Hydraulic engineering data intelligent monitoring method based on digital twinning

Abstract

The invention discloses a hydraulic engineering data intelligent monitoring method based on digital twinning, and particularly relates to the technical field of hydraulic engineering intelligent monitoring, and the method comprises the following steps: obtaining control instruction data and entity hydraulic engineering real-time monitoring data, analyzing the time sequence characteristics and the change trend, and obtaining the real-time monitoring data of the control instruction data and the entity hydraulic engineering real-time monitoring data; detecting whether the control instruction data is returned and participates in system calculation, and identifying a data echo effect; after detection, calculating a short-term influence index and a long-term influence index based on the prediction error convergence degree, respectively evaluating the influence of the echo effect on the current operation state and the future scheduling precision, judging the level of the echo effect to be a low level, a middle level or a high level according to the influence degree, and processing by adopting a corresponding optimization strategy; according to the method, dynamic evaluation and grading judgment of the data echo effect are realized, and the early warning and adaptive capacity of the system is improved; fuzzy logic reasoning is adopted to optimize decisions, and regulation and control intelligence is enhanced; and through a differentiation strategy and a credible data source reconstruction model, the operation stability and the scheduling precision of the water conservancy project are guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent monitoring of water conservancy projects. More specifically, the present invention relates to an intelligent monitoring method for water conservancy project data based on digital twin. Background Art

[0002] With the development of intelligent water conservancy and intelligent dispatching systems, the management and regulation of water conservancy projects are transforming from traditional manual operations and single-point monitoring to digital, intelligent, and automated directions. Due to advantages such as virtual-real fusion, real-time simulation, and intelligent prediction, digital twin technology has become one of the core technologies in modern water conservancy project management. Through the digital twin model, the operating state of the physical system of the water conservancy project can be reflected in real time, assisting decision-makers in formulating scientific regulation strategies and achieving automated control. However, in the actual application process of existing digital twin systems, the system usually highly depends on the feedback data of the physical system as the input for prediction and control, and the optimization and adjustment of control strategies also often rely on historical feedback information. If the system fails to effectively separate the influence of feedback data during operation, it is easy to cause an abnormal cycle between control instructions and feedback data. When the data of control instructions is fed back and re-participates in model calculation and continues to affect the generation of subsequent control strategies, it is easy to trigger the data echo effect.

[0003] The data echo effect will not only cause the continuous accumulation of prediction errors, but may also cause the digital twin model to lose its true perception ability of the actual working conditions, resulting in a decrease in the effectiveness of control strategies, and further leading to abnormal water resource scheduling or unstable operation of hydraulic structures, increasing the operation risk of the system. The existing technology lacks a complete set of echo effect detection and suppression mechanisms. Especially in data fusion models where feedback data occupies a high weight, it is difficult to accurately judge whether the feedback data forms a loop and affects system judgment. In addition, existing intelligent regulation systems for water conservancy projects often only rely on error analysis on a single time scale and cannot achieve a comprehensive assessment of the echo effect at two levels: short-term fluctuations and long-term cumulative risks. Therefore, there is an urgent need to design a multi-level and intelligent echo effect detection and suppression method to ensure the scheduling efficiency and safety of the water conservancy project system. Summary of the Invention

[0004] To achieve the above object, the present invention provides the following technical solutions: An intelligent monitoring method for water conservancy project data based on digital twin, comprising the following steps: During the operation of the digital twin system, obtain the control instruction data for feedback regulation of the operation parameters of the water conservancy project, and synchronously collect the real-time monitoring data of the physical water conservancy project; By comparing the timing characteristics and change trends of the control instruction data and the feedback data of the physical water conservancy project, it is judged whether there is a loop phenomenon that the control instruction data is transmitted back and participates in the calculation of the digital twin system again as input data, so as to detect whether the data echo effect occurs; In the case of detecting the existence of the data echo effect, a comprehensive judgment is made based on the preset test parameters. The test parameter is the prediction error convergence degree of the digital twin model, which is obtained by calculating the dynamic deviation value between the current predicted value and the real-time monitored value; According to the prediction error convergence degree, the short-term influence index and the long-term influence index are calculated respectively. The short-term influence index is used to evaluate the influence of the echo effect on the operation state of the water conservancy project within the current regulation cycle, and the long-term influence index is used to evaluate the cumulative influence of the echo effect on the accuracy of subsequent scheduling decisions; According to the short-term influence index and the long-term influence index, the level of the current data echo effect is determined. The echo effect level is divided into low level, medium level and high level according to the influence degree; According to the determined echo effect level, different data echo optimization strategies are adopted for processing.

[0005] In a preferred implementation manner, it is judged whether there is a loop phenomenon that the control instruction data is transmitted back and participates in the system calculation again as input data, so as to detect whether the data echo effect occurs. The specific operation steps are as follows: Step a, collect the control instruction data set U(t) generated by the digital twin system at time t. U(t) includes control parameters u1, u2, u3, which correspond to the water gate opening, the flood discharge volume and the gate opening and closing frequency respectively; Step b, record the monitoring data set S(t + Δt) fed back by the physical system of the water conservancy project at time t + Δt. S(t + Δt) includes the water level h, the flow velocity v and the gate state θ corresponding to the control parameters u1, u2, u3 respectively; Step c, calculate the similarity D1 between U(t) and S(t + Δt) through the dynamic time warping algorithm. The formula is: ; n is the total number of control parameter dimensions, which is also the total number of monitoring data dimensions, with a value of 3. max(|U|,|S|) is the normalization coefficient to ensure that D1 is dimensionless and eliminate the influence of parameter dimensions. i is the number of parameter types; Step d, at time t + 2Δt, calculate the similarity D2 between the control instruction data U'(t + 2Δt) generated after inputting the feedback data S(t + Δt) into the digital twin system and the original U(t). The Pearson correlation coefficient R calculation formula is: ; and Respectively represent the per - dimension means of each control parameter in U(t); Step e, when the product of D1 and D2 is greater than the set standard threshold, it is determined that there is a phenomenon of control instruction data feedback, and it is recognized that a data echo effect occurs.

[0006] In a preferred implementation, the specific steps for calculating the short - term impact index include: Define a time window τ to evaluate the time range, and divide it into multiple equally spaced time nodes to form a short - term time series. At each time node, collect three parameter values predicted by the digital twin system, including the water level prediction value, the flow velocity prediction value, and the gate opening and closing state prediction value; at the same time, collect three monitored parameter values feedback by the physical water conservancy project system, including the water level monitored value, the flow velocity monitored value, and the gate opening and closing state monitored value; At each time node, subtract the water level monitored value from the water level prediction value and calculate the absolute value of the difference; then divide this absolute value by the water level monitored value plus a very small positive number to avoid division by zero errors. Using the same method, calculate the errors of the flow velocity prediction value and the gate opening and closing state prediction value respectively. For each time node, set three proportionality coefficients for the water level, flow velocity, and gate state, respectively reflecting their importance to the overall system safety or performance, and perform a weighted average of the water level error, flow velocity error, and gate state error according to the proportionality coefficients to obtain the comprehensive error value at this time node; Within the entire time window τ, calculate the average value of the comprehensive error values of all time nodes, then for the comprehensive error value of each time node, find the square of the difference between it and the average value, sum up all the squared differences, divide by the total number of time nodes to obtain the error fluctuation amplitude within the short - term time period, and finally multiply the error fluctuation amplitude by a preset sensitivity adjustment coefficient to obtain the short - term impact index.

