Intelligent hydraulic engineering flood discharge anti-seepage dam safety assessment method and system

Abstract

The invention discloses an intelligent hydraulic engineering flood discharge anti-seepage dam safety assessment method and system. The method comprises the steps of data collection, data preprocessing, dam safety assessment model construction, dam safety assessment model parameter search and dam safety real-time assessment. The invention belongs to the technical field of safety evaluation, and the scheme comprehensively considers the mean value, the standard deviation and the similarity of weight distribution, designs a screening threshold and combines dam monitoring subsets, trains a weak evaluator, adjusts the weak evaluator based on an error rate and calculates a weight, introduces category weighting and historical performance to update data weight, and improves the safety of dam monitoring. Combining the weak evaluators based on the weights to obtain a dam safety evaluation model; a first control factor is designed based on the dynamic position correlation coefficient, a second control factor is designed based on the search stage, the individual position is updated according to the random number and the control factors, the optimal parameter is determined, effective integration and analysis of dam monitoring data are achieved, and the accuracy and applicability of dam safety assessment are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of safety assessment, and specifically refers to an intelligent safety assessment method and system for flood discharge and anti-seepage dams of water conservancy projects. Background Art

[0002] The dam safety assessment method uses advanced artificial intelligence technology to conduct a comprehensive and accurate assessment of the flood discharge capacity and anti-seepage performance of the dam, providing a reliable guarantee and scientific basis for the safe operation of water conservancy projects, helping to take measures in advance to prevent possible safety problems and ensure the stability of water conservancy facilities and the safety of the surrounding environment. However, the existing dam safety assessment method has the problem that the dam monitoring data is large and complex, easily interfered by noise data, difficult to effectively integrate and analyze, and unable to focus on important dam monitoring data, resulting in inaccurate dam safety assessment results; the existing dam safety assessment method is difficult to adapt to dams of different types, sizes and environmental conditions, lacks the ability to adaptively adjust the rapidly changing distribution of dam monitoring data, resulting in reduced accuracy and applicability of the model for dam safety assessment. Summary of the invention

[0003] In view of the above situation, in order to overcome the defects of the prior art, the present invention provides an intelligent water conservancy project flood discharge and anti-seepage dam safety assessment method and system. In view of the problems in the existing dam safety assessment method that the dam monitoring data is large and complex, easily interfered by noise data, difficult to effectively integrate and analyze, and unable to focus on important dam monitoring data, resulting in inaccurate dam safety assessment results, this scheme comprehensively considers the mean, standard deviation and similarity of the weight distribution, designs screening thresholds and combines dam monitoring subsets, trains weak evaluators, adjusts weak evaluators based on error rates and calculates weights, introduces category weighting and historical performance to update data weights, combines weak evaluators based on weights to obtain a dam safety assessment model, screens and distinguishes the importance of a large amount of data, reduces the impact of noise data, and continuously adjusts And combined weak evaluators, effective integration and analysis of dam monitoring data are achieved, and the accuracy and efficiency of dam safety assessment are improved. The existing dam safety assessment methods are difficult to adapt to dams of different types, sizes and environmental conditions, and lack the ability to adaptively adjust the rapidly changing distribution of dam monitoring data, which leads to reduced accuracy and applicability of the model for dam safety assessment. This scheme uses individual positions as representatives of dam safety assessment model parameters, designs the first control factor based on the dynamic position correlation coefficient, designs the second control factor based on the search stage, updates the individual positions according to random numbers and control factors, and determines the optimal parameters, so that the model can explore different parameters more flexibly during the search process to adapt to dams in different situations, thereby improving the accuracy and applicability of dam safety assessment.

[0004] The technical solution adopted by the present invention is as follows: The present invention provides an intelligent water conservancy project flood discharge and anti-seepage dam safety assessment method, the method comprising the following steps:

[0005] Step S1: data collection;

[0006] Step S2: data preprocessing;

[0007] Step S3: constructing a dam safety assessment model;

[0008] Step S4: searching for parameters of dam safety assessment model;

[0009] Step S5: Real-time assessment of dam safety.

[0010] Furthermore, in step S1, the data collection is to collect historical dam monitoring data; the historical dam monitoring data includes structural data, environmental data, operating status data and water conservancy project status; the structural data includes dam geometric parameters, material properties and construction information; the environmental data includes meteorological data, hydrological data and geological data; the operating status data includes dam deformation data, seepage data and stress-strain data; the water conservancy project status includes safety and danger, and the water conservancy project status is used as a data label.

