A reservoir dam safety risk grade evaluation method

Abstract

This invention discloses a method for assessing the safety risk level of a reservoir dam, specifically relating to the field of reservoir dam risk assessment technology. It involves acquiring real-time stress data of the reservoir dam, spillway discharge capacity data, and reservoir water level change data, extracting and predicting abnormal features from these data, and generating relevant outliers. A fuzzy comprehensive evaluation method is used to perform weighted analysis on multiple risk factors to calculate the dam's comprehensive risk score. The comprehensive risk score is compared with a risk level classification threshold to determine the reservoir dam's safety risk level. Based on the assessment results, corresponding early warning information and emergency response measures are output to ensure the safe operation of the reservoir dam. This method can monitor the safety status of the reservoir dam in real time, provide timely early warning of potential risks, and offer scientific decision support for reservoir management personnel.

Description

Technical Field

[0001] This invention relates to the field of reservoir dam risk assessment technology, specifically to a method for assessing the safety risk level of a reservoir dam. Background Technology

[0002] The safety risk assessment of a reservoir dam involves comprehensively analyzing the dam's structure, operational status, external environment, and potential disaster influencing factors to evaluate its risk level for potential safety accidents. The purpose of the assessment is to identify potential safety hazards, pinpoint potential destructive risks, and provide managers with decision-making support based on the assessment results, assisting in the implementation of appropriate preventative and reinforcement measures. Risk level assessments typically include a detailed analysis of the dam's structural stability, spillway capacity, infrastructure aging status, and disaster resistance capabilities to determine its safety level and to predict and prepare for potential risks.

[0003] The existing technology has the following shortcomings:

[0004] During heavy rainfall or flooding events, the stress on reservoir dams, the spillway's discharge capacity, and the reservoir's water level can fluctuate dramatically, increasing safety risks. In such situations, existing assessment methods often lack real-time dynamic monitoring data and emergency response mechanisms, making it difficult to cope with sudden severe weather events and provide timely and accurate risk warnings. This, in turn, affects dam management decisions and disaster prevention and mitigation efforts. Summary of the Invention

[0005] The purpose of this invention is to provide a method for assessing the safety risk level of reservoir dams, in order to address the shortcomings of the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for assessing the safety risk level of a reservoir dam, comprising:

[0007] Obtain real-time force data of a reservoir dam at n time points and establish a corresponding dataset F = [F1, F2, ..., F...]. i ..., F n ], where F i This represents the force value of the reservoir dam at the i-th time, where n is the total number of time points;

[0008] Obtain the force value F i The corresponding spillway discharge capacity data and reservoir water level change data were used to extract abnormal features and construct a water level anomaly prediction model.

[0009] The output of the water level anomaly prediction model is defined as the force value F. i The corresponding abnormal weight coefficients are used to generate a comprehensive risk score for the reservoir dam after fusion calculation.

[0010] The fuzzy comprehensive evaluation method is used to perform weighted analysis on multiple risk factors, determine the classification threshold of risk level, compare it with the comprehensive risk score, classify the safety risk level of reservoir dam, and output corresponding early warning information and emergency response measures according to the risk level.

[0011] Preferably, the acquisition of real-time stress data includes monitoring by installing stress sensors, displacement sensors, temperature sensors, and acceleration sensors at different locations on the reservoir dam.

[0012] Preferably, the abnormal frequency value of the spillway discharge capacity is generated after analyzing the frequency of the spillway discharge capacity value exceeding a preset threshold. The generation method is as follows:

[0013] Acquire spillway discharge capacity data within a fixed time range. Define m time points within this range, representing the spillway discharge capacity value at each moment. Express the discharge capacity data as Q = [Q1, Q2, ..., Q...]. i ... Q m ], where Q i Let Q represent the discharge capacity value at time i. The maximum design threshold for the spillway's discharge capacity is set as Q. max For each time point i, for each Q i Determine whether the following conditions are met: Set the number of times the threshold is exceeded to N. exceed This refers to the number of times when the discharge capacity exceeds the threshold, which is obtained by counting the number of times the abnormal conditions are met. Among them, A(Q) i Q max ) is an indicator function, when Q i Q max Returns 1 if the overflow occurs, otherwise returns 0, and calculates the abnormal frequency value F of the spillway's discharge capacity. exceed The calculation formula is: Where m is the total number of time points, that is, the number of time points within the total monitoring duration.

