Emergency rescue method for leakage danger of earth-rock dam based on multi-model collaboration

Abstract

The emergency rescue method for seepage hazards of earth-rock dams based on multi-model collaboration includes: obtaining the seepage points of earth-rock dams; collecting the stress data of the seepage points of earth-rock dams, inputting it into the early warning model, and outputting the early warning value of the seepage points; obtaining the water flow velocity data and the impact force data of the water-carried substances at the seepage points, combining the early warning value of the seepage points and inputting it into the seepage point danger early warning model, and outputting the danger value of the seepage points; performing real-time priority plugging sorting through the seepage point priority model; obtaining the stress data of the seepage points of the earth-rock dams after plugging, collecting the real-time flow velocity and the impact force of the water-carried substances at the plugged seepage points, inputting it into the pre-built plugging model, and outputting the plugging stress value, and determining the early warning for the plugged seepage points to be released through the preset threshold library. The present invention realizes the rational arrangement of the priority order of emergency rescue resources according to the urgency of different situations of the seepage points of earth-rock dams, and realizes efficient and rapid emergency rescue.

Description

Technical Field

[0001] The present invention relates to the technical field of rapid emergency repair of earth-rock dams, and specifically relates to a method for emergency repair of seepage hazards of earth-rock dams based on multi-model collaboration. Background Art

[0002] In the field of water conservancy projects, earth-rock dams, as important flood control facilities, are of crucial safety and stability. However, during the long-term operation of earth-rock dams, affected by various factors such as water flow scouring, geological condition changes, and construction quality, seepage hazards occur from time to time. Once seepage occurs, if not treated promptly and effectively, it may trigger serious disasters such as dam breaches, posing a huge threat to the lives and property safety of people in surrounding areas. Therefore, it is extremely urgent to develop an efficient and accurate method for rapid emergency repair of seepage hazards of earth-rock dams.

[0003] Traditional earth-rock dam seepage emergency repair technologies mainly include manually patrolling to find leakage points, and then relying on experience to judge the severity of seepage and adopting a single plugging method, such as simply piling up sandbags or simple grouting. This method has many disadvantages. In terms of finding leakage points, manual patrol is inefficient and difficult to quickly and accurately locate all leakage points. Especially when the dam area is large or the seepage situation is complex, it is easy to miss. In terms of emergency repair decision-making, relying solely on experience lacks a scientific basis and cannot accurately evaluate the actual risk degree of seepage, which may lead to unreasonable allocation of emergency repair resources. Moreover, traditional plugging methods lack pertinence. The same or similar methods are used for leakage points of different risk levels, and the plugging effect is not good. When facing the impact of complex water flow and carried objects, the plugging structure is easily damaged and cannot effectively prevent the deterioration of seepage.

[0004] Chinese Patent Application with Publication No. CN115187876A (Patent Name: Unmanned Aerial Vehicle-Borne Automatic Survey and Early Warning Method and Device for Seepage Hazards of Earth-Rock Dams) discloses using an unmanned flight platform to continuously capture infrared images of the earth-rock dam site with a certain overlap; inputting the infrared images into a trained deep convolutional neural network model, and the deep convolutional neural network model predicts and outputs the seepage types of the dam. There are problems that it only focuses on using the unmanned flight platform to capture infrared images and predicting the seepage types of the dam through the deep convolutional neural network model and giving early warnings, and it is unable to accurately collect and analyze the stress data of seepage points of earth-rock dams, and it is difficult to accurately evaluate the seepage risk degree. In terms of water flow, the influence of water flow velocity data and the impact force data of water flow carried objects on seepage hazards is not considered, and it is impossible to scientifically evaluate the risk values of multiple seepage points and rank the emergency repair priorities. When facing complex seepage situations, it is impossible to reasonably allocate emergency repair resources and determine the emergency repair order.

[0005] Meanwhile, based on the existing technology, some simple monitoring devices are adopted to assist in finding leakage points, but the overall emergency rescue process is still not perfect. For example, only the approximate location of the leakage point can be monitored, the force on the leakage point and the detailed information of the surrounding water flow cannot be accurately obtained, and the detection means for the substances carried by the water flow are also relatively rough, making it impossible to accurately judge the magnitude of its impact force. In terms of danger warning and priority determination, there is a lack of systematic model construction, making it difficult to scientifically rank the dangers of multiple leakage points and unable to provide a reasonable decision-making order for the emergency rescue work. In terms of evaluating the plugging effect, there is no effective model to judge whether it is stable and reliable after plugging, and potential risks cannot be detected in a timely manner.

