Earth and rockfill dam seepage safety monitoring, forecasting and early warning system and method

Abstract

The invention relates to an earth and rockfill dam seepage safety monitoring, forecasting and early warning system and method, and belongs to the technical field of dam safety monitoring and artificial intelligence, and the system comprises a data acquisition module, a dam safety monitoring and forecasting model module, a dam safety early warning analysis module, and an early warning setting and pushing module. On the basis of a statistical model, engineering data such as geological exploration, detection and test are collected and analyzed according to rainwater conditions and dam safety monitoring data, the equivalent reservoir water level of a leakage channel is deduced, a seepage pressure monitoring and forecasting model considering the lagging time of the leakage channel is established, seepage pressure four-level monitoring and early warning indexes are constructed, and the seepage pressure monitoring and early warning efficiency is improved. On-line early warning of dam seepage safety is achieved, and the safety state of the dam is evaluated in real time.

Description

Technical Field

[0001] The invention relates to a system and method for monitoring and forecasting early warning of earth-rock dam seepage safety, and belongs to the field of dam safety monitoring technology and artificial intelligence technology. Background Art

[0002] As an important water conservancy project facility, the safety status of dams directly affects the safety of people's property, social security and stability, and economic development in the region. An effective dam safety monitoring, forecasting and early warning system can obtain the seepage data of the dam body in real time, and predict potential risks through model analysis, which helps to timely discover hidden dangers and take corresponding measures to ensure the safe operation of the dam. Taking the upstream water level, downstream water level, rainfall, etc. as independent variables, and the dam seepage pressure water level as the dependent variable, the linear and nonlinear relationship expressions of the dependent and independent variables are established, the dam seepage safety monitoring and early warning indicators are formulated, the dam seepage safety situation is diagnosed in real time, and the future dam seepage status is predicted. These are important theoretical methods and technical means to ensure the safe operation of the dam. Therefore, the development of efficient monitoring and early warning models and systems has important theoretical and practical significance.

[0003] The Chinese invention patent with the publication number CN113221215A discloses a dynamic visualization monitoring and analysis method for seepage of earth-rock dam based on BIM model. The invention is applicable to the field of dam monitoring. The technical scheme of the invention is: obtain real-time monitoring data through the buried measuring points at the upstream, dam foot and each section of the dam under the infiltration line; calculate the measuring point pressure head according to the seepage pressure of the measuring point, and obtain the coordinates of the z-axis projection point of the measuring point on the infiltration line when the spatial coordinates (x, y, z) of the measuring point are determined; generate the infiltration line on the BIM model and visualize it according to the monitoring data and the infiltration line generation method in the previous step; use the particle swarm algorithm to optimize the inversion of the permeability coefficient of each material partition of the dam, generate the seepage field of the dam body and visualize it on the BIM model of the dam body section; perform range check on the real-time monitoring seepage head data, the permeability coefficient of each partition obtained by inversion in the previous step, and the seepage field data obtained by forward calculation. If the system preset critical value is exceeded, an alarm message is generated at the corresponding problem point.

[0004] However, in the existing technology, it is difficult to consider the influence of the lag time of the concentrated leakage channel of the dam body and foundation, and there are problems such as low accuracy, insufficient prediction and early warning capabilities of nonlinear response under complex seepage conditions, and low intelligence, making it difficult to achieve real-time dam safety monitoring and early warning. To this end, a model method and system for monitoring, forecasting and early warning of earth-rock dam seepage safety considering the lag time of the leakage channel is proposed.

[0005] Therefore, there is a need for a seepage safety monitoring, forecasting and early warning system and method for earth-rock dams. Based on the automated monitoring data of dam safety, the influence of the lag time of the concentrated leakage channel in the dam body and foundation is considered to improve the accuracy, nonlinear response forecasting and early warning capabilities under complex seepage conditions, and the degree of intelligence, so as to realize real-time dam safety monitoring and early warning. Summary of the invention

[0006] The technical problem to be solved by the present invention is: in order to overcome the shortcomings of the existing technology, a system and method for monitoring, forecasting and warning of earth-rock dam seepage safety is provided based on dam safety automated monitoring data, taking into account the influence of the lag time of the concentrated leakage channel of the dam body and foundation, improving the accuracy, nonlinear response prediction and warning capabilities under complex seepage conditions and the degree of intelligence, so as to realize real-time dam safety monitoring and warning.

