Dam deformation remote control cruise monitoring vehicle and prediction and early warning method

Abstract

The invention relates to the field of water conservancy and hydropower engineering equipment, and provides a dam deformation remote control cruise monitoring vehicle and a prediction and early warning method. Wherein the cruise monitoring vehicle comprises a laser radar device, an RTK device, a communication device and electronic equipment, and realizes automatic cruise and data acquisition and transmission. The prediction and early warning method comprises the implementation steps that firstly, dam crest point cloud data of a dam is collected, and a dam crest three-dimensional scene model is constructed; thirdly, a path multi-objective optimization function is constructed based on dam crest monitoring points, a genetic algorithm is used for path optimization, and a cruise monitoring path is formulated; then, collecting positioning data of the monitoring points, calculating settlement values and settlement rates of the monitoring points, and obtaining settlement indexes and deformation rates of the dam; and finally, according to the settlement index and the deformation rate, judging whether to give an early warning, and predicting a future settlement value and the deformation rate. According to the unmanned vehicle and the method, the precision and efficiency of dam safety monitoring can be effectively improved, and a scientific basis is provided for dam maintenance and management.

Description

Technical Field

[0001] The invention relates to the field of water conservancy and hydropower engineering equipment, and in particular to a remote-controlled cruise monitoring vehicle for dam deformation and a prediction and early warning method. Background Art

[0002] Levees are important infrastructure in the fields of flood control, irrigation and hydropower generation. Their safety is directly related to the safety of people's lives and property and the sustainable development of the regional economy. With the continuous expansion and complexity of water conservancy projects, the operation of levees may be affected by factors such as changes in water pressure, changes in geological conditions, climate change and material aging, which may cause levee deformation phenomena such as displacement, settlement or crack expansion, and even lead to levee failure. Therefore, levee deformation monitoring is of great significance. As an important part of levee safety management, it is related to the safety of engineering and the safety of life and property of downstream people. Regular deformation monitoring of levees can provide data support for the maintenance, reinforcement and repair of levees, ensure their long-term stable operation, and ensure the safety and reliability of water conservancy projects.

[0003] At present, the existing technologies mainly monitor dams through leveling, trigonometric height measurement and synthetic aperture radar (InSAR) technology; among them, the leveling method measures the height difference by using a level instrument to infer the elevation change, and has high measurement accuracy, but the field workload is large and the visibility conditions are high; the trigonometric height measurement method can measure complex terrain, but the accuracy is easily affected by atmospheric refraction; InSAR technology can perform large-area, high-precision and rapid monitoring, but the data processing is complex, and it is easily affected by many factors such as noise interference, terrain undulations and atmospheric delay effects during the monitoring process, thereby reducing the accuracy of the monitoring results.

[0004] To this end, the present invention proposes a remote-controlled cruise monitoring vehicle for dam deformation and a prediction and early warning method. The method is based on a monitoring vehicle with remote control and cruise functions. Through the high-precision positioning device on board, positioning data is obtained and analyzed in real time to achieve prediction and early warning of dam deformation and ensure the safe operation of the dam. Summary of the invention

[0005] The purpose of the present invention is to provide a remote control cruise monitoring vehicle for dam deformation and a prediction and early warning method to solve the problems of point-based monitoring of dam deformation, large workload of manual monitoring, high cost of burying multiple instruments, and low efficiency of prediction and early warning analysis. In order to solve the problems existing in the prior art, the present invention first collects the cloud data of the dam top through the laser radar device of the cruise monitoring vehicle to construct a three-dimensional scene model of the dam top; then, a multi-objective optimization function of the path is constructed based on the dam top monitoring point, and a genetic algorithm is used to optimize the path to formulate the monitoring path of the cruise monitoring vehicle; then, the positioning data of the monitoring point is collected through the RTK device of the cruise monitoring vehicle, and the settlement value and settlement rate of the monitoring point are calculated to obtain the settlement index and deformation rate of the dam; finally, whether to issue an early warning is determined according to the settlement index and deformation rate, and the future settlement value and deformation rate are predicted. The unmanned vehicle and method can effectively improve the accuracy and efficiency of dam safety monitoring and provide a scientific basis for the maintenance and management of the dam.

[0006] A dam deformation prediction and early warning method, comprising:

[0007] S1. Determine the monitoring section of the dam according to the dam crest information of the target dam;

[0008] S2. Collect the monitoring section of the dam The dam vertex cloud data is assembled and corrected to obtain a dam vertex cloud set; the scene points are clustered based on the point space spacing and reflection intensity of each scene point in the dam vertex cloud set to generate a scene point cloud cluster and perform geometric feature fitting to obtain a dam top three-dimensional scene model;

[0009] S3. Based on the 3D scene model of the dam crest, set the dam crest monitoring points, calculate the monitoring point spacing between the dam crest monitoring points, and construct a path multi-objective optimization function; perform iterative optimization through a genetic algorithm to obtain a monitoring planning path;

[0010] S4. Collect the positioning coordinate data of the engineering reference point, and calculate the coordinate positioning deviation based on the reference coordinate data;

[0011] S5. Collect monitoring point positioning data, obtain the monitoring point positioning data within a predetermined historical time, and calculate the monitoring point settlement value and the monitoring point settlement rate; combine the monitoring point positioning data to obtain the embankment settlement index; construct a spatial deformation curve based on the monitoring point settlement value and the monitoring point positioning data, and calculate the spatial deformation rate; combine the monitoring point settlement rate to obtain the embankment deformation rate;

[0012] S6. Determine whether to issue an early warning based on the dam settlement index and the dam deformation rate; obtain historical dam crest deformation data, calculate the deformation space correlation, and construct a dam crest space deformation map; combine the graph convolutional neural network, the long short-term memory network and the prediction head to predict the future settlement value and future settlement rate in the future.

