An end-cloud collaborative rockfall three-dimensional track full life cycle monitoring method and system

Abstract

This invention discloses an edge-cloud collaborative method and system for monitoring the entire lifecycle of three-dimensional rockfall trajectories, belonging to the field of geological disaster monitoring technology. By processing raw radar echo data and detecting moving targets, the data scale is reduced, potential moving target points are obtained, and feature analysis is performed on these potential moving target points to eliminate interfering target points, thus obtaining effective rockfall target points and forming an anti-interference capability for rockfall disaster monitoring. Furthermore, based on the effective rockfall target points, the three-dimensional spatial motion trajectory of the effective rockfall targets in the real three-dimensional terrain is output, achieving accurate modeling of rockfall movement. This makes the predicted threat area more precise, significantly improving the accuracy of rockfall risk prediction. Threat level determination is also performed, generating corresponding decomposed early warning commands to ensure the reliability of the early warning and avoid false alarms.

Description

Technical Field

[0001] This invention belongs to the field of geological disaster monitoring technology, specifically relating to a method and system for end-to-end cloud collaborative three-dimensional tracking of rolling stones throughout their entire life cycle. Background Technology

[0002] Currently, geological hazard monitoring for rockfalls mainly relies on three types of technologies, but all have serious limitations in complex field environments. The first type is contact-based monitoring technologies, such as active protective nets and landslide meters. These are essentially passive "point" or "line" protection, only triggering an alarm when a rockfall hits a specific physical sensor. This results in large blind spots, equipment fragility, and high maintenance costs. The second type is video surveillance and machine vision technology. This method is greatly constrained by ambient lighting conditions (such as nighttime, rain, fog, and strong light), and the algorithm is easily affected by vegetation movement and cloud shadows, leading to a high false alarm rate. Furthermore, high-definition video streams have extremely high requirements for transmission bandwidth and edge computing resources, which can cause early warning delays in remote mountainous areas. The third category is traditional speed-measuring radar (such as Doppler radar). Although it has a certain anti-interference capability, it can only provide radial velocity information of the target and cannot effectively distinguish between rolling stones and moving interference such as flying birds and shaking trees, thus resulting in false warnings. More importantly, traditional radar lacks the ability to fuse and process target distance, angle, and elevation information, and cannot output the precise trajectory and landing area of ​​rolling stones in real three-dimensional terrain, resulting in a lack of data foundation for subsequent risk quantification and accurate early warning.

[0003] As mentioned above, how to provide an edge-cloud collaborative three-dimensional trajectory full lifecycle monitoring method and system for rockfall disasters that can perform anti-interference monitoring and improve the accuracy and reliability of rockfall risk prediction has become an urgent research topic in this field. Summary of the Invention

[0004] The purpose of this invention is to provide a cloud-edge collaborative method and system for monitoring the entire lifecycle of a rolling stone three-dimensional trajectory, in order to solve the above-mentioned problems existing in the prior art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for end-to-end cloud collaborative monitoring of the entire lifecycle of a rolling stone three-dimensional trajectory, including: A microwave radar signal is emitted to the target slope to acquire the original echo signal of the target slope within the sampling window. The original echo signal is preprocessed and digitally converted to obtain the original radar echo data and sent to the edge computing node. In the edge computing node, moving target detection is performed on the original radar echo data to detect multiple potential moving target points. Doppler oscillation feature analysis, spatiotemporal continuity feature analysis, and motion state feature analysis are performed on each potential moving target point to remove interfering target points from each potential moving target point to obtain effective rolling stone target points. The effective rolling stone target points are then integrated into a set of effective rolling stone target points. Obtain a preset quaternion rotation matrix and digital elevation model, extract the polar coordinates of each effective rolling stone target point from the effective rolling stone target point set, convert the polar coordinates of each effective rolling stone target point into three-dimensional coordinates based on the quaternion rotation matrix and the digital elevation model, form an effective rolling stone target three-dimensional point set, perform spatial clustering calculation on the effective rolling stone target three-dimensional point set to obtain the centroid coordinates of multiple effective rolling stone targets, and form the three-dimensional spatial motion trajectory of each effective rolling stone target based on the centroid coordinates of the effective rolling stone targets, and send the three-dimensional spatial motion trajectory of each effective rolling stone target to the cloud monitoring center; In the cloud monitoring center, based on the three-dimensional spatial motion trajectory of each effective rolling stone target, the effective rolling stone target motion equations are established for each effective rolling stone target. The effective rolling stone target motion equations are then used to predict the threat area of ​​each effective rolling stone target to obtain the predicted threat area of ​​each effective rolling stone target. Finally, the predicted threat area of ​​each effective rolling stone target is used to determine the threat level of each effective rolling stone target. Based on the threat level of each effective rolling stone target, a corresponding graded early warning instruction is generated and sent to the corresponding threat area management platform. The three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target are then visualized.

[0006] In one possible design, a microwave radar signal is emitted towards the target slope to acquire the raw echo signal of the target slope within a sampling window. The raw echo signal is then preprocessed and digitized to obtain raw radar echo data, which is then sent to an edge computing node. This includes: Microwave radar signals are transmitted toward the target slope through a pre-set radar sensing component, and the real-time raw echo signal of the target slope is collected through the radar sensing component, wherein the target slope is a slope threatened by falling rocks. A preset sampling window is obtained, and real-time raw echo signals at multiple times are collected within the sampling window and sorted in chronological order to form raw echo signals. The original echo signal is amplified with low noise, mixed and filtered to obtain a preprocessed original echo signal, and then digitally converted by analog-to-digital conversion sampling to form the original radar echo data. The raw radar echo data is sent to an edge computing node via wireless communication, wherein the edge computing node is a data processor configured to correspond to the radar sensing component.

[0007] In one possible design, the edge computing node performs moving target detection on the raw radar echo data, detecting multiple potential moving target points, including: The preset data standard structure in the edge computing node is obtained, and the original radar echo data is structured using the data standard structure to obtain a radar point cloud dataset. The radar point cloud dataset includes multiple radar data points, and each radar data point includes range information, radial velocity information, horizontal azimuth information, elevation angle information, echo intensity information, and timestamp. A fast Fourier transform is performed on each radar data point in the radar point cloud dataset to obtain the Doppler frequency corresponding to each radar data point, and the Doppler frequencies of each radar data point are used to generate the radar point cloud Doppler velocity spectrum. A preset moving target velocity threshold is obtained, and the moving target velocity threshold is used to identify moving targets in the Doppler velocity spectrum. Each radar data point in the Doppler velocity spectrum with a Doppler frequency higher than the moving target velocity threshold is taken as a pre-potential moving target point, and each radar data point in the Doppler velocity spectrum with a Doppler frequency not higher than the moving target velocity threshold is taken as an interference data point. Based on each of the pre-potential moving target points, the corresponding echo intensity detection threshold is calculated using constant false alarm rate (CFAR) technology. The echo intensity information of each pre-potential moving target point is then filtered using the echo intensity detection threshold. Pre-potential moving target points with echo intensity higher than the corresponding echo intensity detection threshold are designated as potential moving target points, while pre-potential moving target points with echo intensity lower than the corresponding echo intensity detection threshold are designated as interference data points.