[0011] Furthermore, in step S2, the data preprocessing is to perform data cleaning, data conversion, data normalization and data set construction on the collected data; data cleaning includes processing noise, outliers, duplicate values ​​and missing values; data conversion is to convert the data into vector form; data normalization is to unify the data range based on the maximum and minimum normalization method; and data set construction is to construct a training data set and a test data set based on the data after data cleaning, data conversion and data normalization.

[0012] Furthermore, in step S3, the construction of the dam safety assessment model specifically includes the following steps:

[0013] Step S31: Initialize the data weights and set the initial weight of each dam monitoring data in the training data set to , Q is the number of levee monitoring data in the training dataset;

[0014] Step S32: training a weak evaluator, including the following steps:

[0015] Step S321: Design a screening threshold, comprehensively consider the mean and standard deviation of the current weight distribution of the dam monitoring data, and the similarity between the dam monitoring data, and calculate the screening threshold. The formula used is as follows:

[0016] ;

[0017] In the formula, is the screening threshold during the zth training, μ z and σ z are the mean and standard deviation of the weight distribution of dam monitoring data during the zth training, z is the training number index, x q and x p are the qth and pth dam monitoring data respectively, p and q are the dam monitoring data indexes, is the Euclidean distance between the qth and pth dam monitoring data, a1, a2, and a3 are the first adjustment parameter, the second adjustment parameter, and the third adjustment parameter, respectively;

[0018] Step S322: construct a weak evaluator, combine the dam monitoring data with weights greater than the screening threshold into a dam monitoring subset, train and learn the dam monitoring subset based on the weights of the current dam monitoring data, and construct a decision tree as a weak evaluator;

[0019] Step S33: adjust the weak evaluator, calculate the error rate of the weak evaluator on the dam monitoring subset, if the error rate is greater than or equal to 0.5 and the dam monitoring subset is equal to the training data set, stop the iteration; otherwise, if the error rate is greater than or equal to 0.5 and the dam monitoring subset is not equal to the training data set, set the screening threshold equal to 0, and go to step S322; otherwise, if the error rate is less than 0.5, go to step S34;

[0020] Step S34: Calculate the weights, and calculate the weights of the weak evaluators based on the error rate; the formula used is as follows:

[0021] ;

[0022] In the formula, f z is the weight of the weak evaluator obtained from the zth training, b z is the error rate during the zth training;

[0023] Step S35: Update the data weights, introduce category weights and historical performance, and normalize the updated data weights. The formula used is as follows:

[0024] ;

[0025] In the formula, and are the weights of the qth dam monitoring data in the z+1th and zth trainings, respectively; α, β, and γ are the feedback factor, the category weight factor, and the historical impact factor, respectively. belongs to y q The number of dam monitoring data of each category, is the classification label of the qth dam monitoring data by the weak evaluator obtained by the zth training, y qis the true label of the qth dam monitoring data, is the number of misclassifications of the qth dam monitoring data in the historical evaluation, This is the highest misclassification value in the history of all dam monitoring data;

[0026] Step S36: Combination, pre-set the maximum number of training times. If the maximum number of training times is reached, the weak evaluators are combined based on the weights to obtain the dam safety assessment model; otherwise, the number of training times is increased by 1 and returns to step S32 to continue training the weak evaluators; the formula used to combine the weak evaluators is as follows:

[0027] ;

[0028] Where H is the dam safety assessment model, z max is the maximum number of training times, G z is the weak estimator obtained from the z-th training, and sign(·) is the sign function.

[0029] Furthermore, in step S4, the dam safety assessment model parameter search specifically includes the following steps:

[0030] Step S41: initialization, establishing a parameter search space for the first adjustment parameter a1, the second adjustment parameter a2, the third adjustment parameter a3, the feedback factor α, the category weight factor β and the historical impact factor γ in the dam safety assessment model, randomly initializing individual positions in the parameter search space, using the individual positions as representatives of the dam safety assessment model parameters, and using the exponential loss function of the dam safety assessment model trained based on the individual positions for the test sample set as the individual fitness value;

[0031] Step S42: Designing control factors, including the following steps:

[0032] Step S421: Design the first control factor based on the dynamic position correlation coefficient, and the formula used is as follows:

[0033] ;

[0034] ;

[0035] In the formula, and are the dynamic position correlation coefficients during the k-th and k-1-th searches, respectively. is the first control factor during the k-th search, k is the search count index, and are the maximum and minimum values ​​of the first control factor, respectively. and are the average and worst individual positions of all individual positions during the k-th search. The worst individual position is the individual position with the largest fitness value. and are the average and worst individual positions of all individuals in the initial stage, respectively;