[0013] Preferably, the specific implementation process of obtaining the dam vertex cloud set and generating the scene point cloud cluster includes:

[0014] The monitoring section of the dam is collected by remotely controlling the cruise monitoring vehicle. The dam vertex cloud data is obtained; the acquisition timestamp of the dam vertex cloud data is obtained, and the plurality of groups of the dam vertex cloud data are sorted according to the acquisition timestamp; the first dam vertex cloud data is obtained according to the dam vertex cloud data with the earliest acquisition timestamp; and the other The difference in the acquisition timestamp between the dam vertex cloud data of the group and the first dam vertex cloud data is obtained. Combined with the running speed and running direction of the cruise monitoring vehicle, calculate The coordinate difference is collected and The dam vertex cloud data of the group are calibrated; the first dam vertex cloud data and the calibrated dam vertex cloud data are preprocessed and spliced ​​to obtain the dam vertex cloud set; according to the coordinates of each scene point in the dam vertex cloud set, the distance between different scene points is calculated to obtain the point space spacing; the laser reflection angle and the reflection intensity of the scene point are obtained, the direct reflection intensity is obtained through the reflection property of the object, and the direct intensity difference between the scene points is calculated; the scene points are clustered in combination with the point space spacing and the direct intensity difference; two scenes whose point space spacing is less than a predetermined spacing threshold and whose direct intensity difference is less than a predetermined intensity threshold are divided into the same cluster to obtain several groups of scene point cloud clusters; the scene point cloud clusters are geometrically fitted to obtain the dam top three-dimensional scene model.

[0015] Preferably, the specific implementation process of constructing the path multi-objective optimization function includes:

[0016] The coordinates of the dam top monitoring points are obtained, and the distances between the different dam top monitoring points are calculated to obtain the monitoring point spacing; according to the dam top monitoring points and the monitoring point spacing, a path multi-objective optimization function is constructed, including a monitoring point density optimization function and a path length optimization function; the specific calculation formula of the path multi-objective optimization function is:

[0017] ;

[0018] in, represents the optimal solution function for the cruise path; Indicates the cruising path of the cruise monitoring vehicle; Represents the path multi-objective optimization function; represents the monitoring point density optimization function; represents the path length optimization function; express The adjustment factor of express The adjustment factor of

[0019] The multi-objective optimization function of the path is iteratively optimized through a genetic algorithm to obtain the monitoring planning path.

[0020] Preferably, the specific implementation process of obtaining the dam settlement index includes:

[0021] The cruise monitoring vehicle arrives at the dam top monitoring point according to the monitoring planning path, collects the position and elevation information, and performs position correction through the coordinate positioning deviation to obtain the monitoring point positioning data; obtains the monitoring point positioning data of the dam top monitoring point within the predetermined historical time, and obtains the settlement value and the settlement rate of the monitoring point through the elevation difference of each dam top monitoring point within the predetermined historical time; calculates the embankment settlement index by combining the position distribution of the dam top monitoring point and the settlement value of the monitoring point, and the specific formula is expressed as follows:

[0022] ;

[0023] ;

[0024] in, represents the settlement index of the dam; Indicates The settlement value of each monitoring point of the dam crest monitoring point; Indicates The weight coefficient of each of the dam crest monitoring points; Indicates the number of monitoring points on the dam crest; represents a power function with a natural constant as base; represents the minimum function; Indicates Coordinate vectors of the dam top monitoring points; The target dam The coordinate vectors of the core points on the dam crest, including the center point and the edge point of the dam crest; represents the scaling factor; || ||2 represents the two-norm function.

[0025] Preferably, the specific process of obtaining the dam deformation rate includes:

[0026] Based on the monitoring point settlement value of the dam top monitoring point on the monitoring planning path and the monitoring point positioning data, the spatial deformation curve is generated by a linear interpolation method, and the spatial deformation rate of each dam top monitoring point is calculated; the embankment deformation rate is calculated by combining the monitoring point settlement rate and the spatial deformation rate.

[0027] Preferably, the specific process of determining whether to issue an early warning according to the dam settlement index and the dam deformation rate includes:

[0028] If the dam settlement index is greater than or equal to the predetermined settlement threshold, a red warning is issued to the management personnel; if the dam settlement index is less than the predetermined settlement threshold, it is determined whether the dam deformation rate exceeds the threshold; if the dam deformation rate is greater than or equal to the predetermined rate threshold, a yellow warning is issued to the management personnel; if the dam deformation rate is less than the predetermined rate threshold, the target dam is continuously monitored.

[0029] Preferably, the specific process of predicting the future sedimentation value and the future sedimentation rate at a future time includes:

[0030] Acquire the historical dam crest deformation data of the target dam and perform preprocessing operations to obtain a historical dam crest deformation sequence; the historical dam crest deformation sequence includes the position coordinates of the monitoring point, the settlement value of the monitoring point and the settlement rate of the monitoring point;

[0031] According to the position coordinates of different monitoring points in the historical dam crest deformation sequence, the deformation spatial correlation between different dam crest monitoring points is calculated; the dam crest spatial deformation map is constructed by combining the spatial correlation and the dam crest monitoring points; the dam crest spatial deformation map includes deformation nodes and edges; the deformation nodes are deformation feature sequences of the dam crest monitoring points, including the settlement values ​​of the monitoring points at different times and the settlement rates of the monitoring points; the deformation nodes are connected by the edges, and the weights of the edges are the spatial correlation; the dam crest spatial deformation map is input into a graph convolutional neural network to capture the spatial dependency between the deformation feature sequences of different deformation nodes, and obtain the dam crest deformation spatial features;

[0032] Inputting the spatial characteristics of the dam crest deformation into a long short-term memory network, capturing the temporal correlation of the deformation feature sequences of different dam crest monitoring points, and obtaining the temporal characteristics of the dam crest deformation;

[0033] The time series characteristics of the dam crest deformation are input into the prediction head to obtain the future settlement value and the future settlement rate of each dam crest monitoring point of the target dam at a future time.

[0034] A remote-controlled cruise monitoring vehicle for dam deformation comprises a vehicle body, a driving device, driving wheels, electronic equipment, a battery, an automatic charging interface, a laser radar device, an RTK device and a communication device; the electronic equipment comprises a memory, a processor and a dam deformation prediction and early warning program stored in the memory and executable on the processor; when the processor executes the dam deformation prediction and early warning program, the dam deformation prediction and early warning method as described above is implemented.

