Barrier-dam breach form prediction method based on case statistics and breaching process
Through the method based on case statistics and collapse process, the key parameters in the prediction of the dam burst morphology are determined, which solves the problem of insufficient accuracy of the prediction of vertical collapse degree and collapse widening, and achieves high accuracy of the dam burst morphology. Forecasting, especially when the degree of collapse is greater than 1/2, takes into account the river width and improves the prediction accuracy of the collapse widening.
Patent Information
- Application Number
- PCT/CN2023/128389
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-05-18
- Filing Date
- 2023-10-31
- Publication Date
- 2024-11-21
AI Technical Summary
When predicting the breach of the dam, the prediction of vertical collapse degree and breach width is insufficient in the prior art, and the impact of the dam structure and river channel width cannot be effectively considered.
Using a method based on case statistics and the collapse process, the basic parameters of the collapse case are obtained and classified, the key parameters affecting the degree of vertical collapse are determined, and the logistic regression expression is constructed to predict the degree of vertical collapse, and the degree of vertical collapse is calculated based on the collapse degree and the structure of the dam The top width and bottom width of the collapse vent are introduced, and the river width parameters are introduced to improve the accuracy of the prediction of collapse width.
Improve the accuracy of prediction of vertical collapse degree and prediction accuracy of collapse widening, and can accurately determine whether the vertical collapse degree of the dam is greater than 1/2 of the dam height, and consider the river channel width when the collapse degree is greater than 1/2, and improve the The prediction accuracy of the breach widening is achieved.
Smart Images

Figure CN2023128389_21112024_PF_FP_ABST
Abstract
Description
Prediction method of landslide dam breach morphology based on case statistics and breach process Technical Field
[0001] The present invention relates to the technical field of geological disaster emergency response, and in particular to a method for predicting the breach morphology of a landslide dam based on case statistics and breach processes. Background Art
[0002] A landslide dam is a loose accumulation of debris that blocks a river channel and forms a barrier lake after an earthquake or heavy rainfall triggers a landslide. Due to the large variation in particle size distribution within the dam, its overall stability and erosion resistance are low, posing a significant risk of failure. Once breached, the resulting flood would cause severe damage to downstream villages and projects. Predicting the peak flow rate and assessing the damage after a landslide dam occurs can help develop effective emergency response plans. Given the need for rapid predictions, parametric assumption models are commonly used, with the final breach shape of the landslide dam being a key parameter that directly influences prediction accuracy.
[0003] Regarding the degree of vertical collapse, emergency plans are often formulated based on the estimated fraction of vertical collapse during on-site disposal. There is a lack of systematic determination methods, and existing predictions do not take into account the changes in the dam body and reservoir capacity after disposal. Similarly, the parameters in the few existing linear fitting formulas are mainly obtained through accuracy comparison, lacking a theoretical basis for parameter selection, and do not consider the impact of the vertical structure of the dam body on the degree of vertical collapse.
[0004] Furthermore, compared to formulas for predicting the widening of earth-rockfill dams, the prediction of the widening of landslide dam breaches requires further refinement. While a few widening formulas based on landslide dam cases exist, these are often partial failures, with a different degree of breach than earth-rockfill dams. Furthermore, landslide dams often occur in narrow, deep channels. When the vertical breach is significant, the widening is also limited by the river width. Existing research has not yet considered this parameter.
[0005] In summary, when predicting the breach morphology of a landslide dam, the existing technology has problems with the lack of theoreticality and accuracy in predicting the vertical breach extent and breach width.
[0006] Summary of the Invention
[0007] The present invention aims to solve the problems of insufficient theoretical and accuracy in existing methods for predicting the vertical breach degree of landslide dams, and proposes a method for predicting the breach morphology of landslide dams based on case statistics and breach processes.
[0008] The technical solution adopted by the present invention to solve the above technical problems is:
[0009] A method for predicting the breach morphology of a landslide dam based on case statistics and breach processes, the method comprising:
[0010] Obtaining natural breach cases from the case library and the corresponding basic parameters and vertical breach degree of each breach case. The basic parameters include: dam height, dam length, aspect ratio, dam volume, dam material composition, upstream flow, and reservoir capacity data;
[0011] Classify basic parameters and vertical failure degree, and determine the key parameters affecting the vertical failure degree from the basic parameters based on correlation analysis, cross-correlation analysis and failure process analysis;
[0012] The vertical failure degree is divided into two categories: less than or equal to 1 / 2 dam height and greater than 1 / 2 dam height. Based on the key parameters and vertical failure degree of each failure case, an expression for calculating the probability of vertical failure degree greater than 1 / 2 dam height is constructed based on logistic regression analysis.
[0013] The key parameters of the landslide dam to be predicted are obtained, and whether the vertical failure degree of the landslide dam to be predicted is greater than 1 / 2 of the dam height is determined based on the expression according to the category of the key parameters of the landslide dam to be predicted.
