Dam filling quality detection system and method

Abstract

The invention discloses a dam filling quality detection system and method, and relates to the technical field of intelligent construction. The system comprises a mobile detection platform based on an unmanned vehicle, a plurality of nondestructive detection instruments which are arranged on the platform and have different detection functions, a rolling monitoring terminal arranged on a vibrating roller, and a data processing analysis center which is in communication connection with the platform, the instruments and the terminal, the data processing analysis center comprises a data receiving association module, a feature data extraction module, a compactness estimation module and an estimation result output module which are in communication connection in sequence, and through function cooperation of the modules and application of a machine learning algorithm, the compactness estimation value of the position where each platform arrives can be obtained; and drawing to obtain a compactness distribution diagram of the dam filling bin surface, so that the fundamental transformation of dam filling quality detection from a traditional afterward, single-point and destructive mode to a real-time, comprehensive, lossless and intelligent mode can be realized, and the engineering construction efficiency and quality are effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of intelligent construction technology, specifically relating to a dam filling quality detection system and method. Background Technology

[0002] Dam filling construction is a crucial step in water conservancy and hydropower projects, and the density of the fill material directly affects the stability, seepage prevention, and long-term safety of the dam. Therefore, timely and accurate quality testing of the fill surface is essential.

[0003] Currently, the quality control and testing methods for dam filling (especially new materials such as cemented gravel) mainly rely on the following approaches: (1) The most typical traditional destructive testing methods are the pit-filling method or the pit-filling sand method. This method requires manually digging test pits on the already compacted surface and calculating the density by measuring the volume of the test pit and the dry and wet mass of the excavated material. Although this method is regarded as the benchmark, it has inherent defects that cannot be overcome: First, it is extremely inefficient, that is, a single point test takes 30-60 minutes, which seriously slows down the construction progress and is seriously incompatible with the construction rhythm of high-intensity, rapid and continuous filling in large-scale projects; Second, it has poor representativeness, that is, due to the high time cost, the test can only be carried out in the form of sampling (for example, one group of 300 square meters), which cannot fully reflect the quality status of the entire surface and there is a risk of missing quality; Third, it is destructive and interfering, that is, the testing process itself destroys the complete filling body, which needs to be backfilled and repaired later, and the testing area cannot continue to be constructed during the repair period, which interferes with continuous operation. (2) Attempts at single non-destructive testing techniques: In order to overcome the drawbacks of traditional methods, the industry has tried non-destructive testing equipment such as nucleus-free density meters. However, when these techniques are applied to non-uniform materials such as cemented sand and gravel containing large-diameter aggregates, they have revealed significant limitations. For example, the measurement results of nucleus-free density meters are extremely weakly correlated with the results of the pit filling method (correlation coefficient as low as -0.1), the data dispersion is large, and the reliability is insufficient. The fundamental reason is that single non-destructive testing techniques are easily affected by factors such as material inhomogeneity and water content fluctuations, making it difficult to establish a universal and accurate measurement model.

[0004] Furthermore, existing methods for inspecting the quality of dam filling, whether traditional destructive testing or single non-destructive testing, are all post-construction inspections, meaning that inspectors only enter the site after the compaction process is completed. This model results in a significant lag in quality feedback. Once a non-conforming area is discovered, not only must subsequent construction be interrupted, but equipment and personnel must also be organized to return for supplementary compaction. This leads to complex management and coordination, high rework costs, and an inability to achieve a real-time quality control closed loop of "construction, inspection, and adjustment simultaneously."

[0005] In summary, how to provide a new dam filling quality inspection solution that can be carried out simultaneously with rapid filling construction, adapt to the characteristics of cemented sand and gravel materials, and achieve real-time reliable compaction assessment of the entire filling surface is a topic that urgently needs to be studied by those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a dam filling quality inspection system and method to solve the problems of low inspection efficiency, poor representativeness of inspection results, destructive and interference-related issues in traditional destructive testing methods, and the inability of existing single non-destructive testing methods to adapt to the characteristics of cemented sand and gravel materials.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: In the first aspect, a dam filling quality inspection system is provided, including a mobile inspection platform based on an unmanned vehicle, multiple non-destructive testing instruments with different detection functions installed on the mobile inspection platform, a compaction monitoring terminal installed on a vibratory roller, and a data processing and analysis center that is respectively connected to the mobile inspection platform, the multiple non-destructive testing instruments and the compaction monitoring terminal. The mobile detection platform is used to move autonomously on the dam fill surface after being compacted by the vibratory roller operation, and transmit the platform position data to the data processing and analysis center in real time. The non-destructive testing instrument is used to collect non-destructive testing data corresponding to any position on the dam filling surface in real time when the mobile testing platform is located at any position, and transmit the collected results to the data processing and analysis center in real time. The compaction monitoring terminal is used to collect compaction process data in real time and transmit the collected results to the data processing and analysis center in real time. The data processing and analysis center includes a data receiving and association module, a feature data extraction module, a density estimation module, and an estimation result output module that are connected in sequence. The data receiving and association module is used to receive and associate in real time the platform location data from the mobile detection platform, the non-destructive testing data from the multiple non-destructive testing instruments, and the compaction process data from the compaction monitoring terminal; The feature data extraction module is used to determine the locations that each platform has been visited on the dam filling surface based on the platform location data, and to extract the corresponding multi-dimensional non-destructive testing features from the non-destructive testing data most recently collected by the multiple non-destructive testing instruments at the corresponding locations for each platform that has been visited, and to extract the corresponding multi-dimensional compaction process features from the compaction process data collected by the compaction monitoring terminal when the vibratory roller most recently passed the corresponding location; The compaction estimation module is used to input the corresponding multidimensional non-destructive detection features and multidimensional compaction process features into the compaction estimation model pre-trained based on machine learning algorithms for the locations that the platform has visited, and output the corresponding compaction estimation value. The estimation result output module is used to draw and display the density distribution map of the dam filling surface based on the density estimates of the locations visited by each platform.