[0035] Preferably, the driving device, the electronic device and the battery are located inside the vehicle body; the laser radar device, the RTK device and the communication device are located on the top of the vehicle body; and the automatic charging port is located at the rear of the vehicle body.

[0036] Preferably, the communication device is equipped with a 4G / 5G network, and receives and transmits instructions and data; the laser radar device is used for road modeling and obstacle avoidance during operation; the RTK device is used for high-precision positioning, including two information: position and elevation; the electronic device is used for data processing and backup operations; the automatic charging interface is used to charge the cruise monitoring vehicle.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] 1. The present invention collects the dam top cloud data of the dam monitoring section through a remote-controlled cruise monitoring vehicle, and performs data correction in combination with the collection timestamp, which significantly improves the spatial accuracy and time consistency of the dam top cloud data; through the calculation and cluster analysis of point space spacing and direct intensity difference, it can accurately identify and classify scene point cloud clusters, effectively reducing noise interference; through the construction of a three-dimensional scene model through geometric feature fitting, it can truly and comprehensively reflect the structural characteristics of the dam top, provide high-precision three-dimensional data support for dam monitoring, deformation analysis and engineering management, and improve the efficiency of dam monitoring.

[0039] 2. The present invention collects the location information and elevation information of the dam top monitoring point along the monitoring planning path through a cruise monitoring vehicle, uses the coordinate positioning deviation correction method to improve the accuracy of the monitoring point positioning data, and obtains the settlement value and settlement rate of the monitoring point through historical monitoring data analysis, thereby realizing an accurate assessment of the dam settlement state. The dam settlement index is calculated based on the settlement value and location distribution of the monitoring point, and the dam deformation rate is calculated in combination with the settlement rate of the monitoring point and the spatial deformation rate, which effectively improves the accuracy of dam deformation monitoring.

[0040] 3. The present invention combines the dam settlement index and deformation rate to monitor the deformation state of the dam in real time and issue early warnings in a timely manner. When the settlement index reaches or exceeds the predetermined threshold, a red warning can be quickly issued to ensure that management personnel take immediate countermeasures when the dam is seriously deformed; and when the settlement index does not reach the warning value, by monitoring the dam deformation rate and judging whether it exceeds the threshold, a yellow alarm is further issued or continuous monitoring is carried out, which greatly improves the accuracy of the early warning and the timeliness of the response. In addition, by combining historical deformation data, spatial correlation, graph convolutional neural networks and long short-term memory networks, the spatial dependence and temporal correlation of deformation characteristics can be deeply captured, thereby improving the prediction accuracy, providing more reliable data support for dam safety management, ensuring the safe operation of the dam and reducing the risk of emergencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a general flow chart of a remote control cruise monitoring vehicle for dam deformation and a prediction and early warning method proposed in an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of the dam crest path planning proposed in an embodiment of the present invention;

[0043] Figure 3 A flowchart of dam deformation determination proposed in an embodiment of the present invention;

[0044] Figure 4 A schematic diagram of the internal structure of a remote-controlled cruise monitoring vehicle for dam deformation proposed in an embodiment of the present invention.

[0045] In the figure: 1. Vehicle body; 2. Driving device; 21. Driving wheels; 3. Electronic equipment; 4. Battery; 5. Automatic charging interface; 6. LiDAR device; 7. RTK device; 8. Communication device; 201. Dam top; 202. Monitoring and planning path. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] As an important water conservancy engineering facility, dams are widely used in flood control, irrigation, hydropower generation, water resource regulation and other fields. However, during the long-term operation of dams, due to natural factors such as water level fluctuations, rainfall changes and soil settlement, as well as human factors such as construction quality, insufficient maintenance and load changes, dams may deform to varying degrees, resulting in water leakage and even serious dam breach accidents, posing huge safety hazards to surrounding residents and the ecological environment. Therefore, dam deformation monitoring is a key link to ensure the safe operation of dams. By real-time monitoring of parameters such as displacement, settlement and inclination of dams, abnormal changes in dam structures can be detected in time, and potential safety risks can be warned.

[0048] The present invention proposes a remote control cruise monitoring vehicle for dam deformation and a prediction and early warning method to achieve real-time monitoring of dam deformation and prediction and early warning of dam settlement and uneven deformation. In order to illustrate the effectiveness of the method of the present invention, it will be specifically described in conjunction with the accompanying drawings of this embodiment and the following two embodiments.

[0049] Embodiment 1

[0050] The present application embodiment discloses a dam deformation prediction and early warning method to realize deformation monitoring of the dam top of A earth-rock dam, see Figure 1 The specific implementation steps include: S1. Determine the monitoring section of the dam according to the dam crest information of the target dam; S2. Collect the dam top cloud data of the dam monitoring section, and obtain the dam crest three-dimensional scene model through clustering and geometric feature fitting; S3. Construct a path multi-objective optimization function, optimize it through genetic algorithm, and obtain the monitoring planning path; S4. Collect the positioning coordinate data of the engineering benchmark point, and calculate the coordinate positioning deviation according to the benchmark coordinate data; S5. Calculate the dam settlement index and dam deformation rate based on the historical monitoring point positioning data and the collected monitoring point positioning data; S6. Determine whether to issue an early warning based on the dam settlement index and the dam deformation rate, and predict the future settlement value and future settlement rate in the future.

[0051] Optionally, the image of the crest of earth-rock dam A is collected by a high-definition camera mounted on a drone, and the flatness of the crest of earth-rock dam A is obtained by combining an image processing algorithm; then, based on the flatness, it is determined whether it meets the operation requirements of the cruise monitoring vehicle; if it meets the requirements, the crest of earth-rock dam A is continuously monitored for deformation by the remote-controlled cruise monitoring vehicle; based on the total length of the crest of earth-rock dam A, the monitoring section of the dam is determined, corresponding to the above step S1; for example, the crest of dam A is 200 meters long. The corresponding monitoring section of the dam is 200 m long. .