[0014] Preferably, after performing anomaly analysis on water level and discharge capacity data, anomaly fluctuation values ​​for discharge capacity are generated. The generation method is as follows:

[0015] Acquire the spillway discharge capacity data within a time period S. Set the observation period to include h time points for the discharge capacity data, denoted as H = [H1, H2, ..., H...]. i ... H h ], where H i Let represent the drainage capacity at time i; a sliding window of size k is used to calculate the moving average, which is the moving average of the drainage capacity at time i. The calculation formula is: Calculate the fluctuation range of the discharge capacity at each moment, and express it through the standard deviation σ. i This represents the abnormal fluctuation value ΔH of the calculated discharge capacity when the fluctuation exceeds a threshold. i The calculation expression is:

[0016] Where: ΔH i H is the abnormal fluctuation value of the discharge capacity at time i. i It is the water discharge capacity at time i. It is the moving average of the discharge capacity at time i, and α is a threshold coefficient; the overall abnormal fluctuation value of the discharge capacity is obtained by summing and averaging the abnormal fluctuation values ​​of the discharge capacity at each time during the entire monitoring period.

[0017] Preferably, the abnormal frequency value and abnormal fluctuation value of the spillway discharge capacity are converted into a comprehensive feature vector. The comprehensive feature vector is used as the input of the machine learning model. The machine learning model uses the prediction of the water level anomaly coefficient value label for each set of comprehensive feature vectors as the prediction objective and minimizes the sum of prediction errors for all water level anomaly coefficient value labels as the training objective. The machine learning model is trained until the sum of prediction errors converges and the model training stops. The water level anomaly coefficient value is determined based on the model output. The machine learning model is a multinomial regression model.

[0018] Preferably, the water level anomaly coefficient value predicted by the model is used as the anomaly weight coefficient, and the weighted average sum is calculated with the stress data to obtain the comprehensive risk score of the reservoir dam.

[0019] Preferably, each risk factor is fuzzified, mapping the actual value of each risk factor to a fuzzy linguistic variable, and a membership matrix is ​​constructed; the membership degree of each risk factor is multiplied by the corresponding weight matrix to finally calculate the membership degree value of each risk level, and the safety risk level of the reservoir dam is determined based on the membership degree value.

[0020] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0021] 1. This invention can monitor key risk factors such as the stress data of reservoir dams, spillway discharge capacity, and water level changes in real time, and promptly identify potential safety hazards. By combining a water level anomaly prediction model and fuzzy comprehensive evaluation method, it can comprehensively assess the safety risks of dams, generate accurate comprehensive risk scores, and output corresponding early warning information and emergency response measures according to different risk levels, thereby providing scientific and effective decision support for reservoir management personnel.

[0022] 2. This invention has significant advantages and effects. First, through the fusion analysis of multi-dimensional data, it improves the accuracy of dam safety risk assessment and avoids biases that may result from a single data source. Second, by dynamically adjusting risk level thresholds and weights, it can flexibly respond to different environments and monitoring conditions, ensuring efficient application under various complex situations. Finally, this invention provides a comprehensive, real-time, and intelligent risk assessment solution for the safety management of reservoir dams, significantly improving the safety and prevention capabilities of dam management and reducing the risk of catastrophic accidents. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0024] Figure 1 This is a mind map of the method of the present invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] For examples, please refer to Figure 1 As shown in this embodiment, a method for assessing the safety risk level of a reservoir dam includes:

[0027] Obtain real-time force data of a reservoir dam at n time points and establish a corresponding dataset F = [F1, F2, ..., F...]. i ..., F n ], where F i This represents the force value of the reservoir dam at the i-th time, where n is the total number of time points;

[0028] Obtain the force value F i The corresponding spillway discharge capacity data and reservoir water level change data were used to extract abnormal features and construct a water level anomaly prediction model.

[0029] The output of the water level anomaly prediction model is defined as the force value F. i The corresponding abnormal weight coefficients are used to generate a comprehensive risk score for the reservoir dam after fusion calculation.