[0008] Based on the above-mentioned invention, a new dam filling quality inspection scheme is provided, which can be carried out simultaneously with rapid filling construction, adapt to the characteristics of cemented gravel materials, and achieve real-time reliable compaction assessment of the entire filling surface. This scheme includes a mobile inspection platform based on an unmanned vehicle, multiple non-destructive testing instruments with different detection functions mounted on the platform, a compaction monitoring terminal mounted on a vibratory roller, and a data processing and analysis center that communicates with the platform, instruments, and terminal respectively. The data processing and analysis center includes a data receiving and association module, a feature data extraction module, a compaction estimation module, and an estimation result output module that are sequentially connected. Through their functional collaboration and the application of data association, feature extraction, and machine learning algorithms, the compaction estimates of the locations visited by each platform can be obtained, and a compaction distribution map of the dam filling surface can be generated. This fundamentally transforms dam filling quality inspection from the traditional post-hoc, single-point, and destructive mode to a real-time, comprehensive, non-destructive, and intelligent mode, effectively improving the efficiency and quality of engineering construction and facilitating practical application and promotion.

[0009] In one possible design, the plurality of non-destructive testing instruments include a nucleus-free density meter, an ultrasonic detector, a shock echo detector, a microwave humidity detector, and / or ground-penetrating radar. The nucleus-free density meter is used to detect the wet density, dry density, moisture content, and / or compaction degree of the silo surface material. The ultrasonic detector is used to detect the location, size, and / or material uniformity of internal defects beneath the silo surface. The shock echo detector is used to detect the structural thickness, deep internal defects, and / or delamination layers beneath the silo surface. The microwave humidity detector is used to detect the moisture content and / or humidity of the silo surface material. The ground-penetrating radar is used to detect the layered structure, buried materials, and / or cavities beneath the silo surface.

[0010] In one possible design, the compaction monitoring terminal includes a compaction travel recording module, a compaction pass counting module, and / or a vibration parameter monitoring module. The compaction travel recording module is used to record the real-time position coordinates and real-time travel speed of the vibratory roller. The compaction pass counting module is used to count the number of compaction passes point by point. The vibration parameter monitoring module is used to monitor the vibration frequency.

[0011] In one possible design, the data processing and analysis center also includes a compaction equipment positioning module that is communicatively connected to the data receiving and association module, and the mobile detection platform includes a positioning module and an autonomous navigation module that are communicatively connected. The autonomous navigation module is also communicatively connected to the compaction equipment positioning module and integrates lidar obstacle avoidance and path planning functions. The compaction equipment positioning module is used to analyze and determine the current position of the vibratory roller in real time based on the compaction process data collected in real time by the compaction monitoring terminal, and transmit the current position to the autonomous navigation module in real time. The positioning module is used to obtain the platform's location in real time; The autonomous navigation module is used to control the mobile detection platform to follow the vibratory roller in real time based on the platform's location and the current location, using the lidar obstacle avoidance function and the path planning function, so as to achieve autonomous movement on the dam filling surface and maintain a safe distance from the vibratory roller.

[0012] In one possible design, the data processing and analysis center further includes a supplementary compaction analysis module that is communicatively connected to the estimation result output module and the operating end of the vibratory roller. The supplementary compaction analysis module is used to identify at least one area on the dam fill surface where the compaction does not meet the standard based on the density distribution map and a preset density threshold, and then send the location information of the at least one area to the operating end so as to perform supplementary compaction on the at least one area.

[0013] In one possible design, the unmanned vehicle uses a tracked mobile chassis, and the multiple non-destructive testing instruments are mounted on the suspended belly of the tracked mobile chassis.

[0014] Secondly, a method for testing the quality of dam filling is provided, including: The system receives and correlates platform location data from an unmanned vehicle-based mobile detection platform, non-destructive testing data from multiple non-destructive testing instruments, and compaction process data from a compaction monitoring terminal in real time. The mobile detection platform is used to move autonomously on the dam fill surface after it has been compacted by a vibratory roller. The multiple non-destructive testing instruments are mounted on the mobile detection platform and have different detection functions. These instruments are used to collect non-destructive testing data in real time corresponding to any position on the dam fill surface when the mobile detection platform is located at any position. The compaction monitoring terminal is mounted on the vibratory roller and is used to collect compaction process data in real time. Based on the platform location data, it was determined that each platform on the dam filling surface had been visited. For each location that the platform has visited, the corresponding multidimensional non-destructive testing features are extracted from the non-destructive testing data most recently collected by the multiple non-destructive testing instruments at the corresponding location, and the corresponding multidimensional compaction process features are extracted from the compaction process data collected by the compaction monitoring terminal when the vibratory roller most recently passed the corresponding location. For each location that the platform has visited, the corresponding multidimensional non-destructive detection features and multidimensional compaction process features are imported into a density estimation model pre-trained based on a machine learning algorithm, and the corresponding density estimate is output. Based on the density estimates of the locations visited by each platform, a density distribution map of the dam fill surface is drawn and displayed.

