at which the deformation of the risk area begins to accelerate, respectively. This represents the displacement and time at which the slope begins to deform;

[0114] S4.5. Based on the velocity during the constant velocity deformation stage and the current velocity value in the risk area, the tangent angle value of the deformation region at the current moment is calculated using the tangent angle model. The calculation formula is as follows:

[0115]

[0116] in, The tangent angle value. This represents the current speed value for the region.

[0117] Combine the tangent angle value with Figure 2 The system compares the data with the preset warning rule thresholds shown. If the area is in a blue or yellow warning stage, it proceeds to step S1; if it is in an orange or red warning stage, it proceeds to step S5.

[0118] Specifically, when When the angle is ≤45°, the warning level is alert, and the alarm type is blue; when the angle is <45°, the warning level is alert, and the alarm type is blue. When the temperature is less than 80°, the warning level is alert, and the alarm type is yellow; when the temperature is less than or equal to 80°, the warning level is warning, and the alarm type is yellow. When the temperature is <85°, the warning level is alert, and the alarm type is orange; when the temperature is ≤85°, the warning level is warning. At that time, the warning level is alarm level, and the alarm type is red.

[0119] S5. Automatically switch the working mode of the two-dimensional phase-scanning radar according to the warning level to realize time-division multiplexing of deformation monitoring and moving target monitoring;

[0120] By using time-division multiplexing technology, the radar has both deformation monitoring capabilities and all-scenario moving target monitoring capabilities. The radar switches its operating mode by switching the sweep frequency slope and the radar pulse time interval.

[0121] The specific strategy is as follows: when the risk area is under orange alert and there is no red alert area, insert a moving target monitoring between two deformation monitoring (the area status will still be judged between the two deformation monitoring), and use moving target monitoring to correct the shortcomings of the early warning model;

[0122] When the risk area is under red alert, moving target monitoring is continuously performed between two deformation monitoring. The radar is prepared to switch modes as soon as a landslide occurs. When the speed of the moving target exceeds the warning value, it jumps to step S6.

[0123] Step S6: In moving target monitoring mode, track and predict moving targets, and combine a high-precision digital elevation model to perform collision detection and target re-identification. Specifically:

[0124] S6.1 The two-dimensional phase-scanning radar performs full-scene imaging in moving target monitoring mode. It identifies moving targets in the image based on clustering and detection algorithms, and acquires their distance and angle information. Then, it locates the moving target in the real three-dimensional space centered on the radar using a pre-established transformation lookup table. It then uses the slope information pre-acquired by the three-dimensional lidar to determine the confidence level. If the slope of the target's area is lower than a preset slope threshold, it is considered a false alarm and filtered out. If no moving target is detected in the scene for a preset number of consecutive times, the movement is stopped and the process returns to step S1. For example, a slope lower than 10 degrees can be considered a false alarm, and the confidence level is set to 0, meaning the target is not monitored. If no moving target is detected in the scene three times consecutively, the process jumps to step S1.

[0125] S6.2 Based on the current velocity and position information of the moving target, and combined with the slope and aspect information provided by the high-precision digital elevation model, the distance and azimuth of the target at the next phase-scan radar acquisition time are estimated using the integral method and motion model; simultaneously, collision detection is performed on the predicted trajectory.

[0126] If the slope change in the predicted trajectory is constant or monotonically increasing, a parabolic model is used to calculate the landing point. The system then determines whether the target reaches the landing point; if it does, a collision occurs; otherwise, no collision occurs. The formula for the parabolic trajectory is as follows:

[0127]

[0128] Where m is the longitudinal displacement value and n is the horizontal displacement value. Let g represent the horizontal and vertical velocities, and g represent the acceleration due to gravity.

[0129] If the predicted trajectory shows a minimum or monotonically decreasing slope, a collision is predicted. After a collision, the coefficient of restitution is used to quantify the changes in velocity and energy loss before and after the collision. The calculation formula is as follows:

[0130]

[0131] in, These are the normal and tangential velocity components before the collision. These are the normal and tangential velocity components after the collision. These are the normal and tangential restitution coefficients. The collision restitution coefficient can be calculated from the model.

[0132] Step S6.3: Based on the prediction results of step S6.2, adaptively adjust the scanning mode of the two-dimensional phase-scanning radar: if it is determined that no collision has occurred, switch the radar to narrow beam scanning mode to enhance the radar scanning energy, and point the beam to the predicted position and azimuth for accurate tracking. If the target can continue to be detected, jump to step S6.2 to make predictions for the next moment; if it is determined that a collision has occurred or the target has been lost, switch the radar back to full-scene scanning mode and return to step S6.1 to re-detect and identify moving targets.