[0006] Therefore, to address the above problems, the technical solution of this invention is a method for emergency rescue of earth-rock dam leakage disasters based on multi-model collaboration. By constructing an early warning model, it solves the problem of unscientific judgment of the leakage danger level in traditional and existing technologies. With the help of the leakage point priority model, integrating the danger value and location information, it realizes a reasonable ranking of the emergency rescue of leakage points, changing the previous situation of resource waste or untimely emergency rescue caused by lack of scientific ranking. And by evaluating the plugging effect through the plugging model, it improves the success rate and stability of plugging, overcomes the defects of single and poor plugging effects in traditional plugging methods, greatly improves the efficiency and reliability of the emergency rescue of earth-rock dam leakage disasters, and ensures the safety of the dam. Summary of the Invention

[0007] In view of the above, the present invention proposes a method for emergency rescue of earth-rock dam leakage disasters based on multi-model collaboration, including:

[0008] Obtain at least one or more leakage points of the earth-rock dam;

[0009] Collect the force data of the leakage points of the earth-rock dam and input it into a pre-constructed early warning model to output the early warning value of the leakage points;

[0010] Obtain the current water flow velocity data and the impact force data of the substances carried by the water flow in the watershed area of the leakage points of the earth-rock dam, and input them into the leakage point danger early warning model in combination with the early warning value of the leakage points to output the danger value of the leakage points;

[0011] Obtain the danger value of the leakage points through the leakage point priority model and perform real-time priority plugging ranking;

[0012] Obtain the force data of the leakage points of the earth-rock dam after plugging, collect the real-time water flow velocity and the impact force of the substances carried by the water flow at the plugged leakage points, input them into a pre-constructed plugging model to output the plugging force value, and determine the detachment warning of the plugged leakage points through a preset threshold library;

[0013] Real-time output the danger value of at least one leakage point of the earth-rock dam, adjust the priority order of the leakage points of the earth-rock dam, and conduct emergency rescue of the leakage points of the earth-rock dam with priority;

[0014] Among them, the dangerous value of the leakage point is obtained and set with three or more levels of threshold intervals, and different categories are realized through the corresponding level threshold intervals; the plugging types include but are not limited to grouting, geomembrane sandbags and expansion balls.

[0015] Preferably, the leakage point warning value is output through the warning model, specifically including:

[0016] Obtain the stress data of the earth-rock dam under normal conditions and leakage conditions, perform preprocessing, and extract characteristic parameters; the preprocessing includes removing outliers and filling missing values;

[0017] Through the characteristic parameters, use the logistic regression algorithm to construct an initial warning model, and the formula is: , where is the selected characteristic parameter, is the coefficient of the selected characteristic parameter, is the intercept, is the probability of the occurrence of a leakage point in the earth-rock dam;

[0018] Use the characteristic parameters of the stress data of the earth-rock dam under normal conditions and leakage conditions for training, and determine the characteristic parameter coefficients in the logistic regression algorithm through the maximum likelihood estimation method , and the iterative training meets the verification requirements to form a warning model and output the leakage point warning value.

[0019] Preferably, the water-carried substances include but are not limited to stones, gravels and sediment;

[0020] The water-carried substances are obtained through magnetic imaging technology. The magnetic imaging sensor scans the real-time cross-section of the water flow to obtain the contour information of the real-time cross-section of the leakage point watershed area of the earth-rock dam, and judges the water-carried substances through the contour information.

[0021] Preferably, the acquisition of the impact force data of the water-carried substances specifically includes:

[0022] Predict the volume of stones, gravels and sediment in the unit volume of water content through magnetic imaging technology, and determine the mass of stones, gravels and sediment through the volume , and the formula is: ; where , and are the unit volume ratios of the cross-sections of stones, gravels and sediment, , and are the densities of stones, gravels and sediment, is the unit volume;

[0023] Determine the speed of the water-carried substances through a speedometer ;

[0024] Calculate the water flow impact force, and the formula is: , where Unit water flow impact force, is the unit area, is the water flow impact force, where , where is the speed correction coefficient, is the time.

[0025] Preferably, the construction of the leakage point danger warning model specifically includes:

[0026] Obtain water flow velocity data and the impact force data of water flow carried objects, perform preprocessing, and extract characteristic parameters;

[0027] Divide the characteristic parameters into a training set, a validation set, and a test set;

[0028] Input the training set data into the random forest model for training, randomly extract samples from the training set with replacement, construct a training subset of decision trees, and repeat continuously until the preset maximum depth or the number of samples in the node is less than the minimum number of samples;

[0029] Add the decision trees to the random forest model and perform multi-round training to form an initial leakage point danger warning model;

[0030] Input the validation set and the test set into the initial leakage point danger warning model, perform performance evaluation, and adjust the parameters of the initial leakage point danger warning model to form a leakage point danger warning model.

[0031] Preferably, the output of the leakage point danger value comprehensively outputs the prediction results of the decision trees through the random forest model, and the mean method is used to output the leakage point danger value. The formula is: , where is the prediction result of decision tree , is the leakage point danger value.

[0032] Preferably, the leakage point priority model obtains the leakage point danger values output by the leakage point danger warning model, sorts the danger values from high to low to form a first danger value sequence; the leakage point priority model sorts according to the key points of the dam structure by obtaining the position information of the leakage points to form a second danger value sequence; constructs a leakage point association framework, uses the leakage point danger values in the first danger value sequence as nodes, assigns important information to the position information of the leakage point key points in the second danger value sequence, traverses the leakage point association framework through the ant colony algorithm, and selects the next node according to the pheromone concentration on the path of the leakage point association framework. After multiple rounds of iteration, the target leakage point priority order is formed.

[0033] Preferably, obtain the priority order of the target leakage points, extract the leakage risk values of the first leakage point, and determine the preset threshold range to which it belongs; the preset threshold range includes but is not limited to a high-risk range, a medium-risk range, and a low-risk range.