[0052] Furthermore, the monitoring section of the dam is collected The dam vertex cloud data is assembled and corrected to obtain a dam vertex cloud set; the scene points are clustered in combination with the point space spacing and reflection intensity of each scene point in the dam vertex cloud set to generate a scene point cloud cluster and perform geometric feature fitting to obtain a dam top three-dimensional scene model; corresponding to the above step S2, the specific process includes:

[0053] The laser radar device carried by the remote control cruise monitoring vehicle collects the data of the dam monitoring section. The dam vertex cloud data is obtained; the acquisition timestamp of the dam vertex cloud data is obtained, and the plurality of groups of the dam vertex cloud data are sorted according to the acquisition timestamp; the first dam vertex cloud data is obtained according to the dam vertex cloud data with the earliest acquisition timestamp; and the other The difference in the acquisition timestamp between the dam vertex cloud data of the group and the first dam vertex cloud data is obtained. Combined with the running speed and running direction of the cruise monitoring vehicle, calculate The coordinate difference is collected and The dam vertex cloud data of the group is calibrated; the specific calculation formula of the acquisition coordinate difference is:

[0054] ;

[0055] ;

[0056] in, Indicates The acquisition coordinate difference of the cloud data of the dam vertex Axis coordinate difference; Indicates The acquisition coordinate difference of the cloud data of the dam vertex Axis coordinate difference; Indicates The acquisition time difference of the cloud data of the dam top is set; Indicates the movement speed of the cruise monitoring vehicle; Indicates the running direction of the cruise monitoring vehicle The angle between the axes and ; The coordinates of the dam vertex cloud data include axis, Axis and axis, The positive direction of the axis is the heading of the cruise monitoring vehicle. The positive direction of the axis is perpendicular to axis and is located on the left side of the cruise monitoring vehicle, The positive direction of the axis is perpendicular to Axis and axis and is located on the upper side of the cruise monitoring vehicle, and the center point of the laser radar device is the coordinate origin;

[0057] The first dam vertex cloud data and the calibrated dam vertex cloud data are preprocessed and spliced ​​to obtain the dam vertex cloud set; according to the coordinates of each scene point in the dam vertex cloud set, the distance between different scene points is calculated to obtain the point space spacing; the laser reflection angle and the reflection intensity of the scene point are obtained, the direct reflection intensity is obtained by the reflection property of the object, and the direct intensity difference between the scene points is calculated; the formula of the direct reflection intensity is:

[0058] ;

[0059] in, Indicates The direct reflection intensity of each of the scene points; Indicates the reflection intensity of each of the scene points; Indicates the distance from the center point of the laser radar device to the The distance between the scene points; Indicates the distance from the center point of the laser radar device to the The angle between the connecting line of the scene points and the normal line of the scene points is ;

[0060] The scene points are clustered in combination with the point space spacing and the direct intensity difference; the two scenes whose point space spacing is less than a predetermined spacing threshold and whose direct intensity difference is less than a predetermined intensity threshold are divided into the same cluster to obtain several groups of scene point cloud clusters; geometric feature fitting is performed on the scene point cloud clusters to obtain the dam top three-dimensional scene model.

[0061] In order to clearly illustrate the data format collected by the laser radar device of the cruise monitoring vehicle, refer to Table 1 which exemplarily gives some of the dam vertex cloud data.

[0062] Table 1. Cloud data of some dam vertices

[0063] In Table 1, the scene points are different acquisition points in the dam vertex cloud data, and the laser number represents the number of the laser beam of the laser radar device. The total number of laser numbers is related to the model of the laser radar device.

[0064] The embodiment of the present application collects and corrects the dam top cloud data of the dam monitoring section through a remote control cruise monitoring vehicle, which can accurately obtain the three-dimensional point cloud information of the dam top and ensure the timeliness and accuracy of the data. By sorting and calibrating the point cloud data according to the acquisition timestamp, the data deviation caused by the cruise vehicle movement can be effectively eliminated, and the spatial consistency and reliability of the point cloud data can be improved; combined with the point space spacing and laser reflection intensity, cluster analysis is used to divide the scene points into different clusters, and geometric feature fitting is performed to generate an accurate three-dimensional scene model of the dam top, which is helpful for subsequent dam deformation analysis and early warning.

[0065] Furthermore, based on the dam top three-dimensional scene model, dam top monitoring points are set, the monitoring point spacing between the dam top monitoring points is calculated, and a path multi-objective optimization function is constructed; iterative optimization is performed through a genetic algorithm to obtain a monitoring planning path 202; corresponding to the above step S3, the specific implementation process includes:

[0066] The coordinates of the dam top monitoring points are obtained, and the distances between the different dam top monitoring points are calculated to obtain the monitoring point spacing; according to the dam top monitoring points and the monitoring point spacing, a path multi-objective optimization function is constructed, including a monitoring point density optimization function and a path length optimization function; the specific calculation formula of the path multi-objective optimization function is:

[0067] ;

[0068] ;

[0069] ;

[0070] in, represents the optimal solution function for the cruise path; Indicates the cruising path of the cruise monitoring vehicle; Represents the path multi-objective optimization function; represents the monitoring point density optimization function; represents the path length optimization function; express The adjustment factor of express The adjustment factor of and The value of Inside; Indicates The number of dam crest monitoring points in the cruise path; Indicates The length of a cruise path is obtained by summing the distances between adjacent dam crest monitoring points on the cruise path; Indicates the total number of cruise paths;

[0071] The path multi-objective optimization function is iteratively optimized through a genetic algorithm to obtain the monitoring planning path 202.

[0072] Specifically, considering that the deformation in the middle of the dam crest of the earth-rock dam A is large, the monitoring planning path 202 needs to encrypt the middle path; Figure 2 The monitoring planning path 202 set on the earth-rock dam crest 201 encrypts the middle position. This path encryption method enables repeated measurement of the middle monitoring point to improve the measurement accuracy.