[0015] In one possible design, after receiving the compaction process data from the compaction monitoring terminal in real time, the method further includes: Based on the compaction process data collected in real time by the compaction monitoring terminal, the current position of the vibratory roller is determined in real time and transmitted to the mobile detection platform in real time. The mobile detection platform then uses the laser radar obstacle avoidance function and path planning function to follow the vibratory roller in real time to achieve autonomous movement on the dam filling surface and maintain a safe distance from the vibratory roller.

[0016] In one possible design, after obtaining the density distribution map of the dam fill surface, the method further includes: Based on the density distribution map and the preset density threshold, at least one area on the dam fill surface where the density does not meet the standard is identified. The location information of the at least one area is sent to the operating end of the vibratory roller so as to perform supplementary compaction on the at least one area.

[0017] In one possible design, sending the location information of the at least one region to the operating end of the vibratory mill includes: The current position of the vibratory roller is determined in real time based on the rolling process data collected in real time by the rolling monitoring terminal. Based on the current location and the location information of the at least one area, the shortest travel route of the vibratory roller for compacting the at least one area once is planned; The location information of the at least one area and the shortest travel route of the vibratory roller are sent to the operating end of the vibratory roller.

[0018] The beneficial effects of the above scheme are: (1) This invention provides a new dam filling quality inspection scheme that can be carried out simultaneously with rapid filling construction, adapt to the characteristics of cemented sand and gravel materials, and realize real-time reliable compaction assessment of the entire filling surface. It includes a mobile inspection platform based on unmanned vehicles, multiple non-destructive testing instruments with different detection functions set on the platform, a compaction monitoring terminal set on a vibratory roller, and a data processing and analysis center that communicates with the platform, instruments and terminals respectively. The data processing and analysis center includes a data receiving and association module, a feature data extraction module, a compaction estimation module and an estimation result output module that are connected in sequence. Through their functional cooperation and the application of data association, feature extraction and machine learning algorithms, the compaction estimation values ​​of the locations visited by each platform can be obtained, and a compaction distribution map of the dam filling surface can be drawn. In this way, the dam filling quality inspection can be fundamentally transformed from the traditional ex-post, single-point and destructive mode to a real-time, comprehensive, non-destructive and intelligent mode, effectively improving the efficiency and quality of engineering construction. (2) It can realize a fundamental change in quality inspection from "post-event sampling inspection" to "real-time full inspection". On the one hand, it can break through the efficiency bottleneck. By using unmanned vehicles equipped with multi-sensor arrays to automatically inspect the compacted surface, it can achieve 100% full coverage and no dead angle inspection of the dam filling surface. This completely changes the passive situation of the traditional pit digging and water filling method, which is inefficient and can only perform sparse sampling inspection. The inspection efficiency is increased by more than ten times. On the other hand, it can eliminate the inspection lag and achieve near-synchronous parallel operation with the compaction construction. This makes the inspection process almost imperceptible. The quality assessment results are generated immediately after the compaction operation, shortening the quality feedback time from several hours to a few minutes. This solves the serious quality feedback lag problem caused by the traditional "post-event inspection". (3) It can significantly improve the accuracy and reliability of detection in complex materials. On the one hand, it can overcome the limitations of single technology. By integrating non-destructive testing data from various non-destructive testing technologies such as non-nuclear density meters, ultrasonic detectors, impact echo detectors, and ground-penetrating radar, it can construct multi-dimensional non-destructive testing features. By utilizing information complementarity, it can effectively overcome the inherent defects of data dispersion and poor reliability of any single non-destructive testing technology when facing non-uniform, large aggregate materials such as cemented gravel. On the other hand, it can introduce process parameter correlation. It innovatively uses the characteristics of the compaction process (such as the number of compaction passes, vibration frequency, and travel speed) as input to the machine learning model, and establishes a direct and quantitative correlation between construction technology and final quality. This makes the density estimation not only based on the state of the material, but also considers the process of forming this state, which greatly improves the accuracy and engineering applicability of the evaluation model. (4) It can construct an intelligent quality control closed loop of "detection-evaluation-decision-execution". On the one hand, it can realize intelligent diagnosis and decision-making, automatically generate an intuitive density distribution map, and automatically identify substandard areas based on preset thresholds, thus realizing the accurate location of quality problems; on the other hand, it can guide precise compaction, and can directly push the location information of substandard areas and even the optimal compaction route to the vibratory roller operation end, guiding the operator to perform precise and efficient compaction, forming a complete quality control closed loop, and elevating quality control from "passively discovering problems" to a new intelligent level of "active prevention and real-time correction". (5) It can greatly improve construction efficiency. That is, through real-time detection and precise guidance for pressure replenishment, large-scale rework and construction interruption are avoided, and the overall construction period is effectively shortened. It is expected to shorten the construction period by 15-20%. For large-scale hydropower projects, the economic benefits brought by early power generation are extremely considerable. (6) It can effectively reduce the overall cost. On the one hand, the cost of cemented gravel material itself is lower than that of conventional concrete; on the other hand, this solution avoids the waste of labor, equipment and materials caused by rework, and at the same time reduces the risk of later maintenance, and the total life cycle cost is significantly reduced. (7) It can promote technological progress in the industry. That is, this solution provides a reliable quality assurance means for the large-scale promotion and application of new dam construction materials such as cemented gravel, solves the key quality testing problem that restricts its development, and has profound significance for promoting dam construction technology innovation and industry progress. (8) It can improve safety and environmental friendliness. The whole process of non-destructive testing avoids structural damage to the dam body, and the unmanned vehicle automatic operation reduces the risk of manual operation next to large machinery and improves construction safety. At the same time, the technology supports the full utilization of local materials, reduces the excavation and waste disposal of material yards, conforms to the concept of green construction, and is easy to apply and promote. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are 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.