[0034] Preferably, different plugging types are adopted according to the preset threshold range; for the high-risk range of the preset threshold range, the concrete rapid-setting grouting method is used, and the concrete rapid-setting slurry is injected into the leakage channel through high-pressure equipment to quickly solidify and form a high-strength plugging structure; for the medium-risk range, the method of laying geomembrane combined with piling sandbags is adopted; for the low-risk range, an expansion ball carried by a drone is used for expansion plugging.

[0035] Preferably, the plugging model is constructed by a multi-layer perceptron neural network model. The plugging force value is output through the multi-layer perceptron neural network model, and it is determined whether the plugged leakage point is out of warning according to the preset threshold library. Specifically, it includes:

[0036] Obtain the force data of the leakage point of the earth-rock dam after plugging, the real-time flow rate of the plugged leakage point, and the impact force of the water-borne substances, and perform standardization processing to form standardized data;

[0037] Input the standardized data into the input layer of the multi-layer perceptron neural network model, perform weighted summation through the hidden layer of the multi-layer perceptron neural network model, and output the plugging force value through the output layer;

[0038] Compare the plugging force value with the out-of-warning threshold in the preset threshold library. If the plugging force value is less than the out-of-warning threshold, the plugged part is in a safe range, and it is determined that the plugged leakage point is out of warning; if the plugging force value is greater than or equal to the out-of-warning threshold, there is a risk, and it is promoted to the high-risk range of the preset threshold range for processing.

[0039] Compared with the prior art, the technical solution of the present application has the following technical effects:

[0040] The present invention constructs an early warning model, uses the dam force data under normal and leakage conditions of the earth-rock dam, determines the characteristic parameter coefficients through preprocessing and logistic regression algorithm training, and forms an early warning value of the leakage point output by the early warning model, solving the problem that it is difficult to scientifically judge the leakage risk of the earth-rock dam in the traditional way, being able to accurately conduct early warning of the leakage situation, providing a key risk assessment basis for subsequent emergency rescue work, and improving the pertinence and timeliness of the emergency rescue work.

[0041] By considering the problem of water-carried objects not involved in traditional technologies, the present invention acquires information on water-carried objects through magnetic imaging technology, determines the speed in combination with a velocimeter, calculates the impact force data of water-carried objects, and inputs the water flow velocity data, the impact force data of water-carried objects, and the leakage point warning value into the leakage point danger warning model, solving the situation where the influence of water flow factors on leakage emergencies could not be comprehensively considered in the past, comprehensively and accurately evaluating the danger degree of leakage points, enabling the emergency rescue personnel to clearly understand the urgency of each leakage point, and reasonably arranging emergency rescue resources.

[0042] This application uses the leakage point priority model to sort the danger values output by the leakage point danger warning model, and in combination with the leakage point location information, traverses the leakage point association framework through the ant colony algorithm to determine the target priority order, overcoming the drawback of the lack of scientific planning for the treatment order of leakage points in traditional emergency rescue work, realizing the orderly and efficient progress of leakage point emergency rescue, ensuring the priority treatment of leakage points with high danger degrees and key positions, effectively improving the emergency rescue efficiency, and reducing the risk of dam breakage.

[0043] This application constructs a plugging model based on a multi-layer perceptron neural network model, inputs the force data of the leakage point of the earth-rock dam after plugging, the real-time flow velocity of the plugged leakage point, and the impact force data of water-carried objects for processing and outputs the plugging force value, and compares it with the preset threshold library. It solves the problem of the lack of an effective evaluation mechanism after traditional plugging, can timely and accurately judge whether the plugged part is safe and reliable, and take further measures for the plugged points with risks in a timely manner, ensuring the durability of the plugging effect and enhancing the safety of the earth-rock dam.

[0044] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, so as to be implemented in accordance with the content of the specification, and in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following will be described in detail with reference to the preferred embodiments of this application and the accompanying drawings.

[0045] According to the following detailed description of the specific embodiments of this application in combination with the accompanying drawings, those skilled in the art will understand the above and other purposes, advantages, and features of this application more clearly. Description of the Drawings

[0046] In order to more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. In all the drawings, similar elements or parts are generally marked with similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0047] Figure 1 This is a schematic diagram of the process for the emergency repair method of seepage hazards in earth-rock dams based on multi-model collaboration of the present invention;

[0048] Figure 2 This is a flow chart for the priority ranking of the emergency repair method of seepage hazards in earth-rock dams based on multi-model collaboration of the present invention. Specific implementation manners

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. In the following description, specific details such as specific configurations and components are provided only to assist in a comprehensive understanding of the embodiments of the present application. Therefore, those skilled in the art should clearly understand that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present application. Additionally, descriptions of known functions and structures are omitted for clarity and conciseness in the embodiments.

[0050] It should be understood that the term "one embodiment" or "the present embodiment" mentioned throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the present application. Therefore, the appearances of the term "one embodiment" or "the present embodiment" throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0051] In addition, the present application may repeat reference numerals and / or letters in different examples. This repetition is for the purpose of simplicity and clarity, and in itself does not indicate the relationship between the various embodiments and / or settings discussed.

