A dam level displacement monitoring method and system based on digital twinning

Abstract

The application relates to the technical field of displacement monitoring, and discloses a dam level displacement monitoring method and system based on digital twinning, which comprises the following steps: acquiring environment data and level displacement data; performing Kalman filtering and drift detection according to the environment data and the level displacement data to obtain a drift correction index; processing the drift correction index by using a pre-trained support vector machine model to obtain a correction parameter set; fusing and weighting adjusting the correction parameter set and an initial data set to obtain a corrected data set; extracting a feature vector from the corrected data set and calculating a displacement change rate to generate a displacement evaluation report; and comparing and verifying the displacement evaluation report with a preset monitoring standard to obtain a level displacement description. The method effectively improves the accuracy and reliability of monitoring data by using the digital twinning technology.

Description

Technical Field

[0001] This invention relates to the field of displacement monitoring technology, and in particular to a method and system for monitoring the level displacement of a dam based on digital twins. Background Technology

[0002] Currently, dams, as core facilities of water conservancy projects, have structural stability that directly affects the safety of people's lives and property, requiring precise monitoring of level displacement changes to mitigate safety risks. However, dams are located in complex natural environments, where factors such as temperature, humidity, and wind can easily interfere with sensors, placing high demands on the accuracy and real-time performance of monitoring data.

[0003] In one existing technology, dam level displacement monitoring relies on data collected by a single sensor. After simple filtering, the data is compared with a preset threshold. Some schemes combine historical data for static correction, but lack an adaptive adjustment mechanism for dynamic environments.

[0004] However, this approach cannot quantify the degree of interference from environmental factors on the sensors, making it difficult to correct data drift in real time, resulting in biased monitoring results. Furthermore, it fails to integrate multi-source data for comprehensive analysis, making it impossible to accurately assess the dam's structural stability and ultimately affecting the reliability of safety decisions. Therefore, existing technologies suffer from significant environmental interference and insufficient data accuracy in dam level displacement monitoring. Summary of the Invention

[0005] This invention provides a method and system for monitoring dam level displacement based on digital twins, in order to solve the problems of insufficient accuracy and poor reliability of dam level displacement monitoring results caused by environmental interference and data drift in the existing technology.

[0006] In a first aspect, to address the aforementioned problems, the present invention provides a method for monitoring the leveling displacement of a dam based on digital twins, characterized by comprising:

[0007] Obtain environmental data and leveling displacement data to obtain the initial dataset;

[0008] Kalman filtering is applied to the initial dataset to obtain a denoised dataset. If the interference coefficient in the denoised dataset exceeds a preset interference threshold, drift detection is performed on the denoised dataset to obtain a drift correction index.

[0009] The drift correction index is processed using a pre-trained support vector machine model, and the output is a set of correction parameters.

[0010] By merging the set of correction parameters and the initial dataset, and performing a weighted adjustment, the corrected dataset is obtained.

[0011] Extract the feature vector of the level displacement from the corrected dataset, calculate the displacement change rate, and generate a displacement assessment report.

[0012] The displacement assessment report is compared and verified with the preset monitoring standards to obtain a description of the horizontal displacement.

[0013] Preferably, the step of acquiring environmental data and leveling displacement data to obtain an initial dataset includes:

[0014] Environmental data and leveling displacement data are collected in real time through a sensor network and integrated to obtain a dataset; wherein, the environmental data includes temperature, humidity, and wind conditions;

[0015] The collected dataset is screened for outliers using a preset range threshold, and potential drift points are marked to obtain an intermediate dataset.

[0016] For potential drift points in the intermediate dataset, historical environmental monitoring records are obtained for comparison and verification to obtain a corrected dataset;

[0017] Based on the corrected dataset and combined with real-time environmental information, an initial dataset is obtained.

[0018] Preferably, the initial dataset is subjected to Kalman filtering to obtain a denoised dataset. If the interference coefficients in the denoised dataset exceed a preset interference threshold, drift detection is performed on the denoised dataset to obtain a drift correction index, including:

[0019] Based on the initial dataset, Kalman filtering is used to smooth the various environmental parameters, and recursive estimation is performed to generate a smooth parameter set;

[0020] For the smoothing parameter set, outlier screening is performed using the preset range threshold to obtain a filtered dataset. For outliers in the filtered dataset, the authenticity of the outliers is confirmed by comparing them with historical environmental records to obtain a comprehensive parameter set.

[0021] Based on the comprehensive parameter set, data points exceeding the preset interference threshold are extracted, and the data points are compared with historical records to obtain deviation analysis results;

[0022] The consistency of the deviation analysis results is verified, and a drift correction index is generated.

[0023] Preferably, the step of processing the drift correction index using a pre-trained support vector machine model to output a set of correction parameters includes:

[0024] Acquire historical correction data, input the drift correction index and the historical correction data into a pre-trained support vector machine model, and calculate the preliminary correction parameters;

[0025] Obtain the current environmental parameters, and optimize the preliminary correction parameters based on the current environmental parameters to obtain the optimized correction parameters;

[0026] The optimized correction parameters are verified against the preset input threshold, and a set of correction parameters is output.

[0027] Preferably, the step of fusing the correction parameter set and the initial dataset, and then performing a weighted adjustment to obtain the corrected dataset includes:

[0028] The monitoring data in the initial dataset are assigned preset weight values ​​and sorted by priority to obtain a sorted dataset;

[0029] The set of correction parameters is fused with the sorted dataset to obtain an intermediate fused dataset;

[0030] If the weight values ​​of the intermediate fusion dataset exceed the preset weight threshold, dynamic adjustment and optimization are performed to obtain the corrected dataset.

[0031] Preferably, the step of extracting the feature vector of the leveling displacement from the corrected dataset, calculating the displacement change rate, and generating a displacement assessment report includes:

[0032] Based on the leveling displacement data in the corrected dataset, classification processing is performed to obtain a displacement feature set;

[0033] By combining a preset time range, a continuity analysis is performed on the displacement feature set to calculate the displacement change rate;

[0034] If the displacement change rate exceeds a preset change threshold, a potential risk level is identified, and the displacement change rate and the potential risk level are integrated to generate a displacement assessment report.

[0035] Preferably, the step of comparing and verifying the displacement assessment report with a preset monitoring standard to obtain a description of the level displacement includes:

[0036] Based on the displacement assessment report, displacement data points are extracted and classified to obtain a comparison dataset.

[0037] The comparison dataset is compared with the preset monitoring standard to obtain the comparison results;

[0038] The comparison results are summarized and processed to generate a description of the level displacement.

[0039] Secondly, the present invention provides a dam leveling displacement monitoring system based on digital twins, characterized in that it includes:

[0040] The data acquisition module is used to acquire environmental data and leveling displacement data to obtain the initial dataset;

[0041] The filtering module is used to perform Kalman filtering on the initial dataset to obtain a denoised dataset. If the interference coefficient in the denoised dataset exceeds a preset interference threshold, the denoised dataset is subjected to drift detection to obtain a drift correction index.