[0020] Figure 1 This is a schematic diagram of the dam filling quality inspection system provided in an embodiment of this application.

[0021] Figure 2 This is a flowchart illustrating the dam filling quality testing method provided in this embodiment of the application. Detailed Implementation

[0022] 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 embodiments can be obtained based on these embodiments 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.

[0023] It should be understood that although the terms "first" and "second", etc., may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, the first object may be referred to as the second object, and similarly, the second object may be referred to as the first object, without departing from the scope of the exemplary embodiments of the invention.

[0024] 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, or A and B exist simultaneously. Another example is A, B and / or C, which can mean that any one of A, B, and C or any combination thereof exists. 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 or A and B exist simultaneously. 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.

[0025] Example 1 like Figure 1As shown, the dam filling quality inspection system provided in this embodiment includes, but is not limited to, a mobile inspection platform based on an unmanned vehicle, multiple non-destructive testing instruments with different detection functions installed on the mobile inspection platform, a compaction monitoring terminal installed on a vibratory roller, and a data processing and analysis center that is respectively connected to the mobile inspection platform, the multiple non-destructive testing instruments and the compaction monitoring terminal. The unmanned vehicle is used to enable the autonomous movement of the entire mobile detection platform and the multiple non-destructive testing instruments on the dam fill surface after it has been compacted by the vibratory roller (in dams or large hydraulic structures, for ease of construction organization and management, a large structure is divided into several smaller blocks that are easy to construct continuously in one go; these blocks are called "fill sections"; and the surface of the construction area where concrete is being poured or fill compacted within this section is called the "fill surface"). Considering that the dam fill surface is uneven after compaction and that the multiple non-destructive testing instruments need to face the dam fill surface to perform non-destructive testing, in order to enable the unmanned vehicle to adapt to the uneven compaction surface and achieve the purpose of "moving and detecting simultaneously", preferably, the unmanned vehicle adopts a tracked mobile chassis, and the multiple non-destructive testing instruments are installed in the middle of the tracked mobile chassis with the suspension in the middle. Furthermore, the vibratory roller is a commonly used device in hydropower and water conservancy projects. It is used to compact materials such as cemented sand and gravel on the dam filling surface to form structures such as the dam body. The compaction monitoring terminal is specifically installed on the vibratory roller and integrates many acquisition modules to collect real-time data on the compaction process, such as the real-time position coordinates and real-time travel speed of the vibratory roller, as well as the number of compaction passes and vibration frequency.

[0026] The mobile detection platform is used to autonomously move on the dam fill surface after being compacted by the vibratory roller operation, and transmits the platform position data to the data processing and analysis center in real time. The mobile detection platform specifically includes, but is not limited to, a communication-connected positioning module and an autonomous navigation module. The positioning module is used to acquire the platform position in real time, and may specifically employ, but is not limited to, an RTK-GPS (Real-Time Kinematic; Global Positioning System) module to achieve high-precision positioning (e.g., positioning accuracy ±2cm). The autonomous navigation module integrates, but is not limited to, existing technologies such as lidar obstacle avoidance and path planning (which can be implemented based on existing A* algorithms or fast random tree generation), and can, based on the current position and target position, apply the lidar obstacle avoidance function and the path planning function to control the mobile detection platform in real time to achieve autonomous movement on the dam fill surface. Furthermore, the data transmission method between the mobile detection platform and the data processing and analysis center can be, but is not limited to, through an Internet of Things (IoT) network. Specifically, the IoT network can, but is not limited to, adopt a hybrid network mode combining LoRaWAN (Low Power Wide Area Network, a wireless communication technology protocol designed specifically for the Internet of Things) technology with 4G / 5G mobile networks. That is, after the platform location data is aggregated through a LoRaWAN gateway, it can be transmitted to the remote data processing and analysis center in real time and stably through a 4G / 5G network.