[0052] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, B exists alone, and A and B exist simultaneously. The term " / and" in this article describes another association relationship of associated objects, indicating that two relationships can exist. For example, A / and B can represent: A exists alone, and A and B exist simultaneously. Additionally, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0053] The term "at least one" in this article is merely a description of the association relationship of associated objects, indicating that three relationships can exist. For example, at least one of A and B can represent: A exists alone, A and B exist simultaneously, and B exists alone.

[0054] It should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion.

[0055] Embodiment 1

[0056] This embodiment mainly describes a method for emergency rescue of seepage hazards of earth-rock dams based on multi-model collaboration, such as Figure 1 shown, including:

[0057] Obtain at least one or more seepage points of the earth-rock dam;

[0058] Collect the force data of the seepage points of the earth-rock dam and input it into a pre-constructed early warning model to output the early warning value of the seepage points;

[0059] Obtain the current water flow velocity data and the impact force data of the water-borne substances in the watershed area of the seepage points of the earth-rock dam, and input them into the seepage point danger early warning model in combination with the early warning value of the seepage points to output the danger value of the seepage points;

[0060] Obtain the danger value of the seepage points through the seepage point priority model and conduct real-time priority plugging sorting;

[0061] Obtain the force data of the seepage points of the earth-rock dam after plugging, collect the real-time flow velocity and the impact force of the water-borne substances at the plugged seepage points, input them into a pre-constructed plugging model to output the plugging force value, and determine the warning for the plugged seepage points to break away through a preset threshold library;

[0062] Real-time output the danger value of at least one seepage point of the earth-rock dam, adjust the priority order of the seepage points of the earth-rock dam, and conduct emergency rescue of the seepage points of the earth-rock dam with priority;

[0063] Among them, three or more levels of threshold intervals are set for obtaining the danger value of the seepage points, and different categories are realized through the corresponding level threshold intervals. The plugging types include but are not limited to grouting, geotextile sandbags, and expansion balls.

[0064] Furthermore, the output of the early warning value of the seepage points by the early warning model specifically includes:

[0065] Obtain the force data of the dam under normal and seepage conditions of the earth-rock dam, conduct preprocessing, and extract characteristic parameters; where the preprocessing includes removing outliers and filling in missing values;

[0066] Construct an initial early warning model using the logistic regression algorithm through the characteristic parameters, and the formula is: , where are the selected characteristic parameters, is the coefficient of the selected characteristic parameter, is the intercept, is the probability of the risk of leakage points in the earth-rock dam;

[0067] Using the characteristic parameters of the dam stress data under normal and leakage conditions of the earth-rock dam for training, the characteristic parameter coefficients in the logistic regression algorithm are determined by the maximum likelihood estimation method Iterative training meets the verification requirements, forming an early warning model and outputting the leakage point early warning value.

[0068] Furthermore, the waterborne substances include, but are not limited to, stones, gravel, and sediment;

[0069] The waterborne substances are obtained through magnetic imaging technology. The real-time cross-section of the water flow is scanned by a magnetic imaging sensor to obtain the contour information of the real-time cross-section of the leakage point watershed area of the earth-rock dam, and the waterborne substances are judged through the contour information.

[0070] Furthermore, the acquisition of the impact force data of the waterborne substances specifically includes:

[0071] Predict the volume of stones, gravel, and sediment in the unit volume of water flow content through magnetic imaging technology, and determine the mass of stones, gravel, and sediment through the volume The formula is: where , and are the unit volume ratios of the cross-sections of stones, gravel, and sediment, , and are the densities of stones, gravel, and sediment, is the unit volume;

[0072] Determine the speed of the waterborne substances through a speedometer ;

[0073] Calculate the water impact force, and the formula is: where is the unit water impact force, is the unit area, is the water impact force, where where is the speed correction coefficient, is the time.

[0074] Furthermore, the construction of the leakage point danger early warning model specifically includes:

[0075] Obtain the water flow velocity data and the impact force data of the waterborne substances, perform preprocessing, and extract characteristic parameters;

[0076] Dividing the characteristic parameters into a training set, a validation set, and a test set;

[0077] Inputting the training set data into a random forest model for training, randomly sampling with replacement from the training set to construct a training subset of decision trees, and repeating continuously until the preset maximum depth or the number of samples at a node is less than the minimum number of samples;

[0078] Adding the decision trees to the random forest model and performing multiple rounds of training to form an initial leakage point hazard warning model;

[0079] Inputting the validation set and the test set into the initial leakage point hazard warning model for performance evaluation, and adjusting the parameters of the initial leakage point hazard warning model to form a leakage point hazard warning model.

[0080] Further, the output of the leakage point hazard value comprehensively outputs the prediction results of the decision trees through the random forest model, and uses the mean method to output the leakage point hazard value. The formula is: , where is the prediction result of decision tree , and

[0081] is the leakage point hazard value.

[0082] Further, the leakage point priority model obtains the leakage point hazard values output by the leakage point hazard warning model, sorts them from high to low in terms of hazard values to form a first hazard value sequence; the leakage point priority model sorts according to the key points of the dam structure by obtaining the position information of the leakage points to form a second hazard value sequence; constructs a leakage point association framework, takes the leakage point hazard values in the first hazard value sequence as nodes, assigns the position information of the leakage point key points in the second hazard value sequence as important information, traverses the leakage point association framework through the ant colony algorithm, and selects the next node according to the pheromone concentration on the path of the leakage point association framework, and performs multiple rounds of iteration to form the target leakage point priority order.