[0008] In one possible design, Doppler oscillation characteristic analysis, spatiotemporal continuity characteristic analysis, and motion state characteristic analysis are performed on each of the potential moving target points to eliminate interfering target points from the potential moving target points, obtaining effective rolling stone target points, and integrating the effective rolling stone target points into an effective rolling stone target point set, including: Feature extraction is performed on each of the potential moving target points to obtain the temporal target feature vector corresponding to each potential moving target point. The temporal target feature vector includes the velocity oscillation feature, spatial density feature, echo intensity feature, spectral peak feature, and motion trend feature of the potential moving target point within the sampling window. For each potential moving target point, the zero-crossing rate and displacement integral value of the corresponding potential moving target point are calculated based on the velocity oscillation characteristics in the corresponding time-series target feature vector. Obtain a preset lower threshold for zero-crossing rate and an upper threshold for displacement integral. Use the lower threshold for zero-crossing rate and the upper threshold for displacement integral to filter potential moving target points corresponding to each time-series target feature vector. Mark the potential moving target points corresponding to the time-series target feature vectors whose zero-crossing rate is higher than the lower threshold for zero-crossing rate and whose displacement integral value is lower than the upper threshold for displacement integral as in-situ oscillation points, and use the in-situ oscillation points as the first interference target points. Obtain a preset neighborhood radius for each point cluster, perform spatial clustering analysis on each potential moving target point based on the neighborhood radius to form multiple target point clusters, and select potential moving target points that do not belong to any of the target point clusters as isolated noise points; A preset effective point cluster quantity threshold is obtained, and the number of potential moving target points in each target point cluster is counted. Target point clusters with a number of potential moving target points lower than the effective point cluster quantity threshold are identified as micro target point clusters, and each potential moving target point in each micro target point cluster is marked as a micro target point. A time continuity test is performed on each of the isolated noise points and each of the tiny target points, so that each of the isolated noise points and each of the tiny target points that pass the time continuity test is used as the second interference target point; Extract the corresponding velocity oscillation features and motion trend features from the temporal target feature vectors corresponding to each potential moving target point, and predict the motion state of each potential moving target point based on the velocity oscillation features and the motion trend features to obtain the current motion state prediction value of each potential moving target point. Motion state extraction is performed on each of the potential moving target points to obtain the current motion state measurement value of each potential moving target point. The error between the current motion state prediction value and the current motion state measurement value of each potential moving target point is calculated to obtain the motion error of each potential moving target point. A preset true motion error threshold is obtained, and the motion state of each potential motion target point is determined by the motion error of each potential motion target point using the true motion error threshold. Potential motion target points with motion errors higher than the true motion error threshold are marked as false motion points, and the false motion points are used as third interference target points. The first, second, and third interference target points are integrated and deduplicated to obtain multiple interference target points. Each interference target point is removed from the potential moving target points, and the remaining potential moving target points are used as effective rolling stone target points. The effective rolling stone target points are then integrated to form a set of effective rolling stone target points.

[0009] In one possible design, a preset quaternion rotation matrix and digital elevation model are obtained. The polar coordinates of each valid rolling stone target point are extracted from the set of valid rolling stone target points. Based on the quaternion rotation matrix and the digital elevation model, the polar coordinates of each valid rolling stone target point are converted into three-dimensional coordinates, forming a set of valid rolling stone target three-dimensional points. Spatial clustering calculations are performed on this set to obtain the centroid coordinates of multiple valid rolling stone targets. Based on the centroid coordinates of the valid rolling stone targets, a three-dimensional spatial motion trajectory for each valid rolling stone target is formed. The three-dimensional spatial motion trajectories of each valid rolling stone target are then sent to a cloud monitoring center. This includes: The radar parameters of the radar sensing component and the preset geodetic coordinate system are obtained. A corresponding radar coordinate system is established based on the radar parameters of the radar sensing component. The quaternion rotation matrix between the radar coordinate system and the geodetic coordinate system is calculated based on the radar parameters. The radar parameters include the intrinsic parameters of the radar sensing component, the installation position parameters of the radar sensing component, and the installation attitude angle parameters of the radar sensing component. The polar coordinates of each effective rolling stone target point are extracted from the effective rolling stone target point set, and the polar coordinates of each effective rolling stone target point are converted into the corresponding rectangular coordinates of the target point, wherein the rectangular coordinates of the target point are the coordinates of the effective rolling stone target point in the radar coordinate system; Based on the quaternion rotation matrix, the rectangular coordinates of each effective rolling stone target point are converted into the corresponding two-dimensional coordinates of the target point to form a set of effective rolling stone target two-dimensional points, wherein the two-dimensional coordinates of the target point are the planar coordinates of the effective rolling stone target point in the geodetic coordinate system. A preset digital elevation model is obtained. The two-dimensional coordinates of the target points in the effective rolling stone target two-dimensional point set are input into the digital elevation model to extract the slope elevation data corresponding to each target point two-dimensional coordinate. The slope elevation data corresponding to each target point two-dimensional coordinate is added to each target point two-dimensional coordinate to obtain the corresponding effective rolling stone target three-dimensional coordinate. The effective rolling stone target three-dimensional coordinates are integrated to obtain the effective rolling stone target three-dimensional point set. Obtain a preset rolling stone neighborhood search radius, calculate the three-dimensional Euclidean distance between each effective rolling stone target point in the effective rolling stone target three-dimensional point set, and use the rolling stone neighborhood search radius to perform distance screening on the three-dimensional Euclidean distance between each effective rolling stone target point, so as to complete spatial clustering of two adjacent effective rolling stone target points whose three-dimensional Euclidean distance does not exceed the neighborhood search radius, and obtain multiple spatial cluster point clusters. Each of the spatial clusters of points is taken as an effective rolling stone target. The corresponding geometric centroid point is identified for each effective rolling stone target. The three-dimensional coordinates of the geometric centroid point corresponding to each effective rolling stone target in the geodetic coordinate system are taken as the centroid coordinates of each effective rolling stone target. For each effective rolling stone target, the centroid coordinates at each moment within the sampling window are obtained, and the centroid coordinates at each moment are connected sequentially in chronological order to form the three-dimensional spatial motion trajectory of each effective rolling stone target. A rolling stone ID number is added to the three-dimensional spatial motion trajectory of each effective rolling stone target. The three-dimensional spatial motion trajectories of each effective rolling stone target are integrated based on the rolling stone ID number of each effective rolling stone target. The integrated three-dimensional spatial motion trajectories of each effective rolling stone target are then sent to the cloud monitoring center via wireless communication.

[0010] In one possible design, in the cloud monitoring center, based on the three-dimensional spatial motion trajectory of each effective rolling stone target, an effective rolling stone target motion equation is established for each effective rolling stone target. This effective rolling stone target motion equation is then used to predict the threat area of ​​each effective rolling stone target, resulting in a predicted threat area for each target. Finally, a threat level determination is performed on the predicted threat area of ​​each effective rolling stone target to determine its threat level, including: In the cloud monitoring center, the effective rolling stone target motion equations are established for each effective rolling stone target based on the three-dimensional spatial motion trajectory of each effective rolling stone target. Based on the effective rolling stone target motion equations of each effective rolling stone target, the corresponding predicted rolling stone landing speed, predicted rolling stone impulse, predicted rolling stone rebound height and predicted rolling stone landing point coordinates are calculated for each effective rolling stone target. Obtain the preset threat area radius, take the predicted boulder landing point coordinates as the center and the threat area radius as the radius, calculate the circular planar area corresponding to each effective boulder target on the horizontal plane, and take the circular planar area corresponding to each effective boulder target as the predicted threat area of ​​each effective boulder target; Obtain a preset threat level comparison table, and determine the threat level of the predicted rolling stone landing speed, the predicted rolling stone impulse, the predicted rolling stone rebound height, and the predicted threat area based on the threat level comparison table, so as to obtain the threat level of each of the effective rolling stone targets.

