Intelligent decision-making system for identification of dynamic stress sensitive areas and efficiency compensation of hydropower units

Abstract

The present invention discloses an intelligent decision-making system for identifying dynamic stress sensitive areas and compensating for efficiency of hydropower units, which belongs to the field of intelligent monitoring and optimization control of hydropower equipment. The system obtains operating status data, environmental data and task instruction data of key parts of the hydropower unit in real time through a data acquisition module; the sensitive area identification and analysis module uses time-frequency domain analysis and sensitivity coefficient calculation to identify dynamic stress sensitive areas; the risk classification module determines the operating risk interval under different operating parameter combinations based on cluster analysis and outlier detection; the execution decision module combines the operating risk value and power generation efficiency to screen the optimal operating parameter group; the efficiency compensation module generates a compensation priority sequence based on the sensitivity coefficient, monitors and triggers the compensation mechanism in real time; the present invention realizes the intelligent operation and management of hydropower units, significantly improves the safety, power generation efficiency and economic benefits of the equipment, and provides new ideas and methods for technological progress and sustainable development of the hydropower industry.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent monitoring and optimization control of hydropower equipment, and particularly relates to an intelligent decision-making system for dynamic stress sensitive area identification and efficiency compensation of hydropower units. Background Art

[0002] Hydropower units are the core equipment of hydropower stations, used to convert water energy into electricity. During operation, their key parts are subject to dynamic stress, which can easily form stress-sensitive areas, leading to fatigue damage and cracks, affecting the unit's reliability and lifespan. Operating conditions are complex and changeable, requiring improved power generation efficiency while ensuring safety. However, existing technologies suffer from the following technical problems:

[0003] Inaccurate identification of dynamic stress sensitive areas: Existing methods have difficulty in accurately identifying stress concentration areas and sensitivity levels;

[0004] Inadequate operational risk assessment: There is a lack of comprehensive means to assess operational risks in stress-sensitive areas, making it impossible to accurately divide risk intervals;

[0005] Difficulty in optimizing operating parameters: There is a lack of scientific and reasonable decision-making methods to balance safety and efficiency, making it difficult to select the optimal operating parameter combination;

[0006] Lack of dynamic compensation mechanism: The existing technology lacks a mechanism to dynamically adjust operating parameters based on real-time monitoring data to compensate for risks and improve efficiency. To this end, we propose an intelligent decision-making system for identifying dynamic stress sensitive areas and compensating for efficiency of hydropower units. Summary of the Invention

[0007] The purpose of the present invention is to provide an intelligent decision-making system for identifying dynamic stress sensitive areas and compensating for efficiency of hydropower units, so as to solve the problems raised in the above-mentioned background technology.

[0008] To achieve the above-mentioned objectives, the present invention provides the following technical solution: an intelligent decision-making system for identifying dynamic stress sensitive areas and compensating for efficiency of hydropower units, comprising:

[0009] Data acquisition module: collects operating status data of key parts of the hydropower unit, as well as environmental data and task instruction data;

[0010] Sensitive area identification and analysis module: Analyzes the stress distribution of key parts and determines the stress concentration area; performs time-frequency domain analysis on the dynamic stress data of the stress concentration area to obtain the sensitivity coefficient, and identifies the stress sensitive area based on it;

[0011] Risk classification module: Analyzes the correlation between dynamic stress and various operating parameters to determine analysis parameters; draws a scatter plot of dynamic stress and analysis parameters, performs cluster analysis and outlier detection on the data points in the scatter plot, and determines the operating risk range of stress-sensitive areas under different operating parameter combinations;

[0012] Execution decision module: Calculates interval risk values ​​for low, medium, and high risk intervals, builds an operating condition database to screen preliminary operating condition parameter groups; analyzes the operating condition risk values ​​of the preliminary operating condition parameter groups, calculates the selection value based on power generation efficiency, and determines the implementation operating condition parameter group accordingly;

[0013] Efficiency compensation module: Analyzes the sensitivity coefficients of operating parameters to operating risk values ​​and power generation efficiency, and generates priority sequences for risk compensation and efficiency compensation accordingly; monitors the operating risk values ​​and efficiency values ​​of the operating parameter group and triggers the corresponding compensation mechanism.

[0014] Preferably, the specific implementation content of the data acquisition module is:

[0015] Strain gauge sensors and acceleration sensors are installed at key locations of the hydropower unit to collect real-time operating status data of these key locations, including dynamic stress and vibration data. Key locations include the runner blade root, guide vane pivot, main shaft bearing, main shaft bearing seat, volute inlet and outlet, and top cover support points.

[0016] Monitor the environmental data of the upstream water of the hydropower unit through water level sensors and flow sensors. The environmental data includes: water level and flow information;

[0017] By connecting with the power grid dispatching system, the total power generation task instruction data for the current period can be obtained.

[0018] Preferably, the specific process of the sensitive area identification and analysis module is:

[0019] Import the 3D solid model of the hydropower unit into the finite element analysis software for fluid-solid-thermal multi-physics field coupling simulation; set several typical operating conditions; calculate the VonMises stress distribution of each key part under these typical operating conditions, extract areas with a stress concentration factor ≥ K1, and mark them as stress concentration areas, where K1 is the critical value of the set stress concentration factor;

[0020] For each stress concentration area, the dynamic stress data of the stress concentration area is collected in real time, and the collected dynamic stress data is analyzed in the time-frequency domain, specifically:

[0021] Time Domain Analysis: Calculating Mean Stress , standard deviation , peak factor CF;

[0022] Frequency domain analysis: Obtain the main frequency component fdom through FFT transformation, and identify whether the main frequency component has resonance related to the rotation frequency and blade frequency;

[0023] Using the formula: , and obtain the sensitivity coefficient S; where SCF is the stress concentration coefficient of the stress concentration area, is the resonance index. If fdom coincides with the integral multiple of the rotation frequency and blade frequency, Take 1, otherwise take 0; w1, w2, w3 are preset weight coefficients;

[0024] A threshold value S1 of the sensitive region coefficient is preset. If the sensitivity coefficient S of the stress concentration region is greater than or equal to S1, the stress concentration region is marked as a stress sensitive region.