[0014] Furthermore, the expression for calculating the probability that the vertical failure degree is greater than 1 / 2 of the dam height is as follows:
[0015] Where p is the probability that the vertical failure degree is greater than 1 / 2 of the dam height, β0 is a constant term, and β1, β2, ..., β n are partial regression coefficients, x1, x2, …, x n The category of the key parameter.
[0016] Furthermore, the key parameters are the aspect ratio and the dam material. The expression for calculating the probability that the vertical failure degree is greater than 1 / 2 of the dam height is as follows:
[0017] Where x1 is the type of dam material, and x2 is the type of aspect ratio.
[0018] Furthermore, the method further comprises:
[0019] When it is determined that the vertical failure degree of the predicted landslide dam is greater than 1 / 2 of the dam height, it is determined whether the dam body material is composed of loess and clay. If so, the dam height is used as the vertical failure degree of the predicted landslide dam; otherwise, 2 / 3, 3 / 4 or 3 / 5 of the dam height is used as the vertical failure degree of the predicted landslide dam.
[0020] Furthermore, the method further comprises:
[0021] When it is determined that the vertical failure degree of the landslide dam to be predicted is not greater than 1 / 2 of the dam height, the slope ratio of the landslide dam to be predicted is obtained. If the slope ratio is within a first preset range, 1 / 3 of the dam height is used as the vertical failure degree of the landslide dam to be predicted. If the slope ratio is within a second preset range, 1 / 2 of the dam height is used as the vertical failure degree of the landslide dam to be predicted.
[0022] Furthermore, the method further comprises:
[0023] When the landslide dam to be predicted is a binary stacked structure or a ternary stacked structure, the elevation of the top surface of the bottom layer of the binary stacked structure or the ternary stacked structure is obtained, and it is determined whether the elevation corresponding to the vertical failure degree of the landslide dam to be predicted is less than the elevation of the top surface of the bottom layer. If so, the vertical failure degree of the landslide dam to be predicted is re-determined.
[0024] Furthermore, the re-determining of the vertical failure degree of the landslide dam to be predicted specifically includes:
[0025] If the bottom layer of the binary stacked structure is composed of intact rock layers and the top layer is composed of broken soil and rock or isolated pieces of broken rock, the ratio of the height of the top layer to the dam height shall be used as the re-determined vertical failure degree;
[0026] If the bottom layer of the binary stacked structure is composed of broken rock and soil, and the top layer is composed of isolated broken rock, then a value between the vertical failure degree and 2 / 3 of the dam height shall be selected as the re-determined vertical failure degree;
[0027] If the bottom layer of the ternary stack structure is composed of intact rock layers, the middle layer is composed of isolated pieces of gravel, and the top layer is composed of broken stone and soil, the ratio of the height of the top layer to the middle layer and the height of the dam will be used as the re-determined vertical failure degree.
[0028] Furthermore, the method further comprises:
[0029] Obtain the slope ratio, length, height and actual discharge capacity of the landslide dam to be predicted;
[0030] Determine the vertical failure depth of the predicted landslide dam based on its vertical failure degree and dam height, and calculate the breach cross-sectional area of the predicted landslide dam after its failure based on its dam body slope ratio, dam length and vertical failure depth;
[0031] The breach top width and breach bottom width of the predicted landslide dam after its breach are calculated based on the breach cross-sectional area of the predicted landslide dam, the actual downstream storage capacity and the vertical breach depth.
[0032] Furthermore, when the vertical breach degree of the predicted landslide dam is less than or equal to 1 / 2 of the dam height, the calculation formulas for the breach top width and breach bottom width after the predicted landslide dam breach are as follows:
[0033] b mtE′=13.858W′ 0.4577 H′;
[0034] b mb E′=5.1707W′ 0.4964 H′;
[0035] Where b mt is the width of the breach top, b mb is the bottom width of the breach, E′ is the cross-sectional area of the breach, W′ is the actual discharge capacity, and H′ is the vertical breach depth.
[0036] Furthermore, when the vertical breach degree of the predicted landslide dam is greater than 1 / 2 of the dam height, the calculation formulas for the breach top width and breach bottom width after the predicted landslide dam breach are as follows:
[0037] b mt E′=28.818W′ -0.088 B 1.549 H′;
[0038] b mb E′=6.991W′ 0.357 B 0.406 H′;
[0039] Where b mt is the width of the breach top, b mb is the bottom width of the breach, E′ is the cross-sectional area of the breach, B is the width of the river channel, W′ is the actual downstream storage capacity, and H′ is the vertical breach depth.