[0011] In one possible design, based on the threat level of each of the effective rolling stone targets, a corresponding graded early warning instruction is generated. This instruction is then sent to the corresponding threat area management platform, and the three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each target, and the threat level of each target are visualized, including: Obtain the jurisdiction area of ​​each regional management platform, and solve the intersection of the jurisdiction area of ​​each regional management platform with the predicted threat area of ​​each effective rolling stone target to obtain the intersection jurisdiction area and the intersection predicted threat area. Based on the jurisdiction of each intersection, the corresponding regional management platform is selected, and the regional management platform corresponding to each jurisdiction is used as the threat area management platform. Based on the predicted threat areas of each intersection, the corresponding effective rolling stone targets are selected. Based on the threat level of the effective rolling stone targets corresponding to each predicted threat area, the graded early warning instructions corresponding to each predicted threat area are generated. The tiered early warning instructions are transmitted one by one to the threat area management platforms of each overlapping jurisdiction through wireless communication. The three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target are compiled into a rolling stone threat motion view, which is then visualized in the cloud monitoring center.

[0012] Secondly, this invention provides an edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring system, including: The radar signal transceiver unit is used to transmit microwave radar signals to the target slope to acquire the original echo signal of the target slope within the sampling window, preprocess and digitize the original echo signal to obtain the original radar echo data and send it to the edge computing node. The rolling stone target calculation unit is used in the edge computing node to perform moving target detection on the original radar echo data, detect multiple potential moving target points, perform Doppler oscillation feature analysis, spatiotemporal continuity feature analysis and motion state feature analysis on each potential moving target point, so as to remove interfering target points from each potential moving target point, obtain effective rolling stone target points, and integrate each effective rolling stone target point into an effective rolling stone target point set; The rolling stone trajectory calculation unit is used to obtain a preset quaternion rotation matrix and digital elevation model, extract the target point polar coordinates corresponding to each effective rolling stone target point from the effective rolling stone target point set, convert the target point polar coordinates of each effective rolling stone target point into target point three-dimensional coordinates based on the quaternion rotation matrix and the digital elevation model, form an effective rolling stone target three-dimensional point set, perform spatial clustering calculation on the effective rolling stone target three-dimensional point set to obtain the centroid coordinates of multiple effective rolling stone targets, form the three-dimensional spatial motion trajectory of each effective rolling stone target based on the centroid coordinates of the effective rolling stone targets, and send the three-dimensional spatial motion trajectory of each effective rolling stone target to the cloud monitoring center; The threat level calculation unit is used in the cloud monitoring center to establish the effective rolling stone target motion equations of each effective rolling stone target based on the three-dimensional spatial motion trajectory of each effective rolling stone target, and to use the effective rolling stone target motion equations to predict the threat area of ​​each effective rolling stone target, so as to obtain the predicted threat area of ​​each effective rolling stone target, and to determine the threat level of each effective rolling stone target based on the predicted threat area of ​​each effective rolling stone target. The early warning instruction issuing unit is used to generate corresponding graded early warning instructions based on the threat level of each effective rolling stone target, and send the graded early warning instructions to the corresponding threat area management platform, and visualize the three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target.

[0013] Thirdly, the present invention provides an electronic device comprising a memory, a processor, and a transceiver connected in sequence and communication, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the end-to-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method as described in the first aspect or any possible design of the first aspect.

[0014] Fourthly, the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, perform the end-to-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method described in the first aspect or any possible design of the first aspect.

[0015] Fifthly, the present invention provides a computer program product containing instructions that, when the instructions are executed on a computer, cause the computer to perform the end-to-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method as described in the first aspect or any possible design of the first aspect.

[0016] Beneficial Effects: This invention provides a cloud-edge collaborative method and system for monitoring the entire lifecycle of a rolling stone's three-dimensional trajectory. The method includes: first, transmitting microwave radar signals to the target slope to acquire the original echo signal within a sampling window; preprocessing and digitizing the original echo signal to obtain original radar echo data, which is then sent to an edge computing node; second, in the edge computing node, performing moving target detection on the original radar echo data to detect multiple potential moving target points; and performing Doppler oscillation characteristic analysis on each potential moving target point. Spatiotemporal continuity feature analysis and motion state feature analysis are used to eliminate interfering target points from the potential moving target points to obtain effective rolling stone target points, and these effective rolling stone target points are integrated into an effective rolling stone target point set. Then, a preset quaternion rotation matrix and digital elevation model are obtained, and the polar coordinates of the target points corresponding to each effective rolling stone target point are extracted from the effective rolling stone target point set. Based on the quaternion rotation matrix and the digital elevation model, the polar coordinates of the target points of each effective rolling stone target point are converted into three-dimensional coordinates of the target points, forming an effective rolling stone target three-dimensional point set. The system performs spatial clustering calculations on the three-dimensional point set of the effective rolling stone targets to obtain the centroid coordinates of multiple effective rolling stone targets. Based on these centroid coordinates, it forms the three-dimensional spatial motion trajectory of each effective rolling stone target and sends this trajectory to a cloud monitoring center. In the cloud monitoring center, based on the three-dimensional spatial motion trajectories of each effective rolling stone target, it establishes the effective rolling stone target motion equations. These equations are then used to predict the threat area of ​​each effective rolling stone target, resulting in a predicted threat area. The predicted threat area is then used to determine the threat level of each effective rolling stone target. Finally, based on the threat level of each effective rolling stone target, a corresponding graded early warning command is generated and sent to the corresponding threat area management platform. The system also visualizes the three-dimensional spatial motion trajectory, predicted threat area, and threat level of each effective rolling stone target. By processing the raw radar echo data and detecting moving targets, the data size was reduced, potential moving target points were obtained, and feature analysis was performed on these potential moving target points to eliminate interfering target points, thus obtaining effective rockfall target points and forming an anti-interference capability for rockfall disaster monitoring. Furthermore, based on the effective rockfall target points, the three-dimensional spatial motion trajectory of the effective rockfall targets in the real three-dimensional terrain was output, realizing accurate modeling of rockfall movement. This made the predicted threat area more accurate, significantly improving the accuracy of rockfall risk prediction. Threat level determination was also performed, and corresponding decomposed early warning instructions were generated to ensure the reliability of the early warning and avoid false alarms. Attached Figure Description

[0017] Figure 1 A schematic diagram of the process for end-to-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method provided in an embodiment of the present invention; Figure 2 This is a functional structure diagram of the edge-cloud collaborative rolling stone three-dimensional trajectory full life cycle monitoring system provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in conjunction with the accompanying drawings and descriptions of the embodiments or the prior art. Obviously, the following description of the structure of the accompanying drawings is only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. It should be noted that the description of these embodiments is for the purpose of helping to understand the present invention, but does not constitute a limitation of the present invention.

[0019] It should be understood that although the terms first, second, etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another. For example, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit, without departing from the scope of the exemplary embodiments of the invention.

[0020] It should be understood that the term "and / or" that may appear in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can mean: A exists alone, B exists alone, and A and B exist simultaneously. The term " / and" that may appear in this document describes another relationship between related objects, indicating that two relationships can exist. For example, A / and B can mean: A exists alone, and A and B exist alone. In addition, the character " / " that may appear in this document generally indicates that the related objects before and after it are in an "or" relationship.