[0007] The technical solution adopted by the present invention to solve the above problems is: a seepage safety monitoring, forecasting and early warning system for earth-rock dams, including a data acquisition module, a dam safety monitoring and forecasting model module, a dam safety early warning analysis module, and an early warning setting and pushing module; The data acquisition module is used to collect reservoir water level, rainfall and dam seepage monitoring data; The dam safety monitoring and prediction model module is used to construct a dam seepage safety monitoring and prediction model based on reservoir water level, rainfall and dam seepage monitoring data to predict the seepage safety status; The dam safety early warning analysis module is used to construct seepage safety monitoring and early warning indicators; The dam safety monitoring and forecasting model module and the dam safety early warning analysis module visualize the modeling parameters and results; The warning setting and push module is used to set the warning mode of each seepage pressure measuring point to manual warning or automatic warning, count and display the warning results, set the warning push object, and send the warning information through the communication network.

[0008] Preferably, the dam seepage safety monitoring and forecasting model includes two types: a statistical model and a machine learning model.

[0009] A seepage safety monitoring and prediction model method for earth-rock dams comprises the following steps: S1: Obtain historical monitoring data on water level, rainfall, and seepage of the earth-rock dam reservoir, and collect data on geological surveys, geophysical detection, grouting tests, etc. of the dam; S2: Using monitoring, geological survey, testing and test data, calculate the equivalent reservoir water level taking into account the lag time of the leakage channel; S3: Based on the historical monitoring data in step S1 and the equivalent reservoir water level of the leakage channel in step S2, a seepage monitoring and prediction model for the earth-rock dam is constructed, and the prediction model is used to predict the seepage safety status.

[0010] Preferably, the hysteresis effect in step S2 is described by a normal distribution function, thereby obtaining the equivalent reservoir water level expression of the leakage channel as follows:

[0011] Where Hd is the equivalent reservoir water level considering the lag time of the leakage channel; is the normal distribution weight function; H(t) is the reservoir water level; α1 is the adjustment parameter; x1 is the number of days of water level lag; x2 is the distribution parameter of water level lag effect; is a step function.

[0012] Preferably, in the equivalent reservoir water level of the seepage channel, the expression of hysteresis days derived by using Darcy's law of fractured media is:

[0013] Where α is the adjustment parameter; v is the flow velocity; b is the crack width; μ is the viscosity coefficient of water; L is the seepage diameter; and P(t) is the seepage pressure water level at the measuring point.

[0014] Preferably, the step function expression in the equivalent reservoir water level of the seepage channel is:

[0015] Where Hc is the critical reservoir water level elevation for seepage state change, which is related to the elevation of the leakage channel.

[0016] Preferably, the earth-rock dam seepage monitoring and prediction model expression in step S3 is as follows:

[0017] Where h is the predicted value of the seepage pressure water level; hH is the reservoir water level component, hP is the rainfall component, hθ is the time-effect component, Hd0 is the equivalent reservoir water level function of the continuous medium, Hd is the equivalent water level function of the leakage channel, Pd is the equivalent rainfall, θ is 1 / 100 of the cumulative number of days from the start date of modeling, b0~b5 are regression coefficients, and b0 is a constant term.

[0018] Preferably, the equivalent rainfall expression is as follows:

[0019] In the formula, is the normal distribution weight function; P(t) is the rainfall at time t, α2 is the adjustment parameter; x3 is the undetermined rainfall lag days, x4 is the rainfall effect distribution parameter of the undetermined coefficient, β is the infiltration transformation index, and β=0.4 is taken.

[0020] Preferably, the coefficients of the earth-rock dam seepage monitoring and prediction model are solved using multivariate linear regression and kernel extreme learning machine respectively.

[0021] A method for monitoring and warning the safety of earth-rock dam seepage, comprising the following steps: Sa: According to the established forecast model, the predicted value of seepage pressure is generated by inputting the forecast values ​​of rainfall and reservoir water level into the forecast model; Sb: Based on the predicted value of seepage pressure, the confidence interval method is used to construct the seepage pressure monitoring and early warning index; Sc: Based on historical data and real-time monitoring data, continuously update monitoring data, forecast models and early warning indicators, regularly maintain and update models, maintain model sensitivity, realize online early warning of dam seepage safety, and evaluate the dam safety status in real time; The seepage pressure monitoring and early warning indicators in step Sb include first-level, second-level, third-level, and fourth-level early warning indicators, corresponding to confidence levels α=0.3%, 1%, 5%, and 10%, respectively.