[0007] In a preferred implementation, the specific steps for calculating the long - term impact index include: Define a long - term observation window T, where T is much larger than the short - term window τ, divide it into multiple equally spaced time nodes to form a long - term time series. At each time node, collect three parameter values predicted by the digital twin system, including the water level prediction value, the flow velocity prediction value, and the gate opening and closing state prediction value, and at the same time, collect three monitored parameter values feedback by the physical water conservancy project system, including the water level monitored value, the flow velocity monitored value, and the gate opening and closing state monitored value; At each time node, subtract the water level monitored value from the water level prediction value and calculate the absolute value of the difference; then divide this absolute value by the water level monitored value plus a very small positive number to avoid division by zero errors. Using the same method, calculate the errors of the flow velocity prediction value and the gate opening and closing state prediction value respectively.

[0008] For each time node, three proportional coefficients of water level, flow rate and gate status are set to reflect their influence on the long-term stability of the overall system. The water level error, flow rate error and gate status error are weighted averaged according to the proportional coefficients to obtain the comprehensive error value of the time node. The comprehensive error values ​​of all time nodes in the long-term time period are accumulated and multiplied by the time interval to obtain the comprehensive error integral value, which reflects the degree of error accumulation in the long-term period; The integrated error integral value is divided by the product of the total time length of the long-term period and the set maximum allowable error value, so that the long-term impact index is a dimensionless value. Through normalization processing, it is ensured that the index can reflect the comparability between different water conservancy project systems, and finally the long-term impact index is obtained.

[0009] In a preferred embodiment, when determining the data echo effect level, fuzzy logic is used to infer based on the short-term impact index and the long-term impact index.

[0010] In a preferred embodiment, the fuzzy logic uses the principle: The short-term impact index and long-term impact index corresponding to the occurrence of data echo effect are obtained and used as the input variables of fuzzy logic together. The corresponding level of data echo effect is used as the output variable of fuzzy logic. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the degree of adaptation of the corresponding levels of different data echo effects under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to obtain the corresponding level of data echo effect.

[0011] In a preferred embodiment, different data echo optimization strategies are used for processing, including: in a low-level echo effect state, applying a data filtering algorithm to eliminate abnormal feedback data; in a mid-level echo effect state, adjusting the data fusion weight parameters to reduce the model's dependence on feedback data; in a high-level echo effect state, resetting the model input data source, forcibly introducing an external trusted monitoring data source for verification, and rebuilding the digital twin model state to completely eliminate the echo effect.

[0012] In a preferred embodiment, at an advanced echo effect level, the state of the digital twin model is reconstructed by introducing a third-party trusted data source. The trusted data source includes data from drone inspections, satellite remote sensing, and manual inspections. The reconstruction process includes: pausing the existing model, deleting the current model feedback link information, importing trusted data to initialize modeling parameters, generating a new digital twin model, and verifying the model validity through historical verification data.

[0013] In a preferred embodiment, in the low-level echo effect level state, the adopted filtering algorithm is adaptive Kalman filtering.

[0014] In a preferred embodiment, in the medium-level echo effect state, weight optimization adjustment is performed. Specifically: Normalize the short-term impact index and the long-term impact index respectively, and convert them into percentage expressions; Calculate the average value of the short-term impact index and the long-term impact index as the comprehensive echo effect degree; Preset an adjustment coefficient for determining the reduction amplitude of the feedback data fusion weight. Multiply the comprehensive echo effect degree by the preset adjustment coefficient to obtain a reduction value. Subtract the reduction value from the initial fusion weight of the feedback data, and use the result as the new feedback data fusion weight; The fusion weight of the predicted data is equal to one minus the new feedback data fusion weight.

[0015] The beneficial effects of the present invention are: By introducing an analysis method combining the short-term impact index and the long-term impact index, the present invention realizes multi-dimensional and full-cycle dynamic evaluation of the data echo effect in the digital twin system, effectively solving the problems of detection lag or inaccuracy caused by relying solely on a single time scale analysis in the prior art. The short-term impact index can timely capture the prediction fluctuations and operation anomalies caused by data echo in the current regulation cycle of the system, while the long-term impact index comprehensively measures the error accumulation and model deviation trend caused by the echo effect during the long-term operation of the system. Through the collaborative judgment of the dual indicators, the sensitivity and early warning ability of the system to the data echo effect are improved, providing a more timely and accurate risk identification means for the water conservancy project scheduling management, and significantly enhancing the safety and stability of the system operation.

[0016] By adopting the fuzzy logic reasoning method, taking the short-term impact index and the long-term impact index as input variables, and making comprehensive judgments according to the preset fuzzy rules, the present invention scientifically determines the data echo effect level and selects different coping strategies accordingly, overcoming the technical bottlenecks of unclear echo effect level determination and single processing measures in the existing methods. Fuzzy logic can effectively handle the fuzziness and uncertainty of the echo effect influence degree, reasonably divide the low-level, medium-level and high-level echo effect states, and ensure that the system takes optimal intervention measures under different risk levels. This mechanism significantly improves the adaptability and autonomous regulation ability of the digital twin system to complex working conditions, realizes the intelligent management and refined control of the water conservancy project operation state, reduces the frequency of human intervention, and improves the scientificity and reliability of system decision-making.