[0025] Preferably, the risk classification module determines the analysis parameters; and the specific process of drawing the dynamic stress-analysis parameter scatter plot is:

[0026] Collect the historical dynamic stress data of the stress-sensitive areas of the hydropower units in the last three months, as well as the corresponding operating condition data;

[0027] For each stress-sensitive area, the Pearson correlation coefficient method is used to calculate the correlation coefficient between dynamic stress and various working condition parameters. A correlation threshold is set, and the working condition parameters with an absolute value of the correlation coefficient greater than the corresponding threshold are extracted and recorded as analysis parameters.

[0028] For each analysis parameter, a coordinate system is established with dynamic stress as the ordinate and analysis parameter as the abscissa, and a scatter plot is drawn based on the historical dynamic stress data and the corresponding working condition data to obtain a stress-parameter scatter plot.

[0029] Preferably, the process of the risk classification module performing cluster analysis on the data points in the scatter plot to determine the operating risk interval is as follows:

[0030] The data in the stress-parameter scatter plot are normalized and mapped to the interval {0, 1}. The DBSCAN clustering algorithm is used to cluster the stress-parameter scatter plot to obtain several data clusters.

[0031] Calculate the number of data points in each data cluster to obtain the cluster point count, preset the cluster point count threshold, and if the cluster point count is greater than the corresponding threshold, calculate the dynamic stress standard deviation and the standard deviation of the analysis parameters corresponding to the data points in the data cluster; set the dynamic stress dispersion threshold and the analysis parameter dispersion threshold;

[0032] If the standard deviation of the dynamic stress within the cluster is less than the dynamic stress dispersion threshold, and the standard deviation of the analysis parameter is less than the analysis parameter dispersion threshold, then the low-risk area is judged;

[0033] The mean stress value of the stress-sensitive area during safe operation in the last three months is calculated and recorded as the safe stress mean value. At the same time, the mean dynamic stress value corresponding to the data points in the data cluster is calculated and recorded as the cluster stress mean value. The deviation between the cluster stress mean value and the safe stress mean value is calculated to obtain the cluster stress deviation value. A cluster stress deviation threshold value is preset. If the cluster stress deviation value is less than or equal to the corresponding threshold value, the area covered by the data cluster is determined to be a low-risk area.

[0034] Preferably, the risk classification module performs outlier detection on the data points in the scatter plot and determines the operation risk interval in the following process:

[0035] Use the LOF outlier detection algorithm to calculate the outlier factor of each data point and preset the outlier factor threshold; determine the data point whose outlier factor is greater than the corresponding threshold as a preliminary outlier;

[0036] For each preliminary outlier point, calculate the absolute difference between its dynamic stress value and the mean of all dynamic stresses in the scatter plot; if the difference is greater than or equal to K times the standard deviation of the dynamic stress of the scatter plot, then the point is confirmed as an actual outlier; where K is a preset coefficient;

[0037] Draw a circle with a preset radius centered on each actual outlier, and mark the area corresponding to the data points contained in the circle as a high-risk area;

[0038] After excluding the low-risk and high-risk intervals, the remaining middle area is the medium-risk interval.

[0039] Preferably, the specific process of executing the decision module to analyze the interval risk values ​​of the low, medium and high risk intervals is:

[0040] Calculate the mean dynamic stress corresponding to each data point in the low-risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , get the risk value of the low risk interval ; Where a1 and a2 are preset weight coefficients;

[0041] Calculate the mean dynamic stress corresponding to each data point in the risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , and get the risk value in the medium risk interval ; Where b1 and b2 are preset weight coefficients;

[0042] Calculate the mean dynamic stress corresponding to each data point in the high-risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , get the risk value of the high-risk interval ; Where c1 and c2 are preset weight coefficients.

[0043] Preferably, the decision module is executed to build a working condition database to screen a preliminary working condition parameter group; the working condition risk value of the preliminary working condition parameter group is analyzed, and the selection value is calculated in combination with the power generation efficiency, and the specific process of determining the implementation working condition parameter group is as follows:

[0044] Establishing an operating condition database, which stores selectable operating condition parameter groups corresponding to different water heads, flows, and total power generation, and each operating condition parameter group corresponds to a power generation efficiency XL;

[0045] Obtain the water head, flow rate, and total power generation corresponding to the current period, substitute them into the operating condition database, and determine all feasible operating condition parameter groups through database query operations, which are recorded as the primary operating condition parameter groups;

[0046] For each operating parameter group in the preliminary selected operating parameter group, substitute each operating parameter in the operating parameter group into the corresponding stress-parameter scatter plot, and extract the corresponding interval risk value according to the risk interval it falls into;

[0047] Assign different weight coefficients to the interval risk value corresponding to each working condition parameter, then multiply the interval risk value corresponding to each working condition parameter by the corresponding weight coefficient and add them together to obtain the working condition risk value GF;

[0048] After normalizing the operating risk value and power generation efficiency corresponding to the operating parameter group, the operating condition selection value XPZ is obtained using the formula: XPZ=XL×d1-GF×d2, where d1 and d2 are preset weight coefficients;

[0049] The operating condition parameter group with the largest operating condition selection value in the preliminary operating condition parameter group is selected as the implementation operating condition parameter group for the current period.