[0027] The non-destructive testing (NDT) instruments are used to collect NDT data in real time at any position on the dam fill surface when the mobile testing platform is located at any position, and transmit the collected results to the data processing and analysis center in real time. The multiple NDT instruments are used to achieve complementary NDT data through a combined detection array. To better achieve this complementarity, preferably, the multiple NDT instruments include, but are not limited to, a nucleus density meter, an ultrasonic detector, an impact echo detector, a microwave humidity detector, and / or ground-penetrating radar. Specifically, the nucleus density meter is used to detect the wet density, dry density, moisture content, and / or compaction degree of the fill surface material (specifically, it uses electromagnetic wave principles to quickly scan and non-destructively detect these data). The ultrasonic detector is used to detect the location, size, and / or material uniformity of internal defects below the fill surface (specifically, it uses the emission of high-frequency sound waves). The shock echo detector is used to detect the structural thickness, deep internal defects, and / or delamination layers beneath the warehouse surface (specifically, it uses the principle of transient mechanical impact to excite stress waves and analyze their resonant frequencies to non-destructively detect these data). The microwave humidity detector is used to detect the moisture content and / or humidity of the warehouse surface material (specifically, it uses the principle of emitting microwaves and analyzing changes in dielectric constant to non-destructively detect these data). The ground-penetrating radar is used to detect the layered structure, buried objects, and / or cavities beneath the warehouse surface (specifically, it uses the principle of emitting high-frequency electromagnetic waves, receiving reflected signals, and performing deep imaging to non-destructively detect these data). Furthermore, the nucleus-free density meter, the ultrasonic detector, the shock echo detector, the microwave humidity detector, and the ground-penetrating radar can all be implemented using existing corresponding non-destructive testing instruments; and the data transmission method between the non-destructive testing instruments and the data processing and analysis center can also be, but is not limited to, through the Internet of Things (IoT) network.

[0028] The compaction monitoring terminal is used to collect compaction process data in real time and transmit the collected results to the data processing and analysis center in real time. Specifically, the compaction monitoring terminal includes, but is not limited to, a compaction travel recording module, a compaction pass counting module, and / or a vibration parameter monitoring module. The compaction travel recording module is used to record the real-time position coordinates and real-time travel speed of the vibratory roller (specifically, this can be achieved using a GPS locator and speed sensor). The compaction pass counting module is used to count the number of compaction passes point-by-point (specifically, this can be based on historical position coordinate data, by spatially gridding the dam fill surface, and counting the number of compaction passes on each grid unit based on the processing results). The vibration parameter monitoring module is used to monitor the vibration frequency (specifically, this can be achieved using an accelerometer or vibration sensor). Furthermore, the data transmission between the compaction monitoring terminal and the data processing and analysis center can also be achieved, but is not limited to, through the Internet of Things (IoT) network.

[0029] The data processing and analysis center includes, but is not limited to, a data receiving and association module, a feature data extraction module, a density estimation module, and an estimation result output module, all connected in sequence. Furthermore, the data processing and analysis center may, but is not limited to, be a server or cloud platform deployed in the field command center.

[0030] The data receiving and association module is used to receive and associate in real time the platform location data from the mobile detection platform, the non-destructive testing data from the multiple non-destructive testing instruments, and the compaction process data from the compaction monitoring terminal. Since the platform location data contains a receiving timestamp and spatial coordinate information, the non-destructive testing data contains a receiving timestamp, and the compaction process data contains a receiving timestamp and spatial coordinate information, these data from different sources can be associated using a unified timestamp and / or spatial coordinate. This ensures that subsequent analysis can be based on datasets from the same time and / or the same location, ultimately guaranteeing the accuracy of the estimation results.

[0031] The feature data extraction module is used to determine the locations visited by each platform on the dam filling surface based on the platform location data, and to extract corresponding multi-dimensional non-destructive testing features from the most recent non-destructive testing data collected by the multiple non-destructive testing instruments at the corresponding locations for each platform visited, and to extract corresponding multi-dimensional compaction process features from the compaction process data collected by the compaction monitoring terminal when the vibratory roller most recently passed the corresponding location. The locations visited by the platforms refer to the locations reached by the mobile detection platform. The multi-dimensional non-destructive testing features specifically include, but are not limited to, the wet density, dry density, moisture content, compaction degree, water content and / or humidity of the filling surface material, as well as descriptions of internal defect locations, internal defect sizes, material uniformity, structural thickness, deep internal defects, delamination layers, layered structures, buried objects, and / or voids below the filling surface. These features can be conventionally extracted based on the non-destructive testing data. The multidimensional compaction process features include, but are not limited to, the total number of compaction passes, the average vibration frequency during compaction, and / or the average travel speed during compaction. These features can be conventionally extracted based on the compaction process data.

[0032] The compaction estimation module is used to input the corresponding multidimensional non-destructive testing features and multidimensional compaction process features into a compaction estimation model pre-trained based on a machine learning algorithm for each location already visited by the platform, and output the corresponding compaction estimate. The machine learning algorithm is a core artificial intelligence algorithm that specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills, reorganize existing knowledge structures to continuously improve their performance, and is the fundamental way to make computers intelligent. Specifically, the machine learning algorithm preferably uses, but is not limited to, graph neural networks, support vector machines, K-nearest neighbors, stochastic gradient descent, multivariate linear regression, multilayer perceptrons, decision trees, backpropagation neural networks, or radial basis function networks, etc., to quickly and accurately find patterns in the data. Therefore, based on a certain amount of sample data (the model input consists of multidimensional non-destructive detection features and multidimensional compaction process features at a certain location, and the model output consists of the actual density value at that location measured by traditional destructive detection methods), the density estimation model can be trained and validated using conventional calibration and verification modeling methods (the specific process includes model calibration and verification, i.e., first comparing the model simulation results with the measured data, and then adjusting the model parameters according to the comparison results to make the simulation results match the actual results).