[0017] The present invention designs a hierarchical and differentiated data echo optimization strategy for different levels of echo effect, especially in the intermediate and advanced echo effect states, by dynamically adjusting the data fusion weights and introducing external trusted data sources, the digital twin system model can be accurately corrected and quickly reconstructed. Specifically, in the intermediate state, by optimizing the fusion ratio of feedback data and prediction data, the system's dependence on echo data is reduced, and the autonomy and accuracy of model prediction are maintained; in the advanced state, third-party data sources such as drone inspections, satellite remote sensing and manual inspections are introduced to reconstruct the digital twin model state and eliminate the model error diffusion problem caused by the echo effect. This solution ensures the stable operation and control accuracy of the water conservancy project system in a complex environment, effectively avoids prediction inaccuracies and scheduling anomalies caused by the echo effect, and improves the overall intelligence and automation level of project management. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a schematic diagram of the intelligent monitoring method for water conservancy project data based on digital twins in the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Reference Figure 1 The following embodiments are obtained: Embodiment 1: The present invention is proposed under the background of the continuous improvement of the current demand for intelligent management of water conservancy projects and the widespread application of digital twin technology. With the expansion of the scale of water conservancy projects and the increase in management complexity, the traditional water conservancy project operation monitoring and control methods have gradually exposed problems such as delayed response, insufficient data processing capabilities, and decision-making relying on manual experience, which are difficult to meet the needs of modern water conservancy projects for efficient, safe and intelligent operation. Digital twin technology provides a new intelligent management solution for water conservancy projects with its ability of high integration of virtual and reality, real-time dynamic simulation and intelligent analysis and prediction. However, in the actual application process, the digital twin system is highly dependent on the feedback data of the physical system, which easily causes the echo effect of data transmission and re-participation in model calculation, resulting in an abnormal cycle between the system control instructions and the feedback data, which in turn causes problems such as accumulation of prediction errors and failure of decision-making, seriously affecting the safety and control efficiency of water conservancy projects. Therefore, there is an urgent need for an intelligent monitoring method that can effectively detect and suppress the data echo effect in the digital twin system to ensure the stable operation and scientific control of the water conservancy project system.

[0021] The present invention aims to provide a systematic and intelligent data echo effect detection and optimization mechanism by constructing an intelligent monitoring method for water conservancy project data based on a digital twin system. By collecting the operation data of the physical system of the water conservancy project in real time, combining with the prediction results of the digital twin model, dynamically analyzing the convergence degree of the system prediction error, calculating the short-term impact index and the long-term impact index, comprehensively determining the data echo effect level, and taking corresponding optimization strategies according to different levels, it effectively reduces the dependence of the system on feedback data and avoids system control instability and decision-making deviation caused by the data echo effect. This method can realize the intelligent and precise monitoring and control of the operation state of water conservancy projects, improve the safety and reliability of system operation, and is widely applicable to various water conservancy project management scenarios such as reservoir operation, river regulation, and canal irrigation.

[0022] The method disclosed in the present invention not only realizes the early detection and timely intervention of the data echo effect, but also improves the adaptability of the system to complex working conditions through fuzzy logic reasoning, and introduces diversified external credible data sources as the basis for model reconstruction, enhancing the robustness and emergency response ability of the digital twin system under extreme hydrological conditions, and has high engineering application value and popularization prospects.

[0023] The digital twin system refers to a comprehensive intelligent monitoring and control platform constructed in the field of water conservancy projects by integrating physical entity systems, virtual simulation models, and real-time data acquisition and processing technologies. This system uses the digital twin model as the core tool to provide functions such as real-time state perception, decision-making analysis, and feedback regulation through dynamic simulation and prediction of the operation process of water conservancy projects. The digital twin model is the core component in the digital twin system, specifically manifested as a virtual simulation model constructed based on mathematical models, physical laws, engineering parameters, and historical operation data, which is used to accurately reflect and predict the operation state and behavior of the physical system of water conservancy projects. The physical system of water conservancy projects refers to the actual physical engineering facilities including reservoirs, dams, canals, sluice gates, etc. These facilities undertake functions such as water resource scheduling, flood control, irrigation, and water supply in the actual environment. The real-time operation state is collected by sensors, monitoring equipment, etc. and transmitted to the digital twin system to realize the dynamic mapping and interaction between the virtual and the real.

[0024] During the operation of the digital twin system, obtain the control instruction data for feedback regulation of the operation parameters of the water conservancy project, and synchronously collect the real-time monitoring data of the physical water conservancy project; during the operation of the digital twin system, continuously obtain the control instruction data for feedback regulation of the operation parameters of the water conservancy project, and synchronously collect the real-time monitoring data of the physical water conservancy project. The control instruction data is generated by the digital twin model based on the prediction and analysis of the operation state of the water conservancy project system, including the real-time setting of key regulation parameters such as the opening of the sluice gate, the flood discharge volume, and the opening and closing frequency of the gate. The real-time monitoring data of the physical water conservancy project is collected through a multi-source sensor network deployed in hydraulic structures, channels, and water bodies, covering core feedback data such as water level, flow velocity, and gate status. This step ensures the closed-loop data interaction between the digital twin system and the physical system, providing a basic data source for subsequent analysis of the relationship between control instruction data and feedback data. By realizing the real-time synchronous collection of control instructions and feedback monitoring data, the system can comprehensively master the dynamic correspondence between the current project operation state and the control response, laying a data foundation for accurately judging the data echo effect.

[0025] By comparing the temporal characteristics and change trends of the control instruction data and the feedback data of the physical water conservancy project, determine whether there is a loop phenomenon where the control instruction data is transmitted back and participates in the digital twin system calculation again as input data, so as to detect whether the data echo effect occurs; by comparing the temporal characteristics and change trends of the control instruction data and the feedback data of the physical water conservancy project, determine whether there is a loop phenomenon where the control instruction data is transmitted back and participates in the digital twin system calculation again as input data, so as to detect whether the data echo effect occurs. This process uses the dynamic time warping algorithm to analyze the time series matching degree of the control instruction and feedback data, and at the same time combines the Pearson correlation coefficient to evaluate the consistency of the two sets of data in the numerical change trend. When the similarity between the two reaches the set threshold and shows a closed-loop trend in time delay, it indicates that the control instruction data may be transmitted back in the feedback link and has a re-input effect on the digital twin model, resulting in a high correlation and synchronization between the prediction result and the physical feedback data. This step is of great significance in the digital twin system, which can effectively identify the echo effect problem caused by improper feedback path design or excessive system feedback dependence, prevent the system from falling into a misjudgment loop of mutual reinforcement between prediction and feedback, and improve the robustness and reliability of the intelligent regulation system of the water conservancy project.

[0026] In the case of detecting the existence of the data echo effect, a comprehensive determination is made based on preset test parameters. The test parameters are the prediction error convergence degree of the digital twin model, which is obtained by calculating the dynamic deviation value between the current predicted value and the real-time monitored value. When the system detects that there may be a data echo effect, it is necessary to further comprehensively judge the severity of the echo effect to decide whether to intervene and what optimization strategy to adopt. The present invention proposes to use the prediction error convergence degree as the key test parameter, and by calculating the dynamic deviation between the predicted value and the real-time monitored value of the digital twin system, analyze the change of the prediction error in the time dimension. If the prediction error continues to increase or shows an unstable trend in multiple time periods, it indicates that the prediction ability of the system may be interfered by the data echo effect, resulting in continuous error accumulation. If the prediction error quickly converges to a small range in a short time, it indicates that the system can still effectively adjust the prediction strategy and is not seriously affected. The system can set different error convergence thresholds as the criteria for judging the influence degree of the echo effect. The significance of this step is to establish a quantitative evaluation system for the influence degree of the echo effect, enabling the system to judge whether to adjust the operation strategy according to the real-time calculation results and improving the adaptive control ability of the digital twin system.