[0050] Preferably, the efficiency compensation module analyzes the sensitivity coefficients of the operating condition parameters to the operating condition risk value and the power generation efficiency, and generates the risk compensation and efficiency compensation priority sequence accordingly. The specific process is as follows:

[0051] For each operating parameter, calculate the average impact of each 1% change in the operating parameter p on the operating risk value using the formula: , and obtain the risk sensitivity coefficient ; Where i = 1, 2, ... n; n is the number of historical samples; 、 The change in risk value of the i-th sample and the initial value; 、 is the change of the corresponding working condition parameter and its initial value;

[0052] Using the same method of calculating risk sensitivity, we can calculate the average impact of each 1% change in operating parameters on the power generation efficiency XL and obtain the efficiency sensitivity coefficient. ;

[0053] Define the compensation priority of the working condition parameters. The specific process is as follows:

[0054] Risk compensation: take Get the working parameters and follow Arrange the sizes of in descending order to obtain the risk compensation priority sequence; To preset risk sensitivity thresholds;

[0055] Efficiency compensation: take Get the working parameters and follow The size of is sorted in descending order to obtain the efficiency compensation priority sequence; is the preset efficiency sensitivity threshold.

[0056] Preferably, the efficiency compensation module monitors the operating risk value and efficiency value of the operating parameter group in real time and triggers the corresponding compensation mechanism. The specific process is as follows:

[0057] Real-time monitoring of the operating condition risk value GF and power generation efficiency XL of the current operating condition parameter group; when the operating condition risk value is greater than or equal to the preset operating condition risk threshold, select from the risk compensation priority sequence The highest operating condition parameter is used as the risk adjustment parameter;

[0058] If the risk adjustment parameter is positively correlated with the working condition risk value, the parameter is increased for compensation, otherwise it is reduced for compensation;

[0059] The magnitude of risk compensation is, using the formula: , we get the risk compensation margin FBC, where is the preset working condition risk threshold, is the current value of the working condition parameter, is the preset risk compensation coefficient, with a value between (0, 1);

[0060] When the power generation efficiency is less than or equal to the preset power generation efficiency threshold, select The maximum operating condition parameter is adjusted as the efficiency adjustment parameter;

[0061] If the efficiency adjustment parameter is positively correlated with the operating condition risk value, the parameter is increased for compensation, otherwise it is reduced for compensation;

[0062] The magnitude of the efficiency compensation is, using the formula: , and obtain the efficiency compensation amplitude XBC, where is the preset power generation efficiency threshold, is the preset efficiency compensation coefficient, with a value between (0, 1);

[0063] If risk compensation conflicts with efficiency compensation;

[0064] If safety compensation is preferred, select from the risk compensation priority sequence The maximum operating parameter is used as the risk adjustment parameter; from the efficiency compensation priority sequence, select The minimum operating condition parameter is adjusted as the efficiency adjustment parameter; and the corresponding amplitude is adjusted;

[0065] If efficiency compensation is prioritized, select The minimum operating condition parameter is used as the risk adjustment parameter; from the efficiency compensation priority sequence, select The maximum operating condition parameter is adjusted as the efficiency adjustment parameter; and the corresponding amplitude is adjusted;

[0066] If the emphasis is on the balance between safety and efficiency, then the risk compensation sequence is selected <0, and in the efficiency compensation sequence >0 operating parameters (pa, pb) for adjustment;

[0067] The adjustment range is: , we can get the compensation range FBC of risk adjustment parameter and XBC of efficiency adjustment parameter under the balance between safety and efficiency, where and are the initial values ​​of the risk adjustment parameter and the efficiency adjustment parameter under the balance between safety and efficiency; It is a preset small adjustment step.

[0068] Compared with the prior art, the present invention has the following beneficial effects:

[0069] (1) The intelligent decision-making system and system for dynamic stress sensitive area identification and efficiency compensation of hydropower units achieves accurate identification of dynamic stress sensitive areas of hydropower units by combining time-frequency domain analysis and sensitivity coefficient calculation; this method not only considers the static distribution of stress concentration areas, but also captures the dynamic change characteristics of dynamic stress, especially the influence of resonance on sensitivity, through dual analysis in time domain and frequency domain; it can more accurately identify areas prone to fatigue damage and cracks, so as to take measures in advance and effectively extend the service life of equipment;

[0070] In terms of risk assessment, the present invention adopts the DBSCAN clustering algorithm and the LOF outlier detection algorithm to conduct an in-depth analysis of the relationship between dynamic stress and operating parameters; the risk intervals are divided through cluster analysis, and high-risk abnormal operating conditions are identified in combination with outlier detection. This comprehensive assessment method can comprehensively and dynamically reflect the operating risks of the unit, provide scientific and accurate risk warnings for operating personnel, and significantly improve the safety and reliability of hydropower unit operation.

[0071] (2) The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of the hydropower unit can dynamically select the optimal combination of operating parameters by comprehensively evaluating the operating risk value and power generation efficiency, thereby maximizing power generation efficiency while ensuring equipment safety. This intelligent decision-making mechanism not only considers the short-term operating efficiency of the equipment, but also takes into account the long-term health status of the equipment.

[0072] The present invention also introduces a dynamic compensation mechanism that can automatically adjust operating parameters based on real-time monitoring data; by calculating the sensitivity coefficients of operating parameters to risk values ​​and power generation efficiency, the system generates a risk compensation and efficiency compensation priority sequence, and automatically triggers the compensation mechanism according to preset strategies; this dynamic adjustment capability enables the hydropower unit to quickly respond to changes in operating status, effectively cope with complex and changeable operating conditions, and significantly improve the robustness and flexibility of the system.