[0021] Example: like Figure 1 As shown, the first aspect of this embodiment provides an edge-cloud collaborative method for monitoring the entire lifecycle of Rolling Stone's three-dimensional trajectory, which may include, but is not limited to, the following steps: S1. Transmit microwave radar signals to the target slope to obtain the original echo signal of the target slope within the sampling window, preprocess and digitize the original echo signal to obtain the original radar echo data and send it to the edge computing node; S2. In the edge computing node, moving target detection is performed on the original radar echo data to detect multiple potential moving target points. Doppler oscillation feature analysis, spatiotemporal continuity feature analysis and motion state feature analysis are performed on each potential moving target point to remove interfering target points from each potential moving target point to obtain effective rolling stone target points. The effective rolling stone target points are then integrated into a set of effective rolling stone target points. S3. Obtain a preset quaternion rotation matrix and digital elevation model, extract the polar coordinates of each effective rolling stone target point from the effective rolling stone target point set, convert the polar coordinates of each effective rolling stone target point into three-dimensional coordinates based on the quaternion rotation matrix and the digital elevation model, form an effective rolling stone target three-dimensional point set, perform spatial clustering calculation on the effective rolling stone target three-dimensional point set to obtain the centroid coordinates of multiple effective rolling stone targets, form the three-dimensional spatial motion trajectory of each effective rolling stone target based on the centroid coordinates of the effective rolling stone targets, and send the three-dimensional spatial motion trajectory of each effective rolling stone target to the cloud monitoring center; S4. In the cloud monitoring center, based on the three-dimensional spatial motion trajectory of each effective rolling stone target, the effective rolling stone target motion equations of each effective rolling stone target are established. The effective rolling stone target motion equations of each effective rolling stone target are used to predict the threat area of ​​each effective rolling stone target to obtain the predicted threat area of ​​each effective rolling stone target. The threat level of the predicted threat area of ​​each effective rolling stone target is determined to obtain the threat level of each effective rolling stone target. S5. Based on the threat level of each effective rolling stone target, generate a corresponding graded early warning instruction, send the graded early warning instruction to the corresponding threat area management platform, and visualize the three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target.

[0022] In one possible implementation, step S1 involves transmitting a microwave radar signal to the target slope to acquire the raw echo signal of the target slope within a sampling window. The raw echo signal is then preprocessed and digitized to obtain raw radar echo data, which is then sent to the edge computing node. This step can be broken down into, but is not limited to, the following steps S11-S14, specifically including: S11. A microwave radar signal is transmitted toward the target slope through a pre-set radar sensing component, and the real-time raw echo signal of the target slope is collected through the radar sensing component, wherein the target slope is a slope threatened by falling rocks. S12. Obtain a preset sampling window, collect real-time raw echo signals at multiple times within the sampling window, and sort them in chronological order to form raw echo signals; S13. The original echo signal is amplified with low noise, mixed and filtered to obtain a preprocessed original echo signal, and the preprocessed original echo signal is digitally converted by analog-to-digital conversion sampling to form the original radar echo data. S14. The raw radar echo data is sent to an edge computing node via wireless communication, wherein the edge computing node is a data processor configured corresponding to the radar sensing component.

[0023] In specific application scenarios, one or more multi-dimensional radar sensing components are deployed on mountain slopes threatened by geological disasters, on the opposite bank of roadbeds, or at key monitoring points. The core of this component is a frequency-modulated continuous wave (FMCW) radar or pulse Doppler radar employing a multiple-input multiple-output (MIMO) system. During deployment, precise geographic location calibration (via GPS / BeiDou) and attitude calibration (via tilt sensors) are required to ensure that the beam center of the radar sensing component points to the target monitoring area. Simultaneously, self-tests of each radar sensing component are performed, including checks and parameter loading of the antenna array, RF front-end, signal processing board, power supply (such as solar cells and batteries), and communication (LoRa / 5G module) systems.

[0024] The radar controls its transmitting antenna array to transmit a series of mutually orthogonal (e.g., orthogonal in time, frequency, or coding) radio frequency detection signals according to a pre-designed MIMO waveform sequence. The MIMO system, with relatively few antenna elements, forms a "virtual array" at a virtual level that far exceeds the number of actual physical antennas, thereby greatly improving the radar's angular resolution. These signals radiate into the hillside area in the form of clustered waves, forming a fan-shaped or cone-shaped three-dimensional beam covering the predetermined monitoring range, achieving continuous illumination of a "surface" or "volume" area, rather than the "line" scanning of traditional radar.

[0025] In one possible implementation, step S2, where the edge computing node performs moving target detection on the original radar echo data to detect multiple potential moving target points, can be decomposed into steps S21-S24, specifically including: S21. Obtain the preset data standard structure in the edge computing node, and use the data standard structure to perform structured processing on the original radar echo data to obtain a radar point cloud dataset. The radar point cloud dataset includes multiple radar data points, and each radar data point includes range information, radial velocity information, horizontal azimuth information, elevation angle information, echo intensity information, and timestamp. S22. Perform a fast Fourier transform on each radar data point in the radar point cloud dataset to obtain the Doppler frequency corresponding to each radar data point, and generate a radar point cloud Doppler velocity spectrum from the Doppler frequencies of each radar data point. S23. Obtain a preset moving target velocity threshold, and use the moving target velocity threshold to perform moving target identification on the Doppler velocity spectrum, so that each radar data point in the Doppler velocity spectrum with a Doppler frequency higher than the moving target velocity threshold is taken as a pre-potential moving target point, and each radar data point in the Doppler velocity spectrum with a Doppler frequency not higher than the moving target velocity threshold is taken as an interference data point; S24. Based on each of the pre-potential moving target points, calculate the corresponding echo intensity detection threshold using constant false alarm rate (CFAR) technology, and use the echo intensity detection threshold corresponding to each of the pre-potential moving target points to filter the echo intensity information of each pre-potential moving target point, so that the pre-potential moving target points whose echo intensity in the echo intensity information is higher than the corresponding echo intensity detection threshold are taken as potential moving target points, and the pre-potential moving target points whose echo intensity in the echo intensity information is not higher than the corresponding echo intensity detection threshold are taken as interference data points.

[0026] It should be noted that in the end-to-cloud collaborative rolling stone three-dimensional trajectory full life cycle monitoring method provided in this embodiment, moving target detection is performed on the original radar echo data, which can filter out most of the static background (such as mountains) and extremely slow-moving objects (such as slowly growing plants). In other words, steps S21-S24 can initially separate the "moving" and "stationary" echoes and remove interference data, reduce the computational load of subsequent feature analysis, and improve the identification efficiency of effective rolling stone target points.