[0042] The model processing module is used to process the drift correction index using a pre-trained support vector machine model and output a set of correction parameters.

[0043] The data fusion module is used to fuse the correction parameter set and the initial dataset, and after weighted adjustment, obtain the corrected dataset;

[0044] The analysis and evaluation module is used to extract the feature vector of the level displacement from the corrected dataset, calculate the displacement change rate, and generate a displacement evaluation report.

[0045] The verification output module is used to compare and verify the displacement assessment report with the preset monitoring standard to obtain a description of the horizontal displacement.

[0046] Thirdly, the present invention also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the dam level displacement monitoring method based on digital twin as described above.

[0047] Fourthly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the dam level displacement monitoring method based on digital twin as described above.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] (1) This invention collects environmental and level displacement data through sensors, and combines the digital twin system with Kalman filtering to quantify environmental interference, breaking through the limitations of traditional single filtering, eliminating short-term fluctuations, and providing high-quality basic data for subsequent correction.

[0050] (2) This invention uses a pre-trained support vector machine model to match drift correction index and correction parameters, and integrates weighted adjustment to overcome the defects of traditional static correction, solve the problem of correction lag, and significantly improve data accuracy;

[0051] (3) This invention forms a closed-loop mechanism by feature extraction, evaluation report and standard comparison, breaking the limitations of traditional single data monitoring, accurately judging structural risks, and providing reliable support for dam safety decision-making. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the dam leveling displacement monitoring method based on digital twin provided in the first embodiment of the present invention;

[0053] Figure 2 This is a schematic diagram of the structure of a dam level displacement monitoring system based on digital twin provided in the second embodiment of the present invention. Detailed Implementation

[0054] 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, and 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.

[0055] Reference Figure 1 The first embodiment of the present invention provides a method for monitoring the leveling displacement of a dam based on digital twins, comprising the following steps:

[0056] S1, acquire environmental data and leveling displacement data to obtain the initial dataset;

[0057] S2, perform Kalman filtering on the initial dataset to obtain a denoised dataset. If the interference coefficient in the denoised dataset exceeds a preset interference threshold, perform drift detection on the denoised dataset to obtain a drift correction index.

[0058] S3, The drift correction index is processed using a pre-trained support vector machine model, and the correction parameter set is output.

[0059] S4, after merging the correction parameter set and the initial dataset and performing weighted adjustment, the corrected dataset is obtained;

[0060] S5. Extract the feature vector of the level displacement from the corrected dataset, calculate the displacement change rate, and generate a displacement assessment report.

[0061] S6. The displacement assessment report is compared and verified with the preset monitoring standard to obtain the horizontal displacement description.

[0062] In step S1, environmental data and leveling displacement data are acquired to obtain an initial dataset, including:

[0063] S11, Real-time collection of environmental data and level displacement data through a sensor network, and integration to obtain a collected dataset; wherein, the environmental data includes temperature, humidity, and wind conditions;

[0064] S12, use a preset range threshold to screen outliers in the collected dataset, mark potential drift points, and obtain an intermediate dataset;

[0065] S13, for potential drift points in the intermediate dataset, obtain historical environmental monitoring records for comparison and verification to obtain a corrected dataset;

[0066] S14. Based on the calibration dataset and combined with real-time environmental information, an initial dataset is obtained.

[0067] In step S11, environmental data and level displacement data are collected in real time through a sensor network and integrated to obtain a collected dataset; wherein, the environmental data includes temperature, humidity, and wind conditions.

[0068] It should be noted that the sensor network consists of digital temperature sensors, capacitive humidity sensors, ultrasonic anemometers, and high-precision optical leveling sensors, deployed at monitoring points X1 upstream, X2 midstream, and X3 downstream of the dam. Real-time environmental data acquisition via the sensor network refers to using digital temperature sensors to collect temperature, capacitive humidity sensors to collect humidity, and ultrasonic anemometers to collect wind speed, with a collection frequency of once every 10 minutes. Real-time leveling data acquisition refers to using high-precision optical leveling sensors to simultaneously collect data from each monitoring point, at the same frequency as the environmental data. The integrated dataset refers to associating the two types of data by monitoring point and collection time; each data entry includes the monitoring point number, collection time, temperature, humidity, wind speed, and leveling displacement value. The preset data standard range is: temperature -10℃ to 50℃, which covers common extreme temperatures of dams and can prevent sensor accuracy drift; humidity 20% to 90%, which can prevent electrostatic interference and equipment moisture; wind force 0 to 12, which is the internationally recognized level for quantifying wind force impact; level displacement -5mm to 5mm, based on the normal deformation value of dam design, and data outside the range are marked as items to be verified.

[0069] For example, some records in the collected data are X1, 08:00:00, 25℃, 60%, level 3, 1.2mm; X2, 08:00:00, 24℃, 62%, level 2, 1.1mm; X3, 08:00:00, 55℃, 85%, level 4, 6.0mm. Because the temperature is greater than 50℃ and the leveling displacement is greater than 5mm, these are marked as items to be verified.

[0070] In step S12, outlier screening is performed on the collected dataset using a preset range threshold to mark potential drift points and obtain an intermediate dataset.

[0071] It should be noted that the preset range thresholds are consistent with those in step S11, specifically: temperature -10℃ to 50℃, humidity 20% to 90%, and wind force 0 to 12. Referring to the Beaufort scale, level 0 corresponds to a calm dam monitoring scenario, and level 12 corresponds to an extreme typhoon environment. Wind forces exceeding level 12 are considered disaster-level, at which point routine dam monitoring is often suspended, and sensors are easily damaged by strong winds, rendering the data without practical reference value. The level displacement is -5mm to 5mm. This threshold is set in conjunction with the common operating environment of dams and the working accuracy of sensors, avoiding inaccurate data deviations caused by extreme environments. The outlier screening involves checking the temperature, humidity, wind force, and level displacement values ​​of each record in the collected dataset. If any value exceeds the corresponding preset range, the record is marked as a potential drift point. The intermediate dataset is a structured collection that retains the original data dimensions. Each data point includes the monitoring point number, collection time, temperature, humidity, wind force, level displacement value, and potential drift point marker, providing a complete data foundation for subsequent verification of the authenticity of drift points.

[0072] For example, some records in the intermediate dataset are as follows: X1, 08:00:00, 25℃, 60%, Level 3, 1.2mm, unmarked; X2, 08:00:00, 24℃, 62%, Level 2, 1.1mm, unmarked; X3, 08:00:00, 55℃, 85%, Level 4, 6.0mm, marked as a potential drift point; X1, 08:10:00, -12℃, 58%, Level 3, 1.3mm, marked as a potential drift point. Since X1 is a fixed monitoring point upstream of the dam, the temperature of 25℃ at 08:00 is within the preset range, so it is not marked. The temperature drops below -10℃ at 08:10, so it is marked. X3 has a temperature exceeding 50℃ and a level displacement exceeding 5mm, so it is marked.