[0030] The fuzzy comprehensive evaluation method is used to perform weighted analysis on multiple risk factors, determine the classification threshold of risk level, compare it with the comprehensive risk score, classify the safety risk level of reservoir dam, and output corresponding early warning information and emergency response measures according to the risk level.

[0031] Stress data refers to information such as the forces, pressures, and stresses that a reservoir dam experiences at various times. The main purpose of obtaining this data is to provide a foundation for subsequent safety risk assessments. Stress data is typically obtained through the following methods:

[0032] Various sensors, including stress sensors, displacement sensors, temperature sensors, and acceleration sensors, are installed at key parts of the reservoir dam (such as the dam body, dam foundation, and dam face). These sensors are used to monitor the stress on the dam body at different times in real time.

[0033] Stress sensors: used to monitor the stress distribution at various points within the dam body, helping to assess whether the dam body is subjected to excessive or uneven pressure.

[0034] Displacement sensors are used to measure the deformation of the dam body, especially dam settlement, horizontal displacement, and crack propagation, which can directly affect the stability of the dam.

[0035] Accelerometer: Used to monitor the vibration of the dam body, especially in the event of an earthquake or other sudden event, it can reflect the vibration or stress wave propagation of the dam body.

[0036] The frequency of data collection (e.g., per second, per minute) should be set according to specific monitoring needs. When assessing the real-time stress condition of a reservoir dam, it is necessary to obtain stress data at multiple time points, so it is essential to ensure the timeliness and accuracy of the monitoring data.

[0037] The collected force data is uploaded to the central processing system via data transmission equipment (such as a wireless network or wired communication system) to ensure timely and stable data transmission and storage. The data storage system needs to have a large storage capacity and efficient data retrieval capabilities for subsequent processing and analysis.

[0038] After obtaining real-time force data at n time points, these data need to be organized into an effective data set F = [F1, F2, ..., F...]. i ..., F n ], where F i This represents the force value of the reservoir dam at time i, where n is the total number of times, for further analysis and processing. Specifically:

[0039] Data Identification and Classification: Force data at each moment (e.g., moment i) needs to be associated with a timestamp or other unique identifier (such as date, hours, minutes, and seconds) to ensure data uniqueness and traceability. For example, each data item in the dataset may contain the following fields:

[0040] Timestamp: Indicates the specific time of data collection, such as "2025-04-21 12:00:00".

[0041] Stress values: Stress data of the dam at this moment, which may include multiple sub-items, such as:

[0042] Total stress value: The total force borne by the entire dam body.

[0043] Local stress value: The stress value at a specific location, which helps to determine the local stress condition of the dam body.

[0044] Stress distribution: The stress values ​​of each part can be calculated from sensor data to determine the stress changes in each part.

[0045] Sensor location: Identify the location of the sensors used for data acquisition (e.g., upstream, downstream, or foundation of the dam) to facilitate comparison of stress conditions at different locations.

[0046] Data formatting: Stress data needs to be stored in a uniform format for subsequent processing. Common data formats include structured database formats (such as SQL databases) or unstructured formats used for big data analysis (such as JSON, CSV, etc.).

[0047] Data temporal sequence: In order to reflect the stress change trend of the reservoir dam in different time periods, the data needs to be sorted in chronological order so that the data at each moment can clearly show the dynamic changes of stress.

[0048] Ensure that the force data at each moment is complete and consistent, without any missing or erroneous data. If data loss or transmission failure occurs, data compensation or resampling techniques should be used to correct the data.

[0049] In practical applications, the stress data of reservoir dams may change continuously. To understand the dam's stress situation in real time, it is necessary to ensure the real-time nature and frequent updates of the data. Common methods include:

[0050] Data is collected from various sensors in real time and transmitted to the processing center through a data transmission system to update the data set.

[0051] The data monitoring system monitors stress data in real time to see if it exceeds the normal range and issues alarms promptly. If abnormal changes occur in the stress data (such as overload, severe vibration, etc.), the system can provide immediate feedback to help managers respond.

[0052] In the safety risk assessment of reservoir dams, in addition to stress data, spillway discharge capacity data and reservoir water level change data are crucial factors. Especially when abnormal changes occur in the reservoir water level, they can directly affect the stability of the dam.

