Hydroelectric generating set health state assessment method based on multi-dimensional data

Abstract

The invention discloses a hydroelectric generating set health state assessment method based on multi-dimensional data, and the method comprises the steps: S1, obtaining the operation data of a hydroelectric generating set, and carrying out the data preprocessing; s2, sensing data extraction is carried out on the preprocessed hydroelectric generating set operation data; s3, performing anomaly detection on the operation sensing data of the hydroelectric generating set; s4, carrying out sensing abnormal frequency statistics on the operation sensing abnormal data of the hydroelectric generating set; s5, thermal efficiency and operation water conservancy efficiency calculation is carried out on the operation sensing data of the hydroelectric generating set, and operation performance index analysis is carried out; s6, performing health state calculation according to the analyzed hydroelectric generating set operation performance index data; and S7, constructing a health assessment model to assess the health state of the hydroelectric generating set, uploading assessment data to the cloud platform, and executing a preset health early warning task. According to the method, the health state is calculated by combining the operation performance index and the sensing abnormal frequency data, so that more comprehensive health assessment is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of hydropower unit status assessment and monitoring, and in particular to a hydropower unit health status assessment method based on multidimensional data. Background Art

[0002] As global energy demand continues to grow, hydropower, as a renewable energy source, has received widespread attention due to its clean and sustainable characteristics.

[0003] As the core equipment of a hydropower station, the health status of a hydropower unit directly affects power generation efficiency, equipment safety, and operating costs. Therefore, timely and accurate assessment of the health status of a hydropower unit is of great significance for ensuring its stable operation and optimizing maintenance strategies.

[0004] Traditional methods for assessing the health status of hydropower units mainly rely on experience and regular inspections. These methods often have certain limitations, which include insufficient real-time monitoring of equipment status, limited data analysis capabilities, and difficulty in achieving intelligent management; data from different sources (such as sensor data, maintenance records, etc.) have different characteristics and scales, and data fusion and processing face challenges. Summary of the invention

[0005] In order to solve the above problems, the present invention provides a method for evaluating the health status of a hydropower unit based on multidimensional data. The specific technical solution is as follows:

[0006] S1: Obtain the operating data of the hydropower unit and perform data preprocessing;

[0007] The real-time operation data of the hydropower unit includes key parameters such as output power, water flow, pressure and temperature, etc., ensuring that the collected data is comprehensive enough to provide sufficient information basis for subsequent data analysis. At the same time, it ensures real-time monitoring of the operation status of the hydropower unit, timely grasps the dynamic changes of the unit, and facilitates rapid response to potential problems. The rich data sources provide reliable decision-making support for the management and help formulate more effective operation and maintenance strategies;

[0008] Data preprocessing includes filling missing values, eliminating outliers and standardization. By preprocessing the collected data, the accuracy and reliability of the data can be improved, providing a high-quality data foundation for subsequent analysis. The preprocessed data is neater and more structured, which facilitates subsequent rapid calculation and analysis and reduces the time cost of data processing.

[0009] S2: extracting sensor data from the preprocessed hydropower unit operation data, including extraction of hydropower unit operation temperature sensor data and hydropower unit operation pressure sensor data, to obtain hydropower unit operation sensor data;

[0010] By extracting operating sensor data, various operating indicators of hydropower units can be monitored in real time and potential problems can be discovered in a timely manner.

[0011] S3: Performing anomaly detection on the operating sensor data of the hydropower unit to obtain the operating sensor abnormality data of the hydropower unit;

[0012] Through anomaly detection, sensor data that exceeds the normal range can be identified, equipment failures or abnormal operating conditions can be discovered early, and the risk of failure can be reduced.

[0013] S4: Performing sensor abnormality frequency statistics on the sensor abnormality data of the hydropower unit operation, and filtering and obtaining abnormal data whose abnormality frequency is higher than the set threshold according to the set abnormality frequency threshold;

[0014] By counting the frequency of sensor anomalies, operation and maintenance personnel can identify sensors with high-frequency anomalies as key monitoring objects, thereby optimizing the allocation of maintenance resources.

[0015] S5: Calculate the thermal efficiency and hydraulic efficiency of the hydropower unit operation sensor data, and analyze the operation performance index to obtain the hydropower unit operation performance index data;

[0016] By calculating the operating thermal efficiency and hydraulic efficiency, the operating performance of the hydropower unit can be comprehensively evaluated, the causes of efficiency loss can be identified, and a basis can be provided for subsequent optimization. The efficiency data obtained can provide decision-making support for the management, guide the operation and maintenance strategies of the equipment, and improve the overall power generation efficiency. By analyzing the efficiency data, the operation and maintenance team can implement targeted improvement measures to improve the operating economy of the hydropower unit.

[0017] S6: performing health status calculation based on the analyzed operating performance index data of the hydropower unit to obtain operating health status data of the hydropower unit;

[0018] Combining the operation performance indicators and sensor abnormal frequency data to calculate the health status can achieve a more comprehensive health assessment and consider multi-dimensional influencing factors.

[0019] S7: construct a health assessment model based on the operating health status data of the hydropower unit to perform health status assessment of the hydropower unit, upload the health status assessment data to the hydropower unit management cloud platform, and execute the preset corresponding hydropower unit health warning task.

[0020] By evaluating the operating data according to the health assessment model, real-time monitoring of the status of the hydropower unit can be achieved, and the response speed to the health status of the equipment can be improved. The assessment data can be uploaded to the hydropower unit management cloud platform to achieve centralized management and sharing of data, improve the transparency and traceability of information, and automatically generate health warnings through the cloud platform management system to guide operation and maintenance personnel to take timely maintenance measures, reduce failure rates, improve equipment reliability, and improve the hydropower unit's ability to process data from different sources.

[0021] Furthermore, in step S1, the data is preprocessed, and the specific process is as follows:

[0022] S101: Fill missing values ​​in the operation data of the hydropower unit to obtain filled operation data of the hydropower unit;

[0023] By filling in missing values, we can effectively reduce the incompleteness of data and ensure the accuracy and reliability of subsequent analysis. Filling in missing values ​​avoids sample loss due to missing data, retains more valid data, and enhances the effectiveness of the data set.

[0024] S102: performing outlier processing on the filling operation data of the hydropower unit to obtain abnormal correction operation data of the hydropower unit;

[0025] By processing outliers, we can identify and eliminate unreasonable data points, improve the overall quality of data, and ensure the scientific nature of subsequent analysis. Outlier processing helps eliminate data anomalies caused by equipment failure, sensor errors, or external interference, thereby improving the prediction accuracy of the model. The processed data can more realistically reflect the operating status of the hydropower unit, provide a more reliable basis for operation and maintenance personnel, and guide maintenance and management decisions.

[0026] S103: performing standardization processing on the abnormal correction operation data of the hydropower unit to obtain the operation data of the hydropower unit to be processed;

[0027] Standardization can eliminate dimensional differences between data, enable comparison of data with different characteristics at the same scale, and improve data uniformity. In machine learning and data analysis, standardization can accelerate the convergence of models, improve the efficiency and accuracy of algorithms, and reduce computational burden. Standardized data can reduce the sensitivity of models to abnormal data and extreme values, and improve the generalization and robustness of models on new data.

[0028] Furthermore, in step S3, abnormality detection is performed on the operating sensor data of the hydropower unit, including:

[0029] S301: Perform temperature anomaly detection on the operating temperature sensor data of the hydropower unit to obtain the operating temperature sensor anomaly data of the hydropower unit;

[0030] By performing anomaly detection on temperature sensor data, temperature anomalies such as overheating or insufficient cooling can be discovered in a timely manner, reducing the risk of equipment damage. Temperature anomaly detection can ensure that the equipment operates within a safe range, prevent equipment failures or accidents caused by excessive temperature, ensure the safety of hydropower units, and promptly identify temperature anomalies, providing a basis for operation and maintenance personnel to take appropriate measures and optimize equipment operation and maintenance strategies.

[0031] S302: Performing pressure anomaly detection on the operating pressure sensor data of the hydropower unit to obtain the operating pressure sensor anomaly data of the hydropower unit;

[0032] Anomaly detection of pressure sensor data helps ensure that the hydropower unit operates within the appropriate pressure range and avoids equipment failure or efficiency loss due to pressure anomalies. Timely detection of pressure anomalies can enhance the monitoring capabilities of the entire hydropower unit and ensure the coordinated work of various systems. By identifying and recording pressure anomalies, the operation and maintenance team can better schedule maintenance work and reduce downtime and maintenance costs.