[0040] The beneficial effects of the present invention are as follows: the method for predicting the breach morphology of a landslide dam based on case statistics and breach process described in the present invention screens the key parameters for predicting the vertical breach degree through breach cases and based on correlation analysis, cross-column analysis, and breach process analysis, thereby improving the accuracy of key parameter screening and thus improving the accuracy of predicting the vertical breach degree. Furthermore, an expression is constructed based on the key parameters of the breach case, and based on the expression, it is possible to accurately predict whether the vertical breach degree of the landslide dam is greater than 1 / 2 of the dam height. For special landslide dams, the vertical breach degree of the landslide dam is further determined by the dam body material, slope ratio, and dam body structure, further improving the accuracy of the prediction of the vertical breach degree. In addition, the breach width of the landslide dam is calculated based on a formula created based on the predicted vertical breach degree and case analysis, and when the vertical breach degree is greater than 1 / 2 of the dam height, the parameter river channel width is introduced to improve the accuracy of the breach width prediction. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] FIG1 is a flow chart of a method for predicting the breach morphology of a landslide dam based on case statistics and breach processes according to an embodiment of the present invention;
[0042] FIG2 is a schematic diagram of a cross-sectional analysis of the aspect ratio, dam material, and vertical failure degree according to an embodiment of the present invention;
[0043] FIG3 is a schematic diagram of a collapse process analysis according to an embodiment of the present invention;
[0044] FIG4 is a schematic diagram of a process for predicting breach widening according to an embodiment of the present invention;
[0045] FIG5 is a schematic structural diagram of the breach according to an embodiment of the present invention. DETAILED DESCRIPTION
[0046] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings and examples.
[0047] The present invention aims to improve the accuracy of dam breach morphology prediction and provides a dam breach morphology prediction method based on case statistics and breach process. Referring to FIG1 , the method includes the following steps:
[0048] Step 1: Obtain natural breach cases from the case library and the basic parameters and vertical breach degree corresponding to each breach case. The basic parameters include: dam height H, dam length L, length-to-height ratio L / H, dam volume V, dam material composition, upstream flow Q, and reservoir capacity data W;
[0049] In this example, 49 natural landslide dam failure cases with detailed basic parameters at home and abroad since the 1970s were collected, and the corresponding basic parameters and vertical failure degrees were obtained respectively.
[0050] Step 2: Classify the basic parameters and vertical failure degree, and determine the key parameters affecting the vertical failure degree from the basic parameters based on correlation analysis, cross-correlation analysis and failure process analysis;
[0051] In this embodiment, the basic parameters such as dam height H, dam length L, length-to-height ratio L / H, dam volume V, upstream inflow Q and reservoir data W are divided into Class III (3) or Class IV (4) (reservoir capacity refers to volume classification value) according to the risk level boundary value. The dam material is divided into Class I (1), II (2), III (3), IV (4) and V (5) according to the fine-grained group (clay, loess), sand-soil mixture (mainly sand), crushed stone, gravel, sand (mainly gravel and crushed stone), block crushed stone (particle size d < 200 mm), and boulder / large block stone (particle size d is in meters).
[0052] In addition, the following classification is made for the description of basic parameter profiles in individual cases:
[0053] (1) If there is no specific value for the upstream flow Q, it is described as "very small", "small" or "large", and the flow is assigned to Class I (1), II (2), IV (4); if there is heavy rainfall, ice and snow melting, or there are other dam failures upstream, the upstream flow Q is assigned to Class IV (4);
[0054] (2) For some dams that collapsed naturally after treatment (excluding artificially induced dam failures), the treated dams will be reclassified as "new dams" based on the treatment situation. If the data is incomplete and it is determined that the treatment did not cause a change in the category of the dam's basic parameters, the category of the data before the treatment will prevail.
[0055] (3) The vertical failure degree is divided into four categories according to the following: less than or equal to 1 / 2 dam height, approximately equal to 1 / 2 dam height, greater than 1 / 2 dam height and less than dam height, and equal to dam height (total failure).
[0056] In this embodiment, after classification, the data of the 49 landslide dam failure cases collected above are shown in the following table: Note: The dam body characteristics in the table are all post-treatment data. * " stands for post-burst disposal of natural disasters.
[0057] By conducting correlation analysis, cross-correlation analysis and collapse process analysis based on the data of collapse cases, the key parameters affecting the degree of vertical collapse can be determined from the basic parameters.
[0058] Among them, correlation analysis (parameter determination basis 1):
[0059] The results of Kendall's stau-b correlation analysis are shown in the following table: Note: ***, **, * represent significance levels of 1%, 5%, and 10%, respectively.
[0060] The table above shows that the key parameters for vertical failure severity are dam material (-) and aspect ratio (+), with dam material having a more pronounced impact. Experience shows that dams with poor scour resistance and steeper slopes are more susceptible to effective scour and higher vertical failure severity than dams with stronger scour resistance and gentler slopes. While the negative correlation with dam material is reasonable, the positive correlation with aspect ratio requires further verification.
[0061] Contingency cross analysis (parameter determination based on 2)
[0062] Please refer to Figure 2, which shows a cross-sectional analysis of aspect ratio, dam material, and vertical failure severity. The positive correlation of aspect ratio is primarily due to the fact that, within Category III aspect ratio (Figure 2a), 71.40% of the total failure cases correspond to dams composed of Category I materials (Figure 2b), which almost entirely failed (100%). Examples include the Yigong Weir and the Milingala Weir. Aspect ratio is highly correlated with dam material (-0.537 (p = 0.006***)). Analysis of the impact of aspect ratio should be based on the assumption that materials are identical. However, due to the varying materials used in these cases, and the fact that most dams have wide and gentle cross-sections, differentiation is difficult. Therefore, model testing is necessary to analyze the impact of aspect ratio on dams within the same category. The validity of the results of correlation analyses with other parameters can also be verified based on failure events.