[0133] This invention fully considers the possibility of drastic displacement during slope deformation monitoring. By using a high-precision coordinate fusion and transformation relationship lookup table, the three-dimensional deformation prior information acquired by the lidar is introduced into the radar phase unwrapping process, which effectively solves the unwrapping error when the Nyquist sampling theorem is not satisfied, and significantly improves the accuracy and reliability of deformation inversion.

[0134] Secondly, the tangent angle warning level obtained from deformation time sequence analysis is innovatively linked with the radar working mechanism, realizing the autonomous switching and time-division multiplexing of the two-dimensional phase-scanning radar between deformation monitoring and moving target monitoring modes based on risk level. This solves the problems of single function and rigid mode switching of traditional equipment, and realizes unmanned intelligent monitoring of the entire time and process of "before-middle-after landslide".

[0135] Finally, during the landslide movement, the digital elevation model and prediction algorithm were used to predict the three-dimensional trajectory and determine the collision of the moving target, and the radar was guided to switch to a narrow beam for precise tracking. This greatly improved the ability to capture, continuously track and warn of fast-moving targets, and provided critical decision-making time for disaster prevention and mitigation.

[0136] This invention also discloses a slope risk monitoring system based on three-dimensional laser and two-dimensional phase-scanning radar, including an interconnected memory and a processor. The memory stores a computer program, which, when run by the processor, executes the steps of the method described above. The slope risk monitoring system of this invention corresponds to the monitoring method described above and also possesses the advantages described therein.

[0137] The present invention can implement all or part of the processes in the methods of the above embodiments, or it can be implemented by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above method embodiments. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable storage medium includes: any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. The memory is used to store computer programs and / or modules. The processor implements various functions by running or executing the computer programs and / or modules stored in the memory, and by calling data stored in the memory. The memory may include high-speed random access memory, as well as non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0138] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar, characterized in that, Including the following steps: S1. Obtain the digital elevation model of the slope using a three-dimensional lidar and obtain the radar imaging results of the slope using a two-dimensional phase-scanning radar. S2. Perform coordinate system-1 and projection transformation on the digital elevation model and the radar imaging results to generate a transformation relationship lookup table; S3. Based on the digital elevation model, the radar imaging results, and the transformation relationship lookup table, obtain the true deformation value of the target; S4. Perform time-series analysis on the target's actual deformation value to determine the slope condition, and calculate the tangent angle when accelerated deformation is identified to determine the warning level. S5. Automatically switch the working mode of the two-dimensional phase-scanning radar according to the warning level to realize time-division multiplexing of deformation monitoring and moving target monitoring; S6. In the moving target monitoring mode, the moving target is tracked and predicted, and collision judgment and target re-identification are performed in combination with the digital elevation model. The specific process of step S4 is as follows: S4.1 Analyze the actual deformation value of the target, identify pixels whose deformation exceeds the preset threshold and cluster them to form potential risk areas; S4.2 Filter the deformation and velocity values ​​of the identified risk areas; in the spatial dimension, obtain the displacement curve of the regional deformation; in the time dimension, obtain the filtered displacement and velocity curves. S4.3 Identify the time node that begins acceleration in the filtered displacement curve; identify the time node that begins acceleration in the velocity curve; and take the average of the time nodes that begin acceleration in the displacement curve and the identified time nodes that begin acceleration in the velocity curve to obtain the average value of the time nodes that begin acceleration. S4.

4. Based on the displacement and time corresponding to the average time node when acceleration begins, and the initial displacement and time when the slope begins to deform, calculate the velocity value of the constant velocity deformation stage. S4.

5. Based on the velocity value during the constant velocity deformation stage and the current velocity value of the risk area, calculate the tangent angle value at the current moment, and compare the tangent angle value with the preset warning rule threshold to determine the current warning level of the risk area.

2. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 1, characterized in that, The specific process of step S3 is as follows: S3.1 Extracting the displacement values ​​of 3D LiDAR point cloud through multi-period point cloud data registration; S3.2 Based on radar imaging results, multiple differential interferograms are obtained using differential interferometry. S3.3 Using the three-dimensional displacement value as prior information, the three-dimensional phase is unwrapped using the minimum cost flow method to recover the true phase; S3.4 Perform atmospheric phase correction and deformation inversion to obtain the true deformation value of the slope and integrate it into the digital elevation model.

3. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 2, characterized in that, The specific process of step S3.3 is as follows: S3.3.

1. Perform time-dimension unwrapping on the permanent scattering points in the differential interferogram; S3.3.2 Based on the transformation relationship lookup table, the displacement values ​​of the three-dimensional lidar point cloud are projected onto the radar line of sight, and then spatial filtering is used to operate on the point cloud data in the two-dimensional phase-scanning radar pixel unit to obtain the prior deformation value and prior confidence. S3.3.