[0073] The embodiment of the present application can effectively improve the efficiency and accuracy of dam crest monitoring through a multi-objective optimization method of the path based on the spacing between monitoring points. The use of genetic algorithms for path planning optimization can optimize the length of the cruise path while ensuring a reasonable distribution of monitoring point density, ensuring that the monitoring process can cover all monitoring points.

[0074] Furthermore, the cruise monitoring vehicle collects the positioning coordinate data of the engineering benchmark point through the onboard device, calculates the coordinate positioning deviation according to the benchmark coordinate data of the engineering benchmark point, and Figure 1 The S4 step.

[0075] Further, the monitoring point positioning data is collected, the monitoring point positioning data within a predetermined historical time is obtained, and the monitoring point settlement value and the monitoring point settlement rate are calculated; the dam settlement index is obtained by combining the monitoring point positioning data; corresponding Figure 1 The specific implementation process of step S5 includes:

[0076] The cruise monitoring vehicle arrives at the dam top monitoring point according to the monitoring planning path 202, collects position and elevation information through the carried RTK device, and performs position correction through the coordinate positioning deviation to obtain the monitoring point positioning data; obtains the monitoring point positioning data of the dam top monitoring point within a predetermined historical time, and obtains the settlement value and the settlement rate of the monitoring point through the elevation difference of each dam top monitoring point at the predetermined historical time; combines the position distribution of the dam top monitoring point and the settlement value of the monitoring point to calculate the embankment settlement index, which is specifically expressed as follows:

[0077] ;

[0078] ;

[0079] in, represents the settlement index of the dam; Indicates The settlement value of each monitoring point of the dam crest monitoring point; Indicates The weight coefficient of each of the dam crest monitoring points; Indicates the number of monitoring points on the dam crest; represents a power function with a natural constant as base; represents the minimum function; Indicates Coordinate vectors of the dam top monitoring points; The target dam The coordinate vectors of the core points on the dam crest, including the center point and the edge point of the dam crest; represents the scaling factor, which is obtained by fitting the experimental data; || ||2 represents the two-norm function.

[0080] In order to clearly illustrate the positioning data of the dam top monitoring points collected by the RTK device of the cruise monitoring vehicle, refer to Table 2 which exemplarily gives the positioning data of some dam top monitoring points.

[0081] Table 2. Location data of some dam top monitoring points

[0082] The dam settlement index proposed in the embodiment of the present application can accurately obtain the settlement of the dam in real time and dynamically evaluate the safety of the dam. The use of cruise monitoring vehicles to collect elevation data and correct the position of the dam top monitoring point can effectively eliminate the errors caused by coordinate deviations and improve the accuracy and reliability of the data; by calculating the settlement value and settlement rate of the monitoring point, the deformation trend of the dam within a predetermined historical time period can be obtained. Combined with the spatial distribution and settlement of the dam monitoring points, the calculation of the settlement index can provide a quantitative evaluation indicator for the overall settlement trend of the dam.

[0083] Furthermore, according to the settlement value of the monitoring point and the positioning data of the monitoring point, a spatial deformation curve is constructed, and the spatial deformation rate is calculated; combined with the settlement rate of the monitoring point, the deformation rate of the dam is obtained; corresponding Figure 1 The specific implementation process of step S5 includes:

[0084] Based on the monitoring point settlement value and the monitoring point positioning data of the dam top monitoring point on the monitoring planning path 202, the spatial deformation curve is generated by a linear interpolation method, and the spatial deformation rate of each dam top monitoring point is calculated; the dam deformation rate is calculated by combining the monitoring point settlement rate and the spatial deformation rate. The specific formula is expressed as follows:

[0085] ;

[0086] ;

[0087] in, represents the deformation rate of the dam; Indicates The spatial deformation rate of each of the dam top monitoring points; No. The settlement rate of each monitoring point of the dam top monitoring point; Indicates the number of monitoring points on the dam crest; Indicates the total number of monitoring planning paths 202; Indicates The settlement value of each monitoring point of the dam crest monitoring point; Indicates The monitoring plan path 202 described in Article The settlement value of the monitoring point of the dam crest monitoring point preceding the dam crest monitoring point; Indicates The monitoring plan path 202 described in Article the settlement value of the monitoring point of the dam crest monitoring point next to the dam crest monitoring point; Indicates Coordinate vectors of the dam top monitoring points; Indicates The monitoring plan path 202 described in Article The coordinate vector of the previous dam crest monitoring point of the dam crest monitoring point; Indicates The monitoring plan path 202 described in Article The coordinate vector of the next dam crest monitoring point of the dam crest monitoring point; || ||2 represents the two-norm function.

[0088] The embodiment of the present application generates a spatial deformation curve by combining the settlement value and positioning data of the monitoring point through the linear interpolation method, and calculates the spatial deformation rate, so as to accurately obtain the deformation characteristics and change trend of the dam at different locations. This process can not only monitor the stability of the dam in real time, but also help improve the accuracy and response speed of dam monitoring by calculating the deformation rate of the dam, and ensure the safety and stability of the dam in long-term operation.

[0089] Further, according to the dam settlement index and the dam deformation rate, it is determined whether to issue an early warning, corresponding to Figure 1 Step S6 of Figure 3 The specific process includes:

[0090] If the dam settlement index is greater than or equal to the predetermined settlement threshold, a red warning is issued to the management personnel; if the dam settlement index is less than the predetermined settlement threshold, it is determined whether the dam deformation rate exceeds the threshold; if the dam deformation rate is greater than or equal to the predetermined rate threshold, a yellow warning is issued to the management personnel; if the dam deformation rate is less than the predetermined rate threshold, the target dam is continuously monitored.

[0091] The embodiment of the present application can timely and accurately evaluate the safety status of the dam by combining the dam settlement index and deformation rate, and effectively prevent potential safety hazards. By setting a predetermined settlement threshold and deformation rate threshold, a rapid response can be made when the dam undergoes abnormal settlement or deformation, ensuring that management personnel can take necessary preventive measures at the first time, thereby improving the safety and stability of the dam.