[0063] Analysis of the rupture process (parameter determination based on 3)
[0064] To further analyze the impact of the aspect ratio on similar dams and to validate the validity of analysis results related to other parameters based on the breach phenomenon, a generalized model test was conducted. Refer to Figure 3. Referring to the topography of the Jinsha River section where the Baige landslide dam is located, the river channel from 1.5 km upstream to 2 km downstream of the dam was modeled at a horizontal scale of 1:200 and a vertical scale of 1:150. The dam was located between cross-sections CS7 and CS8. Four water level gauges were installed upstream and downstream to record water level changes (asterisks in the figure). Cameras were placed in front of, at the crest of, and behind the dam to capture the breach process. Based on the actual situation of the landslide dam in the case study, the upstream inflow was set at 3 L / s. The dam material was composed of particles ranging from 0.075 to 4 cm, with a homogeneous median particle size d50 of 6 mm. Regarding the influence of the aspect ratio parameter, the downstream slope ratio was controlled (1:1.3 to 1:2.5) to vary the aspect ratio (3.5 to 4.5). The test groups are shown in the table below. The specific test process is as follows:
[0065] (1) The dam material is divided into two groups according to the density (2.41g / cm 3 ) and dam body moisture content (12%), and the dam body surface was filled in layers, and a 10×10cm grid was drawn;
[0066] (2) Open the water supply valve and use a water level meter and camera to record the process of water rising in the reservoir area and the process of the landslide dam breaking and development;
[0067] (3) Measure the breach parameters after the breach is complete.
[0068] The test results show that, for dams composed of the same material but with different aspect ratios (Groups 1, 2, and 3), the failure process and vertical failure extent differed. The aspect ratio affected the vertical failure extent. When the aspect ratio increased from 3.5 to 4.5, the vertical failure extent decreased from 27 cm to 20 cm, a decrease of 7 cm, equivalent to the prototype's 10.5 m.
[0069] Fine and coarse particles have different initiation states during headward erosion and breach. When the dam face is wide and gentle, the water level drop and flow velocity are low. More coarse particles, after shearing, cannot initiate further movement and remain at the breach base, increasing the roughness of the breach base and inhibiting further breach development. Furthermore, because the river channel downstream of the dam has a gentle slope and is transversely wide, floodwaters flowing through the narrow breach into the downstream channel increase the cross-sectional area of the water flow and sharply decrease the flow velocity. Larger particles settle and accumulate at the breach site at the dam bottom, while finer particles are transported downstream and deposited. When the height is relatively small, the breach coarsening is more pronounced and the riverbed rebalancing ends prematurely, resulting in less vertical breach. Therefore, it is the dam slope head difference, not the dam height and dam material alone, that determines the termination height of erosion—that is, the actual downstream storage capacity, not the total storage capacity—that determines the extent of vertical breach. Once a breach is aborted, the undischarged water has no effect on the breach. The total volume of the dam body also has a minimal impact on the extent of vertical failure. The negative correlation with reservoir capacity, as shown in the correlation analysis, is due to the relationship between reservoir capacity and dam height, which in turn has a numerically negative correlation with the extent of vertical failure. During a failure, upstream flow is negligible compared to the dam burst flow, and its impact on the extent of vertical failure is minimal, only slowly scouring the remaining dam body after the failure has completed.
[0070] According to the above analysis, this embodiment conducts correlation analysis, contingency cross analysis and failure process analysis on the data collected from 49 landslide dam failure cases, and concludes that the main factors affecting the vertical failure degree of the dam body are the aspect ratio (-) and the dam body material (-), that is, the aspect ratio and the dam body material are the key parameters screened from the basic parameters.
[0071] Step 3: The vertical failure degree is divided into two categories: less than or equal to 1 / 2 dam height and greater than 1 / 2 dam height. Based on the key parameters and vertical failure degree of each failure case, an expression for calculating the probability of the vertical failure degree being greater than 1 / 2 dam height is constructed based on logistic regression analysis.
[0072] In this example, the vertical failure degree is treated as a binary variable, with the value y = 0 (vertical failure degree is less than or equal to 1 / 2 dam height) or y = 1 (vertical failure degree is greater than 1 / 2 dam height). Data from 49 failure cases are then imported and logistic regression analysis is used to construct an expression for calculating the probability that the vertical failure degree is greater than 1 / 2 dam height.
[0073] Assuming that there are n independent variables (i.e. key parameters) that affect the value of y, the probability p that the vertical failure degree is greater than 1 / 2 of the dam height under the action of n independent variables is expressed as:
[0074] Where p is the probability that the vertical failure degree is greater than 1 / 2 of the dam height, β0 is a constant term, and β1, β2, ..., β nare partial regression coefficients, x1, x2, …, x n The category of the key parameter.