[0073] (3) The intelligent decision-making system and system for dynamic stress sensitive area identification and efficiency compensation of the hydropower unit. The present invention reduces the dependence on manual experience and reduces the cost of manual intervention through intelligent decision-making and compensation mechanisms. The system can automatically complete the entire process from data collection, sensitive area identification, risk assessment to working condition optimization and dynamic compensation, thereby improving the efficiency and accuracy of operation management. This intelligent management method not only improves the operating efficiency of the equipment, but also reduces equipment failures and operating risks caused by human errors.

[0074] More importantly, the present invention significantly extends the service life of the equipment and reduces equipment maintenance costs by optimizing operating conditions and dynamic compensation mechanisms. At the same time, by maximizing power generation efficiency while ensuring relatively safe operation of the equipment, it achieves full utilization of water resources and improves the economic and environmental benefits of the hydropower station. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is a flowchart of the present invention. DETAILED DESCRIPTION

[0076] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0077] Example 1

[0078] See also Figure 1The present invention provides an intelligent decision-making system for identifying dynamic stress sensitive areas and compensating for efficiency of hydropower units, including: a data acquisition module, a sensitive area identification and analysis module, a risk classification module, an execution decision module, and an efficiency compensation module;

[0079] The data acquisition module is used to collect the operating status data of the key parts of the hydropower unit, as well as the environmental data and task instruction data. The specific process is as follows:

[0080] High-precision strain gauge sensors and acceleration sensors are installed at key locations of the hydropower unit to collect real-time operating status data of these key locations, including dynamic stress and vibration data. Key locations include the runner blade root, guide vane pivot, main shaft bearing, main shaft bearing seat, volute inlet and outlet, and top cover support points.

[0081] Monitor the environmental data of the upstream water of the hydropower unit through water level sensors and flow sensors. The environmental data includes: water level and flow information;

[0082] By connecting with the power grid dispatching system, the total power generation task instruction data for the current period can be obtained.

[0083] The sensitive area identification and analysis module is used to analyze the stress distribution of key parts and mark the stress concentration areas. It performs time-frequency domain analysis on the dynamic stress data of the stress concentration areas to obtain the sensitivity coefficients, and then identifies the stress sensitive areas based on them. The specific process is as follows:

[0084] Import the three-dimensional solid model of the hydropower unit into the finite element analysis software for fluid-solid-thermal multi-physics field coupling simulation; set several typical operating conditions, including rated operating conditions, extreme head operating conditions, and load mutation operating conditions; calculate the VonMises stress distribution of each key part under typical operating conditions, extract areas with stress concentration coefficient ≥ K1, and mark them as stress concentration areas, where K1 is the set critical value of the stress concentration coefficient; the stress concentration coefficient is obtained in the same way as in the prior art, and the method of obtaining the stress concentration coefficient is not described in detail;

[0085] For each stress concentration area, the dynamic stress data of the stress concentration area is collected in real time, and the collected dynamic stress data is analyzed in the time-frequency domain, specifically:

[0086] Time Domain Analysis: Calculating Mean Stress , standard deviation , peak factor CF;

[0087] Frequency domain analysis: Obtain the main frequency component fdom through FFT transformation, and identify whether the main frequency component has resonance related to the rotation frequency and blade frequency;

[0088] Using the formula: , and obtain the sensitivity coefficient S; where SCF is the stress concentration coefficient of the stress concentration area, is the resonance index. If fdom coincides with the integral multiple of the rotation frequency and blade frequency, Take 1, otherwise take 0; w1, w2, w3 are preset weight coefficients;

[0089] A threshold value S1 of the sensitive region coefficient is preset. If the sensitivity coefficient S of the stress concentration region is greater than or equal to S1, the stress concentration region is marked as a stress sensitive region.

[0090] Furthermore, the threshold value of the sensitive area coefficient can be obtained by obtaining the mean value and standard deviation of the sensitivity coefficient in the stress concentration area during safe operation and the mean value and standard deviation of the sensitivity coefficient when a fault occurs in the historical data, and then assigning different weight coefficients to the mean value of the sensitivity coefficient in safe operation and the mean value of the sensitivity coefficient when a fault occurs, and performing weighted calculation to obtain the sensitivity comprehensive mean value. ; Similarly calculate the standard deviation of the sensitivity ; and using the formula: , get the sensitive area coefficient threshold S1; where K2 is the confidence coefficient, the value is between (2-3);

[0091] The risk classification module analyzes the correlation between dynamic stress and various operating parameters to determine the analysis parameters. It then draws a scatter plot of dynamic stress and analysis parameters, performs cluster analysis and LOF outlier detection algorithm analysis on the data points in the scatter plot, and determines the operating risk range of the stress-sensitive area under different operating parameter combinations. The specific process is as follows:

[0092] Collect the historical dynamic stress data of the stress-sensitive areas of the hydropower units in the last three months, as well as the corresponding operating condition data; the operating condition data includes: load, head, flow, etc.;

[0093] For each stress-sensitive area, the Pearson correlation coefficient method is used to calculate the correlation coefficient between dynamic stress and various working condition parameters. A correlation threshold is set, and the working condition parameters with an absolute value of the correlation coefficient greater than the corresponding threshold are extracted and recorded as analysis parameters.

[0094] For each analysis parameter, a coordinate system is established with dynamic stress as the ordinate and analysis parameter as the abscissa. A scatter plot is drawn based on the historical dynamic stress data and the corresponding working condition data to obtain a stress-parameter scatter plot.

[0095] The data in the stress-parameter scatter plot are normalized and mapped to the interval {0, 1}. The DBSCAN clustering algorithm is used to cluster the stress-parameter scatter plot to obtain several data clusters.