[0036] Step S422: Design the second control factor based on the search phase, using the following formula:

[0037] ;

[0038] In the formula, is the second control factor during the k-th search, and are the maximum and minimum values ​​of the second control factor, k max is the maximum number of searches;

[0039] Step S43: Position update. Before each search, a random number is randomly generated for each individual, and the individual position is updated based on the random number and the control factor. The formula used is as follows:

[0040] ;

[0041] In the formula, and are the positions of the uth individual in the k+1th and kth searches, respectively, where u is the individual index. is the optimal individual position during the k-th search. The optimal individual position is the individual position with the smallest fitness value. r1 and r2 are the optimal learning factor and the neighborhood learning factor, respectively. is the random number generated for the u-th individual during the k-th search, is a different one from The random individual position of , UB and LB are the upper and lower limits of the parameter search space respectively;

[0042] Step S44: Determine the optimal parameters, pre-set the fitness threshold, and update the individual fitness value. When there is an individual whose fitness value is less than the fitness threshold, the parameters corresponding to the individual position are the optimal parameters. A dam safety assessment model is constructed based on the optimal parameters to complete the parameter search; otherwise, if the maximum number of searches is reached, go to step S41 to reinitialize; otherwise, add 1 to the number of searches and return to step S42 to continue searching.

[0043] Furthermore, in step S5, the real-time assessment of dam safety is to collect real-time dam monitoring data, pre-process it and input it into a dam safety assessment model constructed based on optimal parameters for processing, and based on the output assessment results, obtain the hydraulic engineering status of the dam under flood discharge and seepage prevention conditions; if the hydraulic engineering status is dangerous, output an early warning signal to alert the staff to ensure the safe and stable operation of the hydraulic engineering flood discharge and seepage prevention dam.

[0044] The present invention provides an intelligent water conservancy project flood discharge and anti-seepage dam safety assessment system, comprising a data acquisition module, a data preprocessing module, a dam safety assessment model building module, a dam safety assessment model parameter search module and a dam safety real-time assessment module;

[0045] The data acquisition module collects historical dam monitoring data and sends the data to the data preprocessing module;

[0046] The data preprocessing module performs data cleaning, data conversion, data normalization and data set construction on the collected data, and sends the data to the dam safety assessment model construction module;

[0047] The module for constructing a dam safety assessment model comprehensively considers the mean, standard deviation and similarity of the weight distribution, designs a screening threshold and combines dam monitoring subsets, trains weak evaluators, adjusts weak evaluators based on error rates and calculates weights, introduces category weighting and historical performance to update data weights, combines weak evaluators based on weights to obtain a dam safety assessment model, and sends data to a dam safety assessment model parameter search module;

[0048] The dam safety assessment model parameter search module uses individual positions as representatives of dam safety assessment model parameters, designs a first control factor based on a dynamic position correlation coefficient, designs a second control factor based on a search phase, updates individual positions according to random numbers and control factors, determines optimal parameters, and sends the data to the dam safety real-time assessment module;

[0049] The real-time dam safety assessment module obtains the hydraulic engineering status of the dam under flood discharge and seepage prevention conditions based on the assessment results output by the dam safety assessment model.

[0050] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0051] (1) In view of the problems in existing dam safety assessment methods, such as large and complex dam monitoring data, easy interference from noise data, difficult to effectively integrate and analyze, and unable to focus on important dam monitoring data, resulting in inaccurate dam safety assessment results, this scheme comprehensively considers the mean, standard deviation and similarity of weight distribution, designs screening thresholds and combines dam monitoring subsets, trains weak evaluators, adjusts weak evaluators based on error rates and calculates weights, introduces category weighting and historical performance to update data weights, combines weak evaluators based on weights to obtain a dam safety assessment model, screens a large amount of data and distinguishes their importance, can effectively deal with the large and complex problems of dam monitoring subsets, reduce the impact of noise data, and screen out more valuable dam monitoring subsets. By continuously adjusting and combining weak evaluators, effective integration and analysis of dam monitoring data is achieved, and the accuracy and efficiency of dam safety assessment are improved.