[0092] Furthermore, the historical dam crest deformation data is obtained, the spatial correlation of deformation is calculated, and the dam crest spatial deformation map is constructed; the future settlement value and future settlement rate in the future time are predicted by combining the graph convolutional neural network, long short-term memory network and prediction head; corresponding Figure 1 The S6 step includes:

[0093] Acquire the historical dam crest deformation data of the target dam and perform preprocessing operations to obtain a historical dam crest deformation sequence; the historical dam crest deformation sequence includes the position coordinates of the monitoring point, the settlement value of the monitoring point and the settlement rate of the monitoring point;

[0094] According to the position coordinates of different monitoring points in the historical dam crest deformation sequence and the settlement values ​​of the monitoring points, the deformation spatial correlation between different dam crest monitoring points is calculated. The specific calculation formula is:

[0095] ;

[0096] in, Indicates and The spatial correlation of the deformation between the dam top monitoring points; Indicates The historical monitoring point settlement value sequence of the dam crest monitoring point and The historical monitoring point settlement value sequence of the dam crest monitoring point The covariance of express The variance of express The variance of represents a power function with a natural constant as base; Indicates and The distance between the monitoring points on the dam top; It is the distance adjustment parameter, which indicates the influence of distance on the degree of deformation. .

[0097] Combining the spatial correlation and the dam top monitoring points, constructing the dam top spatial deformation map; the dam top spatial deformation map includes deformation nodes and edges; the deformation nodes are deformation feature sequences of the dam top monitoring points, including the settlement values ​​of the monitoring points at different times and the settlement rates of the monitoring points; the deformation nodes are connected by the edges, and the weights of the edges are the spatial correlation; the dam top spatial deformation map is input into a graph convolutional neural network to capture the spatial dependency between the deformation feature sequences of different deformation nodes, and obtain the dam top deformation spatial features;

[0098] Inputting the spatial characteristics of the dam crest deformation into a long short-term memory network, capturing the temporal correlation of the deformation feature sequences of different dam crest monitoring points, and obtaining the temporal characteristics of the dam crest deformation;

[0099] The time series characteristics of the dam crest deformation are input into the prediction head to obtain the future settlement value and the future settlement rate of each dam crest monitoring point of the target dam at a future time.

[0100] In order to clearly illustrate the accuracy of the dam crest deformation prediction proposed by the present invention, the prediction results of some dam crest monitoring points are given as examples with reference to Table 3.

[0101] Table 3. Prediction results of some dam top monitoring points

[0102] The embodiment of the present application can accurately capture the spatial and temporal characteristics of the dam crest deformation data by combining the graph convolutional neural network and the long short-term memory network, and effectively predict the future settlement value and settlement rate of the target dam. By obtaining the historical dam crest deformation data and preprocessing it, calculating the spatial correlation between the monitoring points and constructing a spatial deformation map, it can comprehensively reflect the deformation characteristics between different areas of the dam, ensuring that the model can accurately predict the future settlement trend, so as to predict that the deformation may affect the safety of the dam operation, and take engineering and non-engineering measures in advance.

[0103] The embodiment of the present application collects the dam top cloud data of the dam monitoring section through a remote-controlled cruise monitoring vehicle, and combines the dam top monitoring point data to achieve efficient monitoring and early warning of dam deformation. The specific implementation process mainly includes the following steps: (1) spatial modeling of the monitoring section; (2) automatic cruise deformation monitoring; (3) monitoring early warning and deformation prediction. The present invention proposes point cloud clustering modeling, dam settlement index and deformation rate, and dam top spatial deformation map for the above three processes; the point cloud clustering modeling realizes high-precision dam top three-dimensional scene reconstruction and improves the reliability of dam monitoring data; the dam settlement index and deformation rate further improve the accuracy of dam deformation monitoring; the dam top spatial deformation map improves the accuracy of dam deformation prediction and provides technical support for the safety monitoring and management of dams.

[0104] Embodiment 2

[0105] In the first embodiment, the method of the present invention realizes the deformation monitoring of the dam crest of earth-rock dam A. In the embodiment of the present application, a remote control cruise monitoring vehicle for dam deformation proposed by the present invention will be described;

[0106] The electronic device 3 includes a memory, a processor, and a dam deformation prediction and early warning program stored in the memory and executable on the processor. When the processor executes the dam deformation prediction and early warning program, Figure 1 The dam deformation prediction and early warning method.

[0107] The cruise monitoring vehicle provides a real-time and efficient monitoring method for dam deformation monitoring, which can continuously track and warn the deformation of the dam. Through the coordinated work of these devices, the cruise monitoring vehicle can achieve all-round and high-precision deformation monitoring of the dam, provide reliable data support, and provide effective guarantee for the safe management and maintenance of the dam.

[0108] Further, see Figure 4 The driving device 2, the electronic device 3 and the battery 4 of the cruise monitoring vehicle are located inside the vehicle body 1; the laser radar device 6, the RTK device 7 and the communication device 8 are located on the top of the vehicle body 1, and the driving wheel 21 is located at the bottom of the vehicle body 1; the automatic charging interface 5 is located at the rear of the vehicle body 1.

[0109] The embodiment of the present application optimizes the functional layout and working efficiency of the cruise monitoring vehicle through the above-mentioned design, ensures that each device can play the best effect, and ensures the stability and safety of the vehicle body. The drive device, electronic equipment and battery are placed inside the vehicle body, which can effectively protect these key components from the influence of the external environment; the design of placing the laser radar device, RTK device and communication device on the top of the vehicle body can improve the working performance of these devices; the design of the automatic charging interface at the rear of the vehicle body can facilitate the automatic charging of the cruise monitoring vehicle when it is docked or returned to the charging station, avoiding the occupation of the front space.

[0110] Furthermore, the communication device 8 is equipped with a 4G / 5G network, and receives and transmits instructions and data; the laser radar device 6 is used for road modeling and obstacle avoidance during operation; the RTK device 7 is used for high-precision positioning, including two information: position and elevation; the electronic device 3 is used for data processing and backup operations; the automatic charging interface 5 is used to charge the cruise monitoring vehicle.