3. Convert the prior deformation value into phase using the radar wavelength, and then obtain the prior entanglement value; S3.3.

4. Set virtual source and virtual sink points for pixels with known prior entanglement values, and set cost weights based on prior confidence levels; S3.3.

5. Use the minimum cost flow method to solve for the global entanglement value and complete the phase untangling.

4. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 3, characterized in that, The formula for obtaining the prior winding value in step S3.3.3 is as follows: in, This is the prior entanglement value; The phase is the phase after the a priori deformation value is converted.

5. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 1, characterized in that, The formula for calculating the velocity value during the constant velocity deformation stage in step S4.4 is as follows: in, The speed value during the constant speed change phase; These represent the displacement and time corresponding to the average time point at which acceleration begins; This represents the initial displacement and time at which the slope begins to deform.

6. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 5, characterized in that, The specific formula for calculating the tangent angle value at the current moment in step S4.5 is as follows: in, The tangent angle value. This represents the current speed value for the risk area.

7. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 6, characterized in that, The preset warning rules are as follows: when When the angle is ≤45°, the warning level is alert level and the alarm type is blue. When 45° < When the temperature is below 80°, the warning level is alert, and the alarm type is yellow. When 80°≤ When the temperature is below 85°, the warning level is alert level and the alarm type is orange. When 85°≤ At that time, the warning level is alarm level, and the alarm type is red.

8. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 7, characterized in that, The specific process of step S5 is as follows: When the risk area is under orange alert and there is no red alert area, insert a moving target monitoring between two deformation monitoring; When the risk area is under red alert, moving target monitoring is continuously performed between two deformation monitoring sessions, and when the speed of the moving target is detected to exceed the warning value, the process switches to step S6.

9. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to any one of claims 1-4, characterized in that, Step S6 specifically includes: S6.1 Identify moving targets and filter false alarm targets by combining slope information; S6.

2. Based on the current speed, position information, and slope and aspect information of the moving target, predict the target's position at the next moment, and perform a collision judgment to obtain the collision judgment result; S6.

3. Based on the collision judgment results, the scanning mode of the two-dimensional phase-scanning radar is adaptively adjusted.

10. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 9, characterized in that, The specific process of step S6.1 is as follows: The two-dimensional phase-scanning radar performs full-scene imaging in moving target monitoring mode, identifies moving targets in the image, and obtains distance and angle information; Then, based on the transformation relationship lookup table generated in step S2, the moving target is located in the real three-dimensional spatial coordinate system centered on the radar; The confidence level is determined by using the slope information pre-acquired by the 3D LiDAR; if the slope of the area where the target is located is lower than the preset slope threshold, it is considered a false alarm and filtered out. If no moving target is detected in the scene for a preset number of consecutive times, the target movement is determined to have stopped.

11. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 10, characterized in that, The specific process of step S6.2 is as follows: Based on the current velocity and position information of the moving target, and combined with the slope and aspect information provided by the high-precision digital elevation model, the distance and azimuth of the target at the next phase-scan radar acquisition time are estimated using the integral method and motion model; simultaneously, collision detection is performed on the predicted trajectory. If the slope change in the predicted trajectory is constant or monotonically increasing, a parabolic model is used to calculate the landing point and determine whether the target reaches the landing point. If it does, it is a collision; otherwise, it is not a collision. If the slope change in the predicted trajectory reaches a minimum or decreases monotonically, it is determined that a collision will occur.

12. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 11, characterized in that, The parabola model is as follows: Where m is the longitudinal displacement value and n is the horizontal displacement value. Let g represent the horizontal and vertical velocities, and g represent the acceleration due to gravity.

13. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 11, characterized in that, The coefficient of restitution is used to quantify the change in velocity and energy loss before and after a collision. The calculation formula is as follows: in, These are the normal and tangential velocity components before the collision. These are the normal and tangential velocity components after the collision. These are the normal and tangential restitution coefficients.

14. The slope risk monitoring method based on three-dimensional laser and two-dimensional phase-scanning radar according to claim 9, characterized in that, The specific process of step S6.3 is as follows: If it is determined that no collision has occurred, the radar is switched to narrow beam scanning mode to enhance the radar scanning energy, and the beam is pointed to the predicted position and azimuth for precise tracking; if the target can continue to be detected, the process jumps to step S6.2 to predict the next moment. If a collision or target loss is detected, the radar switches back to full-scene scanning mode and returns to step S6.1 to re-detect and identify moving targets.

15. A slope risk monitoring system based on three-dimensional laser and two-dimensional phase-scanning radar, comprising an interconnected memory and a processor, wherein the memory stores a computer program, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1-14.

Images