[0052] (2) The existing dam safety assessment methods are difficult to adapt to dams of different types, sizes and environmental conditions, and lack the ability to adaptively adjust to the rapidly changing distribution of dam monitoring data, which reduces the accuracy and applicability of the model for dam safety assessment. This scheme uses individual positions as representatives of dam safety assessment model parameters, designs the first control factor based on the dynamic position correlation coefficient, and designs the second control factor based on the search stage. The individual positions are updated according to random numbers and control factors to determine the optimal parameters. The dynamic position correlation coefficient helps to adjust the search strategy according to the distribution of individual positions. The design of the search stage can reasonably allocate the search intensity according to different stages, improve the search efficiency, and enable the model to explore different parameters more flexibly during the search process to adapt to dams in different situations, thereby improving the accuracy and applicability of dam safety assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of a process flow of an intelligent water conservancy project flood discharge and anti-seepage dam safety assessment method provided by the present invention;

[0054] Figure 2 A schematic diagram of an intelligent water conservancy project flood discharge and anti-seepage dam safety assessment system provided by the present invention;

[0055] Figure 3 is a schematic flow chart of step S3;

[0056] Figure 4 It is a schematic diagram of the process of step S4.

[0057] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0059] In the description of the present invention, it should be understood that terms such as “upper”, “lower”, “front”, “back”, “left”, “right”, “top”, “bottom”, “inside” and “outside” indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0060] Example 1, see Figure 1 The present invention provides an intelligent water conservancy project flood discharge and anti-seepage dam safety assessment method, the method comprising the following steps:

[0061] Step S1: data collection, collecting historical dam monitoring data;

[0062] Step S2: Data preprocessing, data cleaning, data conversion, data normalization and data set construction for the collected data;

[0063] Step S3: construct a dam safety assessment model, comprehensively consider the mean, standard deviation and similarity of the weight distribution, design a screening threshold and combine the dam monitoring subsets, train weak evaluators, adjust the weak evaluators based on the error rate and calculate the weights, introduce category weighting and historical performance update data weights, and combine the weak evaluators based on the weights to obtain the dam safety assessment model;

[0064] Step S4: searching for parameters of the dam safety assessment model, using individual positions as representatives of the parameters of the dam safety assessment model, designing the first control factor based on the dynamic position correlation coefficient, designing the second control factor based on the search phase, updating the individual position according to the random number and the control factor, and determining the optimal parameters;

[0065] Step S5: Real-time assessment of dam safety. Based on the assessment results output by the dam safety assessment model, the hydraulic engineering status of the dam under flood discharge and seepage prevention conditions is obtained.

[0066] Example 2, see Figure 1This embodiment is based on the above embodiment. In step S1, the historical dam monitoring data includes structural data, environmental data, operating status data and water conservancy project status; the structural data includes dam geometric parameters, material properties and construction information; the environmental data includes meteorological data, hydrological data and geological data; the operating status data includes dam deformation data, seepage data and stress-strain data; the water conservancy project status includes safety and danger, and the water conservancy project status is used as a data label.

[0067] Example 3, see Figure 1 This embodiment is based on the above embodiment. In step S2, data cleaning includes processing noise, outliers, duplicate values ​​and missing values; data conversion is to convert data into vector form; data normalization is to unify the data range based on the maximum and minimum normalization method; and data set construction is to construct a training data set and a test data set based on the data after data cleaning, data conversion and data normalization.

[0068] Example 4, see Figure 1 and Figure 3 Based on the above embodiment, in step S3, building a dam safety assessment model specifically includes the following steps:

[0069] Step S31: Initialize the data weights and set the initial weight of each dam monitoring data in the training data set to , Q is the number of levee monitoring data in the training dataset;

[0070] Step S32: training a weak evaluator, including the following steps:

[0071] Step S321: Design a screening threshold, comprehensively consider the mean and standard deviation of the current weight distribution of the dam monitoring data, and the similarity between the dam monitoring data, calculate the screening threshold, make the screening threshold more reasonable and adaptable, and better capture the distribution characteristics and similarity of the data; the formula used is as follows:

[0072] ;

[0073] In the formula, is the screening threshold during the zth training, μ z and σ z are the mean and standard deviation of the weight distribution of dam monitoring data during the zth training, z is the training number index, x q and x p are the qth and pth dam monitoring data respectively, p and q are the dam monitoring data indexes, is the Euclidean distance between the qth and pth dam monitoring data, a1, a2, and a3 are the first adjustment parameter, the second adjustment parameter, and the third adjustment parameter in the range of (0, 1), respectively;

[0074] Step S322: construct a weak evaluator, combine the dam monitoring data with weights greater than the screening threshold into a dam monitoring subset, train and learn the dam monitoring subset based on the weights of the current dam monitoring data, and construct a decision tree as a weak evaluator, so that the weak evaluator can more accurately reflect key information and improve the accuracy and reliability of the model;