[0111] Specifically, the communication device 8 receives instructions from the management personnel through the 4G / 5G network. After receiving the instructions, the driving device 2 in the vehicle body 1 is started, and the cruise monitoring vehicle begins to move along the predetermined path to the top of the dam. During the cruising process, the laser radar device 6 continuously scans the surrounding environment to detect whether there are any obstacles ahead; if an obstacle is found, the laser radar system will immediately feedback information to the electronic device 3. The electronic device 3 will analyze the location of the obstacle through the processor and instruct the driving device 2 to adjust the path to avoid the obstacle, ensuring that the vehicle body can smoothly travel to the top of the dam. During the cruising process on the dam top, the laser radar device 6 collects point cloud data of the surrounding environment and transmits it to the electronic device 3. The processor will calculate the location of the obstacle based on the information. Figure 1 The dam deformation prediction and early warning program in step S2 models the point cloud data to obtain a three-dimensional scene model of the dam crest. The communication device 8 transmits the data to the management personnel through the 4G / 5G network. The management personnel sets the dam crest monitoring points according to the three-dimensional scene model of the dam crest and sends instructions to the cruise vehicle. The cruise monitoring vehicle executes the received instructions through the processor. Figure 1 The dam deformation prediction and early warning program of step S3 obtains the monitoring planning path 202. The cruise monitoring vehicle drives according to the monitoring planning path 202, collects the positioning data of each dam top monitoring point through the RTK device 7, and the processor executes according to the collected positioning data. Figure 1 The dam deformation prediction and early warning program of step S3 is calibrated and stored in the memory of the electronic device 3. The communication device 8 obtains the historical dam deformation data of the cloud server through the 4G / 5G network and transmits it to the processor; the processor Figure 1 The dam deformation prediction and early warning program of steps S4-S5 predicts and warns the deformation of the target dam.

[0112] In order to clearly illustrate the data format collected by the laser radar device 6 and the RTK device 7 of the cruise monitoring vehicle, refer to Table 4 which exemplarily gives some dam top cloud data, and Table 5 lists some dam top monitoring point data.

[0113] Table 4. Cloud data of some dam vertices

[0114] Table 5. Location data of some dam top monitoring points

[0115] The driving device and battery inside the body of the cruise monitoring vehicle proposed in the embodiment of the present application ensure that the cruise monitoring vehicle can move autonomously in the dam area to perform monitoring tasks; the laser radar device accurately scans the dam surface and the surrounding environment to perform road modeling and obstacle avoidance, thereby ensuring safety and smooth progress during the cruise process; the RTK device provides high-precision positioning information, which can accurately obtain the current position and elevation data of the cruise vehicle, and provide necessary spatial data support for dam deformation monitoring; the processor in the electronic device executes the dam deformation prediction and early warning program, and by analyzing and processing the collected data, it can timely discover potential deformation problems of the dam and issue early warning information; the communication device realizes real-time transmission of instructions and data through the 4G / 5G network, ensuring that the monitoring results can be remotely transmitted to the control center, facilitating remote scheduling and decision-making by management personnel; the automatic charging interface ensures that the cruise monitoring vehicle can continue to operate, avoiding interruption of monitoring tasks due to insufficient power, thereby improving the system's continuous monitoring capability.

[0116] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A dam deformation prediction and early warning method, characterized in that: include: S1. Determine the monitoring section of the dam according to the dam crest information of the target dam; S2. Collect the monitoring section of the dam The dam vertex cloud data is assembled and corrected to obtain a dam vertex cloud set; the scene points are clustered based on the point space spacing and reflection intensity of each scene point in the dam vertex cloud set to generate a scene point cloud cluster and perform geometric feature fitting to obtain a dam top three-dimensional scene model; S3. Based on the 3D scene model of the dam crest, set the dam crest monitoring points, calculate the monitoring point spacing between the dam crest monitoring points, and construct a path multi-objective optimization function; perform iterative optimization through a genetic algorithm to obtain a monitoring planning path; S4. Collect the positioning coordinate data of the engineering reference point, and calculate the coordinate positioning deviation based on the reference coordinate data; S5. Collect monitoring point positioning data, obtain the monitoring point positioning data within a predetermined historical time, and calculate the monitoring point settlement value and the monitoring point settlement rate; combine the monitoring point positioning data to obtain the embankment settlement index; construct a spatial deformation curve based on the monitoring point settlement value and the monitoring point positioning data, and calculate the spatial deformation rate; combine the monitoring point settlement rate to obtain the embankment deformation rate; S6. Determine whether to issue an early warning based on the dam settlement index and the dam deformation rate; obtain historical dam crest deformation data, calculate the deformation space correlation, and construct a dam crest space deformation map; combine the graph convolutional neural network, the long short-term memory network and the prediction head to predict the future settlement value and future settlement rate in the future.

2. A dam deformation prediction and early warning method according to claim 1, characterized in that: The specific implementation process of obtaining the dam vertex cloud set and generating the scene point cloud cluster includes: The monitoring section of the dam is collected by remotely controlling the cruise monitoring vehicle. The dam vertex cloud data is obtained; the acquisition timestamp of the dam vertex cloud data is obtained, and the plurality of groups of the dam vertex cloud data are sorted according to the acquisition timestamp; the first dam vertex cloud data is obtained according to the dam vertex cloud data with the earliest acquisition timestamp; and the other The difference in the acquisition timestamp between the dam vertex cloud data of the group and the first dam vertex cloud data is obtained. Combined with the running speed and running direction of the cruise monitoring vehicle, calculate The coordinate difference is collected and The dam vertex cloud data of the group are calibrated; the first dam vertex cloud data and the calibrated dam vertex cloud data are preprocessed and spliced ​​to obtain the dam vertex cloud set; according to the coordinates of each scene point in the dam vertex cloud set, the distance between different scene points is calculated to obtain the point space spacing; the laser reflection angle and the reflection intensity of the scene point are obtained, the direct reflection intensity is obtained through the reflection property of the object, and the direct intensity difference between the scene points is calculated; the scene points are clustered in combination with the point space spacing and the direct intensity difference; two scenes whose point space spacing is less than a predetermined spacing threshold and whose direct intensity difference is less than a predetermined intensity threshold are divided into the same cluster to obtain several groups of scene point cloud clusters; the scene point cloud clusters are geometrically fitted to obtain the dam top three-dimensional scene model.