[0096] Calculate the number of data points in each data cluster to obtain the cluster point count, preset the cluster point count threshold, and if the cluster point count is greater than the corresponding threshold, calculate the dynamic stress standard deviation and the standard deviation of the analysis parameters corresponding to the data points in the data cluster; set the dynamic stress dispersion threshold and the analysis parameter dispersion threshold;

[0097] If the standard deviation of the dynamic stress within the cluster is less than the dynamic stress dispersion threshold, and the standard deviation of the analysis parameter is less than the analysis parameter dispersion threshold, then the low-risk area is judged;

[0098] Calculate the mean stress value of the stress-sensitive area during safe operation in the last three months, which is recorded as the safe stress mean. Simultaneously, calculate the mean dynamic stress value corresponding to the data points in the data cluster, which is recorded as the cluster stress mean. Calculate the deviation between the cluster stress mean and the safe stress mean to obtain the cluster stress deviation value. Preset a cluster stress deviation threshold. If the cluster stress deviation value is less than or equal to the corresponding threshold, the area covered by the data cluster is determined to be a low-risk area.

[0099] It should be noted that the DBSCAN clustering algorithm is used to cluster the stress-parameter scatter plot to obtain several data clusters. This step is to perform preliminary classification of the data and cluster similar data points together.

[0100] If the number of cluster points corresponding to the data cluster is greater than the corresponding threshold, it means that the cluster is representative and not an isolated small cluster. This step filters out clusters that are too small and may not be representative, avoiding over-analysis of local abnormal or accidental data points;

[0101] When the standard deviation of the dynamic stress within a cluster is less than the dynamic stress dispersion threshold, and the standard deviation of the analysis parameter is less than the analysis parameter dispersion threshold, it means that the fluctuations of the dynamic stress and analysis parameters within this cluster are within an acceptable small range, which preliminarily indicates that the operating status is relatively stable, but it cannot be directly determined as a low-risk area.

[0102] Comparison with historical data: Compare the mean dynamic stress within the cluster with the mean dynamic stress during historical safe operation of the stress-sensitive area (set the allowable deviation value). If the deviation between the mean dynamic stress within the cluster and the mean dynamic stress during historical safe operation is within the allowable range, combined with the previous judgment on the degree of dispersion, it can be comprehensively judged whether the area covered by the data cluster is a low-risk area;

[0103] Use the LOF outlier detection algorithm to calculate the outlier factor of each data point and preset the outlier factor threshold; determine the data point whose outlier factor is greater than the corresponding threshold as a preliminary outlier;

[0104] For each preliminary outlier point, calculate the absolute difference between its dynamic stress value and the mean of all dynamic stresses in the scatter plot; if the difference is greater than or equal to K times the standard deviation of the dynamic stress of the scatter plot, then the point is confirmed as an actual outlier; where K is a preset coefficient;

[0105] Draw a circle with a preset radius centered on each actual outlier, and mark the area corresponding to the data points contained in the circle as a high-risk area;

[0106] After excluding the low-risk and high-risk intervals, the remaining middle area is the medium-risk interval.

[0107] It should be noted that by calculating the outlier factor and setting the threshold through the LOF algorithm, preliminary outliers can be quickly screened out. These points may correspond to sudden abnormalities, equipment failures, or extreme operating conditions in the unit operation;

[0108] Initial outlier identification may result in misjudgment, as high outlier factors for some data points may be due to data fluctuations rather than true abnormal operating conditions. By further calculating the absolute difference between the dynamic stress value of the initial outlier and the overall mean and comparing it with the standard deviation multiple, we can combine the statistical characteristics of the data to filter out actual outliers with significant anomalies from the initial outliers. This step effectively reduces the misjudgment rate and ensures that the subsequent high-risk intervals are more reliable.

[0109] Operating condition data that are close to actual outliers may also have higher operating risks. In the scatter plot, the operating parameter combinations corresponding to these points may cause the unit to be unstable, increasing the probability of equipment damage or failure.

[0110] The execution decision module analyzes the interval risk values ​​of low, medium, and high risk intervals, builds a working condition database to screen the preliminary working condition parameter group; analyzes the working condition risk value of the preliminary working condition parameter group, calculates the selection value based on the power generation efficiency, and determines the implementation working condition parameter group based on this. The specific process is as follows:

[0111] Calculate the mean dynamic stress corresponding to each data point in the low-risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , get the risk value of the low risk interval ; Where a1 and a2 are preset weight coefficients;

[0112] Calculate the mean dynamic stress corresponding to each data point in the risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , and get the risk value in the medium risk interval ; Where b1 and b2 are preset weight coefficients;

[0113] Calculate the mean dynamic stress corresponding to each data point in the high-risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , get the risk value of the high-risk interval ; Where c1 and c2 are preset weight coefficients;

[0114] Establishing an operating condition database, which stores selectable operating condition parameter groups corresponding to different water heads, flows, and total power generation, and each operating condition parameter group corresponds to a power generation efficiency XL;

[0115] Obtain the water head, flow rate, and total power generation corresponding to the current period, substitute them into the operating condition database, and determine all feasible operating condition parameter groups through database query operations, which are recorded as the primary operating condition parameter groups;

[0116] For each operating parameter group in the preliminary selected operating parameter group, substitute each operating parameter in the operating parameter group into the corresponding stress-parameter scatter plot, and extract the corresponding interval risk value according to the risk interval it falls into;

[0117] Assign different weight coefficients to the interval risk value corresponding to each working condition parameter, then multiply the interval risk value corresponding to each working condition parameter by the corresponding weight coefficient and add them together to obtain the working condition risk value GF;

[0118] After normalizing the operating risk value and power generation efficiency corresponding to the operating parameter group, the operating condition selection value XPZ is obtained using the formula: XPZ = XL × d1 - GF × d2, where d1 and d2 are preset weight coefficients. The larger the operating condition selection value, the lower the risk of the stress-sensitive area and the higher the power generation efficiency when selecting the operating parameter group under the current head, flow rate, and total power generation.