[0075] Step S33: adjust the weak evaluator, calculate the error rate of the weak evaluator on the dam monitoring subset, if the error rate is greater than or equal to 0.5 and the dam monitoring subset is equal to the training data set, stop the iteration; otherwise, if the error rate is greater than or equal to 0.5 and the dam monitoring subset is not equal to the training data set, set the screening threshold equal to 0, and go to step S322; otherwise, if the error rate is less than 0.5, go to step S34; flexibly process according to the error rate to ensure continuous optimization and improvement of the model to avoid overfitting or maladaptation; the formula used to calculate the error rate is as follows:

[0076] ;

[0077] Where b z is the error rate during the zth training, is the weight of the qth dam monitoring data in the zth training, is the classification label of the qth dam monitoring data by the weak evaluator obtained by the zth training, y q is the true label of the qth dam monitoring data; is the indicator function, when When , the function value is 1, otherwise the function value is 0;

[0078] Step S34: Calculate the weights, and calculate the weights of the weak evaluators based on the error rate; the formula used is as follows:

[0079] ;

[0080] In the formula, f z is the weight of the weak evaluator obtained from the zth training, b z is the error rate during the zth training;

[0081] Step S35: Update the data weights, introduce category weighting and historical performance, and normalize the updated data weights to make the update of data weights more targeted and dynamic, better reflect the importance and historical performance of the data, and further improve the performance of the model; the formula used is as follows:

[0082] ;

[0083] In the formula, and are the weights of the qth dam monitoring data in the z+1th and zth trainings, respectively; α, β, and γ are the feedback factors, category weight factors, and historical impact factors in the range of (0, 1), respectively. belongs to y q The number of dam monitoring data of each category, is the classification label of the qth dam monitoring data by the weak evaluator obtained by the zth training, y q is the true label of the qth dam monitoring data, is the number of misclassifications of the qth dam monitoring data in the historical evaluation, This is the highest misclassification value in the history of all dam monitoring data;

[0084] Step S36: Combination, pre-set the maximum number of training times. If the maximum number of training times is reached, the weak evaluators are combined based on the weights to obtain the dam safety assessment model; otherwise, the number of training times is increased by 1 and returns to step S32 to continue training the weak evaluators; the formula used to combine the weak evaluators is as follows:

[0085] ;

[0086] Where H is the dam safety assessment model, z max is the maximum number of training times, G z is the weak estimator obtained from the z-th training, and sign(·) is the sign function.

[0087] By performing the above operations, in view of the problems in the existing dam safety assessment methods, such as large and complex dam monitoring data, easy interference from noise data, difficult to effectively integrate and analyze, and unable to focus on important dam monitoring data, resulting in inaccurate dam safety assessment results, this scheme comprehensively considers the mean, standard deviation and similarity of weight distribution, designs screening thresholds and combines dam monitoring subsets, trains weak evaluators, adjusts weak evaluators based on error rates and calculates weights, introduces category weighting and historical performance to update data weights, combines weak evaluators based on weights to obtain a dam safety assessment model, screens a large amount of data and distinguishes their importance, can effectively deal with the large and complex problems of dam monitoring subsets, reduce the impact of noise data, screen out more valuable dam monitoring subsets, and through continuous adjustment and combination of weak evaluators, realizes effective integration and analysis of dam monitoring data, and improves the accuracy and efficiency of dam safety assessment.

[0088] Example 5, see Figure 1 and Figure 4 This embodiment is based on the above embodiment. In step S4, the dam safety assessment model parameter search specifically includes the following steps:

[0089] Step S41: initialization, establishing a parameter search space for the first adjustment parameter a1, the second adjustment parameter a2, the third adjustment parameter a3, the feedback factor α, the category weight factor β and the historical impact factor γ in the dam safety assessment model, randomly initializing individual positions in the parameter search space, using the individual positions as representatives of the dam safety assessment model parameters, and using the exponential loss function of the dam safety assessment model trained based on the individual positions for the test sample set as the individual fitness value;

[0090] Step S42: Designing control factors, including the following steps:

[0091] Step S421: Design a first control factor based on the dynamic position correlation coefficient. The control factor can be dynamically adjusted according to the progress of the search to better guide the search direction. The formula used is as follows:

[0092] ;

[0093] ;

[0094] In the formula, and are the dynamic position correlation coefficients during the k-th and k-1-th searches, respectively. is the first control factor during the k-th search, k is the search count index, and are the maximum and minimum values ​​of the first control factor, respectively. and are the average and worst individual positions of all individual positions during the k-th search. The worst individual position is the individual position with the largest fitness value. and are the average and worst individual positions of all individuals in the initial stage, respectively;