[0033] S303: integrating abnormal sensor data according to the abnormal temperature sensor data and the abnormal pressure sensor data of the hydropower unit operation to obtain the abnormal operation sensor data of the hydropower unit;

[0034] Integrating abnormal temperature and pressure data can help comprehensively evaluate the operating status of hydropower units and avoid the one-sidedness caused by isolated analysis. By integrating data from different sensors, potential correlation problems can be identified. For example, temperature anomalies are related to pressure changes, which provides a deeper perspective for fault analysis. The integrated abnormal data allows operation and maintenance personnel to more effectively identify and locate problems, reduce fault diagnosis time, and improve overall operation and maintenance efficiency. By performing frequency statistics on abnormal sensor data, sensors with high-frequency abnormalities can be identified. These sensors require special attention or maintenance, and resource allocation can be optimized. Statistical analysis based on abnormal frequency can help management formulate more scientific monitoring and maintenance strategies to ensure the normal operation of key equipment and components. Frequency statistics can also help discover long-term trends and patterns, provide reference for future maintenance plans, and promote equipment health management.

[0035] Furthermore, in step S3, temperature anomaly detection is performed on the operating temperature sensor data of the hydropower unit, and the specific process is as follows:

[0036] S3011: extracting turbine temperature sensor data from the operating temperature sensor data of the hydropower unit to obtain turbine temperature sensor data;

[0037] By extracting the temperature sensing data of the turbine, we can gain in-depth understanding of the operating status of the turbine and ensure real-time monitoring of key components. Focusing on the temperature data extraction of the turbine helps reduce data interference and ensure that the analysis results are more targeted and effective, facilitating subsequent analysis and processing, ensuring the basic data quality for subsequent analysis, and providing reliable data support for high temperature statistics and anomaly detection.

[0038] S3012: Perform high temperature statistics according to the turbine temperature sensor data, and obtain the turbine high temperature sensor data based on a set first high temperature threshold;

[0039] Through high temperature statistics, high temperature conditions during the operation of the turbine can be identified, operational abnormalities can be discovered in time, and equipment safety can be ensured. High temperature statistical data provides a basis for the operation and maintenance team to help them give priority to frequent temperature abnormalities in maintenance plans and optimize the allocation of maintenance resources. Timely identification of high temperature conditions can serve as part of the risk warning mechanism, and measures can be taken in advance to prevent equipment damage.

[0040] S3013: Perform frequency statistics on the high-temperature sensor data of the turbine, and obtain low-frequency high-temperature sensor data of the turbine based on a set frequency threshold;

[0041] Through frequency statistics, the frequency of high temperature occurrence can be identified, providing a reference for subsequent anomaly detection and helping operation and maintenance personnel determine whether there are systemic problems. Frequency statistics enable the operation and maintenance team to make more scientific decisions based on data, reduce reliance on experience, and help improve management efficiency. Through frequency statistics, the priority of high temperature anomalies can be clarified, allowing operation and maintenance personnel to concentrate on dealing with high-frequency problems and improve overall equipment reliability.

[0042] S3014: Setting a second high temperature threshold value according to the low-frequency high temperature sensing data of the turbine to obtain high temperature threshold value data;

[0043] Setting a high temperature threshold can provide a clear standard for subsequent temperature monitoring, so that high temperature anomaly detection has a basis to rely on and improve the scientific nature of monitoring. Setting the threshold according to the specific operating conditions of the turbine can achieve personalized management and ensure targeted high temperature warning strategies for different units. Setting a reasonable high temperature threshold can effectively reduce the risk of equipment failure caused by excessive temperature and enhance the safety and stability of the equipment.

[0044] S3015: Perform abnormal high temperature detection on the turbine temperature sensor data according to the high temperature threshold data to obtain abnormal operating temperature sensor data of the hydropower unit;

[0045] By using high temperature thresholds to detect abnormally high temperatures, the hydropower unit can monitor temperature changes in real time during operation and detect abnormalities in a timely manner. The abnormally high temperature detection system can be used as part of a fault warning mechanism to ensure that operation and maintenance personnel can be quickly notified when the temperature is abnormal and take necessary measures. The abnormally high temperature data detected can be used to analyze the causes of the temperature abnormality, thereby optimizing maintenance strategies, reducing equipment downtime, and improving power generation efficiency.

[0046] Furthermore, in step S3, the pressure anomaly detection is performed on the operating pressure sensor data of the hydropower unit, and the specific process is as follows:

[0047] S3021: extracting turbine pressure sensor data from the operating pressure sensor data of the hydropower unit to obtain turbine pressure sensor data;

[0048] By extracting the pressure sensing data of the turbine, we can gain in-depth understanding of the operating status of the turbine and ensure real-time monitoring of key components. Focusing on the extraction of pressure data from the turbine can reduce data interference, ensure that subsequent analysis is more targeted, and facilitate the identification of potential problems. It provides a high-quality and dedicated data foundation for subsequent analysis and enhances the availability and reliability of the data.

[0049] S3022: extracting intake pressure sensing features according to turbine pressure sensing data to obtain intake pressure sensing data;

[0050] Extracting the features of the inlet pressure sensor data helps to identify key factors that affect turbine performance, such as the impact of inlet pressure changes on turbine operation. Feature extraction can reduce data dimensions and focus on the most important features, which helps improve the efficiency of subsequent analysis and model building, lays the foundation for subsequent time series analysis, and ensures that the feature data required for analysis is complete and accurate.

[0051] S3023: collecting time series data according to the air inlet pressure sensor data to obtain air inlet pressure sensor time series data;

[0052] The collection of time series data can reflect the dynamic changes of the inlet pressure and help identify trends and periodic characteristics that change over time. By collecting time series data, it can provide the necessary data basis for subsequent prediction models, enhance the ability to predict future operating conditions, enable analysis to take time factors into account, and enhance understanding and insight into data changes.

[0053] S3024: performing abnormal time series analysis according to the air inlet pressure sensor time series data to obtain abnormal time series data of the air inlet pressure sensor;

[0054] Abnormal analysis of time series data can timely detect abnormal fluctuations in the air inlet pressure and ensure early identification of potential faults. Abnormal time series analysis can improve the accuracy of monitoring, identify data points that do not conform to normal operating modes, and provide a basis for subsequent processing. The analysis results can provide important decision-making support for the operation and maintenance team, helping them to take timely measures to prevent faults.

[0055] S3025: constructing a monitoring model according to the air inlet pressure sensor abnormality time series data to obtain an air inlet pressure sensor abnormality monitoring model;

[0056] By constructing an abnormal monitoring model, intelligent monitoring of the air inlet pressure can be achieved, improving the safety and reliability of equipment operation. The monitoring model constructed based on abnormal time series data makes monitoring more dependent on actual operation data, improving the adaptability and accuracy of the model. This model can analyze the changes in the air inlet pressure in real time, detect abnormalities in time and feedback to the operation and maintenance personnel.

[0057] S3026: performing abnormal pressure sensing monitoring on the operating pressure sensing data of the hydropower unit according to the air inlet pressure sensing abnormality monitoring model to obtain abnormal pressure sensing data of the hydropower unit;

[0058] By applying the abnormal monitoring model, comprehensive monitoring of the pressure sensor data of the hydropower unit can be achieved, the ability to identify abnormal conditions can be improved, and an effective fault warning mechanism can be provided to ensure that an alarm can be quickly issued when the pressure is abnormal, reducing potential risks. The monitoring results can provide the operation and maintenance team with targeted maintenance strategies, reduce equipment failures caused by abnormal pressure, and improve overall power generation efficiency.

[0059] Furthermore, an abnormal time series analysis is performed based on the intake pressure sensor time series data. The specific process is as follows:

[0060] S30241: Draw an air inlet pressure sensor time series graph for the air inlet pressure sensor time series data to obtain an air inlet pressure sensor time series graph;

[0061] By drawing a time series graph, the changing trend of the air inlet pressure can be intuitively displayed, which is convenient for operation and maintenance personnel to quickly understand the dynamic characteristics of the data. Through visualization, it is easier to identify abnormal points and trend changes in the data, laying the foundation for subsequent analysis. The time series graph provides an intuitive basis for decision-making and helps the operation and maintenance team to refer to historical data changes when formulating maintenance plans.

[0062] S30242: Perform residual analysis based on the intake pressure sensor time series diagram to obtain residual data;

[0063] Residual analysis can help identify the differences between model predictions and actual observations, thereby optimizing the model and improving the accuracy of predictions. By analyzing the residuals, outliers or abnormal patterns can be identified, helping operation and maintenance personnel to promptly detect potential faults or abnormal conditions. Residual analysis helps to deeply understand the characteristics of the data and identify potential factors affecting the inlet pressure.

[0064] S30243: Perform seasonal fluctuation analysis based on the intake pressure sensor time series diagram to obtain seasonal fluctuation data;

[0065] Seasonal fluctuation analysis can reveal periodic changes in the air inlet pressure data, providing important background information for understanding equipment operation. Identifying seasonal fluctuations helps the operation and maintenance team consider seasonal factors when formulating maintenance and operation plans, and improves the scientific nature of equipment management. By analyzing seasonal fluctuations, the performance of equipment under different environmental conditions can be more accurately evaluated, helping to formulate targeted optimization measures.