[0033] The estimation result output module is used to draw and display a density distribution map of the dam fill surface based on the density estimates of the locations visited by each platform. Although the density estimates of the locations visited by each platform are discrete data, a spatial interpolation algorithm (such as Kriging interpolation) can be used to generate a continuous density distribution map covering the entire inspected fill surface based on the density estimates of these discrete points. Furthermore, the density distribution map can be displayed through a graphical user interface, for example, as a heatmap: different colors represent different density ranges (e.g., green represents areas with a density ≥ 97%, and red represents areas with a density < 95%), allowing construction personnel to intuitively and quickly grasp the overall quality status.

[0034] Preferably, the data processing and analysis center further includes, but is not limited to, a compaction equipment positioning module that is communicatively connected to the data receiving and association module; the mobile detection platform includes, but is not limited to, a positioning module and an autonomous navigation module that are communicatively connected. The autonomous navigation module is also communicatively connected to the compaction equipment positioning module and integrates, but is not limited to, lidar obstacle avoidance and path planning functions. The compaction equipment positioning module is used to analyze and determine the current position of the vibratory roller in real time based on the compaction process data collected in real time by the compaction monitoring terminal, and transmit the current position to the autonomous navigation module in real time. The positioning module is used to obtain the platform position in real time. The autonomous navigation module is used to control the mobile detection platform to follow the vibratory roller in real time based on the platform position and the current position, applying the lidar obstacle avoidance function and the path planning function to achieve autonomous movement on the dam fill surface and maintain a safe distance from the vibratory roller. Since the compaction process data includes spatial coordinate information, the current position of the vibratory roller can be routinely analyzed and determined in real time. This allows a location behind the vibratory roller at a safe distance (e.g., 10-20 meters) to be used as the target position for the mobile inspection platform / unmanned vehicle. Based on this target position and the platform's position, the lidar obstacle avoidance function and path planning function are applied to control the autonomous movement of the mobile inspection platform on the dam fill surface in real time, achieving safe following of the vibratory roller. This enables near-synchronous operation of compaction and inspection, reducing the lag in quality inspection results. Furthermore, the unmanned vehicle can also be mounted as a trailer behind the vibratory roller.

[0035] Preferably, the data processing and analysis center further includes, but is not limited to, a supplementary compaction analysis module that is communicatively connected to the estimation result output module and the operating end of the vibratory roller. The supplementary compaction analysis module is used to identify at least one area on the dam fill surface where the compaction level is below standard based on the density distribution map and a preset density threshold. Then, it sends the location information of the at least one area to the operating end for supplementary compaction of the at least one area. The density threshold can be, for example, 97%. By comparing the density value of each pixel in the distribution map with the threshold, all areas on the dam fill surface where the compaction level is below standard can be automatically identified. Then, the location information of all these areas, including boundary coordinates and center point coordinates, is packaged and sent to the operating end of the vibratory roller via a wireless network. This guides the operating end to perform supplementary compaction of the at least one area, thus forming a quality control closed loop of "detection-location-guided supplementary compaction". Furthermore, after identifying the at least one region, non-compliant regions that are too small and do not require additional compaction can be filtered out by comparing the region area with the area threshold; it is also necessary to filter out non-compliant regions that are currently between the vibratory roller and the unmanned vehicle (because these regions have not been tested for compaction after recent compaction); and the operating terminal is, for example but not limited to, a tablet computer / vehicle computer located in the cab of the vibratory roller.

[0036] Further preferably, sending the location information of the at least one area to the operating terminal of the vibratory roller includes, but is not limited to, the following steps: first, analyzing and determining the current position of the vibratory roller in real time based on the compaction process data collected in real time by the compaction monitoring terminal; then, planning the shortest travel route of the vibratory roller for compacting the at least one area once based on the current position and the location information of the at least one area; finally, sending the location information of the at least one area and the shortest travel route of the vibratory roller to the operating terminal of the vibratory roller. The planning process of the shortest travel route of the vibratory roller is specifically, but not limited to, using existing path planning algorithms (such as Dijkstra's algorithm). By sending this optimal route along with the location information of the uncomplied areas to the operating terminal, the operator can be directly guided to perform accurate and efficient supplementary compaction, greatly reducing the time spent on blindly searching and repeated compaction, thereby enhancing the quality control closed-loop effect of "detection-location-guided supplementary compaction".

[0037] In summary, the dam filling quality inspection system provided in this embodiment has the following technical advantages: (1) This embodiment provides a new dam filling quality inspection scheme that can be carried out simultaneously with rapid filling construction, adapt to the characteristics of cemented sand and gravel materials, and realize real-time reliable compaction assessment of the entire filling surface. It includes a mobile inspection platform based on unmanned vehicles, multiple non-destructive testing instruments with different detection functions set on the platform, a compaction monitoring terminal set on a vibratory roller, and a data processing and analysis center that communicates with the platform, instruments and terminals respectively. The data processing and analysis center includes a data receiving and association module, a feature data extraction module, a compaction estimation module and an estimation result output module that are connected in sequence. Through their functional cooperation and the application of data association, feature extraction and machine learning algorithms, the compaction estimation values ​​of the locations visited by each platform can be obtained, and a compaction distribution map of the dam filling surface can be drawn. In this way, the dam filling quality inspection can be fundamentally transformed from the traditional ex-post, single-point and destructive mode to a real-time, comprehensive, non-destructive and intelligent mode, effectively improving the efficiency and quality of engineering construction, and facilitating practical application and promotion.