[0083] Further, for obtaining the target leakage point priority order, extracting the leakage point hazard value of the first leakage point and judging the preset threshold interval it belongs to; the preset threshold interval includes but is not limited to a high-risk interval, a medium-risk interval, and a low-risk interval.

[0084] Further, the plugging model is constructed by a multi-layer perceptron neural network model. The plugging force value is output by the multi-layer perceptron neural network model, and it is determined whether the plugging leakage point is out of warning according to a preset threshold library, specifically including:

[0085] Obtain the force data of the leakage point of the earth-rock dam after plugging, the real-time flow velocity of the plugging leakage point and the impact force of the water-borne substances, perform standardization processing to form standardized data;

[0086] Input the standardized data into the input layer of the multi-layer perceptron neural network model, perform weighted summation through the hidden layer of the multi-layer perceptron neural network model, and output the plugging force value through the output layer;

[0087] Compare the plugging force value with the out-of-warning threshold in the preset threshold library. If the plugging force value is less than the out-of-warning threshold, the plugging part is within the safe range, and it is determined that the plugging leakage point is out of warning; if the plugging force value is greater than or equal to the out-of-warning threshold, there is a risk, and it is promoted to the high-risk interval of the preset threshold interval for processing.

[0088] This embodiment details that by constructing an early warning model, accurately outputting the early warning value of the leakage point by using the force data of the earth-rock dam, combining the water flow velocity and the impact force data of the water-borne substances to construct a danger early warning model to output the danger value, it can accurately evaluate the leakage situation, solve the problem that the traditional and existing technologies are not scientific in judging the degree of leakage danger, and with the help of the leakage point priority model, combining the danger value and the position information, realize the reasonable ranking of the leakage point emergency rescue, change the situation of resource waste or untimely emergency rescue caused by the lack of scientific ranking in the past, adopt diversified and adapted plugging methods such as concrete rapid-setting grouting, geotextile membrane sandbag combination, and expansion ball plugging for leakage points of different danger levels, and evaluate the plugging effect through the plugging model, effectively improving the success rate and stability of plugging, overcoming the defects of single and poor effect of traditional plugging methods, greatly improving the efficiency and reliability of the emergency rescue of the earth-rock dam leakage danger, and ensuring the safety of the dam.

[0089] Based on Embodiment 1, obtain the force data of the leakage point of the earth-rock dam and the force data of the leakage point of the earth-rock dam after plugging, which are obtained through a distributed sensor group. The distributed sensor group includes a pressure sensor and a strain sensor. The pressure sensor usually adopts a piezoresistive structure, and its core component is a sensitive element made of semiconductor material. When subjected to an external force, its resistance value will change, and the resistance change is converted into a voltage signal output through a Wheatstone bridge. The strain sensor is based on the principle of resistance strain gauges. The strain gauges are pasted on the elastic element. When the elastic element undergoes strain with the soil body, the resistance value of the strain gauges changes, and then an electric signal proportional to the strain is generated.

[0090] For the layout of earth-rock dams, according to the structural characteristics of earth-rock dams and historical leakage data, sensors are arranged in layers at certain intervals and depths in the dam body, dam foundation, and weak parts predicted to have leakage, such as the dam shoulders and around culverts. For example, in the vertical direction of the dam body, a layer of sensors is set every 1-2 meters, and one sensor is arranged every 3-5 meters in the horizontal direction for each layer. When installing the sensors, it is necessary to ensure that they are in close contact with the soil to accurately sense the stress situation. After the layout is completed, the sensors transmit the pressure and strain signals monitored in real time to the data acquisition box through data lines. The data acquisition box processes the signals, such as amplification, filtering, and analog-to-digital conversion, and then transmits the processed digital signals to the computer control system. In the computer, data processing software is used to analyze and calculate the signals. According to the pre-set mechanical model algorithm, the electrical signals collected by the sensors are converted into actual stress values to obtain the stress data of the leakage points of the earth-rock dam.

[0091] Regarding the acquisition of stress data after plugging, the plugging operation will change the stress state of the soil around the leakage point. The previously arranged distributed sensor group continues to work, and technicians closely monitor the changing trend of the data transmitted by the sensor group. If the pressure or strain values gradually tend to be stable after plugging and fluctuate within the preset safety range, it indicates that the plugging effect is good and the stress of the soil is under control. On the contrary, if the data shows abnormal fluctuations or continues to increase and exceeds the safety threshold, it indicates that there are problems with the plugging and further inspection and treatment are required. In this way, the stress situation of the earth-rock dam after plugging can be evaluated timely and accurately, ensuring the safe and stable operation of the dam.

[0092] This embodiment adopts pressure and strain sensor technology. Its unique structure and careful layout ensure that it can accurately sense the stress of the earth-rock dam. The sensors convert the stress into electrical signals, which are collected, transmitted, and processed to provide key data for emergency rescue. For continuous monitoring after plugging, it can effectively judge the plugging effect and strongly guarantee the safety and stability of the dam.