[0066] S30244: construct an abnormal time detection model according to the residual data and seasonal fluctuation data to obtain an abnormal time detection model;

[0067] Building an abnormal time detection model can realize intelligent monitoring of the air inlet pressure and enhance the safety and reliability of equipment operation. The model is based on residual and seasonal fluctuation data, which can more accurately identify abnormal situations and help the operation and maintenance team make data-driven decisions. The constructed model can monitor data changes in real time, detect anomalies in time, improve the operation and maintenance response speed, and reduce potential losses.

[0068] S30245: performing abnormal time series detection on the air inlet pressure sensor time series data according to the abnormal time detection model to obtain abnormal time series data of the air inlet pressure sensor;

[0069] By applying the abnormal time detection model, real-time monitoring of the air inlet pressure sensor data can be achieved, and abnormal changes can be discovered in time. Anomaly detection can be used as part of the fault warning mechanism to quickly notify operation and maintenance personnel when the temperature or pressure is abnormal, thereby reducing the risk of failure. The detection results provide the operation and maintenance team with targeted maintenance strategies, which helps to concentrate resources on high-risk areas and improve overall operating efficiency.

[0070] Furthermore, in step S4, the sensor abnormality frequency statistics of the hydropower unit operation sensor abnormality data are performed, as follows:

[0071] S401: Performing interval statistics on the sensor abnormality data of the hydropower unit operation, and obtaining the sensor abnormality data of the hydropower unit operation in short intervals based on the set intervals;

[0072] By performing interval statistics on abnormal operation sensor data, the time intervals between abnormal events can be identified, helping to analyze the frequency and patterns of abnormalities. Interval statistics help the operation and maintenance team understand the relationship between different types of abnormal events, making it easier to formulate targeted monitoring and maintenance strategies. Understanding the intervals of abnormal data can help operation and maintenance personnel troubleshoot more effectively and quickly locate the source of the problem.

[0073] S402: classifying the sensor type of the short-interval operation sensor abnormality data of the hydropower unit to obtain the short-interval operation sensor type abnormality data of the hydropower unit;

[0074] Classifying short-interval abnormal data helps to quickly identify the source of the abnormality, facilitate subsequent processing and response, and improve management efficiency. Through classification, we can focus on certain specific types of sensors or components, optimize the allocation of resources and maintenance, and ensure that key issues are handled in a timely manner. The classified data provides the operation and maintenance team with a clearer basis for decision-making, making subsequent analysis and processing more scientific and efficient.

[0075] S403: Perform frequency statistics according to the short-interval operation sensor type abnormal data, and obtain high-frequency short-interval sensor abnormal data based on the set frequency;

[0076] Frequency statistics can help identify abnormal types that occur frequently in short intervals. Focusing on these high-frequency abnormalities will help take corresponding measures in a timely manner. Through frequency statistics, the operation and maintenance team can analyze the occurrence trend of abnormal events, evaluate the health status of equipment and adjust maintenance plans in a timely manner. Based on frequency data, operation and maintenance personnel can optimize monitoring strategies, concentrate resources on monitoring high-frequency abnormal types, and improve overall monitoring efficiency.

[0077] S404: performing sensor anomaly clustering on the hydropower unit operation sensor anomaly data according to the high-frequency short-interval sensor anomaly data to obtain the hydropower unit operation sensor high-frequency anomaly data;

[0078] By clustering high-frequency, short-interval abnormal data, similar abnormal events can be grouped together, which helps to identify potential root causes. Cluster analysis helps to reduce noise in abnormal data, extract more representative and effective information, and improve the accuracy of subsequent analysis. The clustering results can provide operation and maintenance personnel with clear maintenance directions, making it easier to concentrate resources to handle high-frequency anomalies, thereby improving the stability and reliability of equipment.

[0079] Furthermore, in step S5, the operating performance index data of the hydropower unit is obtained, and the specific process is as follows:

[0080] S501: extracting flow meter sensor data and water dragon measuring instrument sensor data according to the operating sensor data of the hydropower unit;

[0081] By extracting data from flow meters and sluice gauges, the key operating parameters of the hydropower unit can be obtained, ensuring accurate monitoring of fluid flow and pressure, ensuring a comprehensive understanding of the operating status of the hydropower unit, and providing the necessary basic data for subsequent efficiency calculations and performance analysis. The extracted sensor data provides an important basis for the operation analysis of the hydropower unit, enabling management to make scientific decisions based on data.

[0082] S502: Calculate the thermal efficiency of the water motor according to the flow meter sensor data and the water dragon measuring instrument sensor data;

[0083] By calculating the operating thermal efficiency of the hydropower generator, the unit's performance in energy utilization can be evaluated and potential areas for improvement can be identified. Analysis of thermal efficiency data can help the operation and maintenance team optimize energy utilization strategies and improve the overall operating efficiency of the unit. Abnormal thermal efficiency values ​​can be used as a fault warning indicator to help promptly detect and deal with potential problems.

[0084] S503: Obtaining turbine output power;

[0085] Obtaining the output power data of the turbine can intuitively reflect the power generation capacity of the unit and help determine its working status. The output power data provides basic parameters for subsequent hydraulic efficiency calculations and performance analysis, ensuring the accuracy and effectiveness of the analysis. Real-time output power monitoring can help operation and maintenance personnel respond quickly and adjust operation strategies to adapt to load changes.

[0086] S504: Calculating the hydraulic efficiency of the hydroelectric machine according to the turbine output power, flow meter sensor data, and water dragon measuring instrument sensor data;

[0087] By combining the data from output power, flow meter and sluice meter, the hydraulic efficiency of the hydropower unit can be accurately calculated and its energy conversion efficiency can be evaluated. The calculation results of the hydraulic efficiency can reveal deficiencies in equipment operation and help the operation and maintenance team identify potential optimization directions. By monitoring changes in hydraulic efficiency, problems with equipment performance degradation can be discovered in a timely manner to ensure the stability and reliability of the hydropower unit.

[0088] S505: analyzing the operation performance index according to the operation thermal efficiency data of the hydroelectric generator and the operation hydraulic efficiency data of the hydroelectric generator unit to obtain the operation performance index data of the hydroelectric generator unit;

[0089] By analyzing the thermal efficiency and hydraulic efficiency data, we can comprehensively evaluate the operating performance of the hydropower units and identify the main factors affecting their efficiency. The analysis results of the operating performance indicators provide improvement suggestions for the operation and maintenance team to help them formulate more effective maintenance and management strategies. Based on the performance indicator analysis, the management can make more scientific long-term plans, optimize resource allocation, and improve the overall power generation efficiency.

[0090] Furthermore, the specific process of the operation performance index analysis is as follows:

[0091] S5051: Calculate the average efficiency and fluctuation of the hydropower unit according to the thermal efficiency data and hydraulic efficiency data of the hydropower unit;

[0092] By calculating the average efficiency of the hydropower unit, we can intuitively understand the overall performance of the unit within a certain time range, providing basic data for subsequent analysis. Evaluating the volatility of efficiency can reveal the instability of unit performance and help operation and maintenance personnel identify potential problems and abnormal conditions. The data on average efficiency and its volatility provide a basis for optimizing the unit operation strategy and help improve the overall efficiency and stability of the unit.

[0093] S5052: Acquire rated efficiency data of the hydropower unit and rated efficiency fluctuation data of the hydropower unit;

[0094] The rated efficiency data provides a standard benchmark for the performance of hydropower units, which is helpful for subsequent comparison and analysis. Obtaining the volatility data of the rated efficiency can help understand the performance of the unit under normal operating conditions and provide reference data for evaluating actual operating conditions. The rated efficiency and its volatility data provide the necessary basis for subsequent efficiency comparison analysis, ensuring the accuracy and reliability of the analysis.

[0095] S5053: performing efficiency performance index comparison with average efficiency data of the hydropower unit according to rated efficiency data of the hydropower unit to obtain efficiency performance index of the hydropower unit;

[0096] By comparing the average efficiency of the hydropower unit with the rated efficiency, the actual operating performance of the unit can be clearly evaluated and the increase or decrease in efficiency can be identified. The comparative analysis helps the operation and maintenance team to discover the deficiencies in the operation of the unit and then formulate corresponding improvement measures to improve the overall efficiency. The comparison of efficiency performance indicators provides data support for the management, making the decision more scientific and reasonable.

[0097] S5054: performing a fluctuation index comparison on the efficiency fluctuation data of the hydropower unit according to the rated efficiency fluctuation data of the hydropower unit to obtain the fluctuation index data of the hydropower unit;

[0098] By comparing the volatility data, we can evaluate the stability of the hydropower unit in operation, identify situations with large volatility, and help operation and maintenance personnel take measures to control them. The analysis results of the volatility indicators provide the operation and maintenance team with improvement suggestions, which helps to formulate more effective operation strategies and improve the stability of the unit. By comparing the volatility indicators, we can provide a reference for long-term monitoring and evaluation of the operating status of the hydropower unit, and facilitate the identification of trends and changes.