[0053] Obtaining spillway discharge capacity data is crucial, as spillways are an important component of reservoir flood control and discharge, and their discharge capacity directly affects the dam's risk prevention and control capabilities under extreme weather conditions. Data on spillway discharge capacity can be obtained through the following methods:

[0054] Spillway design data: The design discharge capacity data of the spillway is obtained from the reservoir engineering design drawings. This data typically includes the design flow rate (unit: m3 / s) and maximum discharge capacity of the spillway. Real-time monitoring data: The actual discharge capacity of the spillway is monitored in real time using equipment such as flow meters, pressure sensors, and water level sensors at the spillway inlet and outlet.

[0055] A threshold is set for the spillway's discharge capacity. When the spillway's discharge capacity exceeds the threshold, it is considered an anomaly. For example, when the spillway's discharge capacity cannot cope with the current flow rate, it can be considered an abnormal state.

[0056] After analyzing the frequency of spillway discharge capacity values ​​exceeding preset thresholds, abnormal frequency values ​​of spillway discharge capacity are generated. The generation method is as follows:

[0057] Acquire spillway discharge capacity data within a fixed time range. Define m time points within the fixed time range, representing the spillway discharge capacity value at each moment. Represent this discharge capacity data as Q = [Q1, Q2, ..., Q...]. i ... Q m ], where Q i This represents the discharge capacity value at time i (unit: m). 3 / s). The maximum design threshold for the spillway's discharge capacity is set as Q. max This is the maximum discharge capacity of the spillway under normal conditions. This threshold can be obtained from the reservoir design documents or the safety standards set in the system.

[0058] For each time point i, if the drainage capacity Q i Exceeded the maximum threshold Q max If the spillway discharge capacity is abnormal at that moment, then it is considered that an anomaly occurred in the spillway discharge capacity. That is, for each Q... i Determine whether the following conditions are met: To analyze the frequency with which the spillway's discharge capacity exceeds a threshold, it is necessary to calculate the number of times the threshold is exceeded. Let N be the number of times the threshold is exceeded. exceedThis refers to the number of times when the discharge capacity exceeds the threshold. This can be obtained by counting the number of times the above abnormal conditions are met. Among them, A(Q) i Q max ) is an indicator function, when Q i Q max Returns 1 if the condition is met, otherwise returns 0. Calculates the abnormal frequency value F of the spillway's discharge capacity. exceed The calculation formula is: Where m is the total number of time points, i.e., the number of times within the total monitoring duration. F exceed This is used to assess whether the spillway's discharge capacity has exceeded safe limits on a long-term and frequent basis, thereby providing decision support for reservoir managers.

[0059] Obtaining reservoir water level change data is crucial, as it is a key parameter for assessing reservoir safety. The primary purpose of acquiring this data is to monitor for abnormal fluctuations in water levels in real time, especially during extreme weather events such as heavy rain and floods. Water level change data can be obtained through the following methods:

[0060] Water level sensors: Water level sensors are installed at different key locations in the reservoir (such as upstream, midstream, and downstream) to monitor changes in water level in real time. Common water level monitoring equipment includes float-type water level gauges, pressure sensors, and radar sensors.

[0061] Real-time data recording and transmission: Water level change data is transmitted to the central data processing system via wireless or wired communication systems. This data can be updated at regular time intervals (such as every minute or hour) to ensure real-time monitoring of reservoir water level changes.

[0062] Historical water level data: In addition to real-time monitoring, historical water level data is also very important, as it helps assess long-term trends and abnormal fluctuations in water level changes. Historical water level data can be used to compare with current water level levels to assess whether the water level is within the normal range.

[0063] Statistical methods (such as mean, standard deviation, skewness, kurtosis, etc.) are used to analyze water level and drainage capacity data to identify outliers. For example, if the water level change exceeds a certain standard deviation, an anomaly may occur.

[0064] Anomaly analysis was performed on water level and discharge capacity data to generate abnormal fluctuation values ​​for discharge capacity. The generation method is as follows:

[0065] Acquire the spillway discharge capacity data within a time period S. Set the observation period to h time points for the discharge capacity data, denoted as H = [H1, H2, ..., H...]. i ... H h ], where H iThis represents the discharge capacity at time i (unit: m). 3 / s).