[0073] In step S13, for potential drift points in the intermediate dataset, historical environmental monitoring records are obtained for comparison and verification to obtain a corrected dataset.

[0074] It should be noted that the historical environmental monitoring records are environmental and level displacement data of the monitoring points to which potential drift points belong, within the same time period ±1 hour over the past 30 days. This eliminates the influence of seasonal and time-of-day differences on the data. Each record includes the monitoring point number, collection time, temperature, humidity, wind force, and level displacement value. The comparison and verification involves independently calculating the mean and twice the standard deviation of each indicator in the historical data, and comparing them with the corresponding indicator values ​​of potential drift points. If any indicator of a potential drift point exceeds the corresponding range, it is confirmed as an abnormal drift point and marked; otherwise, the mark is removed. The calibration dataset includes the monitoring point number, collection time, values ​​of each indicator, and the final label, providing accurate data for subsequent processing.

[0075] For example, in the calibration dataset, some records are: X1, 08:10:00, -12℃, 58%, level 3, 1.3mm, marked as an abnormal drift point; X3, 08:00:00, 55℃, 85%, level 4, 6.0mm, marked as an abnormal drift point; X2, 08:15:00, 51℃, 60%, level 3, 1.4mm, unmarked. Since the indicators of X1 and X3 exceed the historical mean ± 2 standard deviations, X2 is within the range.

[0076] In step S14, an initial dataset is obtained based on the correction dataset and real-time environmental information.

[0077] It should be noted that the real-time environmental information includes the weather conditions and atmospheric pressure at the time of data collection. Weather conditions include sunny, rainy, snowy, and cloudy, and atmospheric pressure ranges from 850 to 1050 hPa. This information is obtained synchronously from the dam's meteorological station and can supplement scene factors not covered in the calibration dataset. The combination method involves associating the real-time environmental information with each record in the calibration dataset according to the collection time, ensuring that the monitoring data at the same time corresponds to the environmental state. If the real-time weather is extreme weather such as heavy rain or heavy snow, it needs to be additionally marked in the record. The initial dataset is a structured collection containing monitoring point number, collection time, temperature, humidity, wind force, level displacement value, final label, and real-time environmental information, providing a complete scene-based foundation for subsequent data processing.

[0078] For example, some records in the initial dataset are: X1, 08:10:00, -12℃, 58%, Level 3, 1.3mm, marked with anomaly drift point, sunny, 1012hPa; X2, 08:00:00, 55℃, 85%, Level 4, 6.0mm, marked with anomaly drift point, heavy rain, 985hPa; X2, 08:15:00, 51℃, 60%, Level 3, 1.4mm, unmarked, cloudy, 1008hPa.

[0079] In step S2, the initial dataset is subjected to Kalman filtering to obtain a denoised dataset. If the interference coefficients in the denoised dataset exceed a preset interference threshold, drift detection is performed on the denoised dataset to obtain a drift correction index, including:

[0080] S21. Based on the initial dataset, Kalman filtering is used to smooth the various environmental parameters, and recursive estimation is performed to generate a smooth parameter set.

[0081] S22, For the smoothing parameter set, outlier screening is performed using the preset range threshold to obtain a filtered dataset. For the outliers in the filtered dataset, the authenticity of the outliers is confirmed by comparing them with historical environmental records to obtain a comprehensive parameter set.

[0082] S23, Based on the comprehensive parameter set, extract data points that exceed the preset interference threshold, compare the data points with historical records, and obtain deviation analysis results;

[0083] S24, perform consistency verification on the deviation analysis results and generate drift correction index.

[0084] In step S21, based on the initial dataset, Kalman filtering is used to smooth each environmental parameter and recursively estimate them to generate a smooth parameter set.

[0085] It should be noted that the environmental parameters mentioned refer to the temperature, humidity, wind force, and level displacement values ​​in the initial dataset. During Kalman filtering, these four parameters are first standardized: temperature is normalized according to the actual measurement range, humidity is standardized as a percentage, wind force is quantized and converted according to levels, and level displacement is scaled to the millimeter level. After eliminating the dimensional differences, the state transition equation and observation equation are constructed.

[0086] The state transition equation is given by X at time k, after Z-score normalization. k =[T k H k W k D k ]^T,T k =Standardized temperature, H k =Standardized humidity, W k =Standardized wind force, D k =Standardized level displacement, where the mean and standard deviation are based on historical data, obtained from statistical analysis of data collected over the past 90 days; the equation is in the form of X. k =F·x k-1 +w k-1 The process noise covariance matrix, based on sensor accuracy, can be set as diag(0.12, 0.052, 0.22, 0.012), corresponding to typical errors in temperature, humidity, wind force, and displacement. Here, F is a 4x4 identity matrix, ensuring stable parameter transmission due to the gradual changes in environmental parameters over a short period. k-1 The process noise vector follows a Gaussian distribution with a mean of 0, used to cover small random fluctuations.

[0087] The observation equation is given by Z, where Z is the observation vector at time k. k =[T obs,k H obs,k W obs,k D obs,k ]^T, where T obs,k H obs,k W obs,k D obs,kThese represent the observed values ​​for temperature, humidity, wind speed, and leveling displacement, respectively—data directly measured by sensors and also the observed parameter values ​​in the initial dataset. It should be noted that the wind speed observations must use wind speed data directly collected by the sensors, such as 3.58 m / s, i.e., converting the corresponding wind force level into wind speed data; the equation form is Z. k =H·X k +v k Where H is a 4x4 identity matrix, and the sensor directly observes four parameters, v k The observation noise follows a Gaussian distribution with a mean of 0, representing measurement error. Subsequently, combining the above equations for recursive estimation, we first predict the theoretical parameter value at time k using the state transition equation, and then use the observation vector Z... k The prediction results are corrected to obtain the final smoothed parameter set. This smoothed parameter set continues the structure of the initial dataset, including the monitoring point number, collection time, and smoothed values ​​of four parameters, maintaining the correspondence between the time and monitoring points of the initial dataset, and providing stable data for subsequent analysis.

[0088] For example, from X k-1 =[25,60,3,1.2]^T, after filtering, X is obtained. k =[24.8,59.9,2.9,1.18]^T; The wind speed data in the smoothed parameter set are recorded as follows: X1, 08:10:00, -11.8℃, 57.9%, level 3, 1.28mm; X3, 08:00:00, 54.7℃, 84.8%, level 4, 5.95mm; X2, 08:15:00, 50.8℃, 59.9%, level 3, 1.39mm, which correspond to the parameters of the same monitoring point and time in the initial dataset, respectively. It is worth noting that the wind force data records in the smoothed parameter set still use wind force level, and are only converted to wind speed data when calculation is required.