[0093] Based on Embodiment 1, the current water flow velocity data of the leakage points of the earth-rock dam is obtained through a current meter. For the current meter, the Doppler effect of sound waves propagating in water is used to measure the water flow velocity. Regarding the layout of the leakage area in the earth-rock dam, according to the distribution of the leakage points of the dam and the water flow path, it is installed at the key positions near and upstream of the leakage points. For example, at the suspected leakage area opening.

[0094] The current meter continuously emits sound wave signals and receives the sound waves reflected by the suspended particles in the water body. By analyzing the frequency change of the reflected wave, the velocity of the water flow relative to the current meter is calculated according to the Doppler formula. The processor inside the instrument processes and stores these data, and sends the real-time water flow velocity data to the data receiving terminal through wired or wireless transmission methods to obtain the current water flow velocity data of the leakage points of the earth-rock dam.

[0095] In this embodiment, by obtaining the water flow velocity at the leakage point, a more accurate danger value is provided, making it more accurate in the priority ranking.

[0096] Based on Embodiment 1, as Figure 2 shown, for obtaining the danger value of the leakage point through the leakage point priority model and performing real-time priority plugging ranking. For the priority model, after receiving the danger value output by the leakage point danger warning model, it will quickly start the ranking program, arrange the danger values of the leakage points from high to low, so as to accurately form the first danger value sequence. In this process, each danger value is clearly marked and distinguished;

[0097] For the collection of the location information of the leakage point, determine the specific location of the leakage point in the dam structure through monitoring equipment, and rank the leakage points according to the importance of their locations based on the design drawings of the dam and the classification criteria of key structural parts, so as to form the second danger value sequence. For example, the leakage points located in the core bearing part of the dam body or near the connection of important water conservancy facilities will be given a higher location importance score;

[0098] When associating with the leakage point correlation framework, use the danger value in the first danger value sequence as the basic node, and at the same time associate the location information in the second danger value sequence as an important attribute of the node; in this framework, each node represents a leakage point, which contains key data such as the danger value and location information; through the ant colony algorithm, simulate the path selection behavior of ants in the process of finding food, and let the virtual "ants" traverse in the leakage point correlation framework. The "ants" will choose the next node according to the pheromone concentration on the path. The higher the pheromone concentration of the path, the greater the probability of being selected. After multiple rounds of iteration, continuously optimize the path selection, and finally form a scientific and reasonable priority order of the target leakage points.

[0099] During the emergency repair process, for the leakage points that have been successfully repaired, the system will automatically downgrade their danger values and reduce their weights in the subsequent priority ranking; when it is found that the water flow velocity changes, the danger degree of the leakage points will be re-evaluated; for the leakage points in the area where the water flow velocity increases, if the velocity increases significantly, their danger values will be upgraded accordingly and adjusted forward in the priority order; on the contrary, the leakage points in the area where the velocity increases slightly or the velocity decreases will be downgraded, ensuring that the emergency repair resources are always preferentially allocated to the leakage points that need to be dealt with most, and maximizing the safety of the earth-rock dam.

[0100] The leakage point priority model of this embodiment synthesizes the comprehensive hazard value and location information, forms an accurate priority order through scientific sorting and ant colony algorithm iteration, can degrade the repaired points and dynamically adjust according to the flow rate, ensure the reasonable allocation of emergency resources, efficiently handle dangerous leakage points, and effectively enhance the timeliness and effectiveness of the safety guarantee of earth-rock dams.

[0101] Based on Embodiment 1, for the leakage point hazard value of the first leakage point for which the priority ranking is extracted, determine the preset threshold interval; the setting of the preset threshold interval is based on in-depth analysis of the historical data of a large number of earth-rock dam leakage cases and combined with a professional water conservancy project risk assessment model. Considering multiple factors such as the magnitude of the force on the leakage point, the speed of the water flow, the impact force of the water-borne substances, and the key degree of the leakage point in the dam structure, statistical methods are used to determine the boundary values of different intervals.

[0102] If a high-risk interval, a medium-risk interval, and a low-risk interval are set for the preset threshold interval, the general characteristics of the high-risk interval are that the force on the leakage point is close to or exceeds the bearing limit of the dam material, the water flow velocity is extremely fast and it carries a large number of heavy objects such as large-sized stones, and the impact force is sufficient to quickly damage the dam structure, and the location is at the core or key weak part of the dam body; for the medium-risk interval, the force and flow velocity are at an intermediate level, and the water-borne substances pose a certain threat; for the low-risk interval, the force and flow velocity are relatively small, the impact force of the carried substances is weak, and the impact on the overall stability of the dam is small.

[0103] Different plugging types are adopted through the preset threshold interval; for the high-risk interval of the preset threshold interval, the concrete quick-setting grouting method is used, and the concrete quick-setting slurry is injected into the leakage channel through high-pressure equipment, and quickly solidifies to form a high-strength plugging structure; for the medium-risk interval, the method of laying geomembrane combined with sandbag piling is adopted; for the low-risk interval, an expansion ball carried by a drone is used for expansion plugging;

[0104] For the high-risk interval, after determining that a certain leakage point is in the high-risk interval by the concrete quick-setting grouting method, the construction team quickly transports the high-pressure equipment to the site. Through the pressure control system of the high-pressure equipment, it can ensure that the concrete quick-setting slurry is injected into the leakage channel at an appropriate pressure and flow rate. While ensuring quick solidification, it has extremely high strength and adhesiveness. When the slurry enters the leakage channel, it quickly fills the gap and solidifies in a very short time, tightly combines with the surrounding soil, forms a solid whole, effectively resists the erosion of high-speed water flow and strong impact force, fundamentally stabilizes the dam structure, and prevents the leakage from deteriorating further.