[0075] In this example, the key parameters affecting the vertical failure degree of the dam body are the aspect ratio and the dam body material. Based on the above 49 cases, a logistic regression analysis was performed to obtain the corresponding constant term and partial regression coefficient. The expression for the probability p that the vertical failure degree is greater than 1 / 2 of the dam height is:
[0076] Where x1 is the type of dam material, and x2 is the type of aspect ratio.
[0077] In the above expression, the constant term and partial regression coefficient were established through logistic regression analysis using 80% of the case data in the case database. The model achieved an accuracy of 80% and a likelihood ratio chi-square value of 23.23 (p = 0.016**), demonstrating significance at the horizontal level and indicating the model is valid. The remaining 20% of the case data (six cases) in the case database were used for validation, with only one error found, resulting in an accuracy rate of 83%. This demonstrates that this method can accurately predict whether the vertical failure degree of a dam body is greater than half the dam height.
[0078] Step 4: Obtain key parameters of the landslide dam to be predicted, and determine whether the vertical failure degree of the landslide dam to be predicted is greater than 1 / 2 of the dam height according to the category of the key parameters of the landslide dam to be predicted and based on the expression.
[0079] In this embodiment, when a landslide occurs, the basic data of the landslide dam can be obtained first based on the type of bank instability, sliding speed (determined by sliding height), scale of instability, width of the river valley, and in conjunction with DEM remote sensing maps and drones, mainly including dam height H, dam length L, length-to-height ratio L / H, dam body volume V, etc. Then, the approximate composition of the dam body material can be quickly determined based on the bank lithology, rock mass integrity and bank instability type. For example, the skeleton of the landslide body that slides along the slope at a low position on an unweathered bank is mainly composed of coarse particles, while the skeleton structure of the landslide body formed by sliding at a high position against the slope of a severely weathered bank is mainly composed of fine particles. In addition, the upstream flow rate Q and reservoir capacity data W can be obtained through the upstream hydrological station. Among them, the length-to-height ratio (L / H) and dam body material are key parameters.
[0080] According to the same classification method as in step 2, the categories of the key parameters of the landslide dam to be predicted are obtained and input into the probability calculation expression to determine whether the vertical failure degree of the landslide dam to be predicted is greater than 1 / 2 of the dam height.
[0081] In order to further improve the prediction accuracy, this embodiment predicts the vertical failure degree by the above method, and then uses case experience summary to assist on-site judgment:
[0082] When it is determined that the vertical failure degree of the predicted landslide dam is greater than 1 / 2 of the dam height, it is determined whether the dam material is Class I, that is, whether it is composed of loess and clay. If so, the dam height is used as the vertical failure degree of the predicted landslide dam. Otherwise, 2 / 3 of the dam height, 3 / 4 of the dam height or 3 / 5 of the dam height is used as the vertical failure degree of the predicted landslide dam, with an error within 1 / 15 of the dam height.
[0083] When it is determined that the vertical failure degree of the landslide dam to be predicted is not greater than 1 / 2 of the dam height, the slope ratio of the landslide dam to be predicted is obtained. If the slope ratio is within a first preset range, 1 / 3 of the dam height is used as the vertical failure degree of the landslide dam to be predicted. If the slope ratio is within a second preset range, 1 / 2 of the dam height is used as the vertical failure degree of the landslide dam to be predicted.
[0084] Specifically, according to the breach cases, the dam body with a vertical breach degree of 1 / 2 of the dam height is basically composed of Class III and V materials (gravel, crushed stone). If the slope of the landslide dam to be predicted is relatively gentle (slightly greater than 1:5), it is estimated to be 1 / 3 of the dam height. On the contrary, if the dam body is pointed and thin, and the slope ratio is about 1:4, similar to the Tanggudong landslide dam, it is estimated to be 1 / 2 of the dam height.
[0085] In this way, the vertical failure degree of dam bodies with a failure degree less than or equal to 1 / 2 of the dam height is divided into 1 / 3 failure and 1 / 2 failure, with an error of approximately 1 / 6 of the dam height H. After the treatment of most landslide dams, the dam height H is less than 60m, and the dam height H of less than or equal to 1 / 2 failure is generally even smaller. Therefore, based on the aforementioned error estimates of 1 / 15 and 1 / 6 of the dam height H, the error in the vertical failure degree is less than 10m, which basically meets the requirements for rapid emergency judgment.
[0086] In addition, to further improve prediction accuracy, for landslide dams with distinct structural characteristics, the above-mentioned method should be used as a reference for actual situation analysis based on model prediction to reduce prediction errors. Specifically, in addition to homogeneous structures, the possible binary or ternary stacked structures of the dam body should be considered.
[0087] In this embodiment, when the predicted landslide dam is a binary or ternary stacked structure, the elevation of the top surface of the bottom layer of the binary or ternary stacked structure is obtained, and it is determined whether the elevation corresponding to the vertical failure degree of the predicted landslide dam is less than the elevation of the top surface of the bottom layer. If so, the vertical failure degree of the predicted landslide dam is re-determined. The elevation corresponding to the vertical failure degree refers to the height of the dam body relative to the ground after the landslide dam fails according to the vertical failure degree, and the elevation of the top surface of the bottom layer refers to the height of the top surface of the bottom layer relative to the ground.