[0089] In step S22, the smoothing parameter set is subjected to outlier screening using the preset range threshold to obtain a filtered dataset. For outliers in the filtered dataset, the authenticity of the outliers is confirmed by comparing them with historical environmental records to obtain a comprehensive parameter set.

[0090] It should be noted that outlier screening involves checking the temperature, humidity, wind speed, and leveling displacement values ​​of each record in the smoothing parameter set one by one. Any value exceeding a threshold is marked as an outlier, and these are integrated into a screening dataset. Historical environmental records are data from the monitoring points associated with the outliers over the past 30 days within the same time period. During comparison, the historical parameter mean and reasonable fluctuation range are calculated: temperature ±3℃, humidity ±8%, wind speed ±2, and leveling displacement ±0.5mm. Outliers whose parameters exceed these ranges are confirmed as genuine anomalies and marked; otherwise, they are unmarked. The comprehensive parameter set continues the structure of the smoothing parameter set, including the monitoring point number, collection time, smoothing parameter value, and final anomaly marker, maintaining correspondence with the preceding data in terms of time and monitoring points, providing accurate support for subsequent dam condition assessment.

[0091] For example, in the comprehensive parameter set, X1, 08:10:00, -11.8℃, 57.9%, Level 3, 1.28mm, is marked as a true anomaly; X3, 08:00:00, 54.7℃, 84.8%, Level 4, 5.95mm, is also marked as a true anomaly; X2, 08:20:00, 51.2℃, 60.1%, Level 3, 1.41mm, is unmarked and determined to be a temporary fluctuation based on historical data.

[0092] In step S23, based on the comprehensive parameter set, data points exceeding the preset interference threshold are extracted, and the data points are compared with historical records to obtain deviation analysis results.

[0093] It should be noted that the preset interference threshold is set based on three standard deviations of historical normal data from the comprehensive parameter set. In statistics, under a normal distribution, three standard deviations can cover approximately 99.73% of normal data. Data exceeding this range is likely due to environmental interference or abnormal fluctuations. This setting effectively distinguishes normal parameter fluctuations from interference signals requiring attention, meeting the accuracy requirements of dam monitoring for anomaly identification. Different parameter thresholds are calculated independently, such as a temperature threshold of ±5℃ and a leveling displacement threshold of ±1mm. When extracting data points, the smoothing parameter value of each record is checked; if any value exceeds the threshold, the data is extracted. Historical records are the monitoring points to which the data point belongs for the past 90 days of data with the same parameters. When comparing, the historical mean and fluctuation range are calculated, and then the deviation rate between the current data point and the historical mean is calculated. The deviation rate is equal to the absolute value of the difference between the current data point and the historical mean, divided by the historical mean, and multiplied by 100%. A deviation rate of less than 10% is considered a slight deviation, 10%-30% is a moderate deviation, and greater than 30% is a significant deviation. The deviation analysis results include monitoring point number, collection time, parameter type, deviation rate, and deviation level, providing a quantitative basis for dam condition assessment.

[0094] For example, in the deviation analysis results, X1, 08:10:00, temperature, deviation rate 45%, significant deviation; X3, 08:00:00, level displacement, deviation rate 32%, significant deviation; X2, 08:20:00, temperature, deviation rate 8%, slight deviation.

[0095] In step S24, the consistency of the deviation analysis results is verified, and a drift correction index is generated.

[0096] It should be noted that consistency verification revolves around two aspects: first, the correlation of deviations of different parameters at the same monitoring point, such as whether humidity fluctuates synchronously when temperature deviates significantly; second, the uniformity of deviations of adjacent monitoring points within the same time period, such as whether there are similar deviations in X4 and X5 when the X3 level displacement deviates significantly. The verification standard is that deviations of two or more parameters at the same monitoring point are considered parameter co-variable deviations, while deviations of only one parameter are considered single parameter deviations; deviations of the same parameter at three or more adjacent monitoring points are considered regional co-variable deviations, with adjacent monitoring points defined according to the dam's pre-set monitoring zones or referring to monitoring points within a physical distance of 50 meters; deviations at only one monitoring point are considered single-point deviations. The drift correction index includes the monitoring point number, verification result, correction coefficient, and correction suggestion. The coefficients are 0.8 for significant deviation, 0.9 for moderate deviation, and 1.0 for slight deviation. The coefficients are used for subsequent parameter correction. The coefficients are set according to the degree of influence of the deviation level on the accuracy of the data. Significant deviation has a large interference with the monitoring results, so the coefficient is smaller to strengthen the correction. Moderate deviation has a moderate interference, so the middle value is taken. Slight deviation has a small interference, so a value close to 1.0 is taken. At the same time, the rationality of the coefficients is verified by referring to the historical correction effect. It is recommended to give single-point deviation suggestions for re-examination of the sensor for the deviation type.

[0097] For example, in the drift correction index, X1 represents the deviation of a single parameter in the calibration result, with a correction coefficient of 0.8, and it is recommended to re-inspect the temperature sensor; X3 represents the regional coordination deviation in the calibration result, with a correction coefficient of 0.7, and it is recommended to investigate the foundation condition of the region.

[0098] In step S3, a pre-trained support vector machine model is used to process the drift correction index, and the output is a set of correction parameters, including:

[0099] S31, Obtain historical correction data, input the drift correction index and the historical correction data into the pre-trained support vector machine model, and calculate the preliminary correction parameters;

[0100] S32, obtain the current environmental parameters, and optimize the preliminary correction parameters based on the current environmental parameters to obtain optimized correction parameters;

[0101] S33, verify the optimized correction parameters with the preset input threshold, and output the correction parameter set.

[0102] In step S31, historical correction data is obtained, and the drift correction index and the historical correction data are input into a pre-trained support vector machine model to calculate preliminary correction parameters.

[0103] It should be noted that the historical correction data refers to the drift correction index and corresponding parameter correction records for the past year, including monitoring point number, historical deviation level, correction coefficient, and final correction value, ensuring data coverage of different seasonal environmental conditions. The pre-trained support vector machine model uses the correction coefficient and deviation level in the drift correction index, and the historical deviation level (or corresponding deviation value) in the historical correction data as input features. A kernel function mapping (such as radial basis function) is used to establish the relationship between features and correction parameters. The model has been validated using historical data, with errors controlled within 5%. During calculation, Z-score standardization is first used to process the input data, and then the model outputs preliminary correction values ​​for four parameters: temperature, humidity, wind force, and level displacement. The model can be configured for multi-output mode or to process each parameter independently. The preliminary correction parameters include the monitoring point number, parameter type, and preliminary correction value, ensuring a correspondence with the drift correction index.