[0119] The operating condition parameter group with the largest operating condition selection value in the preliminary operating condition parameter group is selected as the implementation operating condition parameter group for the current period.

[0120] It should be noted that through the above-mentioned technical path of implementing operating condition risk quantification → database intelligent matching → multi-objective selection and decision-making, the dynamic optimization selection of hydropower unit operating parameters is achieved, and the power generation efficiency is maximized while ensuring the safe operation of the equipment. It provides a scientific basis for the intelligent scheduling of the hydropower system and has significant engineering application value and economic benefits.

[0121] The efficiency compensation module analyzes the sensitivity coefficients of the operating parameters to the operating risk value and power generation efficiency, and generates a priority sequence for risk compensation and efficiency compensation based on this. It monitors the operating risk value and efficiency value of the operating parameter group and triggers the corresponding compensation mechanism when an abnormality occurs. The specific process is as follows:

[0122] Obtain the historical implementation data of the current operating condition parameter group, and calculate the sensitivity coefficient of each operating condition parameter in the current operating condition parameter group to the operating condition risk value and power generation efficiency. The specific process is as follows:

[0123] For each operating parameter, calculate the average impact of each 1% change in the operating parameter p on the operating risk value using the formula: , and obtain the risk sensitivity coefficient ; Where i = 1, 2, ... n; n is the number of historical samples; 、 The change in risk value of the i-th sample and the initial value; 、 is the change of the corresponding working condition parameter and its initial value;

[0124] Using the same method of calculating risk sensitivity, we can calculate the average impact of each 1% change in operating parameters on the power generation efficiency XL and obtain the efficiency sensitivity coefficient. ;

[0125] Define the compensation priority of the working condition parameters. The specific process is as follows:

[0126] Risk compensation: take Get the working parameters and follow Arrange the sizes of in descending order to obtain the risk compensation priority sequence; To preset risk sensitivity thresholds;

[0127] Efficiency compensation: take Get the working parameters and follow The size of is sorted in descending order to obtain the efficiency compensation priority sequence; is the preset efficiency sensitivity threshold;

[0128] Real-time monitoring of the operating risk value GF and power generation efficiency XL of the currently implemented operating parameter group;

[0129] When the working condition risk value is greater than or equal to the preset working condition risk threshold, select The highest operating condition parameter is used as the risk adjustment parameter;

[0130] If the risk adjustment parameter is positively correlated with the working condition risk value, the parameter is increased for compensation, otherwise it is reduced for compensation;

[0131] The magnitude of risk compensation is, using the formula: , we get the risk compensation margin FBC, where is the preset working condition risk threshold, is the current value of the working condition parameter, is the preset risk compensation coefficient, with a value between (0, 1);

[0132] When the power generation efficiency is less than or equal to the preset power generation efficiency threshold, select The maximum operating condition parameter is adjusted as the efficiency adjustment parameter;

[0133] If the efficiency adjustment parameter is positively correlated with the operating condition risk value, the parameter is increased for compensation, otherwise it is reduced for compensation;

[0134] The magnitude of the efficiency compensation is, using the formula: , and obtain the efficiency compensation amplitude XBC, where is the preset power generation efficiency threshold, is the preset efficiency compensation coefficient, with a value between (0, 1);

[0135] If risk compensation conflicts with efficiency compensation;

[0136] If safety compensation is preferred, select from the risk compensation priority sequence The maximum operating parameter is used as the risk adjustment parameter; from the efficiency compensation priority sequence, select The minimum operating condition parameter is adjusted as the efficiency adjustment parameter; and the corresponding amplitude adjustment is performed; the adjustment amplitude calculation method is the same as the amplitude algorithm of risk compensation and the efficiency compensation algorithm;

[0137] If efficiency compensation is prioritized, select The minimum operating condition parameter is used as the risk adjustment parameter; from the efficiency compensation priority sequence, select The maximum operating condition parameter is adjusted as the efficiency adjustment parameter; and the corresponding amplitude is adjusted;

[0138] If the emphasis is on the balance between safety and efficiency, then the risk compensation sequence is selected <0, and in the efficiency compensation sequence >0 operating parameters (pa, pb) for adjustment;

[0139] The adjustment range is: , we can get the compensation range FBC of risk adjustment parameter and XBC of efficiency adjustment parameter under the balance between safety and efficiency, where and are the initial values ​​of the risk adjustment parameter and the efficiency adjustment parameter under the balance between safety and efficiency; To preset a small adjustment step; for example: 0.5%;

[0140] It should be noted that the sensitivity coefficients of various operating parameters to risk values ​​and power generation efficiency are calculated through historical real-time data, and the average impact of each 1% change in parameters on the indicators is quantified to facilitate the identification of key parameters;

[0141] By defining the priority of risk compensation and efficiency compensation, it is easy to quickly lock high-impact parameters, avoid invalid adjustments, and shorten response time;

[0142] By calculating the compensation amplitude, the arbitrariness of manual adjustment can be avoided, ensuring that the parameter adjustment amount is proportional to the index deviation, thus achieving the purpose of effective compensation;

[0143] Three strategies are set: priority safety compensation, priority efficiency compensation, and safety and efficiency balance compensation. The compensation strategy can be determined according to the real-time scenario; priority safety compensation can quickly reduce risks and prevent equipment failures; priority efficiency compensation can respond to scheduling agilely and improve power generation efficiency; the balance strategy realizes the coordinated optimization of dual indicators and adapts to normal operation; the dynamic switching strategy can improve system robustness and scheduling agility, extend equipment life, and optimize resource utilization.