[0099] S5055: Integrate the index data according to the efficiency performance index of the hydropower unit and the fluctuation index data of the hydropower unit to obtain the operation performance index data of the hydropower unit;

[0100] By integrating efficiency performance indicators and fluctuation index data, the operating performance of the hydropower unit can be comprehensively evaluated to form a comprehensive performance analysis report. The integrated data provides management with a clear performance overview, helping them to make decisions quickly and optimize resource allocation. The comprehensive operating performance indicator data helps the operation and maintenance team to continuously monitor and improve unit operation and promote the intelligent management of hydropower units.

[0101] Furthermore, in step S6, the operating health status data of the hydropower unit is obtained, and the specific process is as follows:

[0102] S601: Performing operation fault statistics on the operation performance index data of the hydropower unit and the abnormal frequency data of the operation sensor of the hydropower unit to obtain operation fault data;

[0103] By collecting statistics on operating performance indicator data and sensor abnormal frequency data, the fault type and frequency of hydropower units can be identified, helping operation and maintenance personnel to discover potential problems in a timely manner. Operational fault data provides an important basis for decision-making for the operation and maintenance team, helps to formulate effective maintenance strategies, and reduce the occurrence of faults. By analyzing fault data, the most common fault types can be identified, thereby optimizing the allocation of maintenance resources and improving operation and maintenance efficiency.

[0104] S602: Calculate the efficiency of the hydropower unit operation performance index data to obtain performance efficiency data;

[0105] By calculating the performance efficiency of the hydropower unit, we can intuitively understand the unit's power generation capacity and energy utilization efficiency, and provide basic data for subsequent analysis. The efficiency calculation results can reveal the deficiencies in the unit's operation. The operation and maintenance team can formulate improvement measures based on this to improve the overall operating efficiency. The efficiency data can be compared with historical data or industry standards to evaluate the unit's performance.

[0106] S603: Calculate the change rate according to the flow meter sensor data to obtain flow change rate data;

[0107] The calculation of the flow change rate can reflect the flow changes of the hydropower unit in real time during operation, helping the operation and maintenance personnel to adjust the operation strategy in time. The drastic change of flow indicates potential faults or abnormalities. Calculating the change rate helps to discover and deal with these problems in time. By monitoring the flow change rate, the operation and maintenance team can better control the operating status of the hydropower unit and improve safety and stability.

[0108] S604: Scoring the fault data, the performance efficiency data, and the flow change data to obtain corresponding scoring data;

[0109] By scoring fault data, performance efficiency data, and flow change data, the operating status of the hydropower unit can be comprehensively evaluated to form a comprehensive health scoring system. The scoring mechanism quantifies complex operating conditions into specific scoring data, making it easier for management to understand and monitor the health of the unit. The scoring results provide the operation and maintenance team with a clear basis for decision-making, helping them make more scientific choices in resource allocation and maintenance strategies.

[0110] S605: Calculate the comprehensive health status score according to the fault score data, the performance efficiency score data, and the flow rate change rate score data to obtain the operation health status data of the hydropower unit;

[0111] By comprehensively evaluating fault scores, performance efficiency scores, and flow rate change scores, an overall health status score for the hydropower unit can be formed, which helps operation and maintenance personnel fully understand the operating status of the unit. The health status score can help the operation and maintenance team identify potential high-risk units and take measures in advance to reduce the risk of failures. Through the health status score, operation and maintenance personnel can give priority to units with poor health status and improve maintenance efficiency and unit reliability.

[0112] The beneficial effects of the present invention are as follows:

[0113] 1. The present invention extracts operating sensor data, monitors various operating indicators of the hydropower unit in real time, and implements abnormality detection to detect equipment failures or abnormal operating conditions early and reduce the risk of failures. By performing frequency statistics on sensor anomalies, it helps operation and maintenance personnel identify sensors with high-frequency abnormalities as key monitoring objects and optimize the allocation of maintenance resources.

[0114] 2. The present invention comprehensively evaluates the operating performance of the hydropower unit by calculating the operating thermal efficiency and hydraulic efficiency, identifies the cause of efficiency loss, and calculates the health status by combining the operating performance indicators and the sensor abnormal frequency data to achieve a more comprehensive health assessment.

[0115] 3. The present invention takes into account multi-dimensional influencing factors, and through the establishment of a health assessment model, realizes intelligent assessment of the health status of the hydropower unit, reduces manual intervention, improves the accuracy and efficiency of the assessment, and improves the response speed to the health status of the equipment. By uploading the assessment data to the hydropower unit management cloud platform, centralized management and sharing of data are realized, and the transparency and traceability of information are improved. Through the management system of the cloud platform, health warnings can be automatically generated to guide operation and maintenance personnel to take timely maintenance measures, reduce failure rates, improve equipment reliability, and improve the hydropower unit's ability to process data from different sources. BRIEF DESCRIPTION OF THE DRAWINGS

[0116] Figure 1 It is a schematic diagram of the overall process of the method of the present invention.

[0117] Figure 2 It is a schematic diagram of the data preprocessing process of the present invention.

[0118] Figure 3 It is a schematic diagram of the sensor data anomaly detection process of the present invention.

[0119] Figure 4 It is a schematic diagram of the temperature anomaly detection process of the present invention.

[0120] Figure 5 It is a schematic diagram of the pressure anomaly detection process of the present invention.

[0121] Figure 6 It is a schematic diagram of the abnormal time series analysis process of the present invention.

[0122] Figure 7 It is a schematic diagram of the abnormal frequency statistics flow of the present invention.

[0123] Figure 8 It is a schematic diagram of the flow of obtaining the operating performance index data of a hydropower unit according to the present invention.

[0124] Fig. 9 It is a schematic diagram of the operating performance index analysis process of the present invention.

[0125] Fig.10 It is a schematic diagram of the health status calculation process of the present invention. DETAILED DESCRIPTION

[0126] The following description clearly and completely describes 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.

[0127] In the description of the embodiments of the present invention, it should be noted that the indicated orientation or positional relationship is based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the invention product is usually placed when in use, or the orientation or positional relationship commonly understood by those skilled in the art, or the orientation or positional relationship in which the invention product is usually placed when in use, which is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first" and "second" are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0128] In the description of the embodiments of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a direct connection, or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0129] Example 1

[0130] Embodiment 1 of the present invention discloses a method for evaluating the health status of a hydropower unit based on multidimensional data, such as Figure 1 As shown, the details are as follows:

[0131] S1: Obtain the operating data of the hydropower unit and perform data preprocessing;

[0132] The operating data includes various key parameters (such as power generation, water flow, pressure, temperature, etc.). Specifically, the data can be collected through sensors installed on the hydropower units. The collected data can be transmitted to the central data acquisition system through industrial communication protocols (such as MODBUS or OPC UA). The data acquisition frequency is set to once per minute to ensure sufficient real-time performance.

[0133] For example, the flow meter records the flow value once a minute, the pressure sensor records the pressure value, the temperature sensor records the inlet and outlet water temperatures, etc.; ultimately, these data will be stored in the database to form a structured data set for subsequent processing.

[0134] As a preferred embodiment, Figure 2 As shown, the data preprocessing, the specific process is as follows:

[0135] S101: Fill missing values ​​in the operation data of the hydropower unit to obtain filled operation data of the hydropower unit;

[0136] In the acquired hydropower unit operation data, missing values ​​may occur, such as missing data at certain time points due to sensor failure or communication interruption;

[0137] To solve this problem, specifically, filling can adopt mean interpolation, interpolation method, etc., and select appropriate filling strategy according to data characteristics to make the data more consistent with the actual situation;

[0138] In this embodiment, first, the missing values ​​are filled using linear interpolation, that is, interpolation calculations are performed based on adjacent data points to ensure the continuity of the time series. In addition, for variables whose missing value ratio exceeds a certain threshold, the mean or median can be used for filling. For example, for flow data, if there are 3 missing values ​​in a certain hour, the mean of the two hours before and after can be used to fill in the missing values ​​to generate "hydropower unit filling operation data".

[0139] S102: performing outlier processing (removing noise and outliers) on the filling operation data of the hydropower unit to obtain abnormal correction operation data of the hydropower unit;

[0140] Specifically, the Pandas library in Python can be used to apply deduplication and interpolation methods to the data set to ensure the integrity and accuracy of the data, and finally export a structured "operation data of hydropower units to be processed" file for subsequent processing.

[0141] In this embodiment, in the "hydropower unit filling operation data", there may be abnormal values, such as a sudden increase or decrease in flow rate, which is caused by sensor failure or external interference.