[0095] Step S422: Design a second control factor based on the search stage, and use different search strengths according to different search stages to improve search efficiency; the formula used is as follows:

[0096] ;

[0097] In the formula, is the second control factor during the k-th search, and are the maximum and minimum values ​​of the second control factor, k max is the maximum number of searches;

[0098] Step S43: Position update. Before each search, a random number (0, 1) is randomly generated for each individual. The individual position is updated based on the random number and the control factor, so that the individual position changes continuously in the search space to find the optimal parameter combination and increase the flexibility and adaptability of the search. The formula used is as follows:

[0099] ;

[0100] In the formula, and are the positions of the uth individual in the k+1th and kth searches, respectively, where u is the individual index. is the optimal individual position in the kth search. The optimal individual position is the individual position with the smallest fitness value. r1 and r2 are the optimal learning factor and neighborhood learning factor in the range of [1, 2], respectively. is the random number generated for the u-th individual during the k-th search, is a different one from The random individual position of , UB and LB are the upper and lower limits of the parameter search space respectively;

[0101] Step S44: Determine the optimal parameters, pre-set the fitness threshold, and update the individual fitness value. When there is an individual whose fitness value is less than the fitness threshold, the parameters corresponding to the individual position are the optimal parameters. A dam safety assessment model is constructed based on the optimal parameters to complete the parameter search; otherwise, if the maximum number of searches is reached, go to step S41 to reinitialize; otherwise, add 1 to the number of searches and return to step S42 to continue searching.

[0102] By performing the above operations, the existing dam safety assessment methods are difficult to adapt to dams of different types, sizes and environmental conditions, and lack the ability to adaptively adjust the rapidly changing distribution of dam monitoring data, which leads to reduced accuracy and applicability of the model for dam safety assessment. This scheme uses individual positions as representatives of dam safety assessment model parameters, designs the first control factor based on the dynamic position correlation coefficient, designs the second control factor based on the search stage, updates the individual positions according to random numbers and control factors, and determines the optimal parameters. The dynamic position correlation coefficient helps to adjust the search strategy according to the distribution of individual positions. The design of the search stage can reasonably allocate the search intensity according to different stages, improve the search efficiency, and enable the model to explore different parameters more flexibly during the search process to adapt to dams in different situations, thereby improving the accuracy and applicability of dam safety assessment.

[0103] Example 6, see Figure 1This embodiment is based on the above embodiment. In step S5, the real-time assessment of dam safety is to collect real-time dam monitoring data, and after pre-processing, input it into the dam safety assessment model constructed based on optimal parameters for processing. Based on the output assessment result, the water conservancy project status of the dam under flood discharge and seepage prevention conditions is obtained. If the water conservancy project status is dangerous, an early warning signal is output to remind the staff to ensure the safe and stable operation of the flood discharge and seepage prevention dam of the water conservancy project.

[0104] Embodiment 7, see Figure 2 This embodiment is based on the above embodiment. The present invention provides an intelligent water conservancy project flood discharge and anti-seepage dam safety assessment system, including a data acquisition module, a data preprocessing module, a dam safety assessment model building module, a dam safety assessment model parameter search module and a dam safety real-time assessment module;

[0105] The data acquisition module collects historical dam monitoring data and sends the data to the data preprocessing module;

[0106] The data preprocessing module performs data cleaning, data conversion, data normalization and data set construction on the collected data, and sends the data to the dam safety assessment model construction module;

[0107] The module for constructing a dam safety assessment model comprehensively considers the mean, standard deviation and similarity of the weight distribution, designs a screening threshold and combines dam monitoring subsets, trains weak evaluators, adjusts weak evaluators based on error rates and calculates weights, introduces category weighting and historical performance to update data weights, combines weak evaluators based on weights to obtain a dam safety assessment model, and sends data to a dam safety assessment model parameter search module;

[0108] The dam safety assessment model parameter search module uses individual positions as representatives of dam safety assessment model parameters, designs a first control factor based on a dynamic position correlation coefficient, designs a second control factor based on a search phase, updates individual positions according to random numbers and control factors, determines optimal parameters, and sends the data to the dam safety real-time assessment module;

[0109] The real-time dam safety assessment module obtains the hydraulic engineering status of the dam under flood discharge and seepage prevention conditions based on the assessment results output by the dam safety assessment model.

[0110] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0111] While the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that many changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the invention.