[0104] For example, in the preliminary calibration parameters, X1 is temperature, -11.5℃; X1 is horizontal displacement, 1.25mm; X3 is temperature, 54.2℃; X3 is horizontal displacement, 5.90mm.

[0105] In step S32, the current environmental parameters are obtained, and the preliminary correction parameters are optimized based on the current environmental parameters to obtain optimized correction parameters.

[0106] It should be noted that the current environmental parameters are the temperature, humidity, wind force, and level displacement values ​​collected in real time by sensors at the dam site, updated every 5 minutes to ensure that the data reflects real-time operating conditions. During optimization, the deviation between the preliminary correction parameters and the current environmental parameters is first calculated, and then a weighted correction method is used for adjustment. The current environmental parameters have a weight of 0.3, and the preliminary correction parameters have a weight of 0.7, balancing real-time performance with historical accuracy. The optimized correction parameters continue the structured format of the preliminary correction parameters, including monitoring point number, parameter type, and optimized correction value, maintaining consistency with the drift correction index and the monitoring points and time of the preliminary correction parameters, providing accurate data for subsequent parameter applications.

[0107] For example, in the optimized correction parameters, X1 is temperature, -11.6℃; X1 is horizontal displacement, 1.26mm; X3 is temperature, 54.3℃; X3 is horizontal displacement, 5.92mm.

[0108] In step S33, the optimized correction parameters are verified against the preset input threshold, and the correction parameter set is output.

[0109] It should be noted that the preset input thresholds are set based on the historical normal operating range of the dam's environmental parameters and industry monitoring standards, covering temperatures from -12℃ to 55℃, humidity from 15% to 85%, wind speeds from 0 to 13, and horizontal displacements from -6mm to 6mm, ensuring coverage of extreme but reasonable operating conditions. During verification, each optimized calibration parameter is checked individually. If a parameter is within the threshold, it is considered valid and directly retained; if it exceeds the threshold, it is fine-tuned based on the current environmental parameters. For example, if the optimized temperature is -12.5℃, it is fine-tuned to -12℃ based on the real-time environmental temperature of -11℃ to ensure compliance with the threshold requirements. The calibration parameter set includes the monitoring point number, parameter type, final calibration value, and verification result (qualified / fine-tuned), maintaining correspondence between the monitoring points and time of the optimized calibration parameters, providing the final accurate parameters for dam environmental monitoring.

[0110] For example, in the calibration parameter set, X1 has a temperature of -11.6℃ (qualified) and a level displacement of 1.26mm (qualified); X3 has a temperature of 54.3℃ (qualified) and a level displacement of 5.92mm (qualified); X2 has a temperature of 51.1℃ (after fine-tuning, the original optimized value of 51.5℃ exceeds the threshold).

[0111] In step S4, the calibration parameter set and the initial dataset are fused and weighted to obtain the calibrated dataset, which includes:

[0112] S41, assign preset weight values ​​to the monitoring data in the initial dataset and sort them by priority to obtain a sorted dataset;

[0113] S42, the set of correction parameters is fused with the sorted dataset to obtain an intermediate fused dataset;

[0114] S43, if the weight value of the intermediate fusion dataset exceeds the preset weight threshold, dynamic adjustment and optimization are performed to obtain the corrected dataset.

[0115] In step S41, preset weight values ​​are assigned to the monitoring data in the initial dataset, and priority sorting is performed to obtain a sorted dataset.

[0116] It should be noted that the preset weight values ​​are set based on the degree of influence of the monitoring parameters on dam safety. The weight for level displacement related to structural stability is 0.4, the weight for temperature affecting material performance is 0.3, the weight for wind force related to load safety is 0.2, and the weight for humidity affecting corrosion risk is 0.1. These weight values ​​have been calibrated according to industry safety standards. Prioritization is performed by first sorting data from highest to lowest weight value. For parameters with the same weight, the degree of deviation from the preset threshold is compared; the greater the deviation, the higher the priority, ensuring that data with high impact and high deviation are processed first. The sorted dataset includes monitoring point number, parameter type, weight value, and priority, maintaining correspondence with the monitoring points and time in the initial dataset, providing a basis for subsequent key data processing.

[0117] For example, in the sorted dataset, X3 is horizontal displacement with a weight of 0.4 and a priority of 1; X1 is temperature with a weight of 0.3 and a priority of 2; X2 is wind force with a weight of 0.2 and a priority of 3; and X4 is humidity with a weight of 0.1 and a priority of 4.

[0118] In step S42, the set of correction parameters is fused with the sorted dataset to obtain an intermediate fused dataset.

[0119] It should be noted that the fusion uses monitoring point number and parameter type as the association key to ensure accurate matching of the same parameter information for the same monitoring point. Specifically, it extracts the monitoring point number, parameter type, final correction value, and verification result from the calibration parameter set, integrates them with the weight values ​​and priorities of the corresponding monitoring points and parameters in the sorted dataset, removes duplicate fields, and retains core information. The intermediate fusion dataset continues the structured format, including monitoring point number, parameter type, final correction value, verification result, weight value, and priority. This not only preserves the accurate parameters after calibration but also reflects the data processing priority, maintaining consistency with the calibration parameter set and the sorted dataset in terms of monitoring points and time, providing comprehensive data support for subsequent integrated analysis.

[0120] For example, in the intermediate fusion dataset, X1, temperature, -11.6℃, qualified, 0.3, 2; X3, level displacement, 5.92mm, qualified, 0.4, 1; X2, temperature, 51℃, fine adjustment, 0.3, 3; X4, humidity, 60%, qualified, 0.1, 4.

[0121] In step S43, if the weight value of the intermediate fusion dataset exceeds the preset weight threshold, dynamic adjustment and optimization are performed to obtain the corrected dataset.

[0122] It should be noted that the preset weight threshold is set to 0.35 based on the critical value of the parameter's impact on dam safety. A weight value exceeding this threshold means that the parameter plays a crucial role in safety assessment. During dynamic adjustment, the final correction value of parameters exceeding the threshold is first reviewed. Fine-tuning is then performed based on current environmental parameters. For example, when the level displacement weight exceeds the threshold by 0.4, the correction value is adjusted by referring to real-time foundation settlement data. Specifically, the real-time foundation settlement values ​​of the monitoring point and 2-3 adjacent monitoring points are obtained concurrently. After calculating the average settlement value, linear interpolation is used to merge the average settlement value with the original level displacement correction value to obtain the corrected value. Simultaneously, the weight is optimized according to the degree of parameter deviation; the greater the deviation, the weight can be increased by 0.05-0.1, but the upper limit is no more than 0.5. After optimization, the parameters are verified to meet the preset threshold range. Once qualified, they are integrated into a corrected dataset, including monitoring point number, parameter type, adjusted weight, final correction value, and optimization result. This maintains the correspondence between the monitoring points and time in the intermediate fused dataset, providing accurate core data for dam safety assessment.