[0142] To this end, the Z-score method is used to detect outliers; the specific steps are to calculate the mean and standard deviation of the data set, and then compare the Z-score of each data point with the set threshold (for example, 3) to determine whether it is an outlier; for values ​​identified as abnormal, you can choose to replace them with the mean or use the average of the previous and next values ​​for correction. For example, if the flow data at a certain moment is -50, which is obviously an outlier, it can be replaced with the mean of other valid data in that time period. Finally, the "abnormal correction operation data of hydropower units" is obtained.

[0143] S103: performing standardization processing on the abnormal correction operation data of the hydropower unit (converting the data to the same dimension) to obtain the operation data of the hydropower unit to be processed;

[0144] In this embodiment, the "abnormal correction operation data of the hydropower unit" is standardized to eliminate the influence of different dimensions and ranges on subsequent analysis. The standardization process adopts the Z-score standardization method, subtracting the mean of the data set from each data point and dividing it by the standard deviation, thereby converting the data into a distribution with a mean of 0 and a standard deviation of 1. Specifically, assuming that the mean of the flow data is 100 and the standard deviation is 10, the data point with a flow of 120 at a certain moment is calculated as (120-100) / 10=2 after standardization; after completing the standardization, the "operation data of the hydropower unit to be processed" is generated, which provides a unified input format for subsequent analysis and modeling, ensuring the comparability and validity of the data.

[0145] S2: extracting the operating temperature sensor data of the hydropower unit and the operating pressure sensor data of the hydropower unit from the preprocessed operating data of the hydropower unit, and obtaining the operating temperature sensor data of the hydropower unit and the operating pressure sensor data of the hydropower unit;

[0146] Specifically, multiple temperature sensors and pressure sensors are installed in the monitoring system of the hydropower unit to monitor the inlet water temperature, outlet water temperature and internal pressure of the unit respectively. The data of these sensors are obtained in real time through the data acquisition system.

[0147] Specifically, data transmission can use industrial protocols (such as MODBUS or OPC UA) to transmit data from sensors to the central control system. Assuming that each sensor records data once a minute, a structured data set including temperature and pressure will eventually be generated, such as the fields "timestamp", "inlet water temperature", "outlet water temperature" and "unit pressure". The extracted data will be stored in the database for subsequent analysis.

[0148] S3: Performing anomaly detection on the operating sensor data of the hydropower unit to obtain the operating sensor abnormality data of the hydropower unit;

[0149] As a preferred embodiment, Figure 3 As shown, the anomaly detection is as follows:

[0150] S301: Perform temperature anomaly detection on the operating temperature sensor data of the hydropower unit to obtain the operating temperature sensor anomaly data of the hydropower unit;

[0151] Specifically, the IQR (interquartile range) method may be used to identify temperature anomalies.

[0152] First, the first quartile (Q1) and the third quartile (Q3) of the temperature data are calculated, and then the interquartile range (IQR = Q3-Q1) is calculated; the outlier judgment standard is set as follows: data points lower than Q1-1.5*IQR or higher than Q3+1.5*IQR will be regarded as abnormal data.

[0153] For example, if the inlet water temperature Q1 is 15°C and Q3 is 25°C, the abnormal value is the data below 7.5°C or above 32.5°C. Finally, the "hydropower unit temperature sensor abnormal data" is generated, recording the timestamps and temperature values ​​of all abnormalities.

[0154] As a preferred embodiment, Figure 4 As shown in the figure, the temperature anomaly detection is performed on the operating temperature sensor data of the hydropower unit. The specific process is as follows:

[0155] S3011: extracting turbine temperature sensor data from the operating temperature sensor data of the hydropower unit to obtain turbine temperature sensor data;

[0156] In this embodiment, multiple temperature sensors are installed in the hydropower unit to monitor the temperature of the turbine in real time. These sensors can be connected to the data acquisition system through industrial communication protocols (such as MODBUS or OPC UA). The data acquisition system records the turbine temperature data once a minute, and the data includes the fields "timestamp", "turbine inlet temperature" and "turbine outlet temperature". By writing a Python script, these sensor data are imported into a data frame using the Pandas library, and sorted and cleaned according to the timestamp. In the data cleaning process, invalid or missing data is removed, and finally a structured "turbine temperature sensor data" data set is obtained for subsequent analysis.

[0157] S3012: Perform high temperature statistics according to the turbine temperature sensor data, and obtain the turbine high temperature sensor data based on a set first high temperature threshold;

[0158] In this embodiment, the extracted "turbine temperature sensor data" is used to perform high temperature statistics. A high temperature threshold, such as 65°C, is set as the standard for judging high temperature events. By traversing the temperature data and using conditional filtering, all temperature records exceeding 65°C are extracted. Using the conditional filtering function in Python, a new data frame is created to record the timestamps and temperature values ​​of all high temperature events. For example, if the turbine temperature reaches 70°C and 75°C within a certain time period, the program will include these data points in the "turbine high temperature sensor data" to form a new structured data set.

[0159] S3013: Perform frequency statistics on the high-temperature sensor data of the turbine, and obtain low-frequency high-temperature sensor data of the turbine based on a set frequency threshold;

[0160] In this embodiment, frequency statistics are performed on the "turbine high temperature sensor data" to identify the frequency of high temperature events. The timestamp is converted into a time series of hours or minutes, and the number of high temperature events in each time period is counted. Use Python's groupby function to summarize high temperature events in the same time period. For example, if 3 high temperature events are recorded within a certain hour, the high temperature frequency of that hour is 3. Finally, the "turbine low-frequency high temperature sensor data" is generated to record the time period with a low frequency of high temperature events (for example, less than 1 high temperature event per hour) for subsequent analysis.

[0161] S3014: Setting a second high temperature threshold value according to the low-frequency high temperature sensing data of the turbine to obtain high temperature threshold value data;

[0162] In this embodiment, the high temperature threshold is set according to the "turbine low-frequency high temperature sensor data". First, the data of low-frequency high temperature events is analyzed and a reasonable threshold is selected, for example, the high temperature threshold is set to 63°C (i.e., lower than the initial 65°C). Statistical analysis methods (such as histograms) are used to observe the distribution of temperature data to determine whether the set threshold can effectively distinguish between normal and abnormal high temperatures. Finally, the set high temperature threshold is recorded in a configuration file for use in subsequent high temperature detection.

[0163] S3015: Perform abnormal high temperature detection on the turbine temperature sensor data according to the high temperature threshold data to obtain abnormal operating temperature sensor data of the hydropower unit;

[0164] In this embodiment, the "turbine temperature sensor data" is subjected to abnormal high temperature detection based on the set high temperature threshold data. The detection algorithm is written in Python to traverse the temperature data and determine whether there are data points that exceed the set high temperature threshold (for example, 63°C). If it is found that the turbine temperature at a certain point in time exceeds the threshold, the point is marked as abnormal. During the abnormality detection process, the timestamp and temperature value of each abnormal event are recorded. For example, if the turbine temperature is 65°C at a certain point in time, it is recorded as "abnormal temperature sensor data of the hydropower unit". Finally, an abnormal record containing a timestamp and abnormal temperature value is output to facilitate subsequent fault analysis and early warning.

[0165] S302: Performing pressure anomaly detection on the operating pressure sensor data of the hydropower unit to obtain the operating pressure sensor anomaly data of the hydropower unit;

[0166] Similarly, the IQR method can be used. First, Q1 and Q3 are calculated from the pressure data set, and then the IQR is obtained; according to the set criteria, it is determined which data points are outliers.

[0167] For example, if the Q1 of the unit pressure is 90kPa and Q3 is 110kPa, then the data below 82.5kPa or above 117.5kPa is judged as an abnormal value. Through code implementation (such as the Pandas library in Python), all abnormal pressure data are screened out and their occurrence time is recorded, and finally the "abnormal pressure sensor data of the hydropower unit" is formed.

[0168] As a preferred embodiment, Figure 5 As shown in the figure, the pressure anomaly detection is performed on the operating pressure sensor data of the hydropower unit. The specific process is as follows:

[0169] S3021: extracting turbine pressure sensor data from the operating pressure sensor data of the hydropower unit to obtain turbine pressure sensor data;

[0170] In this embodiment, it is necessary to extract the pressure sensing data related to the turbine from the sensor data of the hydropower unit. These data usually include the output of the pressure sensors at various parts of the turbine, such as the inlet, outlet and pressure measurement values ​​at different levels of the turbine. By using data extraction tools or programming scripts, the data segments related to the turbine can be filtered out from the overall sensor data stream. For example, a data processing program is written in Python to extract the data segments of the turbine pressure sensor by specifying the sensor ID and time range to ensure accurate acquisition of the pressure data during the operation period of the turbine.

[0171] S3022: extracting intake pressure sensing features according to turbine pressure sensing data to obtain intake pressure sensing data;

[0172] In this embodiment, the extracted turbine pressure sensor data needs to be further extracted to extract the pressure characteristics related to the air inlet. The air inlet pressure is usually one of the key indicators for evaluating the performance of the turbine, which affects the stable operation and efficiency of the unit. Through data analysis tools such as MATLAB or Pandas library, the statistical characteristics of the pressure (such as mean, variance, kurtosis, etc.) can be calculated, or the frequency domain characteristics of the pressure signal can be extracted using spectrum analysis methods. For example, for time series data, a MATLAB script can be written to perform sliding window analysis to extract the fluctuation frequency and amplitude of the air inlet pressure for subsequent anomaly detection and model building.