[0027] In one possible implementation, step S2 involves performing Doppler oscillation characteristic analysis, spatiotemporal continuity characteristic analysis, and motion state characteristic analysis on each of the potential moving target points to eliminate interfering target points from the potential moving target points, obtaining effective rolling stone target points, and integrating the effective rolling stone target points into an effective rolling stone target point set. This can be decomposed into, but is not limited to, the following steps S25-S214, specifically including: S25. Perform feature extraction on each of the potential moving target points to obtain the temporal target feature vector corresponding to each of the potential moving target points, wherein the temporal target feature vector includes the velocity oscillation feature, spatial density feature, echo intensity feature, spectral peak feature and motion trend feature of the potential moving target point within the sampling window; S26. For each of the potential moving target points, calculate the zero-crossing rate and displacement integral value of the corresponding potential moving target point based on the velocity oscillation characteristics in the corresponding time-series target feature vector; S27. Obtain a preset lower threshold for zero-crossing rate and an upper threshold for displacement integral. Use the lower threshold for zero-crossing rate and the upper threshold for displacement integral to filter the potential moving target points corresponding to each of the time-series target feature vectors. Mark the potential moving target points corresponding to the time-series target feature vectors whose zero-crossing rate is higher than the lower threshold for zero-crossing rate and whose displacement integral value is lower than the upper threshold for displacement integral as in-situ oscillation points, and use the in-situ oscillation points as the first interference target points. S28. Obtain a preset neighborhood radius of point clusters, perform spatial clustering analysis on each potential moving target point based on the neighborhood radius of the point clusters to form multiple target point clusters, and select potential moving target points that do not belong to any of the target point clusters as isolated noise points; S29. Obtain a preset effective point cluster quantity threshold, perform quantity statistics on the potential moving target points in each target point cluster, and classify the target point clusters with the number of potential moving target points lower than the effective point cluster quantity threshold as micro target point clusters, and mark each potential moving target point in each micro target point cluster as a micro target point. S210. Perform a time continuity test on each of the isolated noise points and each of the micro-target points, so as to use each of the isolated noise points and each of the micro-target points that pass the time continuity test as the second interference target points; S211. Extract the corresponding velocity oscillation features and motion trend features from the temporal target feature vectors corresponding to each potential moving target point, and predict the motion state of each potential moving target point based on the velocity oscillation features and the motion trend features to obtain the current motion state prediction value of each potential moving target point. S212. Extract the motion state of each potential moving target point to obtain the current motion state measurement value of each potential moving target point, and calculate the error between the current motion state prediction value of each potential moving target point and the current motion state measurement value of each potential moving target point to obtain the motion error of each potential moving target point. S213. Obtain a preset real motion error threshold, use the real motion error threshold to determine the motion state of each potential motion target point, so as to mark the potential motion target point whose motion error is higher than the real motion error threshold as a false motion point, and use the false motion point as a third interference target point; S214. Integrate the first interference target point, the second interference target point and the third interference target point and perform deduplication to obtain multiple interference target points. Remove each interference target point from each potential moving target point so that the remaining potential moving target points are used as effective rolling stone target points. Integrate each effective rolling stone target point to form a set of effective rolling stone target points.

[0028] It should be noted that in practical applications, if rain, snow, fog, and / or dust storms occur during target slope identification, a spectral feature analysis process can be introduced after completing Doppler oscillation feature analysis, spatiotemporal continuity feature analysis, and motion state feature analysis to identify these external disturbances. In one possible implementation, the spectral feature analysis process may include, but is not limited to, the following steps S215-S219, specifically: S215. Perform a fast Fourier transform on each of the potential moving target points to obtain the Doppler frequency corresponding to each of the potential moving target points, and generate the Doppler velocity spectrum of the potential moving target points from the Doppler frequencies of each of the potential moving target points. S216. In the Doppler velocity spectrum of the potential moving target points, a corresponding local spectral region is formed with each potential moving target point as the center; S217. For each local spectral region, extract the corresponding spectral width, peak-to-sidelobe ratio, and average energy of the local spectral region; S218. Obtain a preset spectrum feature standard, and make a spectrum standard judgment on the spectrum width, peak-to-sidelobe ratio and average energy of each local spectrum region based on the spectrum feature standard, so as to mark the potential moving target points corresponding to each local spectrum region that does not meet the spectrum feature standard as irrelevant moving points, and use each of the irrelevant moving points as the fourth interference target points; S219. Remove each of the fourth interference target points from the set of effective rolling stone target points to update the set of effective rolling stone target points.

[0029] In one possible implementation, step S3 involves obtaining a preset quaternion rotation matrix and digital elevation model, extracting the polar coordinates of each effective rolling stone target point from the effective rolling stone target point set, converting the polar coordinates of each effective rolling stone target point into three-dimensional coordinates based on the quaternion rotation matrix and the digital elevation model to form an effective rolling stone target three-dimensional point set, performing spatial clustering calculations on the effective rolling stone target three-dimensional point set to obtain the centroid coordinates of multiple effective rolling stone targets, forming the three-dimensional spatial motion trajectory of each effective rolling stone target based on the centroid coordinates of the effective rolling stone targets, and sending the three-dimensional spatial motion trajectory of each effective rolling stone target to the cloud monitoring center. This can be decomposed into, but is not limited to, the following steps S31-S38, specifically including: S31. Obtain the radar parameters of the radar sensing component and the preset geodetic coordinate system, establish the corresponding radar coordinate system based on the radar parameters of the radar sensing component, and calculate the quaternion rotation matrix between the radar coordinate system and the geodetic coordinate system according to the radar parameters. The radar parameters include the intrinsic parameters of the radar sensing component, the installation position parameters of the radar sensing component, and the installation attitude angle parameters of the radar sensing component. S32. Extract the polar coordinates of each effective rolling stone target point from the set of effective rolling stone target points, and convert the polar coordinates of each effective rolling stone target point into the corresponding rectangular coordinates of the target point, wherein the rectangular coordinates of the target point are the coordinates of the effective rolling stone target point in the radar coordinate system; S33. Based on the quaternion rotation matrix, the rectangular coordinates of each effective rolling stone target point are converted into the corresponding two-dimensional coordinates of the target point to form a set of effective rolling stone target two-dimensional points, wherein the two-dimensional coordinates of the target point are the planar coordinates of the effective rolling stone target point in the geodetic coordinate system; S34. Obtain a preset digital elevation model, input the two-dimensional coordinates of the target points in the effective rolling stone target two-dimensional point set into the digital elevation model, extract the slope elevation data corresponding to the two-dimensional coordinates of each target point respectively, and add the slope elevation data corresponding to the two-dimensional coordinates of each target point to the two-dimensional coordinates of each target point to obtain the corresponding effective rolling stone target three-dimensional coordinates, integrate the three-dimensional coordinates of each effective rolling stone target to obtain the effective rolling stone target three-dimensional point set; S35. Obtain the preset rolling stone neighborhood search radius, calculate the three-dimensional Euclidean distance between each effective rolling stone target point in the effective rolling stone target three-dimensional point set, and use the rolling stone neighborhood search radius to perform distance screening on the three-dimensional Euclidean distance between each effective rolling stone target point, so as to complete the spatial clustering of two adjacent effective rolling stone target points whose three-dimensional Euclidean distance does not exceed the neighborhood search radius, and obtain multiple spatial cluster point clusters. S36. Each of the spatial clusters of points is taken as an effective rolling stone target. The corresponding geometric centroid point is identified for each effective rolling stone target. The three-dimensional coordinates of the geometric centroid point corresponding to each effective rolling stone target in the geodetic coordinate system are taken as the centroid coordinates of each effective rolling stone target. S37. For each effective rolling stone target, obtain the centroid coordinates at each moment within the sampling window, and connect the centroid coordinates at each moment in chronological order to form the three-dimensional spatial motion trajectory of each effective rolling stone target. S38. Add a rolling stone ID number to the three-dimensional spatial motion trajectory of each effective rolling stone target, integrate the three-dimensional spatial motion trajectories of each effective rolling stone target based on the rolling stone ID number of each effective rolling stone target, and send the integrated three-dimensional spatial motion trajectory of each effective rolling stone target to the cloud monitoring center through wireless communication.

[0030] Among them, edge-cloud collaboration is achieved through the transmission and reception of signals, data and commands via wireless communication. This wireless communication can be, but is not limited to, LoRa / 5G dual-link communication. The edge computing node and radar sensing component can be integrated into a radar terminal. While meeting the requirements of LoRa / 5G dual-link communication, the radar terminal also supports solar power supply to ensure long-term stable operation in remote areas without electricity.