[0022] Compared with the prior art, the advantages of the present invention are: 1. The present invention proposes a seepage safety monitoring, forecasting and early warning system for earth-rock dams that takes into account the influence of seepage channels. It integrates multiple functions such as data collection and processing, forecast model establishment, early warning indicator construction, model and result visualization, and early warning information release to achieve real-time analysis and early warning of the safety status of seepage pressure; 2. The present invention constructs a leakage channel hysteresis time function based on Darcy's law of fractured media, thereby establishing a seepage monitoring and prediction model that takes into account the leakage channel hysteresis time, thereby improving the accuracy of seepage monitoring and prediction under complex seepage conditions; 3. Based on the proposed seepage monitoring and prediction model, the present invention uses the confidence interval method to propose a seepage monitoring and early warning method that takes into account the lag time of the leakage channel. It can continuously update the monitoring data, prediction model and early warning indicators, realize online early warning of dam seepage safety, and evaluate the safety status of the dam in real time. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a structural schematic diagram of a seepage safety monitoring, forecasting and early warning system for earth-rock dams according to the present invention; Figure 2 A flow chart of a method for monitoring and predicting earth-rock dam seepage safety according to the present invention; Figure 3 It is a process line diagram of rainfall, water level and seepage pressure of the dam in the embodiment; Figure 4 This is a statistical model forecast and warning result diagram in the embodiment; Figure 5 It is a cloud diagram of the optimization of the parameters of the core extreme learning machine model in the embodiment; Figure 6 This is a diagram of the forecast and warning results of the nuclear extreme learning machine model in the embodiment. DETAILED DESCRIPTION

[0024] like Figure 1 As shown, a seepage safety monitoring, forecasting and early warning system for earth-rock dams includes a data acquisition module, a dam safety monitoring and forecasting model module, a dam safety early warning analysis module, and an early warning setting and pushing module; The data acquisition module is used to collect reservoir water level, rainfall and dam seepage monitoring data; The dam safety monitoring and prediction model module is used to construct a dam seepage safety monitoring and prediction model based on reservoir water level, rainfall and dam seepage monitoring data to predict the seepage safety status. The dam seepage safety monitoring and prediction model includes two types: a statistical model and a machine learning model; The dam safety early warning analysis module is used to construct seepage safety monitoring and early warning indicators; The dam safety monitoring and forecasting model module and the dam safety early warning analysis module visualize the modeling parameters and results; The warning setting and push module is used to set the warning mode of each seepage pressure measuring point to manual warning or automatic warning, count and display the warning results, set the warning push object, and send the warning information through the communication network; like Figure 2 As shown, based on the earth-rock dam seepage safety monitoring, forecasting and early warning system, a model method for earth-rock dam seepage safety monitoring and forecasting is provided. This model method takes into account the lag time of the leakage channel and includes the following steps: S1: Obtain historical monitoring data on water level, rainfall, and seepage of the earth-rock dam reservoir, and collect data on geological surveys, geophysical detection, grouting tests, etc. of the dam; S2: Using monitoring, geological survey, testing and test data, calculate the equivalent reservoir water level taking into account the lag time of the leakage channel; S3: Based on the historical monitoring data in step S1 and the equivalent reservoir water level of the leakage channel in step S2, a seepage monitoring and prediction model for earth-rock dam is constructed, and the seepage safety state is predicted using the prediction model; The hysteresis effect in step S2 is described by a normal distribution function, and the equivalent reservoir water level expression of the leakage channel is obtained as follows:

[0025] Where Hd is the equivalent reservoir water level considering the lag time of the leakage channel; is the normal distribution weight function; H(t) is the reservoir water level; α1 is the adjustment parameter; x1 is the number of days of water level lag; x2 is the distribution parameter of water level lag effect; is a step function; In the equivalent reservoir water level of the leakage channel, the expression of hysteresis days derived by using Darcy's law of fractured media is:

[0026] In the formula, α is the adjustment parameter; v is the flow velocity; b is the crack width; μ is the viscosity coefficient of water; L is the seepage diameter; P(t) is the seepage pressure water level at the measuring point; The step function expression in the equivalent reservoir water level of the leakage channel is:

[0027] In the formula, Hc is the critical reservoir water level elevation for seepage state change, which is related to the elevation of the leakage channel; The earth-rock dam seepage monitoring and prediction model expression in step S3 is as follows:

[0028] Where, h is the predicted value of seepage pressure water level; hH is the reservoir water level component, hP is the rainfall component, hθ is the time-effect component, Hd0 is the equivalent reservoir water level function of the continuous medium, Hd is the equivalent water level function of the leakage channel, Pd is the equivalent rainfall, θ is 1 / 100 of the cumulative number of days from the start date of modeling, b0~b5 are regression coefficients, and b0 is a constant term; The equivalent rainfall expression is as follows:

[0029] In the formula, is the normal distribution weight function; P(t) is the rainfall at time t (mm), α2 is the adjustment parameter; x3 is the undetermined rainfall lag days, x4 is the rainfall effect distribution parameter of the undetermined coefficient, β is the infiltration transformation index, and β=0.4 is taken.