[0173] S3023: collecting time series data according to the air inlet pressure sensor data to obtain air inlet pressure sensor time series data;

[0174] In this embodiment, the time series of the inlet pressure sensor data after feature extraction is collected. These data are obtained by recording the output of the inlet pressure sensor in real time or regularly, and usually include timestamps and pressure values. Data collection can use a real-time data stream acquisition system to ensure high-frequency and low-latency data acquisition, or establish a historical database through regular data downloads. For example, use data acquisition software to connect to real-time sensors and store the inlet pressure data as a CSV file for subsequent analysis and modeling.

[0175] S3024: performing abnormal time series analysis according to the air inlet pressure sensor time series data to obtain abnormal time series data of the air inlet pressure sensor;

[0176] In this embodiment, the collected inlet pressure sensor time series data is subjected to anomaly detection and analysis. The anomaly detection method may include a statistical method (such as mean, standard deviation), a machine learning algorithm (such as support vector machine, random forest) or a time series analysis (such as ARIMA model). For example, the anomaly detection algorithm is implemented using Python's Scikit-Learn library to train and predict the inlet pressure data, and identify pressure fluctuations or abnormal events that exceed the normal range for subsequent modeling and monitoring.

[0177] As a preferred embodiment, Figure 6 As shown, the abnormal time series analysis is performed based on the intake pressure sensor time series data. The specific process is as follows:

[0178] S30241: Draw an air inlet pressure sensor time series graph for the air inlet pressure sensor time series data to obtain an air inlet pressure sensor time series graph;

[0179] In this embodiment, it is necessary to draw a time series graph of the air inlet pressure sensor. This can be achieved by using data visualization tools such as Matplotlib in Python or professional data analysis software such as Tableau. The time series graph shows the trend of pressure changes over time, helping us observe the periodicity, trends and anomalies.

[0180] S30242: Perform residual analysis based on the intake pressure sensor time series diagram to obtain residual data;

[0181] In this embodiment, the focus is on the random fluctuation part after removing seasonality and trend. By subtracting the seasonal component and trend from the original time series data, the residual data obtained reflects other changes besides seasonality and trend. For example, seasonal decomposition methods such as STL (Seasonal and Trend decomposition using Loess) or time series decomposition functions can be used.

[0182] S30243: Perform seasonal fluctuation analysis based on the intake pressure sensor time series diagram to obtain seasonal fluctuation data;

[0183] In this embodiment, seasonal components are obtained by seasonal decomposition. Specifically, each seasonal cycle can be analyzed, such as seasonal fluctuations within a year or seasonal changes within a shorter period. For example, in the automotive industry, different fluctuation patterns of the air intake pressure in winter and summer are observed.

[0184] S30244: construct an abnormal time detection model based on residual data and seasonal fluctuation data to obtain an abnormal time detection model;

[0185] In this embodiment, the abnormal time detection model is constructed to automatically identify atypical patterns or abnormal events in the time series. This can be achieved by using statistical methods such as Z-score or machine learning algorithms such as Isolation Forest. The training of the model is usually based on historical data, and the goal is to identify anomalies in new data in the future. For example, the anomaly detection model can identify abnormal pressure values ​​that suddenly appear in the air intake pressure sensor data, indicating that there is a problem with the vehicle or abnormal operation.

[0186] S30245: performing abnormal time series detection on the air inlet pressure sensor time series data according to the abnormal time detection model to obtain abnormal time series data of the air inlet pressure sensor;

[0187] In this embodiment, the abnormal time series detection stage applies the trained anomaly detection model to the actual data to identify and mark the abnormal events that occur. This can be detected by applying the model in real time through the monitoring system, or by periodically checking the anomalies in the historical data in a batch processing manner. For example, a car suddenly has abnormal intake pressure while driving at high speed, which requires timely maintenance or alerting the driver.

[0188] S3025: constructing a monitoring model according to the air inlet pressure sensor abnormality time series data to obtain an air inlet pressure sensor abnormality monitoring model;

[0189] In this embodiment, a monitoring model for monitoring inlet pressure sensor anomalies is constructed based on abnormal time series data. This involves developing a monitoring algorithm or establishing a rule engine to monitor the change pattern and trend of the inlet pressure in real time. For example, a logistic regression model or a neural network model is used to combine historical abnormal data and real-time sensor input to establish a monitoring model so that an alarm can be issued in time or preventive measures can be taken when pressure anomalies occur.

[0190] S3026: Perform abnormal pressure sensing monitoring on the operating pressure sensing data of the hydropower unit according to the air inlet pressure sensing abnormality monitoring model to obtain the abnormal pressure sensing data of the hydropower unit.

[0191] In this embodiment, the constructed air inlet pressure sensor abnormality monitoring model is used to perform real-time monitoring and abnormality detection on the pressure sensor data of the hydropower unit. This includes monitoring the real-time sensor output, comparing it with the model prediction or rules, and identifying the pressure sensor data of the hydropower unit related to the air inlet pressure abnormality. For example, the developed monitoring model is deployed in the production environment to receive and analyze the pressure data in real time, identify and respond to abnormal situations in a timely manner, and ensure the safe operation and stable performance of the unit.

[0192] S303: Integrate abnormal sensor data according to abnormal temperature sensor data and abnormal pressure sensor data of the hydropower unit operation to obtain abnormal operation sensor data of the hydropower unit operation.

[0193] In this embodiment, the abnormal sensing data is integrated based on the identified “abnormal temperature sensing data of the hydropower unit” and “abnormal pressure sensing data of the hydropower unit”.

[0194] Specifically, a new data structure is created, which contains the timestamp of the abnormal data, the sensor type (temperature or pressure), the abnormal value, and the abnormal type.

[0195] For example, if the water inlet temperature is abnormally 35°C and the unit pressure is abnormally 120kPa at a certain moment, the integrated abnormal data will include the timestamp, sensor type (such as "temperature" or "pressure"), the corresponding abnormal value and its description, and finally generate "hydropower unit operation sensor abnormal data" for subsequent analysis and processing.

[0196] S4: Performing sensor abnormality frequency statistics on the sensor abnormality data of the hydropower unit operation, and filtering and obtaining abnormal data whose abnormality frequency is higher than the set threshold according to the set abnormality frequency threshold;

[0197] In this embodiment, the frequency statistics of the integrated "hydropower unit operation sensor abnormal data" are performed to identify high-frequency abnormal situations. The Pandas library in Python can be used to perform statistics based on the number of occurrences of each abnormal type.

[0198] For example, the frequency of abnormal water inlet temperature and abnormal unit pressure data is counted. By setting a threshold, for example, if the frequency of a certain abnormality exceeds the set threshold (such as more than 5 times per hour), it is marked as "high-frequency abnormality". Finally, "hydropower unit operation sensor high-frequency abnormality data" is generated to record the type, number of occurrences and time range of high-frequency abnormalities, providing a basis for subsequent fault analysis and early warning.

[0199] As a preferred embodiment, Figure 7 As shown in the figure, the sensor abnormality frequency statistics of the hydropower unit operation sensor abnormality data are performed, as follows:

[0200] S401: Performing interval statistics on the sensor abnormality data of the hydropower unit operation, and obtaining the sensor abnormality data of the hydropower unit operation in short intervals based on the set intervals;

[0201] In this embodiment, the interval statistics of the abnormal data of the hydropower unit operation sensor are performed to identify abnormal situations that occur frequently in a short period of time. For example, we can set a time window, such as every hour or every ten minutes, to count the abnormal data. By setting a suitable threshold, such as the number of abnormal occurrences exceeding a certain multiple of the normal range, these short-interval abnormal data can be identified. For example, if the number of abnormalities recorded by a sensor within an hour exceeds the set threshold, we can determine that there is an operation abnormality during this period.

[0202] S402: classifying the sensor type of the short-interval operation sensor abnormality data of the hydropower unit to obtain the short-interval operation sensor type abnormality data of the hydropower unit;

[0203] In this embodiment, the sensor type classification of the short-interval operation sensor abnormality data of the hydropower unit is to further analyze the characteristics and impact of each abnormality. For example, we can classify the abnormal data according to the specific manifestation of the abnormality and the sensor type. For example, a sensor may be abnormal due to voltage fluctuations or temperature changes, while another sensor may alarm due to humidity abnormalities. Through this classification, we can more accurately understand the specific impact of each abnormality on the operation of the unit, and then formulate targeted maintenance strategies.