[0066] Calculate the smoothing trend of discharge capacity data. A common method is to use a moving average. Set up a sliding window of size k to calculate the moving average, which is the moving average of the discharge capacity at time i. The calculation formula is: Calculate the fluctuation range of the discharge capacity at each time step. To quantify the fluctuation, it can be represented by the standard deviation. The fluctuation value of the discharge capacity at time step i is σ. i The calculation formula is: Where, σ i It is the fluctuation value of the discharge capacity at time i, which reflects the degree of dispersion between the discharge capacity at that time and its moving average.

[0067] Abnormal fluctuations in discharge capacity reflect the degree of abnormal variation in discharge capacity relative to its smooth trend. To determine whether abnormal fluctuations exist, the scenario where the discharge capacity fluctuation exceeds a threshold can be calculated. The abnormal fluctuation value of discharge capacity is ΔQ. i The calculation expression is:

[0068] Where: ΔH i H is the abnormal fluctuation value of the discharge capacity at time i. i It is the water discharge capacity at time i. This is the moving average of the drainage capacity at time i, and α is a threshold coefficient, typically 2 or 3, used to control sensitivity to fluctuations. If If the fluctuation exceeds α times the standard deviation, the fluctuation at that moment is considered abnormal.

[0069] Abnormal fluctuations in discharge capacity can be used to reflect anomalies over a given period. The overall abnormal fluctuation value is calculated by summing and averaging the abnormal fluctuation values ​​at various points throughout the monitoring period. This average value can be used to assess whether abnormal fluctuations in discharge capacity exist during the monitoring period. A large abnormal fluctuation value indicates that the discharge capacity fluctuations exceeded the normal range during that period, potentially indicating system instability.

[0070] The abnormal frequency values ​​and abnormal fluctuation values ​​of the spillway's discharge capacity are converted into comprehensive feature vectors. These comprehensive feature vectors are then used as input to a machine learning model. The machine learning model uses the predicted water level anomaly coefficient value label for each set of comprehensive feature vectors as its prediction objective and minimizes the sum of prediction errors for all water level anomaly coefficient value labels as its training objective. The machine learning model is trained until the sum of prediction errors converges, at which point the model training stops. The water level anomaly coefficient value is determined based on the model output. The machine learning model is a multinomial regression model.

[0071] The comprehensive risk score of the reservoir dam is calculated by weighting and averaging the anomaly coefficient value of the water level predicted by the model with the stress data.

[0072] Fuzzy comprehensive evaluation is a decision-making method based on fuzzy mathematics, used to handle the comprehensive analysis of multiple uncertain factors. In the safety risk assessment of reservoir dams, fuzzy comprehensive evaluation can perform weighted analysis of multiple risk factors to comprehensively derive the risk level of the dam.

[0073] First, the safety risk assessment of a reservoir dam involves multiple risk factors, each representing a different aspect of dam safety. These risk factors include, but are not limited to:

[0074] Water level anomaly coefficient; spillway discharge capacity (including the frequency and fluctuation of spillway discharge capacity anomalies); dam stress condition (including stress and deformation); meteorological conditions (such as precipitation and temperature changes); geological conditions (such as dam foundation stability and seismic activity); these risk factors can be collected through sensor monitoring, historical data, and weather forecasts.

[0075] The fuzzy comprehensive evaluation method first requires fuzzification of each risk factor. Specifically, the actual value of each risk factor is mapped to a fuzzy linguistic variable, such as a classification like "low," "medium," or "high." These classifications help determine the degree of impact of the risk factors on dam safety.

[0076] For example:

[0077] Water level anomaly coefficient: Based on the actual value, it may be classified as "normal", "high", or "severely high".

[0078] Spillway discharge capacity: Based on flow data, it can be divided into "normal", "low" and "extremely low".

[0079] Dam stress condition: Based on the stress condition, it can be divided into "safe", "relatively dangerous" and "extremely dangerous".

[0080] These fuzzy linguistic variables are defined through membership functions, which map specific numerical values ​​to categories such as "low", "medium", and "high".