[0030] The coefficients of the earth-rock dam seepage monitoring and prediction model are solved by using multiple linear regression (MLR) and kernel extreme learning machine (KELM). Based on the earth-rock dam seepage safety monitoring forecasting and early warning system and earth-rock dam seepage safety monitoring forecasting model method, a method for earth-rock dam seepage safety monitoring and early warning is also provided, comprising the following steps: Sa: According to the established forecast model, the predicted value of seepage pressure is generated by inputting the forecast values ​​of rainfall and reservoir water level into the forecast model; Sb: Based on the predicted value of seepage pressure, the confidence interval method is used to construct the seepage pressure monitoring and early warning index; Sc: Based on historical data and real-time monitoring data, continuously update monitoring data, forecast models and early warning indicators, regularly maintain and update models, maintain model sensitivity, realize online early warning of dam seepage safety, and evaluate the dam safety status in real time; The seepage pressure monitoring and early warning indicators in step Sb include first-level, second-level, third-level, and fourth-level early warning indicators, corresponding to confidence levels α=0.3%, 1%, 5%, and 10%, respectively.

[0031] Embodiment: The specific steps of earth-rock dam seepage safety monitoring, forecasting and early warning are as follows: Step 1: Take the historical monitoring data of reservoir water level, rainfall, and seepage of a certain earth-rock dam from October 15, 2020 to October 15, 2024 as an example for forecast and warning. The process line diagram of rainfall, water level and seepage pressure of the dam is shown in Figure 3 According to data from previous geological surveys, geophysical detection, grouting tests, etc., there are three permeable areas at 20m, 45m and 102m below the dam crest on the right dam shoulder.

[0032] Step 2: Preset parameters such as the elevation of the seepage channel, the length of the seepage path, and the width of the seepage channel, establish the equivalent reservoir water level of the seepage channel, build a seepage monitoring and prediction model for earth-rock dams, and use the prediction model to predict the seepage safety state. The model factor coefficients are solved using multiple linear regression (MLR) and kernel extreme learning machine (KELM). The constructed seepage pressure water level prediction model considering the seepage channel is expressed as follows:

[0033] Step 3: According to the established forecast model, the predicted water level and rainfall are manually input or the system automatically reads the predicted water level and rainfall to generate the predicted value of seepage pressure. The confidence interval method is used to construct the four-level monitoring and early warning indicators of seepage pressure. The predicted value, standard deviation, and first to fourth level early warning indicators of each measuring point are shown in Table 1. The forecast and early warning results of the statistical model are shown in Figure 4 , machine learning model parameter optimization cloud diagram see Figure 5 , forecast and warning results are shown in Figure 6 ; Table 1 Unit: m

[0034] Step 4: When the newly connected seepage pressure monitoring value exceeds the warning indicator of the corresponding level, the system automatically generates warning information and sends it to relevant personnel through the communication network, realizing online automatic warning and real-time assessment of the dam safety status.

[0035] Step 5: Complete the system database development on the SQL Server platform and enter relevant data, use Java language to develop the information platform, use Python language to write the forecast and warning model script, use Json data format to transmit model parameters, and realize data interaction and forecast and warning system development.

[0036] In addition to the above embodiments, the present invention also includes other implementation modes. Any technical solutions formed by equivalent transformation or equivalent replacement should fall within the protection scope of the claims of the present invention.

Claims

1. A seepage safety monitoring, forecasting and early warning system for earth-rock dams, characterized by: It includes data acquisition module, dam safety monitoring and forecasting model module, dam safety early warning analysis module, early warning setting and push module; The data acquisition module is used to collect reservoir water level, rainfall and dam seepage monitoring data; The dam safety monitoring and prediction model module is used to construct a dam seepage safety monitoring and prediction model based on reservoir water level, rainfall and dam seepage monitoring data to predict the seepage safety status; The dam safety early warning analysis module is used to construct seepage safety monitoring and early warning indicators; The dam safety monitoring and forecasting model module and the dam safety early warning analysis module visualize the modeling parameters and results; The warning setting and push module is used to set the warning mode of each seepage pressure measuring point to manual warning or automatic warning, count and display the warning results, set the warning push object, and send the warning information through the communication network.