[0088] Specifically, there are two types of binary stacking structures: one with a complete rock layer at the bottom and fragmented soil and rock or isolated rubble at the top; the other with a fragmented rock and soil at the bottom and isolated rubble at the top. A ternary stacking structure has a complete rock layer at the bottom, isolated rubble in the middle, and fragmented rock and soil at the top.
[0089] For the first type of binary stacking structure, it is determined whether the elevation corresponding to the vertical failure degree of the predicted landslide dam is less than the elevation of the top surface of the bottom layer. If so, the ratio of the height of the top layer to the dam height is used as the re-determined vertical failure degree. Otherwise, the influence of the binary stacking structure is not considered.
[0090] For the second type of binary stacking structure, the elevation corresponding to the predicted vertical failure degree of the landslide dam is determined to be less than the elevation of the bottom layer's top surface. If so, a value between the vertical failure degree and 2 / 3 of the dam height is selected as the re-determined vertical failure degree. Otherwise, the impact of the binary stacking structure is not considered. This is because the dam body's bottom is mainly composed of fine particles, which causes severe subsidence at low flow rates. At the same time, the dam body's top, which is mainly composed of coarse particles, has poor cohesion and severe collapse, forming a large amount of coarse particle sedimentation at the downstream dam site. Compared to a nearly half-failed homogeneous dam body, the fine particles in the lower part of this type of dam body have weak scour resistance, and the degree of undercutting will be greater than 1 / 2 failure. However, due to the silting of coarse particles near the dam axis, it is difficult to start, so the vertical failure degree of this type of dam body does not exceed 2 / 3 of the dam height.
[0091] For a three-layer stacked structure, the vertical failure degree of the predicted landslide dam is determined to be lower than the elevation of the top surface of the bottom layer. If so, the ratio of the height of the top layer to the middle layer and the dam height is used as the re-determined vertical failure degree. Otherwise, the influence of the three-layer stacked structure is not considered. This method can avoid the influence of the special dam structures of two-layer stacked and three-layer stacked structures on the vertical failure degree, further improving the accuracy of the vertical failure degree determination.
[0092] Based on the above solution, this embodiment can also realize the prediction of the breach top width and breach bottom width after the landslide dam is predicted to break.
[0093] It can be understood that considering that breach scouring is a process in which the soil first collapses and then collapses, and the collapsed soil is washed away by the discharged water, this embodiment establishes a sediment transport balance formula under the final state of the breach. When the vertical breach degree is less than or equal to 1 / 2 of the dam height, the amount of soil washed away at the end of the dam breach is related to the actual discharged water (related to the widening) and the actual vertical breach depth.
[0094] In addition, landslide dams often occur in "V"-shaped river channels. In addition to considering the above parameters, the widening of the dam body with a vertical breach degree greater than 1 / 2 of the dam height is also limited by the river channel width B.
[0095] In order to obtain the relationship formula for calculating the breach width when the vertical breach degree is less than or equal to 1 / 2 of the dam height and greater than 1 / 2 of the dam height, this embodiment carried out a series of generalized tests in the figure, converted the test data to the prototype, and combined the collected test data and case data to establish and verify the breach width formula considering multiple scales and types of dam bodies. The test data involved are:
[0096] (1) Flow material combination group: 12 fields with incoming flow (1, 3, 5 L / s) and dam material (d50 = 0.18, 0.27, 1 and 1.6 mm), numbered ①②③④, ⑤⑥⑦⑧, ⑨⑩ (2) Variable structure group: 3 test fields, including binary and ternary stacked structures, numbered (3) Variable length-to-height ratio group: 2 test sites, corresponding to groups 2 and 3 in Table 4, numbered (4) Variable dam shape group: 4 test sites, namely "short and fat", "pointy and thin", "steep on top and gentle on the bottom", and "gentle on top and steep on the bottom", numbered (flow 1 L / s, d50 = 0.18 mm); (5) Variable density group: 2 test fields, including dense and loose fields (2.338 g / cm3 and 2.448 g / cm3), numbered (6) Wang Daozheng et al., Shi Zhenming et al. conducted a flume test and six working conditions including horizontal heterogeneous groups, numbered
[0097] Among them, the dam failure data with a vertical failure degree greater than 1 / 2 of the dam height are as follows:
[0098] The dam failure data with vertical failure degree less than or equal to 1 / 2 of the dam height are as follows:
[0099] Through the above series of field cases and model test data, it can be deduced that when the vertical failure degree of the landslide dam to be predicted is less than or equal to 1 / 2 of the dam height: b mt E′=13.858W′ 0.4577 H′(R 2 =0.96); b mb E′=5.1707W′ 0.4964 H′(R 2 =0.86);
[0100] When the vertical breach degree of the predicted landslide dam is greater than 1 / 2 of the dam height, the calculation formulas for the breach top width and breach bottom width after the predicted landslide dam breach are as follows: mt E′=28.818W′ -0.088 B 1.549 H′; b mbE′=6.991W′ 0.357 B 0.406 H′;
[0101] Where b mt is the width of the breach top, b mb is the bottom width of the breach, E′ is the cross-sectional area of the breach, B is the width of the river channel, W′ is the actual downstream storage capacity, and H′ is the vertical breach depth.