[0123] For example, in the corrected dataset, X3 represents level displacement, with an adjusted weight of 0.42 and a final corrected value of 5.90 mm, which is a satisfactory optimization result; X1 represents temperature, with a weight of 0.3 that does not exceed the threshold, and a final corrected value of -11.6℃, which does not require adjustment; X4 represents humidity, with a weight of 0.1 that does not exceed the threshold, and a final corrected value of 58%, which also does not require adjustment.

[0124] In step S5, the feature vector of the leveling displacement is extracted from the corrected dataset, the displacement change rate is calculated, and a displacement assessment report is generated, including:

[0125] S51, Based on the level displacement data in the corrected dataset, perform classification processing to obtain a displacement feature set;

[0126] S52, Combined with a preset time range, perform a continuity analysis on the displacement feature set to calculate the displacement change rate;

[0127] S53, if the displacement change rate exceeds a preset change threshold, then mark and obtain the potential risk level, integrate the displacement change rate and the potential risk level, and generate a displacement assessment report.

[0128] In step S51, the level displacement data in the corrected dataset is classified to obtain a displacement feature set.

[0129] It should be noted that the classification process revolves around three core dimensions: displacement amplitude, trend of change, and duration. The classification criteria are based on the dam structure safety monitoring specifications. Displacement amplitude is categorized by value: less than or equal to 0.5 mm is considered micro-change, 0.5-1 mm is small change, 1-3 mm is medium change, and greater than 3 mm is large change. The trend of change is determined by the daily average change: less than or equal to 0.05 mm per day is stable, 0.05-0.1 mm per day is slow growth, and greater than 0.1 mm per day is rapid growth. The duration is distinguished by the number of consecutive monitoring days: less than 7 days is short-term, 7-30 days is medium-term, and greater than 30 days is long-term. The processing first extracts level displacement data from the calibrated dataset, including monitoring point number, final displacement value, number of consecutive monitoring days, and daily average change. Then, it is classified according to the above dimensions one by one. Finally, the classification results of each monitoring point are integrated to form a displacement feature set. The set includes monitoring point number, displacement amplitude category, trend of change category, duration category, and original displacement value, providing a feature basis for subsequent dam stability analysis.

[0130] For example, in the displacement feature set, X3 has a large displacement amplitude (5.90 mm), a rapid growth trend (0.12 mm / day), a medium duration (15 days), and an original displacement value of 5.90 mm; X4 has a medium displacement amplitude (2.1 mm), a slow growth trend (0.08 mm / day), a short duration (5 days), and an original displacement value of 2.1 mm.

[0131] In step S52, the displacement feature set is subjected to continuity analysis in combination with a preset time range, and the displacement change rate is calculated.

[0132] It should be noted that the preset time range is set based on the timeliness requirements of dam level displacement monitoring, divided into short-term (7 days), medium-term (30 days), and long-term (90 days). Different ranges correspond to different analysis focuses: short-term focuses on detecting sudden changes, medium-term focuses on trend changes, and long-term focuses on assessing stability. Continuity analysis first checks the data integrity within the corresponding time range in the displacement feature set, counting the number of missing monitoring points. If the missing rate is less than 10%, the data is considered valid; if it exceeds 10%, linear interpolation is used to fill in the missing data, ensuring a reliable analytical basis. The displacement change rate is calculated by taking the difference in displacement values ​​between the first and last two time points of the corresponding time range, dividing it by the number of days in that time range, to obtain the short-term, medium-term, and long-term rates, respectively, in mm per day. The data integrity results are also marked as valid or complete. The final displacement change rate includes the monitoring point number, short-term rate, medium-term rate, long-term rate, and data integrity, maintaining a correspondence with the monitoring points in the displacement feature set, providing quantitative data for assessing dam stability trends.

[0133] For example, in the displacement change rate data, X3 has a short-term rate of 0.12 mm / day, a medium-term rate of 0.11 mm / day, and a long-term rate of 0.09 mm / day, and the data is complete and valid; X4 has a short-term rate of 0.08 mm / day, a medium-term rate of 0.07 mm / day, and a long-term rate of 0.06 mm / day, and the data is complete; X5 has a short-term rate of 0.13 mm / day, a medium-term rate of 0.12 mm / day, and a long-term rate of 0.10 mm / day, and the data is complete and valid (due to one missing monitoring, the missing rate is 5%).

[0134] In step S53, if the displacement change rate exceeds a preset change threshold, a potential risk level is marked and obtained. The displacement change rate and the potential risk level are then integrated to generate a displacement assessment report.

[0135] It should be noted that the preset time range is set based on the timeliness requirements of dam level displacement monitoring, and is divided into short-term (7 days), medium-term (30 days), and long-term (90 days). Different ranges correspond to different analysis focuses: short-term focuses on detecting sudden changes, medium-term focuses on trend changes, and long-term focuses on assessing stability. Continuity analysis first checks the data integrity within the corresponding time range in the displacement feature set, counts the number of missing monitoring points, and considers data valid if the missing rate is less than 10%. If it exceeds 10%, linear interpolation is used to fill in the missing data to ensure the reliability of the analysis basis. The displacement change rate is calculated by dividing the difference in displacement values ​​at the beginning and end of the time range by the number of days. Short-term, medium-term, and long-term rates are calculated separately, in mm per day, while the data integrity results (valid or filled) are marked. The final displacement change rate includes the monitoring point number, short-term rate, medium-term rate, long-term rate, and data integrity, maintaining a correspondence with the monitoring points in the displacement feature set, providing quantitative data for judging the dam stability trend.

[0136] For example, in the displacement change rate X3, the short-term rate is 0.12 mm per day, the medium-term rate is 0.11 mm per day, and the long-term rate is 0.09 mm per day, and the data integrity is valid; X4, the short-term rate is 0.08 mm per day, the medium-term rate is 0.07 mm per day, and the long-term rate is 0.06 mm per day, and the data integrity is complete.

[0137] In step S6, the displacement assessment report is compared and verified with the preset monitoring standard to obtain a description of the level displacement, including:

[0138] S61, Based on the displacement assessment report, extract displacement data points and classify and organize them to obtain a comparison dataset;

[0139] S62, compare the comparison dataset with the preset monitoring standard to obtain the comparison result;

[0140] S63, Summarize and process the comparison results to generate a level displacement description.

[0141] In step S61, based on the displacement assessment report, displacement data points are extracted and classified to obtain a comparison dataset.