[0031] In one possible implementation, in step S4, in the cloud monitoring center, based on the three-dimensional spatial motion trajectory of each effective rolling stone target, an effective rolling stone target motion equation is established for each effective rolling stone target. This effective rolling stone target motion equation is then used to predict the threat area of ​​each effective rolling stone target, thereby obtaining the predicted threat area of ​​each effective rolling stone target. Finally, a threat level determination is made for the predicted threat area of ​​each effective rolling stone target, thereby obtaining the threat level of each effective rolling stone target. This can be decomposed, but is not limited to, the following steps S41-S44, specifically including: S41. In the cloud monitoring center, based on the three-dimensional spatial motion trajectory of each effective rolling stone target, the effective rolling stone target motion equation is established for each effective rolling stone target; S42. Based on the effective rolling stone target motion equations of each effective rolling stone target, calculate the corresponding predicted rolling stone landing speed, predicted rolling stone impulse, predicted rolling stone rebound height and predicted rolling stone landing point coordinates for each effective rolling stone target. S43. Obtain the preset threat area radius, take the predicted boulder landing point coordinates as the center and the threat area radius as the radius, calculate the circular planar area corresponding to each effective boulder target on the horizontal plane, and take the circular planar area corresponding to each effective boulder target as the predicted threat area of ​​each effective boulder target. S44. Obtain a preset threat level comparison table, and determine the threat level of the predicted rolling stone landing speed, the predicted rolling stone impulse, the predicted rolling stone rebound height, and the predicted threat area based on the threat level comparison table, so as to obtain the threat level of each of the effective rolling stone targets.

[0032] In one possible implementation, step S5 involves generating a corresponding graded early warning instruction based on the threat level of each effective rolling stone target, sending the graded early warning instruction to the corresponding threat area management platform, and visually displaying the three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target. This can be broken down into, but is not limited to, the following steps S51-S54, specifically including: S51. Obtain the jurisdiction area of ​​each regional management platform, and solve the intersection of the jurisdiction area of ​​each regional management platform with the predicted threat area of ​​each effective rolling stone target to obtain the solved intersection jurisdiction area and the intersection predicted threat area. S52. Based on each intersection jurisdiction area, select the corresponding area management platform, use the area management platform corresponding to each intersection jurisdiction area as the threat area management platform, select the corresponding effective rolling stone target based on each intersection predicted threat area, and generate the graded early warning instruction corresponding to each intersection predicted threat area based on the threat level of the effective rolling stone target corresponding to each intersection predicted threat area. S53. The graded early warning instructions are sent one by one to the threat area management platform of each overlapping jurisdiction area through wireless communication; S54. The three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target are compiled into a rolling stone threat motion view, and the rolling stone threat motion view is visualized in the cloud monitoring center.

[0033] like Figure 2 As shown, the second aspect of this embodiment provides a hardware system for implementing the edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method described in the first aspect of the embodiment, including: The radar signal transceiver unit is used to transmit microwave radar signals to the target slope to acquire the original echo signal of the target slope within the sampling window, preprocess and digitize the original echo signal to obtain the original radar echo data and send it to the edge computing node. The rolling stone target calculation unit is used in the edge computing node to perform moving target detection on the original radar echo data, detect multiple potential moving target points, perform Doppler oscillation feature analysis, spatiotemporal continuity feature analysis and motion state feature analysis on each potential moving target point, so as to remove interfering target points from each potential moving target point, obtain effective rolling stone target points, and integrate each effective rolling stone target point into an effective rolling stone target point set; The rolling stone trajectory calculation unit is used to obtain a preset quaternion rotation matrix and digital elevation model, extract the target point polar coordinates corresponding to each effective rolling stone target point from the effective rolling stone target point set, convert the target point polar coordinates of each effective rolling stone target point into target point three-dimensional coordinates based on the quaternion rotation matrix and the digital elevation model, form an effective rolling stone target three-dimensional point set, perform spatial clustering calculation on the effective rolling stone target three-dimensional point set to obtain the centroid coordinates of multiple effective rolling stone targets, form the three-dimensional spatial motion trajectory of each effective rolling stone target based on the centroid coordinates of the effective rolling stone targets, and send the three-dimensional spatial motion trajectory of each effective rolling stone target to the cloud monitoring center; The threat level calculation unit is used in the cloud monitoring center to establish the effective rolling stone target motion equations of each effective rolling stone target based on the three-dimensional spatial motion trajectory of each effective rolling stone target, and to use the effective rolling stone target motion equations to predict the threat area of ​​each effective rolling stone target, so as to obtain the predicted threat area of ​​each effective rolling stone target, and to determine the threat level of each effective rolling stone target based on the predicted threat area of ​​each effective rolling stone target. The early warning instruction issuing unit is used to generate corresponding graded early warning instructions based on the threat level of each effective rolling stone target, and send the graded early warning instructions to the corresponding threat area management platform, and visualize the three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target.

[0034] The working process, working details and technical effects of the system provided in this embodiment can be found in the first aspect of the embodiment, and will not be repeated here.

[0035] like Figure 3As shown, the third aspect of this embodiment provides an electronic device, including: a memory, a processor, and a transceiver that are sequentially and communicatively connected, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program and execute the end-to-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method as described in the first aspect of the embodiment.

[0036] For specific examples, the memory may include, but is not limited to, random access memory (RAM), read-only memory (ROM), flash memory, first-in-first-out (FIFO) memory, and / or first-in-last-out (FILO) memory, etc.; specifically, the processor may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor may be implemented using at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), PLA (Programmable Logic Array). The processor may also include a main processor and a coprocessor. The main processor, also known as the CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state.

[0037] In some embodiments, the processor may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the screen. For example, the processor may not be limited to microprocessors of the STM32F105 series, reduced instruction set computer (RISC) microprocessors, x86 architecture processors, or processors with integrated neural network processing units (NPUs). The transceiver may be, but is not limited to, a Wi-Fi transceiver, a Bluetooth transceiver, a General Packet Radio Service (GPRS) transceiver, a ZigBee (a low-power LAN protocol based on the IEEE 802.15.4 standard) transceiver, a 3G transceiver, a 4G transceiver, and / or a 5G transceiver. Furthermore, the device may also include, but is not limited to, a power module, a display screen, and other necessary components.

[0038] The working process, working details and technical effects of the electronic device provided in this embodiment can be found in the first aspect of the embodiment, and will not be repeated here.

[0039] The fourth aspect of this embodiment provides a storage medium that stores instructions containing the end-to-cloud collaborative Rolling Stone 3D trajectory full lifecycle monitoring method described in the first aspect of the embodiment. That is, the storage medium stores instructions, and when the instructions are run on a computer, the end-to-cloud collaborative Rolling Stone 3D trajectory full lifecycle monitoring method described in the first aspect of the embodiment is executed.

[0040] The storage medium refers to a carrier for storing data, which may include, but is not limited to, floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or memory sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0041] The working process, working details and technical effects of the storage medium provided in this embodiment can be found in the first aspect of the embodiment, and will not be repeated here.

[0042] The fifth aspect of this embodiment provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform the edge-cloud collaborative Rolling Stone 3D trajectory full lifecycle monitoring method as described in the first aspect of the embodiment, wherein the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device.

[0043] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for end-to-end cloud collaborative three-dimensional trajectory full lifecycle monitoring of rolling stones, characterized in that, include: A microwave radar signal is emitted to the target slope to acquire the original echo signal of the target slope within the sampling window. The original echo signal is preprocessed and digitally converted to obtain the original radar echo data and sent to the edge computing node. In the edge computing node, moving target detection is performed on the original radar echo data to detect multiple potential moving target points. Doppler oscillation feature analysis, spatiotemporal continuity feature analysis, and motion state feature analysis are performed on each potential moving target point to remove interfering target points from each potential moving target point to obtain effective rolling stone target points. The effective rolling stone target points are then integrated into a set of effective rolling stone target points. Obtain a preset quaternion rotation matrix and digital elevation model, extract the polar coordinates of each effective rolling stone target point from the effective rolling stone target point set, convert the polar coordinates of each effective rolling stone target point into three-dimensional coordinates based on the quaternion rotation matrix and the digital elevation model, form an effective rolling stone target three-dimensional point set, perform spatial clustering calculation on the effective rolling stone target three-dimensional point set to obtain the centroid coordinates of multiple effective rolling stone targets, and form the three-dimensional spatial motion trajectory of each effective rolling stone target based on the centroid coordinates of the effective rolling stone targets, and send the three-dimensional spatial motion trajectory of each effective rolling stone target to the cloud monitoring center; In the cloud monitoring center, based on the three-dimensional spatial motion trajectory of each effective rolling stone target, the effective rolling stone target motion equations are established for each effective rolling stone target. The effective rolling stone target motion equations are then used to predict the threat area of ​​each effective rolling stone target to obtain the predicted threat area of ​​each effective rolling stone target. Finally, the predicted threat area of ​​each effective rolling stone target is used to determine the threat level of each effective rolling stone target. Based on the threat level of each effective rolling stone target, a corresponding graded early warning instruction is generated and sent to the corresponding threat area management platform. The three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target are then visualized.