[0112] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. An intelligent safety assessment method for flood discharge and anti-seepage dams of water conservancy projects, characterized by: The method comprises the following steps: Step S1: data collection, collecting historical dam monitoring data; Step S2: data preprocessing; Step S3: constructing a dam safety assessment model; Step S4: searching for parameters of dam safety assessment model; Step S5: Real-time assessment of dam safety, based on the assessment results output by the dam safety assessment model, the hydraulic engineering status of the dam under the conditions of flood discharge and seepage prevention is obtained; Step S3 includes step S35: updating the data weight, introducing the category weight and historical performance, and normalizing the updated data weight. The formula used is as follows: ； In the formula, and are the weights of the qth dam monitoring data in the z+1th and zth trainings, respectively. z is the training times index, q is the dam monitoring data index, and f z is the weight of the weak evaluator obtained from the z-th training, α, β and γ are the feedback factor, category weight factor and historical influence factor respectively, belongs to y q The number of dam monitoring data of each category, is the classification label of the qth dam monitoring data by the weak evaluator obtained by the zth training, y q is the true label of the qth dam monitoring data, is the number of misclassifications of the qth dam monitoring data in the historical evaluation, This is the highest misclassification value in the history of all dam monitoring data.

2. According to claim 1, a smart water conservancy project flood discharge and anti-seepage dam safety assessment method is characterized by: In step S3, the construction of the dam safety assessment model specifically includes the following steps: Step S31: Initialize the data weights and set the initial weight of each dam monitoring data in the training data set to , Q is the number of levee monitoring data in the training dataset; Step S32: training a weak evaluator; Step S33: adjust the weak evaluator, calculate the error rate of the weak evaluator on the dam monitoring subset, if the error rate is greater than or equal to 0.5 and the dam monitoring subset is equal to the training data set, stop the iteration; otherwise, if the error rate is greater than or equal to 0.5 and the dam monitoring subset is not equal to the training data set, set the screening threshold equal to 0, and go to step S322; otherwise, if the error rate is less than 0.5, go to step S34; Step S34: Calculate the weights, and calculate the weights of the weak evaluators based on the error rate; the formula used is as follows: ； In the formula, f z is the weight of the weak evaluator obtained from the zth training, b z is the error rate during the zth training; Step S35: Update data weight; Step S36: Combination, pre-set the maximum number of training times. If the maximum number of training times is reached, the weak evaluators are combined based on the weights to obtain the dam safety assessment model; otherwise, the number of training times is increased by 1 and returns to step S32 to continue training the weak evaluators; the formula used to combine the weak evaluators is as follows: ； Where H is the dam safety assessment model, z max is the maximum number of training times, G z is the weak estimator obtained from the z-th training, and sign(·) is the sign function.

3. The intelligent safety assessment method for flood discharge and anti-seepage dams of water conservancy projects according to claim 2 is characterized by: In step S32, the training of the weak evaluator specifically includes the following steps: Step S321: Design a screening threshold, comprehensively consider the mean and standard deviation of the current weight distribution of the dam monitoring data, and the similarity between the dam monitoring data, and calculate the screening threshold. The formula used is as follows: ； In the formula, is the screening threshold during the zth training, μ z and σ z are the mean and standard deviation of the weight distribution of dam monitoring data during the zth training, z is the training number index, x q and x p are the qth and pth dam monitoring data respectively, p and q are the dam monitoring data indexes, is the Euclidean distance between the qth and pth dam monitoring data, a1, a2, and a3 are the first adjustment parameter, the second adjustment parameter, and the third adjustment parameter, respectively; Step S322: construct a weak evaluator, combine the dam monitoring data with weights greater than the screening threshold into a dam monitoring subset, train and learn the dam monitoring subset based on the weight of the current dam monitoring data, and construct a decision tree as a weak evaluator.

4. The intelligent safety assessment method for flood discharge and anti-seepage dams of water conservancy projects according to claim 1 is characterized by: In step S4, the dam safety assessment model parameter search specifically includes the following steps: Step S41: initialization, establishing a parameter search space for the first adjustment parameter a1, the second adjustment parameter a2, the third adjustment parameter a3, the feedback factor α, the category weight factor β and the historical impact factor γ in the dam safety assessment model, randomly initializing individual positions in the parameter search space, using the individual positions as representatives of the dam safety assessment model parameters, and using the exponential loss function of the dam safety assessment model trained based on the individual positions for the test sample set as the individual fitness value; Step S42: designing control factors; Step S43: Position update. Before each search, a random number is randomly generated for each individual, and the individual position is updated based on the random number and the control factor. The formula used is as follows: ； In the formula, and are the positions of the u-th individual in the k+1-th and k-th searches, respectively, where u is the individual index. is the worst individual position during the k-th search, is the optimal individual position in the kth search, r1 and r2 are the optimal learning factor and the neighborhood learning factor, respectively. is the random number generated for the u-th individual during the k-th search, is a different one from The random individual position of , UB and LB are the upper and lower limits of the parameter search space, respectively. and are the first control factor and the second control factor in the k-th search respectively; Step S44: Determine the optimal parameters, pre-set the fitness threshold, and update the individual fitness value. When there is an individual whose fitness value is less than the fitness threshold, the parameters corresponding to the individual position are the optimal parameters. A dam safety assessment model is constructed based on the optimal parameters to complete the parameter search; otherwise, if the maximum number of searches is reached, go to step S41 to reinitialize; otherwise, add 1 to the number of searches and return to step S42 to continue searching.