[0027] According to the prediction error convergence degree, calculate the short-term influence index and the long-term influence index respectively. The short-term influence index is used to evaluate the influence of the echo effect on the operation state of the water conservancy project in the current regulation cycle, and the long-term influence index is used to evaluate the cumulative influence of the echo effect on the accuracy of subsequent scheduling decisions. In order to comprehensively evaluate the influence of the data echo effect on the system operation, the present invention proposes to calculate the short-term influence index and the long-term influence index to quantify the influence degree of the data echo effect in different time dimensions. The short-term influence index mainly focuses on the current regulation cycle, that is, to evaluate the stability of the system operation state within a short time range. If the short-term influence index is high, it indicates that within the current regulation cycle, the data echo effect has a greater impact on the system prediction and control, and it is necessary to quickly adjust the data fusion method or optimize the model parameters. The long-term influence index is used to evaluate the cumulative influence of the echo effect on the future operation of the system. If the long-term influence index is at a high level for a long time, it indicates that the data echo effect gradually accumulates during the system operation, which may lead to an increasing prediction error and ultimately affect the scheduling accuracy of the entire water conservancy project. The significance of this step is that through the dual-time scale evaluation, the system can not only identify sudden anomalies in a short time but also identify long-term trend deviations, providing comprehensive data support for subsequent optimization decisions.

[0028] According to the short-term impact index and the long-term impact index, determine the level of the current data echo effect. The echo effect level is divided into low, medium, and high levels according to the degree of impact. After calculating the short-term impact index and the long-term impact index, the system needs to classify the severity of the data echo effect in order to adopt corresponding optimization strategies. The echo effect level is divided into three levels: low, medium, and high. Among them, the low-level echo effect indicates that the data echo has little impact on the system prediction and control, and only needs to appropriately adjust the data fusion ratio to optimize; the medium-level echo effect indicates that the data echo effect has affected the stability of the system regulation, and data filtering or weight adjustment is required to reduce the system's dependence on the feedback data; the high-level echo effect indicates that the system prediction model has been severely disturbed and the error cannot converge normally, and more radical optimization measures must be taken, such as resetting the model input data source or using reliable external data to correct the model. The significance of this step is that the system can adopt targeted optimization measures according to the impact degree of the data echo effect to ensure the stability of the system operation and the reliability of decision-making.

[0029] According to the determined echo effect level, different data echo optimization strategies are adopted for processing. According to the level of the data echo effect, the system will adopt different optimization strategies to reduce the impact of the data echo on the system operation. In the state of low-level echo effect, the system can apply the adaptive Kalman filter algorithm to eliminate abnormal feedback data and fine-tune the data fusion ratio at the same time to reduce the interference of the data echo on the control decision. In the state of medium-level echo effect, the system will adjust the data fusion weight parameter to reduce the participation of the feedback data in the generation process of the control instruction, so that the leading role of the prediction model in the decision-making is enhanced, thereby reducing the interference of the data echo signal on the system. In the state of high-level echo effect, the system will take more thorough optimization measures, including suspending the automatic data fusion of the existing model, resetting the model input data source, and introducing a third-party reliable data source, such as unmanned aerial vehicle inspection data, satellite remote sensing data or manual inspection data, to calibrate the model prediction value to ensure that the system can still accurately reflect the actual operation state of the water conservancy project under extreme conditions. The significance of this step is that for different levels of data echo effect problems, the system can select the most appropriate optimization strategy to minimize the impact of the data echo on the safety of the water conservancy project regulation, and at the same time ensure the adaptive ability and long-term stable operation of the digital twin system.

[0030] Judge whether there is a loop phenomenon that the control instruction data is transmitted back and participates in the system calculation again as input data to detect whether the data echo effect occurs. The specific operation steps are as follows: Step a, collect the control instruction data set U(t) generated by the digital twin system at time t. U(t) includes control parameters u1, u2, u3, which correspond to the water gate opening, the flood discharge volume, and the gate opening and closing frequency respectively. Step b, record the monitoring data set S(t+Δt) fed back by the water conservancy project entity system at time t+Δt. S(t+Δt) includes the water level h, flow velocity v, and gate status θ corresponding to the control parameters u1, u2, and u3 respectively; Step c, calculate the similarity D1 between U(t) and S(t+Δt) through the dynamic time warping algorithm. The formula is: ; n is the total number of control parameter dimensions, which is also the total number of monitoring data dimensions, with a value of 3. max(|U|,|S|) is the normalization coefficient to ensure that D1 is dimensionless and eliminate the influence of parameter dimensions. i is the serial number of the parameter type; Step d, at time t+2Δt, calculate the similarity D2 between the control instruction data U'(t+2Δt) generated after inputting the feedback data S(t+Δt) into the digital twin system and the original U(t). The Pearson correlation coefficient R calculation formula is: ; and respectively represent the per-dimension means of each control parameter in U(t); Step e, when the product of D1 and D2 is greater than the set standard threshold, it is determined that there is a phenomenon of control instruction data backhaul, and it is recognized that a data echo effect occurs.

[0031] In the dynamic interaction process between the digital twin system and the water conservancy project entity system, the digital twin system will continuously correct and generate new control instructions according to the currently collected feedback data. However, when the system fails to effectively isolate the direct coupling relationship between the feedback data and the control instructions, it may cause the system to continuously use the feedback data to reversely affect the generation of control instructions, thereby forming an abnormal cycle between the control instruction data and the feedback data. This cyclic phenomenon, called the data echo effect, will cause the system prediction model to lose the true perception of the actual state of the physical system, make the control instructions and feedback data fall into a closed-loop self-reinforcement, and may seriously lead to the continuous accumulation of system prediction errors, the out-of-control of decision-making effects, and affect the safety and efficiency of water conservancy project scheduling.

[0032] By collecting the control instruction data generated by the digital twin system and the monitoring data fed back by the water conservancy project entity system at preset time intervals and using the time series similarity analysis method, it is possible to effectively detect whether there is a phenomenon of control instruction data being backhauled and re-input into the system. This detection method can not only timely discover the data echo effect, but also provide a judgment basis for subsequent control strategy adjustment and optimization, ensuring the independence and stability of the interaction between the digital twin system and the physical system.

[0033] In the present invention, in order to evaluate the correlation between control instruction data at different time nodes, the system needs to compare the similarity of control instruction data sets generated at different times. Specifically, when the system detects whether the feedback data affects the generation of subsequent control instruction data, it is necessary to analyze the variation law of control parameters generated at different time nodes.