[0105] For the treatment of the medium-risk area, the method of laying geomembrane combined with piling up sandbags is adopted. Geomembranes with high strength and corrosion resistance are selected, and their materials have good flexibility and waterproof performance. Construction workers accurately cut the geomembrane according to the specific position and scope of the leakage point and lay it tightly on the leakage area. During the laying process, ensure that the geomembrane has no wrinkles or damages, and special sealing measures are adopted for the edge part to prevent water from seeping in through the gaps. Subsequently, sandbags are piled up on the geomembrane. There are strict requirements for the filling amount and piling method of sandbags to ensure the stability and compactness of the piled sandbags, forming a solid protective layer, which can not only enhance the impact resistance of the geomembrane but also further block the water flow, providing reliable protection for the dam.

[0106] For the plugging method of the low-risk area, an unmanned aerial vehicle (UAV) is used to carry an expansion ball for operation. The UAV is equipped with a positioning system and a delivery device and can quickly and accurately reach above the leakage point. The expansion ball is made of a special water-absorbing and expanding material. When not encountering water, it is small, light, and convenient for the UAV to carry. When the UAV reaches the designated position, the expansion ball is accurately delivered to the leakage point according to the preset program. After the expansion ball comes into contact with the water flow, it quickly expands and fills the leakage channel. Its expanded shape and size can perfectly fit the channel, effectively preventing the water flow from passing through. Moreover, its material has certain elasticity and durability, and can maintain the plugging effect for a long time, greatly improving the treatment efficiency and convenience of the low-risk leakage point.

[0107] For the plugging ball to quickly plug the leakage channel, it includes:

[0108] Loading the plugging ball into the special delivery device carried by the transport UAV;

[0109] Planning the optimal path through the position coordinates of the leakage point, using visual recognition technology to lock the position of the target point, adjusting the delivery height, and feeding back the position point coordinates and height to the ground control center;

[0110] Sending an instruction to activate the delivery device of the UAV and releasing the plugging ball near the leakage point;

[0111] The plugging ball impacts the leakage point through the internal propulsion device, quickly expands, and forms a water-blocking barrier.

[0112] For the point-to-point delivery of the delivery UAV, the delivery position point and the height of the position point are determined by detecting the current wind speed and wind direction to achieve precise delivery; the inside of the plugging ball includes a plurality of plugging balls. When impacting the leakage point, they quickly disperse to form a plugging net and quickly expand to fill the fine leakage parts and the edge area; the inside of the plugging ball stores the expansion filler, and the expansion filler includes but is not limited to polyacrylate sodium, expanded rubber, or polyurethane gel.

[0113] Through this differential plugging strategy for different threshold intervals in this embodiment, the leakage emergency repair work of earth-rock dams becomes more scientific and efficient. Whether it is the emergency repair in high-risk intervals or the effective treatment in medium-risk and low-risk intervals, it can maximize the safety and stability of the dam, reduce the potential threats of leakage risks to the surrounding environment and people's lives and property, and provide strong technical support and guarantee for the safe operation of water conservancy projects.

[0114] The above are only the preferred embodiments of the present invention, and it does not limit the protection scope of the present invention. For those skilled in the art, the present invention can have various changes and modifications; within the spirit and principle of the present invention, through conventional substitutions or the ability to achieve the same function, without departing from the principle and spirit of the present invention, any changes, modifications, substitutions, integrations, and parameter changes to these embodiments all fall within the protection scope of the present invention.