[0204] S403: Perform frequency statistics according to the short-interval operation sensor type abnormal data, and obtain high-frequency short-interval sensor abnormal data based on the set frequency;

[0205] In this embodiment, frequency statistics are performed based on short-interval operation sensor type abnormal data in order to find out abnormal situations that occur frequently so as to give priority to processing. For example, we can count the frequency of occurrence of each sensor abnormality type within a period of time, such as daily, weekly or monthly statistics. By identifying which abnormality types exceed the preset threshold in frequency, we can determine which problems are most urgent and need to be handled with priority.

[0206] S404: performing sensor anomaly clustering on the hydropower unit operation sensor anomaly data according to the high-frequency short-interval sensor anomaly data to obtain the hydropower unit operation sensor high-frequency anomaly data;

[0207] In this embodiment, the sensor anomaly clustering of the hydropower unit operation sensor anomaly data based on the high-frequency short-interval sensor anomaly data is to identify anomaly groups with similar characteristics or root causes. For example, we can use a clustering algorithm such as k-means to group high-frequency anomaly data and find anomaly data sets with similar patterns. This method can help engineers diagnose and solve problems more effectively, while optimizing maintenance strategies and improving the stability and reliability of the unit.

[0208] S5: Calculate the thermal efficiency and hydraulic efficiency of the hydropower unit operation sensor data, and analyze the operation performance index to obtain the hydropower unit operation performance index data;

[0209] As a preferred embodiment, Figure 8 As shown in the figure, the operating performance index data of the hydropower unit is obtained. The specific process is as follows:

[0210] S501: extracting flow meter sensor data and water dragon measuring instrument sensor data according to the operating sensor data of the hydropower unit;

[0211] In this embodiment, it is necessary to obtain sensor data of the operation of the hydropower unit, including sensor data of flow meters and water dragon measuring instruments. Flow meter sensor data usually involves measuring the speed and amount of water flow, which can be measured by ultrasound, electromagnetic induction or pressure difference. For example, the flow rate is measured by ultrasound, and the flow rate is calculated based on the time it takes for the sound wave to propagate in the fluid. This method is suitable for liquids and gases. Water dragon measuring instruments are usually based on the principle of pressure difference, and the flow rate is inferred by the change in water flow pressure, which is suitable for liquid media.

[0212] S502: Calculate the thermal efficiency of the water motor according to the flow meter sensor data and the water dragon measuring instrument sensor data;

[0213] In this embodiment, the operating thermal efficiency of the hydropower unit can be calculated based on the sensor data obtained from the flow meter and the water faucet measuring instrument. This involves analyzing the energy difference between input and output, and the commonly used methods include thermal power measurement method and thermal efficiency formula. For example, by measuring the heat entering and leaving the hydropower unit, the computer unit can calculate the efficiency of converting the kinetic energy of water into mechanical energy.

[0214] S503: Obtaining turbine output power;

[0215] In this embodiment, obtaining the output power of the turbine is one of the key indicators for evaluating the overall performance of the hydroelectric generator set. This can be achieved by measuring the output current and voltage of the power meter or the generator to calculate the output power.

[0216] S504: Calculating the hydraulic efficiency of the hydroelectric machine according to the output power of the hydro turbine, the flow meter sensor data, and the water dragon measuring instrument sensor data;

[0217] In this embodiment, the hydraulic efficiency of the hydroelectric unit is calculated based on the output power of the turbine, the flow meter sensor data, and the water dragon measuring instrument sensor data. The hydraulic efficiency measures the efficiency of converting water flow energy into mechanical energy, and is usually calculated by comparing theoretical power and actual output power. For example, the specific value of the hydraulic efficiency is derived by combining the flow meter data and the actual output power of the turbine.

[0218] S505: analyzing the operation performance index according to the operation thermal efficiency data of the hydroelectric generator and the operation hydraulic efficiency data of the hydroelectric generator unit to obtain the operation performance index data of the hydroelectric generator unit;

[0219] In this embodiment, the operating thermal efficiency and hydraulic efficiency data of the hydropower unit are combined to perform performance index analysis to evaluate its overall operating performance. This includes comparing efficiency changes under different operating conditions, analyzing the energy consumption and loss of the equipment, etc. For example, through data simulation or real-time monitoring, the reasons for the decline in unit efficiency under certain conditions can be discovered, thereby optimizing the operating strategy.

[0220] As a preferred embodiment, Fig. 9 As shown, the specific process of the operation performance index analysis is as follows:

[0221] S5051: Calculate the average efficiency and fluctuation of the hydropower unit according to the thermal efficiency data and hydraulic efficiency data of the hydropower unit;

[0222] In this embodiment, the thermal efficiency data and hydraulic efficiency data of each operation cycle are obtained from the operation record of the hydropower unit. Thermal efficiency is usually calculated by measuring the ratio between the unit output power and the input thermal energy, while hydraulic efficiency is based on the ratio of the actual power of the water flow to the theoretical power. After obtaining these data, the thermal efficiency and hydraulic efficiency are respectively averaged to obtain the average thermal efficiency and the average hydraulic efficiency. In order to calculate volatility, statistical methods such as standard deviation or coefficient of variation can be used to evaluate the degree of fluctuation of efficiency data. For example, if the standard deviation of thermal efficiency is 0.5% and the standard deviation of hydraulic efficiency is 0.3%, it can be concluded that there is a certain efficiency fluctuation during the operation of the unit. These calculation results provide basic data for subsequent performance index analysis.

[0223] S5052: Acquire rated efficiency data of the hydropower unit and rated efficiency fluctuation data of the hydropower unit;

[0224] In this embodiment, the rated efficiency data of the hydropower unit is obtained by consulting the design specifications and manufacturing standards. These data are usually given in the technical documents of the unit, indicating the theoretical optimal efficiency of the unit under the design conditions. At the same time, the rated efficiency fluctuation data is obtained, which is usually provided by the manufacturer and describes the efficiency fluctuation range of the unit under rated conditions. For example, the rated efficiency is 90%, and the rated efficiency fluctuation data indicates that there is a fluctuation of ±1% in actual operation. These data are helpful in evaluating the stability and reliability of the unit under normal operating conditions.

[0225] S5053: performing efficiency performance index comparison with average efficiency data of the hydropower unit according to rated efficiency data of the hydropower unit to obtain efficiency performance index of the hydropower unit;

[0226] In this embodiment, the average efficiency data of the hydropower unit is compared with the rated efficiency data. The difference between the actual average efficiency and the rated efficiency is calculated and expressed as a percentage. For example, if the actual average efficiency is 88% and the rated efficiency is 90%, then the comparison result of the efficiency performance index is that the actual efficiency is 2% lower than the rated efficiency. This can be achieved by calculating the deviation between the actual efficiency and the rated efficiency, thereby evaluating the degree of compliance between the actual operating efficiency of the unit and the expected design efficiency.

[0227] S5054: performing a fluctuation index comparison on the efficiency fluctuation data of the hydropower unit according to the rated efficiency fluctuation data of the hydropower unit to obtain the fluctuation index data of the hydropower unit;

[0228] In this embodiment, the efficiency fluctuation data is compared with the rated efficiency fluctuation data in step S352. The difference between the actual efficiency fluctuation and the rated fluctuation is compared using the same statistical method, such as standard deviation. For example, if the actual efficiency fluctuation of the unit is ±0.5% and the rated fluctuation is ±0.3%, the fluctuation in actual operation is higher than the rated standard. This comparison can reveal the change in the stability of the unit in actual operation and whether there are additional sources of fluctuation.

[0229] S5055: Integrate the index data according to the efficiency performance index of the hydropower unit and the fluctuation index data of the hydropower unit to obtain the operation performance index data of the hydropower unit;

[0230] In this embodiment, a comprehensive analysis is performed by combining the efficiency performance index and the fluctuation index data. For example, the efficiency performance index and the fluctuation index are integrated through weighted average or other comprehensive evaluation methods to generate a comprehensive operation performance index. For example, a comprehensive performance score is obtained by comprehensively considering the extent to which the actual efficiency is lower than the rated efficiency and the extent to which the actual volatility exceeds the rated volatility. This integration can help evaluate the overall performance of the hydropower unit under actual operating conditions, including efficiency level and stability, and provide data support for equipment maintenance and optimization.

[0231] S6: performing health status calculation based on the analyzed operating performance index data of the hydropower unit to obtain operating health status data of the hydropower unit;

[0232] As a preferred embodiment, Fig.10 As shown, the operation health status data of the hydropower unit is obtained. The specific process is as follows:

[0233] S601: Performing operation fault statistics on the operation performance index data of the hydropower unit and the abnormal frequency data of the operation sensor of the hydropower unit to obtain operation fault data;

[0234] In this embodiment, during the operation of the hydropower unit, various performance indicator data are first collected, such as unit speed, power generation, current, etc. At the same time, the abnormal frequency of the sensors of the hydropower unit is recorded, such as the over-limit alarm of the temperature sensor, the abnormal signal of the vibration sensor, etc. By performing statistical analysis on these data, the frequency and type of fault occurrence are identified. Specifically, the abnormal data recorded by the sensor is classified by time and type, and the number of occurrences and time distribution of each fault are counted. In this way, operating fault data is generated, including the frequency, time period and abnormal information of related sensors of each fault, which provide a basis for subsequent fault diagnosis and health assessment.