[0144] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units, characterized by: include: Data acquisition module: collects operating status data of key parts of the hydropower unit, as well as environmental data and task instruction data; Sensitive area identification and analysis module: Analyzes the stress distribution of key parts and determines the stress concentration area; performs time-frequency domain analysis on the dynamic stress data of the stress concentration area to obtain the sensitivity coefficient, and identifies the stress sensitive area based on it; The process of identifying sensitive areas is as follows: Import the three-dimensional solid model of the hydropower unit into the finite element analysis software for fluid-solid-thermal multi-physics field coupling simulation; Set up several typical working conditions; calculate the VonMises stress distribution of each key part under the typical working conditions, extract the area with stress concentration factor ≥ K1, and mark it as stress concentration area, where K1 is the critical value of the set stress concentration factor; For each stress concentration area, the dynamic stress data of the stress concentration area is collected in real time, and the collected dynamic stress data is analyzed in the time-frequency domain, specifically: Time Domain Analysis: Calculating Mean Stress , standard deviation , peak factor CF; Frequency domain analysis: Obtain the main frequency component fdom through FFT transformation, and identify whether the main frequency component has resonance related to the rotation frequency and blade frequency; Using the formula: , get the sensitivity coefficient S; Where SCF is the stress concentration factor of the stress concentration area, is the resonance index. If fdom coincides with the integral multiple of the rotation frequency and blade frequency, Take 1, otherwise take 0; w1, w2, w3 are preset weight coefficients; A threshold value S1 of the sensitive region coefficient is preset. If the sensitivity coefficient S of the stress concentration region is greater than or equal to S1, the stress concentration region is marked as a stress sensitive region. Risk classification module: Analyzes the correlation between dynamic stress and various operating parameters to determine analysis parameters; draws a scatter plot of dynamic stress and analysis parameters, performs cluster analysis and outlier detection on the data points in the scatter plot, and determines the operating risk range of stress-sensitive areas under different operating parameter combinations; Execution decision module: calculates the interval risk values ​​of low, medium and high risk intervals, builds the working condition database and screens the preliminary working condition parameter group; Analyze the operating risk value of the pre-selected operating parameter group, calculate the selection value based on the power generation efficiency, and determine the implementation operating parameter group accordingly; Efficiency compensation module: Analyzes the sensitivity coefficients of operating parameters to operating risk values ​​and power generation efficiency, and generates priority sequences for risk compensation and efficiency compensation accordingly; monitors the operating risk values ​​and power generation efficiency of the operating parameter group, and triggers the corresponding compensation mechanism.

2. The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units according to claim 1 is characterized by: The data acquisition module includes: Collect operating status data of key parts, including dynamic stress and vibration data; key parts include: runner blade root, guide vane pivot, main shaft bearing, main shaft bearing seat, volute inlet and outlet, and top cover support point; Monitor the environmental data of the upstream water of the hydropower unit, including water level and flow information; By connecting with the power grid dispatching system, the total power generation task instruction data for the current period can be obtained.

3. The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units according to claim 2 is characterized by: The risk classification module determines the analysis parameters; the specific process of drawing the dynamic stress-analysis parameter scatter plot is as follows: Collect the recent historical dynamic stress data of stress-sensitive areas of hydropower units and the corresponding operating condition data; For each stress-sensitive area, the Pearson correlation coefficient method is used to calculate the correlation coefficient between dynamic stress and various operating parameters. Based on the correlation coefficient, it is determined whether to extract the corresponding operating parameters and record them as analysis parameters. The specific implementation method is as follows: the operating parameters with an absolute value of the correlation coefficient greater than the corresponding threshold are extracted and recorded as analysis parameters; For each analysis parameter, a coordinate system is established with dynamic stress as the ordinate and analysis parameter as the abscissa, and a scatter plot is drawn based on the historical dynamic stress data and the corresponding working condition data to obtain a stress-parameter scatter plot.

4. The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units according to claim 3 is characterized by: The process of the risk classification module performing cluster analysis on the data points in the scatter plot to determine the operating risk interval is as follows: Normalize the data in the stress-parameter scatter plot; The DBSCAN clustering algorithm is used to cluster the stress-parameter scatter plot and obtain several data clusters; Calculate the number of data points in each data cluster to obtain the cluster point number, preset the cluster point number threshold, and if the cluster point number is greater than the corresponding threshold, calculate the standard deviation of dynamic stress and the standard deviation of analysis parameters corresponding to the data points in the data cluster; Set the dynamic stress discrete degree threshold and the analysis parameter discrete degree threshold; If the standard deviation of the dynamic stress within the cluster is less than the dynamic stress dispersion threshold, and the standard deviation of the analysis parameter is less than the analysis parameter dispersion threshold, then the low-risk area is judged; Calculate the stress mean value of the stress-sensitive area during the most recent safe operation and record it as the safe stress mean value; At the same time, the dynamic stress mean corresponding to the data points in the data cluster is calculated and recorded as the cluster stress mean; The deviation between the cluster stress mean and the safety stress mean is calculated to obtain the cluster stress deviation value. If the cluster stress deviation value is less than or equal to the stress deviation threshold, the area covered by the data cluster is determined to be a low-risk area.

5. The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units according to claim 4 is characterized by: The risk classification module performs outlier detection on the data points in the scatter plot and determines the operating risk interval in the following process: Calculate the outlier factor of each data point, and determine the data points whose outlier factor is greater than the corresponding threshold as preliminary outliers; For each preliminary outlier, calculate the absolute difference between its dynamic stress value and the mean of all dynamic stresses in the scatter plot; If the difference is greater than or equal to K times the standard deviation of the dynamic stress in the scatter plot, the point is confirmed as an actual outlier; where K is a preset coefficient; Draw a circle with a preset radius centered on each actual outlier, and mark the area corresponding to the data points contained in the circle as a high-risk area; After excluding the low-risk and high-risk intervals, the remaining middle area is the medium-risk interval.