2. The edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method according to claim 1, characterized in that, A microwave radar signal is emitted towards the target slope to acquire the raw echo signal of the target slope within a sampling window. The raw echo signal is preprocessed and digitally converted to obtain raw radar echo data, which is then sent to the edge computing node. This process includes: Microwave radar signals are transmitted toward the target slope through a pre-set radar sensing component, and the real-time raw echo signal of the target slope is collected through the radar sensing component, wherein the target slope is a slope threatened by falling rocks. A preset sampling window is obtained, and real-time raw echo signals at multiple times are collected within the sampling window and sorted in chronological order to form raw echo signals. The original echo signal is amplified with low noise, mixed and filtered to obtain a preprocessed original echo signal, and then digitally converted by analog-to-digital conversion sampling to form the original radar echo data. The raw radar echo data is sent to an edge computing node via wireless communication, wherein the edge computing node is a data processor configured to correspond to the radar sensing component.

3. The edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method according to claim 1, characterized in that, In the edge computing node, moving target detection is performed on the raw radar echo data, detecting multiple potential moving target points, including: The preset data standard structure in the edge computing node is obtained, and the original radar echo data is structured using the data standard structure to obtain a radar point cloud dataset. The radar point cloud dataset includes multiple radar data points, and each radar data point includes range information, radial velocity information, horizontal azimuth information, elevation angle information, echo intensity information, and timestamp. A fast Fourier transform is performed on each radar data point in the radar point cloud dataset to obtain the Doppler frequency corresponding to each radar data point, and the Doppler frequencies of each radar data point are used to generate the radar point cloud Doppler velocity spectrum. A preset moving target velocity threshold is obtained, and the moving target velocity threshold is used to identify moving targets in the Doppler velocity spectrum. Each radar data point in the Doppler velocity spectrum with a Doppler frequency higher than the moving target velocity threshold is taken as a pre-potential moving target point, and each radar data point in the Doppler velocity spectrum with a Doppler frequency not higher than the moving target velocity threshold is taken as an interference data point. Based on each of the pre-potential moving target points, the corresponding echo intensity detection threshold is calculated using constant false alarm rate (CFAR) technology. The echo intensity information of each pre-potential moving target point is then filtered using the echo intensity detection threshold. Pre-potential moving target points with echo intensity higher than the corresponding echo intensity detection threshold are designated as potential moving target points, while pre-potential moving target points with echo intensity lower than the corresponding echo intensity detection threshold are designated as interference data points.

4. The edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method according to claim 1, characterized in that, Doppler oscillation characteristic analysis, spatiotemporal continuity characteristic analysis, and motion state characteristic analysis are performed on each of the potential moving target points to eliminate interfering target points from the potential moving target points, thereby obtaining effective rolling stone target points. These effective rolling stone target points are then integrated into a set of effective rolling stone target points, including: Feature extraction is performed on each of the potential moving target points to obtain the temporal target feature vector corresponding to each potential moving target point. The temporal target feature vector includes the velocity oscillation feature, spatial density feature, echo intensity feature, spectral peak feature, and motion trend feature of the potential moving target point within the sampling window. For each potential moving target point, the zero-crossing rate and displacement integral value of the corresponding potential moving target point are calculated based on the velocity oscillation characteristics in the corresponding time-series target feature vector. Obtain a preset lower threshold for zero-crossing rate and an upper threshold for displacement integral. Use the lower threshold for zero-crossing rate and the upper threshold for displacement integral to filter potential moving target points corresponding to each time-series target feature vector. Mark the potential moving target points corresponding to the time-series target feature vectors whose zero-crossing rate is higher than the lower threshold for zero-crossing rate and whose displacement integral value is lower than the upper threshold for displacement integral as in-situ oscillation points, and use the in-situ oscillation points as the first interference target points. Obtain a preset neighborhood radius for each point cluster, perform spatial clustering analysis on each potential moving target point based on the neighborhood radius to form multiple target point clusters, and select potential moving target points that do not belong to any of the target point clusters as isolated noise points; A preset effective point cluster quantity threshold is obtained, and the number of potential moving target points in each target point cluster is counted. Target point clusters with a number of potential moving target points lower than the effective point cluster quantity threshold are identified as micro target point clusters, and each potential moving target point in each micro target point cluster is marked as a micro target point. A time continuity test is performed on each of the isolated noise points and each of the tiny target points, so that each of the isolated noise points and each of the tiny target points that pass the time continuity test is used as the second interference target point; Extract the corresponding velocity oscillation features and motion trend features from the temporal target feature vectors corresponding to each potential moving target point, and predict the motion state of each potential moving target point based on the velocity oscillation features and the motion trend features to obtain the current motion state prediction value of each potential moving target point. Motion state extraction is performed on each of the potential moving target points to obtain the current motion state measurement value of each potential moving target point. The error between the current motion state prediction value and the current motion state measurement value of each potential moving target point is calculated to obtain the motion error of each potential moving target point. A preset true motion error threshold is obtained, and the motion state of each potential motion target point is determined by the motion error of each potential motion target point using the true motion error threshold. Potential motion target points with motion errors higher than the true motion error threshold are marked as false motion points, and the false motion points are used as third interference target points. The first, second, and third interference target points are integrated and deduplicated to obtain multiple interference target points. Each interference target point is removed from the potential moving target points, and the remaining potential moving target points are used as effective rolling stone target points. The effective rolling stone target points are then integrated to form a set of effective rolling stone target points.