[0102] The above working conditions The data of Hebaige 10.10 landslide dam were not used in the derivation of calculation formula, so the working condition Verification with data from the Hebaige 10.10 landslide dam revealed bottom and top widths of 194 and 278 meters, and 104 and 161 meters, respectively. The actual values are 200 and 240 meters, and 100 and 120 meters, respectively. The error is within 13%-25%, indicating a relatively safe prediction. This method can generally meet the needs of river width prediction when the vertical breach extent, dam dimensions, and river channel topography are known.
[0103] Please refer to FIG4 . After obtaining the above formula, this embodiment predicts the breach widening by the following steps:
[0104] (1) Obtain the slope ratio, length, height and actual discharge capacity of the landslide dam to be predicted;
[0105] (2) Determine the vertical failure depth H′ of the predicted landslide dam based on its vertical failure degree and dam height, and calculate the breach cross-sectional area E′ of the predicted landslide dam after its failure based on its dam body slope ratio, dam length, and vertical failure depth H′;
[0106] Please refer to Figure 5. The cross-sectional area of the breach E′ is the cross-sectional projection area of the breach, that is, the cross-sectional area of the dam that has been washed away. In this embodiment, it is approximately regarded as a trapezoid and is calculated through the dam body slope ratio, dam length and vertical breach depth H′.
[0107] (3) Calculate the top width and bottom width of the breach of the predicted landslide dam based on the breach cross-sectional area of the predicted landslide dam, the actual downstream storage capacity and the vertical breach depth.
[0108] In this embodiment, when the vertical breach degree of the landslide dam to be predicted is less than or equal to 1 / 2 of the dam height, the breach top width and breach bottom width can be calculated by inputting the breach cross-sectional area, actual downstream storage capacity, and vertical breach depth into the corresponding calculation formula. When the vertical breach degree of the landslide dam to be predicted is greater than 1 / 2 of the dam height, the breach top width and breach bottom width can be calculated by inputting the breach cross-sectional area, actual downstream storage capacity, river width, and vertical breach depth into the corresponding calculation formula.
[0109] In summary, the method for predicting the breach morphology of a landslide dam based on case statistics and the breach process described in this embodiment provides a prediction method for the vertical breach degree and lateral width (top width, bottom width) of a landslide dam based on case analysis results and the laws of the breach process. It has good prediction accuracy, and the required parameters are all parameters that are easy and quick to obtain on site, which can meet the requirements of quickly determining the breach size (especially the vertical cut depth) during emergency rescue. It can also provide an effective reference for the calculation of the assumed parameter flood peak model and the risk removal of on-site decision makers. In addition, the key parameters for predicting the vertical breach degree are screened based on breach cases and correlation analysis, cross-column analysis, and breach process analysis, which improves the accuracy of key parameter screening and thus improves the accuracy of vertical breach degree prediction. In addition, an expression is constructed based on the key parameters of the breach case, and based on the expression, it can accurately predict whether the vertical breach degree of the landslide dam is greater than 1 / 2 of the dam height. For special landslide dams, the vertical breach degree of the landslide dam is further determined by the dam body material, slope ratio, and dam body structure, further improving the accuracy of the vertical breach degree prediction. In addition, the breach width of the landslide dam is calculated based on the predicted vertical breach degree and a formula created based on case analysis. When the vertical breach degree is greater than 1 / 2 of the dam height, the parameter river width is introduced to improve the accuracy of the breach width prediction.
Claims
1. Prediction method of dam breach morphology based on case statistics and breach process, It is characterized in that The method comprises: Obtaining natural breach cases in the case library and the basic parameters and vertical breach degrees corresponding to each breach case, wherein the basic parameters include: dam height, dam length, length-to-height ratio, dam volume, dam material composition, upstream flow and reservoir capacity data; Classify the basic parameters and vertical failure degree, and determine the key parameters affecting the vertical failure degree from the basic parameters based on correlation analysis, contingency cross analysis and failure process analysis; The vertical failure degree is divided into two categories, which are less than or equal to 1 / 2 of the dam height and greater than 1 / 2 of the dam height. According to the key parameters and vertical failure degree of each failure case, an expression for calculating the probability of the vertical failure degree being greater than 1 / 2 of the dam height is constructed based on logistic regression analysis. The key parameters of the landslide dam to be predicted are obtained, and whether the vertical breach degree of the landslide dam to be predicted is greater than 1 / 2 of the dam height is determined according to the categories of the key parameters of the landslide dam to be predicted and based on the expression.
2. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 1, It is characterized in that The expression used to calculate the probability that the vertical failure degree is greater than 1 / 2 of the dam height is as follows: Where p is the probability that the vertical failure degree is greater than 1 / 2 of the dam height, β 0 is a constant term, β 1 , β 2 , …, β n is the partial regression coefficient, x 1 、x 2 , …, x n The category of the key parameter.
3. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 2, It is characterized in that The key parameters are the aspect ratio and the dam material. The expression for calculating the probability that the vertical breach degree is greater than 1 / 2 of the dam height is as follows: In the formula, x 1 is the type of dam material, x 2 Category of aspect ratio.
4. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 1, It is characterized in that The method further comprises: When the vertical failure degree of the predicted dam is determined to be greater than 1 / 2 of the dam height, it is determined whether the dam body material is composed of loess and clay. If so, the dam height is used as the vertical failure degree of the predicted dam. Otherwise, 2 / 3 dam height, 3 / 4 dam height or 3 / 5 dam height is taken as the vertical breach degree of the landslide dam to be predicted.
5. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 4, It is characterized in that The method further comprises: When it is determined that the vertical failure degree of the predicted landslide dam is not greater than 1 / 2 of the dam height, the slope ratio of the predicted landslide dam is obtained. If the slope ratio is within a first preset range, 1 / 3 of the dam height is used as the vertical failure degree of the predicted landslide dam. If the slope ratio is within a second preset range, 1 / 2 of the dam height is used as the vertical failure degree of the predicted landslide dam.
6. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 5, It is characterized in that The method further comprises: When the landslide dam to be predicted is a binary stacked structure or a ternary stacked structure, the elevation of the top surface of the bottom layer of the binary stacked structure or the ternary stacked structure is obtained, and it is determined whether the elevation corresponding to the vertical breach degree of the landslide dam to be predicted is less than the elevation of the top surface of the bottom layer. If so, the vertical breach degree of the landslide dam to be predicted is re-determined.
7. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 6, It is characterized in that The re-determination of the vertical failure degree of the landslide dam to be predicted specifically includes: If the bottom layer of the binary stack structure is composed of intact rock layers and the top layer is composed of broken soil and rocks or isolated pieces of broken rocks, the ratio of the height of the top layer to the dam height shall be used as the re-determined vertical failure degree; If the bottom layer of the binary stacked structure is composed of broken rock and soil, and the top layer is composed of isolated broken rocks, a value between the vertical failure degree and 2 / 3 of the dam height is selected as the re-determined vertical failure degree; If the bottom layer of the ternary stack structure is composed of complete rock layers, the middle layer is composed of isolated pieces of gravel, and the top layer is composed of broken stones and soil, then the ratio of the height of the top layer to the middle layer and the height of the dam will be used as the re-determined vertical failure degree.
8. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 7, It is characterized in that The method further comprises: Obtain the slope ratio, length, height and actual discharge capacity of the landslide dam to be predicted; Determine the vertical breach depth of the predicted dam according to the vertical breach degree and dam height of the predicted dam, and calculate the breach cross-sectional area of the predicted dam after the dam is breached according to the dam body slope ratio, dam length and vertical breach depth of the predicted dam; The breach top width and breach bottom width of the predicted dam after its breach are calculated based on the breach cross-sectional area of the predicted dam, the actual downstream storage capacity and the vertical breach depth.
9. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 8, It is characterized in that When the vertical breach degree of the predicted dam is less than or equal to 1 / 2 of the dam height, the calculation formulas for the breach top width and breach bottom width after the predicted dam breach are as follows: mt E′=13.858W′ 0.4577 H′(R 2 =0.96); b mb E′=5.1707W′ 0.4964 H′(R 2 =0.86); Where b mt is the width of the breach top, b mb is the bottom width of the breach, E′ is the cross-sectional area of the breach, W′ is the actual downstream storage capacity, and H′ is the vertical breach depth.
10. The method for predicting the breach morphology of a dam based on case statistics and breach process as claimed in claim 8, It is characterized in that When the vertical breach degree of the predicted dam is greater than 1 / 2 of the dam height, the calculation formulas for the breach top width and breach bottom width after the predicted dam breach are as follows: mt E′=28.818W′ -0.088 B 1.549 H′; b mb E′=6.991W′ 0.357 B 0.406 H′; Where b mt is the width of the breach top, b mb is the bottom width of the breach, E′ is the cross-sectional area of the breach, B is the width of the river channel, W′ is the actual downstream storage capacity, and H′ is the vertical breach depth.
Citation Information
Patent Citations
Predication method of happening position of slump disaster after earthquake based on multiple factors
CN103364830A
Method for determining risk level of cascade reservoir group dam
CN112396305A
Deep learning-based barrier dam stability and outburst parameter rapid prediction method
CN115796376A
Case statistics and outburst process-based dam breach form prediction method
CN116561529A
Distribution prediction system of cultural assets and prediction method of the same using statistical analysis
KR101121026B1
Cited By
Anti-flood safety evaluation system for dike project
CN119761833A
Earth and rockfill dam burst peak flow prediction method based on elite strategy
CN119784194A
Breakage dam body material design method of barrier dam
CN119830570A
A method for designing dam body materials in the event of a landslide dam failure.
CN119830570B
Numerical simulation method for flood routing disaster chain of river blocking by debris flow and dam burst by barricade
CN119885959A