Claims

1. A multi-model collaborative earth-rock dam leakage emergency rescue method, characterized in that: include: Obtain at least one or more leakage points of the earth-rock dam; Collect the stress data of the leakage point of the earth-rock dam, input it into the pre-built early warning model, and output the early warning value of the leakage point; Obtain the current water velocity data and the impact force data of the objects carried by the water flow in the watershed area of ​​the leakage point of the earth-rock dam, combine the leakage point warning value with the input of the leakage point danger warning model, and output the leakage point danger value; The leakage point priority model is used to obtain the leakage point hazard value and perform real-time priority plugging sorting; Obtain the force data of the leakage point of the earth-rock dam after plugging, collect the real-time flow velocity and impact force of the water flow at the plugged leakage point, input them into the pre-built plugging model, output the plugging force value, and determine the plugging leakage point out of warning through the preset threshold library; Outputting the leakage point danger value of at least one leakage point of the earth-rock dam in real time, adjusting the priority order of the leakage points of the earth-rock dam, and carrying out emergency rescue of the leakage points of the earth-rock dam with priority; The leakage point hazard value is obtained to set three or more level threshold intervals, and different categories are realized through corresponding level threshold intervals. The plugging types include but are not limited to grouting, geomembrane sandbags and expansion balls; The outputting of the leakage point warning value through the warning model specifically includes: Obtain the stress data of the earth-rock dam under normal conditions and leakage conditions, perform preprocessing, and extract characteristic parameters; the preprocessing includes removing outliers and filling missing values; Through the characteristic parameters, the initial warning model is constructed using the logistic regression algorithm. The formula is: ,in is the selected characteristic parameter, is the selected characteristic parameter coefficient, is the intercept, is the probability of leakage risk in earth-rock dam; The characteristic parameters of the dam stress data under normal and leakage conditions are used for training, and the characteristic parameter coefficients in the logistic regression algorithm are determined by the maximum likelihood estimation method. , iterative training meets the verification requirements, forms an early warning model, and outputs the leakage point early warning value; The construction of the leakage point hazard early warning model specifically includes: Obtain water flow velocity data and impact force data of objects carried by water flow, perform preprocessing, and extract characteristic parameters; Divide the feature parameters into training set, validation set and test set; The training set data is input into the random forest model for training, samples are randomly extracted with replacement from the training set, and the training subset of the decision tree is constructed. This process is repeated until the preset maximum depth is reached or the number of node samples is less than the minimum number of samples. Add decision trees to the random forest model and conduct multiple rounds of training to form an initial leakage point hazard warning model; Input the validation set and the test set into the initial leakage point hazard warning model to evaluate the performance, adjust the parameters of the initial leakage point hazard warning model, and form a leakage point hazard warning model; The leakage point priority model obtains the leakage point hazard value output by the leakage point hazard warning model, sorts the hazard value from high to low, and forms a first hazard value sequence; the leakage point priority model obtains the location information of the leakage point, sorts it according to the key points of the dam structure, and forms a second hazard value sequence; constructs a leakage point association framework, takes the leakage point hazard value in the first hazard value sequence as a node, and assigns important information to the location information of the key points of the leakage point in the second hazard value sequence, traverses the leakage point association framework through the ant colony algorithm, selects the next node according to the pheromone concentration on the path of the leakage point association framework, and iterates multiple rounds to form a priority order of the target leakage points.

2. The method for emergency rescue of earth-rock dam leakage based on multi-model collaboration according to claim 1 is characterized in that: The objects carried by the water flow include but are not limited to stones, gravel and mud; The objects carried by the water flow are obtained through magnetic imaging technology. The real-time cross-section of the water flow is scanned by a magnetic imaging sensor to obtain the contour information of the real-time cross-section of the watershed area of ​​the leakage point of the earth-rock dam, and the objects carried by the water flow are determined based on the contour information.

3. The method for emergency rescue of earth-rock dam leakage based on multi-model collaboration according to claim 1 is characterized in that: The acquisition of the impact force data of the object carried by the water flow specifically includes: Predict the volume of rocks, gravel and silt per unit volume of water flow using magnetic imaging technology, and determine the mass of rocks, gravel and silt by volume , the formula is: ;in , and is the unit volume ratio of stone, gravel and mud and sand sections, , and is the density of rocks, gravel and sediment, is the unit volume; Determine the speed of the water flow by using a velocimeter ; Calculate the water flow impact force, the formula is: ,in Unit water flow impact force, is the unit area, is the water flow impact force, ,in is the speed correction factor, For time.

4. The method for emergency rescue of earth-rock dam leakage based on multi-model collaboration according to claim 1 is characterized in that: The output of the leakage point danger value is to comprehensively output the prediction results of the decision tree through the random forest model, and the leakage point danger value is output using the mean method, and the formula is: ,in For decision tree The prediction results, It is the danger value of leakage point.

5. The method for emergency rescue of earth-rock dam leakage based on multi-model collaboration according to claim 1 is characterized in that: The priority order of the target leakage points is obtained, the leakage point risk value of the first leakage point is extracted, and the preset threshold interval to which it belongs is determined; the preset threshold interval includes but is not limited to a high-risk interval, a medium-risk interval and a low-risk interval.

6. The method for emergency rescue of earth-rock dam leakage based on multi-model collaboration according to claim 5 is characterized in that: Different types of plugging are adopted in the preset threshold intervals; the high-risk area of ​​the preset threshold interval adopts the concrete rapid-setting grouting method, in which the concrete rapid-setting slurry is injected into the leakage channel through high-pressure equipment, and quickly solidifies to form a high-strength plugging structure; the medium-risk area adopts the method of geomembrane laying combined with sandbag stacking; the low-risk area is expanded and sealed by drones carrying expansion balls.

7. The method for emergency rescue of earth-rock dam leakage based on multi-model collaboration according to claim 1 is characterized in that: The plugging model is constructed by a multi-layer perceptron neural network model, and the plugging force value is output by the multi-layer perceptron neural network model. It is determined whether the plugging leakage point is out of warning according to the preset threshold library, which specifically includes: Obtain the force data of the leakage point of the earth-rock dam after plugging, the real-time flow velocity of the plugged leakage point and the impact force of the water flow carrying objects, perform standardization processing, and form standardized data; The standardized data is input into the input layer of the multi-layer perceptron neural network model, weighted sum is performed through the hidden layer of the multi-layer perceptron neural network model, and the blocking force value is output through the output layer; The plugging force value is compared with the escape warning threshold in the preset threshold library. If the plugging force value is less than the escape warning threshold, the plugging part is in a safe range and the plugging leakage point is determined to be out of warning. If the plugging force value is greater than or equal to the escape warning threshold, there is a risk and it is elevated to the high-risk range of the preset threshold range for processing.

Images