5. The edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method according to claim 1, characterized in that, The process involves acquiring a preset quaternion rotation matrix and digital elevation model, extracting the polar coordinates of each valid rolling stone target point from the set of valid rolling stone target points, converting the polar coordinates of each valid rolling stone target point into three-dimensional coordinates based on the quaternion rotation matrix and the digital elevation model, forming a set of valid rolling stone target three-dimensional points, performing spatial clustering calculations on the set of valid rolling stone target three-dimensional points to obtain the centroid coordinates of multiple valid rolling stone targets, forming the three-dimensional spatial motion trajectory of each valid rolling stone target based on the centroid coordinates of the valid rolling stone targets, and sending the three-dimensional spatial motion trajectory of each valid rolling stone target to the cloud monitoring center, including: The radar parameters of the radar sensing component and the preset geodetic coordinate system are obtained. A corresponding radar coordinate system is established based on the radar parameters of the radar sensing component. The quaternion rotation matrix between the radar coordinate system and the geodetic coordinate system is calculated based on the radar parameters. The radar parameters include the intrinsic parameters of the radar sensing component, the installation position parameters of the radar sensing component, and the installation attitude angle parameters of the radar sensing component. The polar coordinates of each effective rolling stone target point are extracted from the effective rolling stone target point set, and the polar coordinates of each effective rolling stone target point are converted into the corresponding rectangular coordinates of the target point, wherein the rectangular coordinates of the target point are the coordinates of the effective rolling stone target point in the radar coordinate system; Based on the quaternion rotation matrix, the rectangular coordinates of each effective rolling stone target point are converted into the corresponding two-dimensional coordinates of the target point to form a set of effective rolling stone target two-dimensional points, wherein the two-dimensional coordinates of the target point are the planar coordinates of the effective rolling stone target point in the geodetic coordinate system. A preset digital elevation model is obtained. The two-dimensional coordinates of the target points in the effective rolling stone target two-dimensional point set are input into the digital elevation model to extract the slope elevation data corresponding to each target point two-dimensional coordinate. The slope elevation data corresponding to each target point two-dimensional coordinate is added to each target point two-dimensional coordinate to obtain the corresponding effective rolling stone target three-dimensional coordinate. The effective rolling stone target three-dimensional coordinates are integrated to obtain the effective rolling stone target three-dimensional point set. Obtain a preset rolling stone neighborhood search radius, calculate the three-dimensional Euclidean distance between each effective rolling stone target point in the effective rolling stone target three-dimensional point set, and use the rolling stone neighborhood search radius to perform distance screening on the three-dimensional Euclidean distance between each effective rolling stone target point, so as to complete spatial clustering of two adjacent effective rolling stone target points whose three-dimensional Euclidean distance does not exceed the neighborhood search radius, and obtain multiple spatial cluster point clusters. Each of the spatial clusters of points is taken as an effective rolling stone target. The corresponding geometric centroid point is identified for each effective rolling stone target. The three-dimensional coordinates of the geometric centroid point corresponding to each effective rolling stone target in the geodetic coordinate system are taken as the centroid coordinates of each effective rolling stone target. For each effective rolling stone target, the centroid coordinates at each moment within the sampling window are obtained, and the centroid coordinates at each moment are connected sequentially in chronological order to form the three-dimensional spatial motion trajectory of each effective rolling stone target. A rolling stone ID number is added to the three-dimensional spatial motion trajectory of each effective rolling stone target. The three-dimensional spatial motion trajectories of each effective rolling stone target are integrated based on the rolling stone ID number of each effective rolling stone target. The integrated three-dimensional spatial motion trajectories of each effective rolling stone target are then sent to the cloud monitoring center via wireless communication.

6. The edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method according to claim 1, characterized in that, In the cloud monitoring center, based on the three-dimensional spatial motion trajectory of each effective rolling stone target, an effective rolling stone target motion equation is established for each effective rolling stone target. This equation is then used to predict the threat area of ​​each effective rolling stone target, resulting in a predicted threat area for each target. Finally, a threat level determination is made for each predicted threat area to obtain the threat level of each effective rolling stone target, including: In the cloud monitoring center, the effective rolling stone target motion equations are established for each effective rolling stone target based on the three-dimensional spatial motion trajectory of each effective rolling stone target. Based on the effective rolling stone target motion equations of each effective rolling stone target, the corresponding predicted rolling stone landing speed, predicted rolling stone impulse, predicted rolling stone rebound height and predicted rolling stone landing point coordinates are calculated for each effective rolling stone target. Obtain the preset threat area radius, take the predicted boulder landing point coordinates as the center and the threat area radius as the radius, calculate the circular planar area corresponding to each effective boulder target on the horizontal plane, and take the circular planar area corresponding to each effective boulder target as the predicted threat area of ​​each effective boulder target; Obtain a preset threat level comparison table, and determine the threat level of the predicted rolling stone landing speed, the predicted rolling stone impulse, the predicted rolling stone rebound height, and the predicted threat area based on the threat level comparison table, so as to obtain the threat level of each of the effective rolling stone targets.

7. The edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method according to claim 1, characterized in that, Based on the threat level of each valid rolling stone target, a corresponding graded early warning instruction is generated and sent to the corresponding threat area management platform. The three-dimensional spatial motion trajectory of each valid rolling stone target, the predicted threat area of ​​each valid rolling stone target, and the threat level of each valid rolling stone target are then visualized, including: Obtain the jurisdiction area of ​​each regional management platform, and solve the intersection of the jurisdiction area of ​​each regional management platform with the predicted threat area of ​​each effective rolling stone target to obtain the intersection jurisdiction area and the intersection predicted threat area. Based on the jurisdiction of each intersection, the corresponding regional management platform is selected, and the regional management platform corresponding to each jurisdiction is used as the threat area management platform. Based on the predicted threat areas of each intersection, the corresponding effective rolling stone targets are selected. Based on the threat level of the effective rolling stone targets corresponding to each predicted threat area, the graded early warning instructions corresponding to each predicted threat area are generated. The tiered early warning instructions are transmitted one by one to the threat area management platforms of each overlapping jurisdiction through wireless communication. The three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target are compiled into a rolling stone threat motion view, which is then visualized in the cloud monitoring center.

8. A cloud-edge collaborative rolling stone three-dimensional trajectory full lifecycle monitoring system, characterized in that, The method for end-to-end cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring as described in any one of claims 1 to 7 includes: The radar signal transceiver unit is used to transmit microwave radar signals to the target slope to acquire the original echo signal of the target slope within the sampling window, preprocess and digitize the original echo signal to obtain the original radar echo data and send it to the edge computing node. The rolling stone target calculation unit is used in the edge computing node to perform moving target detection on the original radar echo data, detect multiple potential moving target points, perform Doppler oscillation feature analysis, spatiotemporal continuity feature analysis and motion state feature analysis on each potential moving target point, so as to remove interfering target points from each potential moving target point, obtain effective rolling stone target points, and integrate each effective rolling stone target point into an effective rolling stone target point set; The rolling stone trajectory calculation unit is used to obtain a preset quaternion rotation matrix and digital elevation model, extract the target point polar coordinates corresponding to each effective rolling stone target point from the effective rolling stone target point set, convert the target point polar coordinates of each effective rolling stone target point into target point three-dimensional coordinates based on the quaternion rotation matrix and the digital elevation model, form an effective rolling stone target three-dimensional point set, perform spatial clustering calculation on the effective rolling stone target three-dimensional point set to obtain the centroid coordinates of multiple effective rolling stone targets, form the three-dimensional spatial motion trajectory of each effective rolling stone target based on the centroid coordinates of the effective rolling stone targets, and send the three-dimensional spatial motion trajectory of each effective rolling stone target to the cloud monitoring center; The threat level calculation unit is used in the cloud monitoring center to establish the effective rolling stone target motion equations of each effective rolling stone target based on the three-dimensional spatial motion trajectory of each effective rolling stone target, and to use the effective rolling stone target motion equations to predict the threat area of ​​each effective rolling stone target, so as to obtain the predicted threat area of ​​each effective rolling stone target, and to determine the threat level of each effective rolling stone target based on the predicted threat area of ​​each effective rolling stone target. The early warning instruction issuing unit is used to generate corresponding graded early warning instructions based on the threat level of each effective rolling stone target, and send the graded early warning instructions to the corresponding threat area management platform, and visualize the three-dimensional spatial motion trajectory of each effective rolling stone target, the predicted threat area of ​​each effective rolling stone target, and the threat level of each effective rolling stone target.

9. An electronic device, characterized in that, The device includes a memory, a processor, and a transceiver that are connected in sequence. The memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer programs and execute the edge-cloud collaborative rolling stone three-dimensional trajectory full lifecycle monitoring method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or the instructions are executed by the computer, they implement the edge-cloud collaborative rolling stone three-dimensional trajectory full life cycle monitoring method as described in any one of claims 1 to 7.

Images