3. A dam deformation prediction and early warning method according to claim 1, characterized in that: The specific implementation process of constructing the path multi-objective optimization function includes: The coordinates of the dam top monitoring points are obtained, and the distances between the different dam top monitoring points are calculated to obtain the monitoring point spacing; according to the dam top monitoring points and the monitoring point spacing, a path multi-objective optimization function is constructed, including a monitoring point density optimization function and a path length optimization function; the specific calculation formula of the path multi-objective optimization function is: ； in, represents the optimal solution function for the cruise path; Indicates the cruising path of the cruise monitoring vehicle; Represents the path multi-objective optimization function; represents the monitoring point density optimization function; represents the path length optimization function; express The adjustment factor of express The adjustment factor of The multi-objective optimization function of the path is iteratively optimized through a genetic algorithm to obtain the monitoring planning path.

4. A dam deformation prediction and early warning method according to claim 1, characterized in that: The specific implementation process of obtaining the dam settlement index includes: The cruise monitoring vehicle arrives at the dam top monitoring point according to the monitoring planning path, collects the position and elevation information, and performs position correction through the coordinate positioning deviation to obtain the monitoring point positioning data; obtains the monitoring point positioning data of the dam top monitoring point within the predetermined historical time, and obtains the settlement value and the settlement rate of the monitoring point through the elevation difference of each dam top monitoring point at the predetermined historical time; combines the position distribution of the dam top monitoring point and the settlement value of the monitoring point to calculate the embankment settlement index, and the specific formula is expressed as follows: ； ； in, represents the settlement index of the dam; Indicates The settlement value of each monitoring point of the dam crest monitoring point; Indicates The weight coefficient of each of the dam crest monitoring points; Indicates the number of monitoring points on the dam crest; represents a power function with a natural constant as base; represents the minimum function; Indicates Coordinate vectors of the dam top monitoring points; The target dam The coordinate vectors of the core points on the dam crest, including the center point and the edge point of the dam crest; represents the scaling factor; || ||2 represents the two-norm function.

5. A dam deformation prediction and early warning method according to claim 1, characterized in that: The specific process of obtaining the dam deformation rate includes: Based on the monitoring point settlement value of the dam top monitoring point on the monitoring planning path and the monitoring point positioning data, the spatial deformation curve is generated by a linear interpolation method, and the spatial deformation rate of each dam top monitoring point is calculated; the embankment deformation rate is calculated by combining the monitoring point settlement rate and the spatial deformation rate.

6. A dam deformation prediction and early warning method according to claim 1, characterized in that: The specific process of determining whether to issue an early warning according to the dam settlement index and the dam deformation rate includes: If the dam settlement index is greater than or equal to the predetermined settlement threshold, a red warning is issued to the management personnel; if the dam settlement index is less than the predetermined settlement threshold, it is determined whether the dam deformation rate exceeds the threshold; if the dam deformation rate is greater than or equal to the predetermined rate threshold, a yellow warning is issued to the management personnel; if the dam deformation rate is less than the predetermined rate threshold, the target dam is continuously monitored.

7. A dam deformation prediction and early warning method according to claim 1, characterized in that: The specific process of predicting the future sedimentation value and the future sedimentation rate at a future time includes: Acquire the historical dam crest deformation data of the target dam and perform preprocessing operations to obtain a historical dam crest deformation sequence; the historical dam crest deformation sequence includes the position coordinates of the monitoring point, the settlement value of the monitoring point and the settlement rate of the monitoring point; According to the position coordinates of different monitoring points in the historical dam crest deformation sequence, the deformation spatial correlation between different dam crest monitoring points is calculated; the dam crest spatial deformation map is constructed by combining the spatial correlation and the dam crest monitoring points; the dam crest spatial deformation map includes deformation nodes and edges; the deformation nodes are deformation feature sequences of the dam crest monitoring points, including the settlement values ​​of the monitoring points at different times and the settlement rates of the monitoring points; the deformation nodes are connected by the edges, and the weights of the edges are the spatial correlation; the dam crest spatial deformation map is input into a graph convolutional neural network to capture the spatial dependency between the deformation feature sequences of different deformation nodes, and obtain the dam crest deformation spatial features; Inputting the spatial characteristics of the dam crest deformation into a long short-term memory network, capturing the temporal correlation of the deformation feature sequences of different dam crest monitoring points, and obtaining the temporal characteristics of the dam crest deformation; The time series characteristics of the dam crest deformation are input into the prediction head to obtain the future settlement value and the future settlement rate of each dam crest monitoring point of the target dam at a future time.

8. A remote control cruise monitoring vehicle for dam deformation, comprising a vehicle body, a driving device, driving wheels, electronic equipment, a battery, an automatic charging interface, a laser radar device, an RTK device and a communication device; the electronic equipment comprises a memory, a processor and a dam deformation prediction and early warning program stored in the memory and executable on the processor, characterized in that: When the processor executes the dam deformation prediction and early warning program, the dam deformation prediction and early warning method according to any one of claims 1 to 7 is implemented.

9. A remote control cruise monitoring vehicle for dam deformation according to claim 8, characterized in that: The driving device, the electronic device and the battery are located inside the vehicle body; the laser radar device, the RTK device and the communication device are located on the top of the vehicle body; and the automatic charging interface is located at the rear of the vehicle body.

10. The remote control cruise monitoring vehicle for dam deformation according to claim 8, characterized in that: The communication device is equipped with a 4G / 5G network and receives and transmits instructions and data; the laser radar device is used for road modeling and obstacle avoidance during operation; The RTK device is used for high-precision positioning, including two pieces of information: position and elevation; the electronic device is used for data processing and backup operations; The automatic charging interface is used to charge the cruise monitoring vehicle.

Images