[0034] The dimension-wise mean refers to calculating the average value of each type of control parameter in the control instruction data set separately. Since the control instruction data set usually contains multiple different types of parameters, such as the opening degree of the sluice gate, the flood discharge volume, and the opening and closing frequency of the gate, the dimensions and value ranges of these parameters are different, so it is impossible to directly calculate the overall mean value. In order to eliminate the interference between different parameter dimensions, it is necessary to conduct statistical analysis on each parameter separately.

[0035] Specifically, within the time window set by the system, control instruction data at multiple time points are collected. For the parameter of the opening degree of the sluice gate, the system extracts the values of this parameter at all time points, calculates the arithmetic mean of these values, and obtains the dimension-wise mean of the opening degree of the sluice gate. For the parameters of the flood discharge volume and the opening and closing frequency of the gate, the dimension-wise means are calculated respectively using the same method. In this way, each type of control parameter has an independent average value, reflecting the overall level or trend of this parameter within a specific time period.

[0036] In the process of calculating Pearson correlation, the dimension-wise mean plays a role in eliminating data offset and revealing the true correlation between variables. By subtracting the dimension-wise mean, the system can more accurately evaluate the linear correlation degree between control instruction data at different time nodes, avoiding affecting the correlation judgment due to numerical offset or different data scales.

[0037] The present invention can accurately detect whether the data echo effect occurs by comparing the temporal characteristics of control instruction data and feedback data, combining the dynamic time warping algorithm and Pearson correlation analysis. The dynamic time warping algorithm ensures the alignment and effectiveness of data comparison in different time dimensions, and through normalization processing and dimensionless calculation, it ensures the comparability of data of different parameter types. The Pearson correlation analysis makes the correlation determination between different control instruction data sets more objective and accurate by introducing the dimension-wise mean. When the system combines the analysis results of the similarity between control instruction data and feedback data and the results of Pearson correlation analysis, and the product of the two exceeds the set threshold, the system can accurately determine that there is a data echo effect. This judgment criterion takes into account the strong correlation between control and feedback and the feedback loop characteristics of the system itself, effectively avoiding misjudgment and ensuring the high reliability and security of the prediction and control decisions of the digital twin system.

[0038] The specific steps for calculating the short-term impact index include: Define a time window τ to evaluate the time range, which is divided into multiple equally spaced time nodes to form a short-term time series. At each time node, three parameter values predicted by the digital twin system are collected respectively, including the predicted water level value, the predicted flow velocity value, and the predicted gate opening and closing state value; at the same time, three monitored parameter values feedback by the physical water conservancy project system are collected, including the monitored water level value, the monitored flow velocity value, and the monitored gate opening and closing state value. At each time node, subtract the monitored water level value from the predicted water level value and calculate the absolute value of the difference; then divide the absolute value by the monitored water level value plus a very small positive number to avoid division by zero error; use the same method to calculate the errors of the predicted flow velocity value and the predicted gate opening and closing state value respectively. For each time node, set three proportionality coefficients for the water level, flow velocity, and gate state, which respectively reflect their importance to the overall system safety or performance. Weightedly average the water level error, flow velocity error, and gate state error according to the proportionality coefficients to obtain the comprehensive error value of this time node. Within the entire time window τ, calculate the average value of the comprehensive error values of all time nodes. Then, for the comprehensive error value of each time node, calculate the square of the difference between it and the average value. Sum up all the squared differences and divide by the total number of time nodes to obtain the error fluctuation range within the short-term time period. Finally, multiply the error fluctuation range by a preset sensitivity adjustment coefficient to obtain the short-term impact index.

[0039] In the present invention, the calculation method of the short-term impact index dynamically evaluates the deviation volatility between the predicted values of the digital twin system and the feedback values of the physical water conservancy project within the set time window, effectively reflecting the impact of the data echo effect on the operation state of the water conservancy project in a short period. Specifically, the gate opening, flood discharge volume, and gate opening and closing frequency in the control parameters respectively determine the water level, flow velocity, and gate state of the physical water conservancy project system. Among them, the gate opening directly affects the water level regulation in the reservoir area, the flood discharge volume determines the flow velocity change in the downstream river channel, and the gate opening and closing frequency reflects the dynamic state of the hydraulic structure operation and its execution frequency. Therefore, the digital twin system needs to focus on the real-time changes of the three core feedback parameters of water level, flow velocity, and gate state during the prediction control process. The short-term impact index calculates the water level prediction and monitoring error, flow velocity prediction and monitoring error, and gate state prediction and monitoring error respectively, and assigns different proportionality coefficients to these three errors for weighted average at each time node. The proportionality coefficients are assigned by the assignment method. According to the safety requirements and operation management rules of different water conservancy projects, different weight values are pre-assigned to the water level, flow velocity, and gate state to reflect their influence degrees in the system stability. In this way, the short-term impact index can quantify the impact of the data echo effect on the stability and reliability of the water conservancy project regulation system in a short time, providing a scientific basis for timely optimizing control decisions.

[0040] The specific steps for calculating the long-term impact index include: Define a long-term observation window T, where T is much larger than the short-term window τ. Divide it into multiple equally spaced time nodes to form a long-term time series. At each time node, collect three parameter values predicted by the digital twin system, including the water level prediction value, the flow velocity prediction value, and the gate opening and closing state prediction value. At the same time, collect three monitoring parameter values feedback by the physical water conservancy project system, including the water level monitoring value, the flow velocity monitoring value, and the gate opening and closing state monitoring value; At each time node, subtract the water level monitoring value from the water level prediction value and calculate the absolute value of the difference. Then divide this absolute value by the water level monitoring value plus a very small positive number to avoid division by zero errors. Use the same method to calculate the errors of the flow velocity prediction value and the gate opening and closing state prediction value respectively.

[0041] For each time node, set three proportionality coefficients for the water level, flow velocity, and gate state, respectively reflecting their influence degrees on the long-term stability of the overall system. Weight-average the water level error, flow velocity error, and gate state error according to the proportionality coefficients to obtain the comprehensive error value at this time node; Accumulate the comprehensive error values of all time nodes within the long-term time period and multiply by the time interval to obtain the comprehensive error integral value, which reflects the cumulative degree of errors within the long-term period; Divide the comprehensive error integral value by the product of the total time length of the long-term period and the set maximum allowable error value to make the long-term impact index a dimensionless value. Through normalization processing, ensure that this index can reflect the comparability between different water conservancy project systems, and finally obtain the long-term impact index.