[0081] In the fuzzy comprehensive evaluation method, different risk factors have varying degrees of influence on the final evaluation result. Therefore, it is necessary to assign corresponding weights to each risk factor. Given b risk factors, the weight matrix is ​​defined as W = [w1, w2, ..., w...]. b ], where w b Let w1 + w2 + ... + w be the weights of the b-th risk factor, satisfying: w1 + w2 + ... + w b =1. These weights can be determined based on the actual situation through expert scoring, historical data analysis, or other methods.

[0082] The fuzzification result of each risk factor is represented by a membership matrix, where matrix A is a b×p fuzzy matrix, where b is the number of risk factors and p is the number of fuzzy categories of the risk factors (e.g., "low", "medium", "high"). Among them, a bp is the membership degree of the b-th risk factor under the p-th category, indicating the degree to which the factor belongs to a certain fuzzy category. For example, the water level anomaly coefficient may have a membership degree of 0.8 under the "high" category.

[0083] The core of fuzzy comprehensive evaluation is calculating the fuzzy comprehensive evaluation value C, which is formulated as: C = W·A; that is, multiplying the weight matrix W by the membership matrix A yields the comprehensive evaluation matrix C, which represents the membership value of each risk level. Finally, the comprehensive risk level is calculated based on C: C = pc1,c2,…,c p ]; where c p This indicates the degree of membership of the reservoir dam to the p-th risk level.

[0084] To correlate the comprehensive assessment results with the actual risk level, it is necessary to set classification thresholds for the risk level. For example, the risk level can be divided into four levels: normal, low, high, and very high. Based on the comprehensive assessment result C and the classification thresholds, the final risk level is determined.

[0085] For example: if `cnormal` is greater than the membership degree of other levels and exceeds a certain threshold, it is judged as "normal". If the values ​​of `chighc` or `cextreme` are high, it is judged as a "high" or "extremely high" risk level. `cnormal` represents the membership degree of the dam at the normal risk level. The higher the value, the greater the probability that the dam is in a normal state. `chigh` represents the membership degree of the dam at the higher risk level. The higher the value, the greater the probability that the dam has a higher risk. `cextreme` represents the membership degree of the dam at the extremely high risk level. The higher the value, the greater the probability that the dam has a serious risk.

[0086] Based on the safety risk level of the reservoir dam, corresponding early warning information and emergency response measures will be provided:

[0087] Normal risk level: The reservoir dam is operating normally and no special attention is required.

[0088] Warning information: No warning.

[0089] Emergency response measures: Regular inspections and routine maintenance.

[0090] Low risk level: There is a slight risk to the reservoir dam, but no immediate action is required.

[0091] Warning: There is a certain risk; it is recommended to strengthen monitoring.

[0092] Emergency response measures: Strengthen monitoring and conduct regular inspections.

[0093] High risk level: The reservoir dam poses a significant risk and requires attention.

[0094] Warning: Pay attention to the safety of the dam and take timely measures if potential hazards are found.

[0095] Emergency response measures: Immediately conduct structural inspections and adjust water levels, and prepare an emergency response plan.

[0096] Extremely high risk level: The reservoir dam is in a state of severe risk and a catastrophic accident may occur.

[0097] Warning: Extreme risk warning. Take immediate emergency response measures.

[0098] Emergency response measures: Activate the emergency response plan, evacuate surrounding personnel, and carry out reinforcement measures or release water, etc.

[0099] In this application, the fuzzy comprehensive evaluation method, through weighted analysis and fuzzification of multiple risk factors, helps us comprehensively assess the safety risks of reservoir dams and output corresponding early warning information and emergency response measures based on the risk level. Using this method, reservoir managers can promptly grasp the safety status of the dam and take effective measures to prevent disasters.