[0038] Example 2 Based on the technical solution of Embodiment 1, this embodiment also provides a dam filling quality testing method with the same inventive concept as the dam filling quality testing system described in Embodiment 1. This method can be specifically executed by computer equipment with certain computing resources; such as Figure 2 As shown, the dam filling quality inspection method includes, but is not limited to, the following steps S1 to S4.

[0039] S1. Real-time reception and association of platform location data from an unmanned vehicle-based mobile detection platform, non-destructive testing data from multiple non-destructive testing instruments, and compaction process data from a compaction monitoring terminal. The mobile detection platform is used to autonomously move on the dam fill surface after it has been compacted by a vibratory roller. The multiple non-destructive testing instruments are mounted on the mobile detection platform and have different detection functions. The non-destructive testing instruments are used to collect non-destructive testing data corresponding to any position on the dam fill surface in real time when the mobile detection platform is located at any position. The compaction monitoring terminal is mounted on the vibratory roller and is used to collect compaction process data in real time.

[0040] In step S1, preferably, after receiving the compaction process data from the compaction monitoring terminal in real time, the method further includes: analyzing and determining the current position of the vibratory roller in real time based on the compaction process data collected in real time by the compaction monitoring terminal, and transmitting the current position to the mobile detection platform in real time, so that the mobile detection platform can follow the vibratory roller in real time based on the current position and the platform position data, using the lidar obstacle avoidance function and path planning function to achieve autonomous movement on the dam filling surface and maintain a safe distance from the vibratory roller.

[0041] S2. Based on the platform location data, determine the positions that each platform on the dam filling surface has been reached.

[0042] S3. For each location that the platform has visited, extract the corresponding multi-dimensional non-destructive testing features from the non-destructive testing data most recently collected by the multiple non-destructive testing instruments at the corresponding location, and extract the corresponding multi-dimensional compaction process features from the compaction process data collected by the compaction monitoring terminal when the vibratory roller most recently passed the corresponding location.

[0043] S4. For each location that the platform has visited, the corresponding multidimensional non-destructive detection features and multidimensional compaction process features are imported into the density estimation model pre-trained based on machine learning algorithms, and the corresponding density estimation value is output.

[0044] S5. Based on the density estimates of the locations visited by each platform, draw a density distribution map of the dam fill surface and output it for display.

[0045] In step S5, preferably, after drawing the density distribution map of the dam filling surface, the method further includes, but is not limited to, the following steps S61 to S62.

[0046] S61. Based on the density distribution map and the preset density threshold, identify at least one area on the dam fill surface where the density does not meet the standard.

[0047] S62. The location information of the at least one area is sent to the operating end of the vibratory roller so as to perform supplementary compaction on the at least one area.

[0048] In step S62, it is further preferred that the location information of the at least one area be sent to the operating end of the vibratory mill, including but not limited to the following steps S621 to S623.

[0049] S621. Based on the rolling process data collected in real time by the rolling monitoring terminal, the current position of the vibratory roller is determined in real time through analysis.

[0050] S622. Based on the current position and the position information of the at least one area, plan the shortest travel route of the vibratory roller for compacting the at least one area once.

[0051] S623. The location information of the at least one area and the shortest travel route of the vibratory roller are sent to the operating end of the vibratory roller.

[0052] The technical details and effects of this embodiment can be derived conventionally by referring to the technical details and effects of Embodiment 1, and will not be repeated here.

[0053] 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 dam filling quality inspection system, characterized in that, It includes a mobile detection platform based on unmanned vehicles, multiple non-destructive testing instruments with different detection functions installed on the mobile detection platform, a compaction monitoring terminal installed on the vibratory roller, and a data processing and analysis center that is respectively connected to the mobile detection platform, the multiple non-destructive testing instruments and the compaction monitoring terminal. The mobile detection platform is used to move autonomously on the dam fill surface after being compacted by the vibratory roller operation, and transmit the platform position data to the data processing and analysis center in real time. The non-destructive testing instrument is used to collect non-destructive testing data corresponding to any position on the dam filling surface in real time when the mobile testing platform is located at any position, and transmit the collected results to the data processing and analysis center in real time. The compaction monitoring terminal is used to collect compaction process data in real time and transmit the collected results to the data processing and analysis center in real time. The data processing and analysis center includes a data receiving and association module, a feature data extraction module, a density estimation module, and an estimation result output module that are connected in sequence. The data receiving and association module is used to receive and associate in real time the platform location data from the mobile detection platform, the non-destructive testing data from the multiple non-destructive testing instruments, and the compaction process data from the compaction monitoring terminal; The feature data extraction module is used to determine the locations that each platform has been visited on the dam filling surface based on the platform location data, and to extract the corresponding multi-dimensional non-destructive testing features from the non-destructive testing data most recently collected by the multiple non-destructive testing instruments at the corresponding locations for each platform that has been visited, and to extract the corresponding multi-dimensional compaction process features from the compaction process data collected by the compaction monitoring terminal when the vibratory roller most recently passed the corresponding location; The compaction estimation module is used to input the corresponding multidimensional non-destructive detection features and multidimensional compaction process features into the compaction estimation model pre-trained based on machine learning algorithms for the locations that the platform has visited, and output the corresponding compaction estimation value. The estimation result output module is used to draw and display the density distribution map of the dam filling surface based on the density estimates of the locations visited by each platform.