[0042] In the present invention, the calculation of the long-term impact index continuously evaluates the cumulative error situation between the prediction values of the digital twin system and the feedback values of the physical water conservancy project system within the set long-term observation window, reflecting the influence of the data echo effect on the overall operation accuracy and regulation strategy reliability of the system on a longer time scale. The long-term impact index measures the continuous cumulative trend of the system prediction error in the form of error integral, can identify the long-term cumulative phenomenon of the prediction model deviation caused by the data echo effect, and then prompt potential risks such as possible decision-making failure and regulation imbalance in the system. This index eliminates the differences in scale, parameter dimensions, etc. between different water conservancy project systems through normalization processing, ensures the unity and comparability of the analysis results under different application scenarios, and provides a standardized evaluation basis for the governance of the data echo effect and regulation optimization across engineering systems.

[0043] The three proportionality coefficients used in the calculation process of the long-term impact index are basically the same as those used in the short-term impact index, and are respectively used to reflect the importance of the three parameters of water level, flow velocity, and gate status in the long-term stability analysis of the system. These proportionality coefficients are set by the assignment method and determined according to factors such as the management requirements of water conservancy projects, safety levels, and project operation characteristics. For example, in the flood control scheduling scenario, the proportionality coefficient of the water level parameter can be given a higher weight, while in the water supply scheduling or irrigation management, the proportionality coefficients of the flow velocity and gate status may be more prominent. Although the proportionality coefficients remain the same in the short-term and long-term impact indices, in specific applications, the set values of the long-term proportionality coefficients can also be adjusted individually according to different requirements of long-term operation risk assessment to enhance the sensitivity and response ability to the risk of long-term prediction error accumulation.

[0044] When determining the level of the data echo effect, it is obtained by reasoning using fuzzy logic based on the short-term impact index and the long-term impact index. The principle of using fuzzy logic is as follows: Obtain the short-term impact index and the long-term impact index corresponding to the occurrence of the data echo effect, and use them together as the input variables of the fuzzy logic. Take the level corresponding to the data echo effect as the output variable of the fuzzy logic. Perform fuzzy processing on the input variables, convert the values of the input variables into fuzzy sets, perform fuzzy processing on the output variable, convert the output variable into a fuzzy set, formulate fuzzy rules to describe the adaptation degree of different data echo effect levels corresponding to different combinations of data types, and perform reasoning on the fuzzy input variables through the fuzzy rules to obtain the level corresponding to the data echo effect.

[0045] First, perform fuzzy processing on the two input variables of the short-term impact index and the long-term impact index, divide their value ranges into several fuzzy linguistic value intervals, such as low, medium, and high. Each interval corresponds to a certain membership function to reflect the membership relationship of the current input data in different fuzzy intervals. At the same time, the system also performs fuzzy processing on the output level of the data echo effect, sets it as low level, medium level, high level, etc., and respectively establishes fuzzy sets of the output variables. The system formulates a fuzzy rule base according to the water conservancy project regulation logic and expert experience knowledge. This rule base describes the level of the data echo effect that should be determined under different combinations of short-term and long-term impact indices. For example, when the short-term impact index is high and the long-term impact index is also high, it is determined as a high-level echo effect; when the short-term impact index is low and the long-term impact index is medium, it is determined as a medium-level echo effect; if both are low, it is determined as a low-level echo effect.

[0046] In the reasoning stage, the system matches the fuzzy rules according to the fuzzy set membership corresponding to the current input data, and combines the reasoning engine to conduct a comprehensive analysis of all valid rules, and finally obtains the fuzzy output of the data echo effect level. Finally, through the defuzzification process, the fuzzy output variables are converted into clear level judgment results to guide the system to adopt corresponding optimization strategies. This fuzzy logic reasoning process fully reflects the intelligent characteristics of multi-factor comprehensive evaluation, effectively improves the system's judgment ability and response accuracy for complex echo effect phenomena, and ensures the safety and stability of water conservancy project regulation.

[0047] The short-term impact index indicates the immediate impact of the data echo effect on the current operating status of the physical system of the water conservancy project within a relatively short time frame. The index dynamically analyzes the error fluctuation amplitude between the prediction value of the digital twin system and the feedback value of the physical system within a short time window, reflecting the prediction accuracy of the system and the stability of the execution effect of the control command in a short period of time. The short-term impact index mainly measures whether the data echo effect causes abnormal fluctuations in the system state during the current regulation cycle, and reveals whether the system has risk problems such as increased instantaneous prediction deviation, unstable control, or command failure due to data echo. It is mainly used to assist in the rapid diagnosis and immediate adjustment of the current operating status during the regulation of water conservancy projects, ensuring the safety and effectiveness of the system in a short period of time.

[0048] The long-term impact index represents the cumulative impact of the data echo effect on the overall prediction capability of the digital twin system and the accuracy of water conservancy project scheduling decisions over a longer time scale. The index cumulatively analyzes the comprehensive errors between the system prediction value and the feedback value within the long-term observation window, and integrates these errors to reflect whether the prediction model has a trend risk of deviating from the actual state due to the data echo effect in the long-term operation. The long-term impact index mainly measures the degree of accumulation of system prediction errors and whether the long-term control strategy leads to systematic instability or continued expansion of decision-making deviations due to the echo effect. It is mainly used to guide the long-term operation optimization, model correction and decision parameter adjustment of the water conservancy project system, and to ensure the overall stability and control reliability of the system during long-term operation.

[0049] Different data echo optimization strategies are used for processing, including: in the low-level echo effect state, applying data filtering algorithm to eliminate abnormal feedback data; in the intermediate echo effect state, adjusting the data fusion weight parameters to reduce the model's dependence on feedback data; in the high-level echo effect state, resetting the model input data source, forcibly introducing external trusted monitoring data source for verification, and rebuilding the digital twin model state to completely eliminate the echo effect.

[0050] In the state of the high - level echo effect, the state reconstruction of the digital twin model is achieved by introducing a third - party trusted data source. The trusted data source includes data from drone inspections, satellite remote sensing, and manual inspections. The reconstruction process includes: pausing the operation of the existing model, deleting the current model feedback link information, importing trusted data to initialize the modeling parameters, generating a new digital twin model, and performing model validity tests using historical verification data.

[0051] In the state of the low - level echo effect, the filtering algorithm adopted is the adaptive Kalman filter. In the state of the low - level echo effect, using the adaptive Kalman filter algorithm can effectively identify and eliminate abnormal feedback data caused by data echo. The adaptive Kalman filter dynamically adjusts the prediction error covariance and measurement error covariance in the system state estimation, enabling the filter to automatically optimize the filtering parameters according to the real - time changes of the feedback data. When the feedback data shows short - term anomalies or slight deviations due to the low - level echo effect, the adaptive Kalman filter can utilize the error dynamic balance mechanism between the system prediction value and the observed value to automatically reduce the trust level of the abnormal feedback data, suppress the impact of the data echo effect on the system state estimation, thereby ensuring that the control instructions output by the digital twin model remain stable and reliable, and avoiding system control failures caused by small - scale echo interference.