5. The intelligent safety assessment method for flood discharge and anti-seepage dams of water conservancy projects according to claim 4 is characterized by: In step S42, the design control factor specifically includes the following steps: Step S421: Design the first control factor based on the dynamic position correlation coefficient, and the formula used is as follows: ； ； In the formula, and are the dynamic position correlation coefficients during the k-th and k-1-th searches, respectively. is the first control factor during the k-th search, k is the search count index, and are the maximum and minimum values ​​of the first control factor, respectively. is the average of all individual positions during the k-th search, and are the average and worst individual positions of all individuals in the initial stage, respectively; Step S422: Design the second control factor based on the search phase, using the following formula: ； In the formula, is the second control factor during the k-th search, and are the maximum and minimum values ​​of the second control factor, k max is the maximum number of searches.

6. The intelligent safety assessment method for flood discharge and anti-seepage dams of water conservancy projects according to claim 1 is characterized by: In step S1, the data collection is to collect historical dam monitoring data; the historical dam monitoring data includes structural data, environmental data, operating status data and water conservancy project status; the structural data includes dam geometric parameters, material properties and construction information; the environmental data includes meteorological data, hydrological data and geological data; the operating status data includes dam deformation data, seepage data and stress-strain data; the water conservancy project status includes safety and danger, and the water conservancy project status is used as a data label.

7. The intelligent safety assessment method for flood discharge and anti-seepage dams of water conservancy projects according to claim 1 is characterized by: In step S5, the real-time assessment of dam safety is to collect real-time dam monitoring data, pre-process it and input it into a dam safety assessment model constructed based on optimal parameters for processing. Based on the output assessment results, the hydraulic engineering status of the dam under flood discharge and seepage prevention conditions is obtained. If the hydraulic engineering status is dangerous, an early warning signal is output to remind the staff to ensure the safe and stable operation of the flood discharge and seepage prevention dam of the hydraulic engineering.

8. The intelligent safety assessment method for flood discharge and anti-seepage dams of water conservancy projects according to claim 1 is characterized by: In step S2, the data preprocessing is to perform data cleaning, data conversion, data normalization and data set construction on the collected data; data cleaning includes processing noise, outliers, duplicate values ​​and missing values; Data conversion is to convert data into vector form; data normalization is to unify the data range based on the maximum and minimum normalization method; data set construction is to construct training data sets and test data sets based on data after data cleaning, data conversion and data normalization.

9. An intelligent water conservancy project flood discharge and anti-seepage dam safety assessment system, used to implement an intelligent water conservancy project flood discharge and anti-seepage dam safety assessment method as claimed in any one of claims 1 to 8, characterized in that: It includes data acquisition module, data preprocessing module, dam safety assessment model building module, dam safety assessment model parameter search module and dam safety real-time assessment module; The data acquisition module collects historical dam monitoring data and sends the data to the data preprocessing module; The data preprocessing module performs data cleaning, data conversion, data normalization and data set construction on the collected data, and sends the data to the dam safety assessment model construction module; The module for constructing a dam safety assessment model comprehensively considers the mean, standard deviation and similarity of the weight distribution, designs a screening threshold and combines dam monitoring subsets, trains weak evaluators, adjusts weak evaluators based on error rates and calculates weights, introduces category weighting and historical performance to update data weights, combines weak evaluators based on weights to obtain a dam safety assessment model, and sends data to a dam safety assessment model parameter search module; The dam safety assessment model parameter search module uses individual positions as representatives of dam safety assessment model parameters, designs a first control factor based on a dynamic position correlation coefficient, designs a second control factor based on a search phase, updates individual positions according to random numbers and control factors, determines optimal parameters, and sends the data to the dam safety real-time assessment module; The real-time dam safety assessment module obtains the hydraulic engineering status of the dam under flood discharge and seepage prevention conditions based on the assessment results output by the dam safety assessment model.

Images