6. The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units according to claim 5 is characterized by: The specific process of analyzing the interval risk values ​​of low, medium and high risk intervals in the execution decision module is as follows: Calculate the mean dynamic stress corresponding to each data point in the low-risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , get the risk value of the low risk interval ; Where a1 and a2 are preset weight coefficients; Calculate the mean dynamic stress corresponding to each data point in the risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , and get the risk value in the medium risk interval ; Where b1 and b2 are preset weight coefficients; Calculate the mean dynamic stress corresponding to each data point in the high-risk interval , analysis parameter mean , standard deviation of dynamic stress , analysis parameter standard deviation , and using the formula: , get the risk value of the high-risk interval ; Where c1 and c2 are preset weight coefficients.

7. The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units according to claim 6 is characterized by: The execution decision module builds a working condition database to screen the preliminary working condition parameter group; analyzes the working condition risk value of the preliminary working condition parameter group, calculates the selection value based on the power generation efficiency, and determines the specific process of implementing the working condition parameter group based on this: Establishing an operating condition database, which stores selectable operating condition parameter groups corresponding to different water heads, flows, and total power generation, and each operating condition parameter group corresponds to a power generation efficiency XL; Obtain the water head, flow rate, and total power generation corresponding to the current period, substitute them into the operating condition database, and determine all feasible operating condition parameter groups through database query operations, which are recorded as the primary operating condition parameter groups; For each operating parameter group in the preliminary selected operating parameter group, substitute each operating parameter in the operating parameter group into the corresponding stress-parameter scatter plot, and extract the corresponding interval risk value according to the risk interval it falls into; Assign different weight coefficients to the interval risk value corresponding to each working condition parameter, then multiply the interval risk value corresponding to each working condition parameter by the corresponding weight coefficient and add them together to obtain the working condition risk value GF; After normalizing the operating risk value and power generation efficiency corresponding to the operating parameter group, the operating condition selection value XPZ is obtained using the formula: XPZ=XL×d1-GF×d2, where d1 and d2 are preset weight coefficients; The operating condition parameter group with the largest operating condition selection value in the preliminary operating condition parameter group is selected as the implementation operating condition parameter group for the current period.

8. The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units according to claim 7 is characterized by: The efficiency compensation module analyzes the sensitivity coefficients of operating parameters to operating risk values ​​and power generation efficiency, and generates the risk compensation and efficiency compensation priority sequences based on the analysis. The specific process is as follows: For each operating parameter, calculate the average impact of each 1% change in the operating parameter p on the operating risk value using the formula: , and obtain the risk sensitivity coefficient ; Where i = 1, 2, ... n; n is the number of historical samples; 、 The change in risk value of the i-th sample and the initial value; 、 is the change of the corresponding working condition parameter and its initial value; Using the same method of calculating risk sensitivity, we can calculate the average impact of each 1% change in operating parameters on the power generation efficiency XL and obtain the efficiency sensitivity coefficient. ; Define the compensation priority of the working condition parameters. The specific process is as follows: Risk compensation: take Get the working parameters and follow Arrange the sizes of in descending order to obtain the risk compensation priority sequence; To preset risk sensitivity thresholds; Efficiency compensation: take Get the working parameters and follow The size of is sorted in descending order to obtain the efficiency compensation priority sequence; is the preset efficiency sensitivity threshold.

9. The intelligent decision-making system for identifying dynamic stress sensitive areas and compensating efficiency of hydropower units according to claim 8 is characterized by: The efficiency compensation module monitors the operating risk value and power generation efficiency of the operating parameter group in real time and triggers the corresponding compensation mechanism. The specific process is as follows: Real-time monitoring of the operating condition risk value GF and power generation efficiency XL of the current operating condition parameter group; when the operating condition risk value is greater than or equal to the preset operating condition risk threshold, select from the risk compensation priority sequence The highest operating condition parameter is used as the risk adjustment parameter; If the risk adjustment parameter is positively correlated with the working condition risk value, the parameter is increased for compensation, otherwise it is reduced for compensation; The magnitude of risk compensation is, using the formula: , we get the risk compensation margin FBC, where is the preset working condition risk threshold, is the current value of the working condition parameter, is the preset risk compensation coefficient, with a value between (0, 1); When the power generation efficiency is less than or equal to the preset power generation efficiency threshold, select The maximum operating condition parameter is adjusted as the efficiency adjustment parameter; If the efficiency adjustment parameter is positively correlated with the operating condition risk value, the parameter is increased for compensation, otherwise it is reduced for compensation; The magnitude of the efficiency compensation is, using the formula: , and obtain the efficiency compensation amplitude XBC, where is the preset power generation efficiency threshold, is the preset efficiency compensation coefficient, with a value between (0, 1); If risk compensation conflicts with efficiency compensation; If safety compensation is preferred, select from the risk compensation priority sequence The maximum operating condition parameter, used as a risk adjustment parameter; From the efficiency compensation priority sequence, select The minimum operating condition parameter is adjusted as the efficiency adjustment parameter; and the corresponding amplitude is adjusted; If efficiency compensation is prioritized, select The minimum operating condition parameter is used as the risk adjustment parameter; from the efficiency compensation priority sequence, select The maximum operating condition parameter is adjusted as the efficiency adjustment parameter; and the corresponding amplitude is adjusted; If the emphasis is on the balance between safety and efficiency, then the risk compensation sequence is selected <0, and in the efficiency compensation sequence >0 operating parameters (pa, pb) for adjustment; The adjustment range is: , we can get the compensation range FBC of risk adjustment parameter and XBC of efficiency adjustment parameter under the balance between safety and efficiency, where and are the initial values ​​of the risk adjustment parameter and the efficiency adjustment parameter under the balance between safety and efficiency; It is a preset small adjustment step.

Images