[0052] In the state of the medium - level echo effect, weight optimization and adjustment are carried out, specifically as follows: Normalize the short - term impact index and the long - term impact index respectively, and convert them into percentage expressions; Calculate the average value of the short - term impact index and the long - term impact index as the comprehensive echo effect degree; Preset an adjustment coefficient for determining the decrease amplitude of the feedback data fusion weight. Multiply the comprehensive echo effect degree by the preset adjustment coefficient to obtain a decrease value. Subtract the decrease value from the initial fusion weight of the feedback data, and the result is used as the new feedback data fusion weight; The fusion weight of the prediction data is equal to one minus the new feedback data fusion weight.

[0053] The value range of the adjustment coefficient is usually set or dynamically optimized according to the operating characteristics of the water conservancy project system, the sensitivity of the digital twin model, and historical operation data. Generally speaking, the value range of the adjustment coefficient is set between zero and one, where zero means no adjustment of the fusion weight, and one means that when the echo effect reaches the limit state, the system is allowed to reduce the fusion weight of the feedback data by the maximum amplitude to strongly suppress the echo effect. It can be determined in the following ways: one is based on the historical data analysis method. By statistically analyzing the historical operation data of the system, the response ability and regulation effect of the system under different echo effect levels are evaluated to form an empirical adjustment coefficient interval; the other is through the expert knowledge base method. Combining the judgments of experts in the field of water conservancy projects on the system regulation risk and the trust degree of feedback data, a weight adjustment strategy and adjustment coefficient range that meet the actual project requirements are set.

[0054] In the state of medium echo effect, by optimizing and adjusting the fusion weights of the feedback data and the prediction data, the dependence of the system on the echo effect source can be effectively reduced, and the further diffusion and accumulation of the echo effect can be suppressed. Dynamically calculate the decrease amplitude of the fusion weight of the feedback data to achieve flexible control of the trust degree of the feedback data. By reducing the proportion of the feedback data in the fusion process, the decision-making leading role of the prediction model itself is enhanced, the interference of the echo data on the prediction and control accuracy of the system is effectively reduced, and at the same time, the autonomy and robustness of the model are maintained, so as to ensure the stable operation and regulation safety of the digital twin system in the state of medium echo effect.

[0055] Example 2: In the real-time operation monitoring of a large reservoir, the digital twin system is used to predict the water level change of the reservoir and adjust the opening of the sluice. In the normal operation state, the system relies on feedback data at a relatively high proportion for real-time regulation. However, during the rainstorm, the inflow and water level of the reservoir fluctuate violently, and the feedback data changes rapidly. The system detects that the similarity between the control command data and the feedback data abnormally increases. By comparing the time series characteristics of the control command data and the feedback data through the dynamic time warping algorithm, it is found that there is a data echo effect. The system starts the echo effect detection mechanism and calculates the short-term impact index and the long-term impact index based on the prediction error convergence degree. According to the index results, the system determines that the current echo effect level is medium, and then reduces the fusion weight of the feedback data in the process of generating the control command, and enhances the leading role of the prediction model in decision-making. By adjusting the data fusion ratio, the misjudgment risk caused by the feedback data cycle is successfully reduced, and the safety and accuracy of the reservoir flood discharge regulation are guaranteed.

[0056] Example 3: In a multifunctional water conservancy project, the digital twin system monitors the opening and closing status of the sluice gate and the downstream flow regulation effect for a long time. After the system has been running for a period of time, by analyzing the short-term impact index and the long-term impact index, it is found that the long-term prediction error has a continuous poor convergence phenomenon. After detection, the Pearson correlation between the control instruction data and the feedback monitoring data is abnormally increased, and the product judgment value exceeds the set threshold, confirming that the system has entered the advanced echo effect state. In order to avoid the formation of a continuous self-excitation cycle between the model and the feedback system, the system automatically executes the advanced echo effect optimization strategy. The system suspends the existing model feedback link, interrupts the feedback data input, and temporarily introduces a third-party trusted data source for status verification, including drone inspection images, remote sensing water surface change monitoring data, and data obtained from manual inspections. The state of the digital twin model is reconstructed through these external data, and the prediction parameters are calibrated with the assistance of historical verification data, and the normal operation of the model is finally restored, effectively eliminating the echo effect.

[0057] Example 4: In a channel irrigation management system in a certain area, a digital twin model is used to predict and control irrigation water flow to ensure efficient use of farmland water resources. When the system enters the flood season, it continuously monitors the error change trend between the model and the actual feedback data through the short-term impact index and the long-term impact index. The detection found that the short-term impact index increased suddenly, and the long-term impact index remained at a medium level. The system judged that the current echo effect level was intermediate based on fuzzy logic reasoning, so it started the echo effect optimization process. First, the system dynamically adjusts the data fusion weight parameters to reduce the participation of feedback data in predictive control decisions, and adjusts the fusion proportion to be dominated by the prediction model. Secondly, adaptive Kalman filtering is performed on the real-time feedback data to remove abnormal feedback data points detected to prevent them from affecting control decisions. This optimization process ensures that the channel control system can operate stably and efficiently in the face of sudden changes in water conditions, avoiding confusion in water resource control caused by data echo effects.

[0058] The above three implementation methods respectively describe in detail the technical path and actual operation process of the present invention for solving the problem of data echo effect from different water conservancy engineering scenarios such as reservoir scheduling, hub project management and channel irrigation control, highlighting the application value and innovation of the invention solution.

[0059] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0060] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0061] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0062] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described devices and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0063] As described above, the above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claimed rights.

Claims

1. A method for intelligent monitoring of water conservancy project data based on digital twins, characterized in that: The following steps are involved: During the operation of the digital twin system, the control instruction data used to feedback and regulate the operating parameters of the water conservancy project is obtained, and the real-time monitoring data of the physical water conservancy project is collected simultaneously; By comparing the time series characteristics and change trends of the control command data and the feedback data of the physical water conservancy project, it is determined whether there is a cycle phenomenon in which the control command data is transmitted back and re-participates in the calculation of the digital twin system as input data, so as to detect whether the data echo effect occurs; When the data echo effect is detected, a comprehensive judgment is made based on the preset test parameters. The test parameters are the prediction error convergence of the digital twin model, which is obtained by calculating the dynamic deviation between the current prediction value and the real-time monitoring value. According to the convergence degree of prediction error, the short-term impact index and long-term impact index are calculated respectively. The short-term impact index is used to evaluate the impact of the echo effect on the operation status of the water conservancy project in the current regulation cycle, and the long-term impact index is used to evaluate the cumulative impact of the echo effect on the accuracy of subsequent scheduling decisions. According to the short-term impact index and the long-term impact index, the level of the current data echo effect is determined. The echo effect level is divided into low, medium and high levels according to the degree of impact. Different data echo optimization strategies are used for processing according to the determined echo effect level.

2. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 1 is characterized in that: Determine whether there is a loop phenomenon in which the control instruction data is transmitted back and re-participates in the system calculation as input data to detect whether the data echo effect occurs, specifically including the following steps: Step a, collecting the control instruction data set U(t) generated by the digital twin system at time t, where U(t) includes control parameters u1, u2, and u3, which correspond to the sluice opening, flood discharge, and gate opening and closing frequency, respectively; Step b, recording the monitoring data set S(t+Δt) fed back by the physical system of the water conservancy project at time t+Δt, where S(t+Δt) includes the water level h, flow velocity v and gate state θ corresponding to the control parameters u1, u2 and u3 respectively; Step c: Calculate the similarity D1 between U(t) and S(t+Δt) by using the dynamic time warping algorithm. The formula is: ; n is the total number of control parameter dimensions, which is also the total number of monitoring data dimensions, and its value is 3. max(|U|,|S|) is the normalization coefficient, which ensures that D1 is dimensionless and eliminates the influence of parameter dimensions. i is the number of parameter types. Step d: At time t+2Δt, the feedback data S(t+Δt) is input into the digital twin system to generate the control instruction data U'(t+2Δt), and the similarity D2 between U'(t+2Δt) and the original U(t) is calculated using the Pearson correlation coefficient R calculation formula: ; and They represent the dimension-by-dimensional means of each control parameter in U(t); Step e: when the product of D1 and D2 is greater than the set standard threshold, it is determined that there is a control instruction data feedback phenomenon and that a data echo effect occurs.

3. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 2 is characterized in that: The specific steps for calculating the short-term impact index include: Define the time window τ evaluation time range and divide it into multiple equally spaced time nodes to form a short-term time series. At each time node, collect the three parameter values ​​predicted by the digital twin system, including the water level prediction value, flow velocity prediction value and gate opening and closing state prediction value; at the same time, collect the three monitoring parameter values ​​fed back by the physical water conservancy project system, including the water level monitoring value, flow velocity monitoring value and gate opening and closing state monitoring value; At each time point, the water level prediction value is subtracted from the water level monitoring value, and the absolute value of the difference is calculated; then the absolute value is divided by the water level monitoring value plus a very small positive number to avoid zero division errors; the same method is used to calculate the errors of the flow velocity prediction value and the gate opening and closing state prediction value respectively; For each time node, three proportional coefficients of water level, flow rate and gate status are set to reflect their importance to the overall system safety or performance. The water level error, flow rate error and gate status error are weighted averaged according to the proportional coefficients to obtain the comprehensive error value of the time node. Within the entire time window τ, the average value of the comprehensive error values ​​of all time nodes is calculated, and then the square of the difference between the comprehensive error value of each time node and the average value is calculated. After summing up all the squares of the differences, the sum is divided by the total number of time nodes to obtain the error fluctuation amplitude within the short-term time period. Finally, the error fluctuation amplitude is multiplied by the preset sensitivity adjustment coefficient to obtain the short-term impact index.

4. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 3 is characterized in that: The specific steps for calculating the long-term impact index include: Define the long-term observation window T, divide it into multiple equally spaced time nodes to form a long-term time series. At each time node, collect the three parameter values ​​predicted by the digital twin system, including the water level prediction value, flow velocity prediction value and gate opening and closing state prediction value. At the same time, collect the three monitoring parameter values ​​fed back by the physical water conservancy project system, including the water level monitoring value, flow velocity monitoring value and gate opening and closing state monitoring value. At each time point, the water level prediction value is subtracted from the water level monitoring value, and the absolute value of the difference is calculated; then the absolute value is divided by the water level monitoring value plus a very small positive number to avoid zero division errors. The same method is used to calculate the errors of the flow velocity prediction value and the gate opening and closing state prediction value; For each time node, three proportional coefficients of water level, flow rate and gate status are set to reflect their influence on the long-term stability of the overall system. The water level error, flow rate error and gate status error are weighted averaged according to the proportional coefficients to obtain the comprehensive error value of the time node. The comprehensive error values ​​of all time nodes in the long-term time period are accumulated and multiplied by the time interval to obtain the comprehensive error integral value, which reflects the degree of error accumulation in the long-term period; The integrated error integral value is divided by the product of the total time length of the long-term period and the set maximum allowable error value, so that the long-term impact index is a dimensionless value. Through normalization processing, it is ensured that the index can reflect the comparability between different water conservancy project systems, and finally the long-term impact index is obtained.

5. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 4 is characterized in that: When determining the level of data echo effect, fuzzy logic is used to infer based on the short-term impact index and the long-term impact index.

6. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 5 is characterized in that: The principle of fuzzy logic is: The short-term impact index and long-term impact index corresponding to the occurrence of data echo effect are obtained and used as the input variables of fuzzy logic together. The corresponding level of data echo effect is used as the output variable of fuzzy logic. The input variables are fuzzified and the values ​​of the input variables are converted into fuzzy sets. The output variables are fuzzified and the output variables are converted into fuzzy sets. Fuzzy rules are formulated to describe the degree of adaptation of the corresponding levels of different data echo effects under different data type combinations. The fuzzified input variables are inferred through fuzzy rules to obtain the corresponding level of data echo effect.

7. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 6 is characterized in that: Different data echo optimization strategies are used for processing, including: in the low-level echo effect state, applying data filtering algorithm to eliminate abnormal feedback data; in the intermediate echo effect state, adjusting the data fusion weight parameters to reduce the model's dependence on feedback data; in the high-level echo effect state, resetting the model input data source, forcibly introducing external trusted monitoring data source for verification, and rebuilding the digital twin model state to completely eliminate the echo effect.

8. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 7 is characterized in that: At the advanced echo effect level, the digital twin model state is reconstructed by introducing a third-party trusted data source. The trusted data source includes data from drone inspections, satellite remote sensing, and manual inspections. The reconstruction process includes: pausing the existing model, deleting the current model feedback link information, importing trusted data to initialize modeling parameters, generating a new digital twin model, and verifying the model validity through historical verification data.

9. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 8 is characterized in that: In the low-level echo effect state, the filtering algorithm used is adaptive Kalman filtering.

10. The method for intelligent monitoring of water conservancy project data based on digital twins according to claim 9 is characterized in that: In the intermediate echo effect state, weight optimization adjustment is performed, specifically: Normalize the short-term impact index and the long-term impact index respectively and convert them into percentage expressions; Calculate the average of the short-term impact index and the long-term impact index as the comprehensive echo effect level; An adjustment coefficient for determining a decrease in the feedback data fusion weight is preset, and the comprehensive echo effect degree is multiplied by the preset adjustment coefficient to obtain a decrease value, and the decrease value is subtracted from the initial fusion weight of the feedback data, and the result is used as a new feedback data fusion weight; The fusion weight of the predicted data is equal to one minus the fusion weight of the new feedback data.

Images