[0142] It should be noted that when extracting displacement data points, core information is selected from the displacement assessment report, including monitoring point number, monitoring time, actual displacement value, assessment conclusion (stable or warning), and the displacement change rate within the corresponding time range. This ensures that each data point is associated with a complete assessment background, avoiding isolated value extraction. Classification is based primarily on the assessment conclusion, initially dividing data into stable and warning categories. Stable data points are those with a displacement change rate less than or equal to 0.05 mm per day and an assessment conclusion of "stable." Warning data points are those with a displacement change rate greater than 0.05 mm per day or an assessment conclusion of "warning." The warning category is further subdivided by displacement amplitude, following the S51 classification standard, into four categories: slight change, small change, medium change, and large change. The comparison dataset uses a structured format, including monitoring point number, monitoring time, actual displacement value, displacement change rate, assessment conclusion, and classification result, maintaining consistency with the monitoring points and time in the displacement assessment report, providing a clear classification basis for subsequent data comparison and trend verification.

[0143] For example, in the comparison dataset, X3, with a monitoring time of 08:10, has an actual displacement value of 5.90 mm and a displacement change rate of 0.12 mm per day. The assessment conclusion is a warning, and the classification result is a warning class of "large change"; X4, with a monitoring time of 08:20, has an actual displacement value of 2.1 mm and a displacement change rate of 0.07 mm per day. The assessment conclusion is a warning, and the classification result is a warning class of "medium change"; X2, with a monitoring time of 08:30, has an actual displacement value of 0.3 mm and a displacement change rate of 0.03 mm per day. The assessment conclusion is "stable," and the classification result is a stable class of "small change."

[0144] In step S62, the comparison dataset is compared with a preset monitoring standard to obtain the comparison result.

[0145] It should be noted that the preset monitoring standards are based on dam structural safety specifications and are divided into stable and early warning categories. The stable category requires a displacement amplitude of less than or equal to 0.5 mm and a displacement change rate of less than or equal to 0.05 mm per day. The early warning category is further subdivided by displacement amplitude: for minor changes of 0.5-1 mm, the rate must be less than or equal to 0.08 mm per day; for minor changes of 1-3 mm, the rate must be less than or equal to 0.1 mm per day; and for major changes greater than 3 mm, the rate must be less than or equal to 0.12 mm per day. Exceeding these ranges is considered exceeding the standard. During comparison, the classification result (stable or early warning) of the data points in the comparison dataset is first determined. Then, the displacement amplitude and displacement change rate are checked against the corresponding standards to verify whether they meet the requirements. Simultaneously, the consistency of the assessment conclusions is verified. The comparison results include the monitoring point number, the comparison dimension (displacement amplitude, rate), the standard requirement, the actual value, and the comparison conclusion (compliant or exceeding the standard), maintaining correspondence with the monitoring points and time in the comparison dataset, providing direct evidence for determining the dam's safety status.

[0146] For example, in the comparison results, X3, the actual values ​​for displacement amplitude and rate in the comparison dimension are 5.90 mm and 0.12 mm / day, respectively. The standard requires that the rate of large changes be less than or equal to 0.12 mm / day, and the actual value meets the requirement, so the comparison conclusion is correct. X4, the actual values ​​for displacement amplitude and rate in the comparison dimension are 2.1 mm and 0.07 mm / day, respectively. The standard requires that the rate of small changes be less than or equal to 0.1 mm / day, and the actual value meets the requirement, so the comparison conclusion is correct. X5, the actual values ​​for displacement amplitude and rate in the comparison dimension are 3.5 mm and 0.13 mm / day, respectively. The standard requires that the rate of large changes be less than or equal to 0.12 mm / day, and the actual value exceeds the standard, so the comparison conclusion is incorrect.

[0147] In step S63, the comparison results are summarized and processed to generate a level displacement description.

[0148] It should be noted that the summary processing revolves around three core dimensions: statistics of the number of monitoring points, distribution of comparison results, and analysis of details of exceedances, ensuring coverage of the overall situation and key anomalies. First, the total number of monitoring points participating in the comparison is counted. Then, the points are categorized and counted according to the comparison results (compliant or exceedance). For compliant points, the number of minor, small, medium, and large changes under the stable and early warning categories needs to be distinguished. For exceedance points, the specific monitoring point number and the dimension of exceedance (displacement amplitude or rate) must be marked to avoid information omissions. When generating the level displacement description, the overall overview is presented first, including the total number of monitoring points and the compliance rate (number of compliant points divided by the total number multiplied by 100%). Next, the classification distribution is explained, including the proportion of stable and early warning categories and the number of sub-categories. Finally, the details of exceedances are listed, including the exceeding monitoring points, the exceeding dimension, and the difference between the actual value and the standard requirements. The description should be concise and structured, maintaining correspondence with the monitoring points in the comparison results, providing core displacement summary content for the dam safety assessment report.

[0149] For example, the leveling displacement is described as follows: A total of 5 monitoring points were compared, with a compliance rate of 80%, 4 of which were compliant and 1 exceeded the standard. Among the compliant points, 1 was in the stable category (X2), 3 were in the slight change / early warning category (X3 large change, X4 medium change, X1 small change), and 1 exceeded the standard (S5), with a displacement rate of 0.13 mm / day, exceeding the large change rate standard by 0.01 mm / day.

[0150] In summary, this invention provides a method for monitoring dam level displacement based on digital twins, including steps such as collecting environmental and displacement data to construct an initial dataset, processing environmental parameters through Kalman filtering, triggering drift detection to quantify the amplitude, obtaining real-time correction parameters through support vector machines, fusing and adjusting to obtain a corrected dataset, and generating an assessment report through trend analysis. This invention, through a digital twin-driven environmental disturbance correction and displacement trend analysis mechanism, achieves closed-loop management of dam level displacement from data processing to state assessment, effectively improving monitoring accuracy and the accuracy of structural risk assessment, and providing reliable technical support for dam safety decision-making.

[0151] Reference Figure 2 The second embodiment of the present invention provides a dam leveling displacement monitoring system based on digital twins, comprising:

[0152] The data acquisition module is used to acquire environmental data and leveling displacement data to obtain the initial dataset;

[0153] The filtering module is used to perform Kalman filtering on the initial dataset to obtain a denoised dataset. If the interference coefficient in the denoised dataset exceeds a preset interference threshold, the denoised dataset is subjected to drift detection to obtain a drift correction index.

[0154] The model processing module is used to process the drift correction index using a pre-trained support vector machine model and output a set of correction parameters.

[0155] The data fusion module is used to fuse the correction parameter set and the initial dataset, and after weighted adjustment, obtain the corrected dataset;

[0156] The analysis and evaluation module is used to extract the feature vector of the level displacement from the corrected dataset, calculate the displacement change rate, and generate a displacement evaluation report.

[0157] The verification output module is used to compare and verify the displacement assessment report with the preset monitoring standard to obtain a description of the horizontal displacement.

[0158] It should be noted that the dam leveling displacement monitoring system based on digital twin provided in this embodiment of the invention is used to execute all the process steps of the dam leveling displacement monitoring method based on digital twin in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0159] It should be noted that the system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0160] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

[0161] This invention also provides an electronic device. The electronic device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a dam level displacement monitoring program. When the processor executes the computer program, it implements the steps described in the various method embodiments above, for example... Figure 1 The step S11 shown. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the above-described device embodiments, such as the data acquisition module.