2. The earth-rock dam seepage safety monitoring, forecasting and early warning system according to claim 1 is characterized by: The dam seepage safety monitoring and prediction model includes two types: a statistical model and a machine learning model.

3. A model method for monitoring and predicting earth-rock dam seepage safety, characterized in that: The following steps are involved: S1: Obtain historical monitoring data on water level, rainfall, and seepage of the earth-rock dam reservoir, and collect data on geological surveys, geophysical detection, grouting tests, etc. of the dam; S2: Using monitoring, geological survey, testing and test data, calculate the equivalent reservoir water level taking into account the lag time of the leakage channel; S3: Based on the historical monitoring data in step S1 and the equivalent reservoir water level of the leakage channel in step S2, a seepage monitoring and prediction model for the earth-rock dam is constructed, and the prediction model is used to predict the seepage safety status.

4. The earth-rock dam seepage safety monitoring and prediction model method according to claim 3 is characterized by: The hysteresis effect in step S2 is described by a normal distribution function, and the equivalent reservoir water level expression of the leakage channel is obtained as follows: ； Where Hd is the equivalent reservoir water level considering the lag time of the leakage channel; is the normal distribution weight function; H(t) is the reservoir water level; α1 is the adjustment parameter; x1 is the number of days of water level lag; x2 is the distribution parameter of water level lag effect; is a step function.

5. The earth-rock dam seepage safety monitoring and prediction model method according to claim 3 is characterized by: In the equivalent reservoir water level of the leakage channel, the expression of hysteresis days derived by using Darcy's law of fractured media is: ; Where α is the adjustment parameter; v is the flow velocity; b is the crack width; μ is the viscosity coefficient of water; L is the seepage diameter; and P(t) is the seepage pressure water level at the measuring point.

6. The earth-rock dam seepage safety monitoring and prediction model method according to claim 3 is characterized by: The step function expression in the equivalent reservoir water level of the leakage channel is: ； Where Hc is the critical reservoir water level elevation for seepage state change, which is related to the elevation of the leakage channel.

7. The earth-rock dam seepage safety monitoring and prediction model method according to claim 3 is characterized by: The earth-rock dam seepage monitoring and prediction model expression in step S3 is as follows: ； Where h is the predicted value of the seepage pressure water level; hH is the reservoir water level component, hP is the rainfall component, hθ is the time-effect component, Hd0 is the equivalent reservoir water level function of the continuous medium, Hd is the equivalent water level function of the leakage channel, Pd is the equivalent rainfall, θ is 1 / 100 of the cumulative number of days from the start date of modeling, b0~b5 are regression coefficients, and b0 is a constant term.

8. The earth-rock dam seepage safety monitoring and prediction model method according to claim 3 is characterized by: The equivalent rainfall expression is as follows: ； In the formula, is the normal distribution weight function; P(t) is the rainfall at time t, α2 is the adjustment parameter; x3 is the undetermined rainfall lag days, x4 is the rainfall effect distribution parameter of the undetermined coefficient, β is the infiltration transformation index, and β=0.4 is taken.

9. The earth-rock dam seepage safety monitoring and prediction model method according to claim 3 is characterized by: The coefficients of the earth-rock dam seepage monitoring and prediction model are solved by using multivariate linear regression and kernel extreme learning machine respectively.

10. A method for monitoring and early warning of seepage safety in earth-rock dams, characterized in that: The following steps are involved: Sa: According to the established forecast model, the predicted value of seepage pressure is generated by inputting the forecast values ​​of rainfall and reservoir water level into the forecast model; Sb: Based on the predicted value of seepage pressure, the confidence interval method is used to construct the seepage pressure monitoring and early warning index; Sc: Based on historical data and real-time monitoring data, continuously update monitoring data, forecast models and early warning indicators, regularly maintain and update models, maintain model sensitivity, realize online early warning of dam seepage safety, and evaluate the dam safety status in real time; The seepage pressure monitoring and early warning indicators in step Sb include first-level, second-level, third-level, and fourth-level early warning indicators, corresponding to confidence levels α=0.3%, 1%, 5%, and 10%, respectively.

Images