[0100] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0101] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0102] It should be understood that the term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone, where A and B can be singular or plural. Additionally, the character " / " in this document generally indicates an "or" relationship between the preceding and following related objects, but it may also indicate an "and / or" relationship; please refer to the context for specific understanding. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0103] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for assessing the safety risk level of a reservoir dam, characterized in that: include: Obtain real-time stress data of a reservoir dam at n time points and establish a corresponding dataset. ,in, This represents the force value of the reservoir dam at the i-th time, where n is the total number of time points; Get the force value The corresponding spillway discharge capacity data and reservoir water level change data were used to extract abnormal features and construct a water level anomaly prediction model. Among them, the frequency of the spillway's discharge capacity exceeding a preset threshold is analyzed to generate an abnormal frequency value for the spillway's discharge capacity. The generation method is as follows: acquire spillway discharge capacity data within a fixed time range, set a total of m time points within the fixed time range, representing the spillway's discharge capacity value at each time point, and represent the discharge capacity data as... ,in, This represents the discharge capacity value at time i, and the maximum design threshold for the spillway discharge capacity is set as follows: For each time point i, for each Determine whether the following conditions are met: Set the number of times the threshold is exceeded. This refers to the number of times when the discharge capacity exceeds the threshold, which is obtained by counting the number of times the abnormal conditions are met. ;in, It is an indicator function, when Returns 1 if the overflow occurs, otherwise returns 0, to calculate the abnormal frequency value of the spillway's discharge capacity. The calculation formula is: Where m is the total number of time points, that is, the number of time points within the total monitoring duration; Anomaly analysis was performed on water level and discharge capacity data to generate abnormal fluctuation values ​​for discharge capacity. The generation method was as follows: Discharge capacity data of the spillway within a time period of S was obtained. Within the observation period, there were h time points for discharge capacity data, denoted as [data point name missing]. ,in Let represent the drainage capacity at time i; a sliding window of size k is used to calculate the moving average, which is the moving average of the drainage capacity at time i. The calculation formula is: ; Calculate the fluctuation range of the discharge capacity at each moment, and express it through the standard deviation. This indicates the abnormal fluctuation value of the discharge capacity when the fluctuation exceeds a threshold. The calculation expression is: ;in: It is the abnormal fluctuation value of the discharge capacity at time i. It is the water discharge capacity at time i. It is the moving average of the discharge capacity at time i, and α is a threshold coefficient; the overall abnormal fluctuation value of the discharge capacity is obtained by summing and averaging the abnormal fluctuation values ​​of the discharge capacity at each time during the entire monitoring period. The abnormal frequency values ​​and abnormal fluctuation values ​​of the spillway's discharge capacity are converted into comprehensive feature vectors. These comprehensive feature vectors are then used as input to a machine learning model. The machine learning model uses the prediction of water level anomaly coefficient values ​​for each set of comprehensive feature vectors as its prediction objective and minimizes the sum of prediction errors for all water level anomaly coefficient value labels as its training objective. The machine learning model is trained until the sum of prediction errors converges, at which point the model training stops. The water level anomaly coefficient values ​​are determined based on the model output. The machine learning model is a multinomial regression model. The output of the water level anomaly prediction model is defined as the force value. The corresponding abnormal weight coefficients are used to generate a comprehensive risk score for the reservoir dam after fusion calculation. The fuzzy comprehensive evaluation method is used to perform weighted analysis on multiple risk factors, determine the classification threshold of risk level, compare it with the comprehensive risk score, classify the safety risk level of reservoir dam, and output corresponding early warning information and emergency response measures according to the risk level. The risk factors include the water level anomaly coefficient, the spillway discharge capacity, and the dam body stress condition; the fuzzification result of each risk factor is represented by a membership matrix, where matrix A is a b×p fuzzy matrix, where b is the number of risk factors and p is the number of fuzzy categories of the risk factors.

2. The method for assessing the safety risk level of a reservoir dam according to claim 1, characterized in that: The acquisition of real-time stress data includes monitoring by installing stress sensors, displacement sensors, temperature sensors, and acceleration sensors at different locations on the reservoir dam.

3. The method for assessing the safety risk level of a reservoir dam according to claim 1, characterized in that: The comprehensive risk score of the reservoir dam is calculated by weighting and averaging the anomaly coefficient value of the water level predicted by the model with the stress data.

4. The method for assessing the safety risk level of a reservoir dam according to claim 3, characterized in that: Each risk factor is fuzzified, and its actual value is mapped to a fuzzy linguistic variable. A membership matrix is ​​then constructed. The membership degree of each risk factor is multiplied by the corresponding weight matrix to calculate the membership degree value of each risk level. The safety risk level of the reservoir dam is then determined based on the membership degree value.

Images