[0162] For example, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the electronic device.

[0163] The electronic device may be a desktop computer, laptop, handheld computer, or smart tablet, etc. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the above components are merely examples of electronic devices and do not constitute a limitation on the electronic device. It may include more or fewer components than described above, or combine certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, etc.

[0164] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the electronic device, connecting all parts of the electronic device via various interfaces and lines.

[0165] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0166] Wherein, if the modules / units integrated in the electronic device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0167] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0168] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A method for monitoring the leveling displacement of a dam based on digital twins, characterized in that, include: Acquire environmental data and leveling displacement data to obtain the initial dataset; Kalman filtering is applied to the initial dataset to obtain a denoised dataset. If the interference coefficient in the denoised dataset exceeds a preset interference threshold, drift detection is performed on the denoised dataset to obtain a drift correction index. The drift correction index is processed using a pre-trained support vector machine model, and a set of correction parameters is output. By merging the set of correction parameters and the initial dataset, and performing a weighted adjustment, the corrected dataset is obtained. Extract the feature vector of the level displacement from the corrected dataset, calculate the displacement change rate, and generate a displacement assessment report. The displacement assessment report is compared and verified with the preset monitoring standards to obtain a description of the horizontal displacement. Wherein, the initial dataset is subjected to Kalman filtering to obtain a denoised dataset. If the interference coefficients in the denoised dataset exceed a preset interference threshold, drift detection is performed on the denoised dataset to obtain a drift correction index, including: Based on the initial dataset, Kalman filtering is used to smooth the various environmental parameters, and recursive estimation is performed to generate a smooth parameter set; For the smoothing parameter set, outlier screening is performed using a preset range threshold to obtain a filtered dataset. For outliers in the filtered dataset, the authenticity of the outliers is confirmed by comparing them with historical environmental records to obtain a comprehensive parameter set. Based on the comprehensive parameter set, data points exceeding the preset interference threshold are extracted, and the data points are compared with historical records to obtain deviation analysis results; The consistency of the deviation analysis results is verified, and a drift correction index is generated. The consistency verification includes: verification of the correlation between deviations of different parameters at the same monitoring point, and verification of the uniformity of deviations of adjacent monitoring points within the same time period. The verification criteria are: if two or more parameters at the same monitoring point show deviations, it is determined to be a parameter co-variance deviation; if only one parameter shows a deviation, it is determined to be a single parameter deviation. Adjacent monitoring points are those within a preset monitoring zone or those within a physical distance of 50 meters. If three or more adjacent monitoring points show the same parameter deviation, it is determined to be a regional co-variance deviation; if only one monitoring point shows a deviation, it is determined to be a single-point deviation. The drift correction index includes the monitoring point number, verification result, correction coefficient, and correction suggestion. The correction coefficient is determined according to the deviation level: the correction coefficient for significant deviation is 0.8, the correction coefficient for moderate deviation is 0.9, and the correction coefficient for slight deviation is 1.

0. Correction suggestions are generated according to the deviation type: the correction suggestion for single-point deviation is to re-examine the sensor.

2. The dam leveling displacement monitoring method based on digital twin according to claim 1, characterized in that, The acquisition of environmental data and leveling displacement data yields an initial dataset, including: Environmental data and leveling displacement data are collected in real time through a sensor network and integrated to obtain a dataset; wherein, the environmental data includes temperature, humidity, and wind conditions; Anomalies are screened in the collected dataset using a preset range threshold, and potential drift points are marked to obtain an intermediate dataset. For potential drift points in the intermediate dataset, historical environmental monitoring records are obtained for comparison and verification to obtain a corrected dataset; Based on the corrected dataset and combined with real-time environmental information, an initial dataset is obtained.

3. The dam leveling displacement monitoring method based on digital twin according to claim 1, characterized in that, The pre-trained support vector machine model is used to process the drift correction index, and the output is a set of correction parameters, including: Acquire historical correction data, input the drift correction index and the historical correction data into a pre-trained support vector machine model, and calculate the preliminary correction parameters; Obtain the current environmental parameters, and optimize the preliminary correction parameters based on the current environmental parameters to obtain the optimized correction parameters; The optimized correction parameters are verified against the preset input threshold, and a set of correction parameters is output.

4. The dam leveling displacement monitoring method based on digital twin according to claim 1, characterized in that, The fused set of correction parameters and the initial dataset, after weighted adjustment, yield the corrected dataset, which includes: The monitoring data in the initial dataset are assigned preset weight values ​​and sorted by priority to obtain a sorted dataset; The set of correction parameters is fused with the sorted dataset to obtain an intermediate fused dataset; If the weight values ​​of the intermediate fusion dataset exceed the preset weight threshold, dynamic adjustment and optimization are performed to obtain the corrected dataset.

5. The dam leveling displacement monitoring method based on digital twin according to claim 1, characterized in that, The process of extracting the feature vector of the leveling displacement from the corrected dataset, calculating the rate of displacement change, and generating a displacement assessment report includes: Based on the leveling displacement data in the corrected dataset, classification processing is performed to obtain a displacement feature set; By combining a preset time range, a continuity analysis is performed on the displacement feature set to calculate the displacement change rate; If the displacement change rate exceeds a preset change threshold, a potential risk level is identified, and the displacement change rate and the potential risk level are integrated to generate a displacement assessment report.

6. The dam leveling displacement monitoring method based on digital twin according to claim 1, characterized in that, The step of comparing and verifying the displacement assessment report with preset monitoring standards to obtain a description of the level displacement includes: Based on the displacement assessment report, displacement data points are extracted and classified to obtain a comparison dataset. The comparison dataset is compared with the preset monitoring standard to obtain the comparison results; The comparison results are summarized and processed to generate a description of the level displacement.

7. A dam leveling displacement monitoring system based on digital twin, characterized in that, For implementing the method as described in any one of claims 1-6, comprising: The data acquisition module is used to acquire environmental data and leveling displacement data to obtain the initial dataset; The filtering module is used to perform Kalman filtering on the initial dataset to obtain a denoised dataset. If the interference coefficient in the denoised dataset exceeds a preset interference threshold, the denoised dataset is subjected to drift detection to obtain a drift correction index. The model processing module is used to process the drift correction index using a pre-trained support vector machine model and output a set of correction parameters. The data fusion module is used to fuse the correction parameter set and the initial dataset, and after weighted adjustment, obtain the corrected dataset; The analysis and evaluation module is used to extract the feature vector of the level displacement from the corrected dataset, calculate the displacement change rate, and generate a displacement evaluation report. The verification output module is used to compare and verify the displacement assessment report with the preset monitoring standard to obtain a description of the horizontal displacement.

Images