2. The dam filling quality inspection system as described in claim 1, characterized in that, The plurality of non-destructive testing instruments include a nucleus-free density meter, an ultrasonic detector, a shock echo detector, a microwave humidity detector, and / or ground-penetrating radar. The nucleus-free density meter is used to detect the wet density, dry density, moisture content, and / or compaction of the silo surface material. The ultrasonic detector is used to detect the location, size, and / or material uniformity of internal defects below the silo surface. The shock echo detector is used to detect the structural thickness, deep internal defects, and / or delamination layers below the silo surface. The microwave humidity detector is used to detect the moisture content and / or humidity of the silo surface material. The ground-penetrating radar is used to detect the layered structure, buried objects, and / or cavities below the silo surface.

3. The dam filling quality inspection system as described in claim 1, characterized in that, The compaction monitoring terminal includes a compaction travel recording module, a compaction pass counting module, and / or a vibration parameter monitoring module. The compaction travel recording module is used to record the real-time position coordinates and real-time travel speed of the vibratory roller. The compaction pass counting module is used to count the number of compaction passes point by point. The vibration parameter monitoring module is used to monitor the vibration frequency.

4. The dam filling quality inspection system as described in claim 1, characterized in that, The data processing and analysis center also includes a compaction equipment positioning module that is communicatively connected to the data receiving and association module. The mobile detection platform includes a positioning module and an autonomous navigation module that are communicatively connected. The autonomous navigation module is also communicatively connected to the compaction equipment positioning module and integrates laser radar obstacle avoidance function and path planning function. The compaction equipment positioning module is used to analyze and determine the current position of the vibratory roller in real time based on the compaction process data collected in real time by the compaction monitoring terminal, and transmit the current position to the autonomous navigation module in real time. The positioning module is used to obtain the platform's location in real time; The autonomous navigation module is used to control the mobile detection platform to follow the vibratory roller in real time based on the platform's location and the current location, using the lidar obstacle avoidance function and the path planning function, so as to achieve autonomous movement on the dam filling surface and maintain a safe distance from the vibratory roller.

5. The dam filling quality inspection system as described in claim 1, characterized in that, The data processing and analysis center also includes a supplementary compaction analysis module that is communicatively connected to the estimation result output module and the operating end of the vibratory roller. The supplementary compaction analysis module is used to identify at least one area on the dam filling surface where the compaction does not meet the standard based on the density distribution map and a preset density threshold, and then send the location information of the at least one area to the operating end so as to perform supplementary compaction on the at least one area.

6. The dam filling quality inspection system as described in claim 1, characterized in that, The unmanned vehicle uses a tracked mobile chassis, and the multiple non-destructive testing instruments are mounted on the suspended belly of the tracked mobile chassis.

7. A method for testing the quality of dam filling, characterized in that, include: The system receives and correlates platform location data from an unmanned vehicle-based mobile detection platform, non-destructive testing data from multiple non-destructive testing instruments, and compaction process data from a compaction monitoring terminal in real time. The mobile detection platform is used to move autonomously on the dam fill surface after it has been compacted by a vibratory roller. The multiple non-destructive testing instruments are mounted on the mobile detection platform and have different detection functions. These instruments are used to collect non-destructive testing data in real time corresponding to any position on the dam fill surface when the mobile detection platform is located at any position. The compaction monitoring terminal is mounted on the vibratory roller and is used to collect compaction process data in real time. Based on the platform location data, it is determined that each platform on the dam filling surface has been visited. For each location that the platform has visited, the corresponding multidimensional non-destructive testing features are extracted from the non-destructive testing data most recently collected by the multiple non-destructive testing instruments at the corresponding location, and the corresponding multidimensional compaction process features are extracted from the compaction process data collected by the compaction monitoring terminal when the vibratory roller most recently passed the corresponding location. For each location that the platform has visited, the corresponding multidimensional non-destructive detection features and multidimensional compaction process features are imported into a density estimation model pre-trained based on a machine learning algorithm, and the corresponding density estimate is output. Based on the density estimates of the locations visited by each platform, a density distribution map of the dam fill surface is drawn and displayed.

8. The dam filling quality testing method as described in claim 7, characterized in that, After receiving the compaction process data from the compaction monitoring terminal in real time, the method further includes: Based on the compaction process data collected in real time by the compaction monitoring terminal, the current position of the vibratory roller is determined in real time and transmitted to the mobile detection platform in real time. The mobile detection platform then uses the laser radar obstacle avoidance function and path planning function to follow the vibratory roller in real time to achieve autonomous movement on the dam filling surface and maintain a safe distance from the vibratory roller.

9. The method for detecting the quality of dam filling as described in claim 7, characterized in that, After obtaining the density distribution map of the dam fill surface, the method further includes: Based on the density distribution map and the preset density threshold, at least one area on the dam fill surface where the density does not meet the standard is identified. The location information of the at least one area is sent to the operating end of the vibratory roller so as to perform supplementary compaction on the at least one area.

10. The dam filling quality testing method as described in claim 9, characterized in that, Sending the location information of the at least one area to the operating end of the vibratory mill includes: The current position of the vibratory roller is determined in real time based on the rolling process data collected in real time by the rolling monitoring terminal. Based on the current location and the location information of the at least one area, the shortest travel route of the vibratory roller for compacting the at least one area once is planned; The location information of the at least one area and the shortest travel route of the vibratory roller are sent to the operating end of the vibratory roller.

Images