[0235] S602: Calculate the efficiency of the hydropower unit operation performance index data to obtain performance efficiency data;

[0236] In this embodiment, the efficiency calculation is performed on the performance index data of the hydropower unit. First, the actual power generation, input water flow and other relevant performance indicators of the unit are collected. The operating efficiency of the unit is obtained by calculating the ratio of the actual power generation to the theoretical maximum power generation. For example, if the actual power generation of the unit is 5MW, and the theoretical maximum power is 6MW according to the water flow and equipment parameters, the efficiency is 5 / 6 or 83.33%. In this way, the performance efficiency data of the unit can be obtained, which is used to measure the gap between the actual operating performance of the unit and its theoretical optimal performance.

[0237] S603: Calculate the change rate according to the flow meter sensor data to obtain flow change rate data;

[0238] In this embodiment, the flow rate change rate is calculated based on the flow meter sensor data. First, the flow data within a certain time interval is recorded. For example, the flow value is recorded once every hour. Then the flow difference between adjacent time points is calculated. The specific operation is: set a time window (such as 24 hours), and calculate the change between the current flow value and the flow value at the previous time point at each time point. Through these calculations, the flow rate change data is obtained. These data can reveal the fluctuation of the flow and help analyze the impact of flow changes on the operation of the unit.

[0239] S604: Scoring the fault data, the performance efficiency data, and the flow change data to obtain corresponding scoring data;

[0240] In this embodiment, scoring is performed based on fault data, performance efficiency data, and flow rate change data. First, a standard is determined for fault scoring, for example, a score is assigned to each fault based on the frequency and severity of the fault. For performance efficiency, an efficiency scoring standard is set, for example, a lower score is given when the efficiency is below a certain threshold. The flow rate change rate also needs to be scored, and a lower score is given when the flow fluctuation is large. By integrating these three types of scoring data, fault scoring data, performance efficiency scoring data, and flow rate change scoring data can be obtained.

[0241] S605: Calculate the comprehensive health status score according to the fault score data, the performance efficiency score data, and the flow rate change rate score data to obtain the operation health status data of the hydropower unit;

[0242] In this embodiment, the comprehensive health status score is calculated based on the fault score data, the performance efficiency score data, and the flow rate change score data. A weight is set for each score data, and the weight is assigned according to its importance in the unit health status assessment. For example, the fault score accounts for 40%, the performance efficiency score accounts for 30%, and the flow rate change score accounts for 30%. After multiplying these score data by their respective weights, the sum is obtained to obtain the comprehensive health status score. This comprehensive score can reflect the overall health status of the hydropower unit.

[0243] S7: constructing a health assessment model according to the health status data of the hydropower unit to perform health status assessment on the hydropower unit, uploading the health status assessment data to the hydropower unit management cloud platform, and executing a preset health warning task for the corresponding hydropower unit;

[0244] In this embodiment, a health assessment model is constructed based on the health status data of the hydropower unit. First, long-term health status score data are collected and analyzed to identify the relationship between the health status and faults, performance efficiency and flow rate change rate. A model is established using statistical methods or machine learning algorithms (such as regression analysis or decision trees), and the health status score data is input to predict the future health status of the unit. The model can include training and verification stages of historical data, and the prediction accuracy can be improved by continuously optimizing the model parameters. The constructed health assessment model will be able to evaluate the health status of the unit in real time during actual operation and provide a basis for maintenance decisions.

[0245] The present invention is not limited to the above-mentioned specific embodiments, but extends to any new features or any new combination disclosed in this specification, as well as any new method or process steps or any new combination disclosed.

Claims

1. A method for evaluating the health status of a hydropower unit based on multidimensional data, characterized in that: include: S1: Obtain the operating data of the hydropower unit and perform data preprocessing; S2: extracting sensor data from the preprocessed hydropower unit operation data to obtain hydropower unit operation sensor data; S3: Performing abnormality detection on the operating sensor data of the hydropower unit to obtain abnormal operating sensor data of the hydropower unit; S4: performing sensor abnormality frequency statistics on the sensor abnormality data of the hydropower unit operation, and filtering and obtaining abnormal data whose abnormality frequency is higher than the set threshold according to the set abnormality frequency threshold; S5: Calculate the thermal efficiency and hydraulic efficiency of the hydropower unit operation sensor data, and analyze the operation performance index to obtain the hydropower unit operation performance index data; S6: Calculate the health status of the hydropower unit based on the analyzed operating performance index data to obtain the operating health status data of the hydropower unit; S7: construct a health assessment model based on the operating health status data of the hydropower unit to perform health status assessment of the hydropower unit, upload the health status assessment data to the hydropower unit management cloud platform, and execute the preset corresponding hydropower unit health warning task.

2. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 1 is characterized in that: In step S1, the data is preprocessed, and the specific process is as follows: S101: Fill missing values ​​in the operation data of the hydropower unit to obtain filled operation data of the hydropower unit; S102: performing outlier processing on the filling operation data of the hydropower unit to obtain abnormal correction operation data of the hydropower unit; S103: Standardize the abnormal correction operation data of the hydropower unit to obtain the operation data of the hydropower unit to be processed.

3. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 1 is characterized in that: In step S3, abnormality detection is performed on the operating sensor data of the hydropower unit, including: S301: Perform temperature anomaly detection on the operating temperature sensor data of the hydropower unit to obtain the operating temperature sensor anomaly data of the hydropower unit; S302: Performing pressure anomaly detection on the operating pressure sensor data of the hydropower unit to obtain the operating pressure sensor anomaly data of the hydropower unit; S303: Integrate abnormal sensor data according to abnormal temperature sensor data and abnormal pressure sensor data of the hydropower unit operation to obtain abnormal operation sensor data of the hydropower unit operation.

4. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 3 is characterized in that: In step S3, the temperature anomaly detection is performed on the operating temperature sensor data of the hydropower unit. The specific process is as follows: S3011: extracting turbine temperature sensor data from the operating temperature sensor data of the hydropower unit to obtain turbine temperature sensor data; S3012: Perform high temperature statistics according to the turbine temperature sensor data, and obtain the turbine high temperature sensor data based on a set first high temperature threshold; S3013: Perform frequency statistics on the high-temperature sensor data of the turbine, and obtain low-frequency high-temperature sensor data of the turbine based on a set frequency threshold; S3014: Setting a second high temperature threshold value according to the low-frequency high temperature sensing data of the turbine to obtain high temperature threshold value data; S3015: Perform abnormal high temperature detection on the turbine temperature sensor data according to the high temperature threshold data to obtain abnormal operating temperature sensor data of the hydropower unit.

5. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 3 is characterized in that: In step S3, the pressure anomaly detection is performed on the operating pressure sensor data of the hydropower unit. The specific process is as follows: S3021: extracting turbine pressure sensor data from the operating pressure sensor data of the hydropower unit to obtain turbine pressure sensor data; S3022: extracting intake pressure sensing features according to turbine pressure sensing data to obtain intake pressure sensing data; S3023: collecting time series data according to the air inlet pressure sensor data to obtain air inlet pressure sensor time series data; S3024: performing abnormal time series analysis according to the air inlet pressure sensor time series data to obtain abnormal time series data of the air inlet pressure sensor; S3025: constructing a monitoring model according to the air inlet pressure sensor abnormality time series data to obtain an air inlet pressure sensor abnormality monitoring model; S3026: Perform abnormal pressure sensing monitoring on the operating pressure sensing data of the hydropower unit according to the air inlet pressure sensing abnormality monitoring model to obtain the abnormal pressure sensing data of the hydropower unit.

6. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 5 is characterized in that: The abnormal time series analysis is performed based on the intake pressure sensor time series data. The specific process is as follows: S30241: Draw an air inlet pressure sensor time series graph for the air inlet pressure sensor time series data to obtain an air inlet pressure sensor time series graph; S30242: Perform residual analysis based on the intake pressure sensor time series diagram to obtain residual data; S30243: Perform seasonal fluctuation analysis based on the intake pressure sensor time series diagram to obtain seasonal fluctuation data; S30244: construct an abnormal time detection model according to the residual data and seasonal fluctuation data to obtain an abnormal time detection model; S30245: Perform abnormal time series detection on the air inlet pressure sensor time series data according to the abnormal time detection model to obtain the air inlet pressure sensor abnormal time series data.

7. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 1 is characterized in that: In step S4, the sensor abnormality frequency statistics of the hydropower unit operation sensor abnormality data are performed as follows: S401: Performing interval statistics on the sensor abnormality data of the hydropower unit operation, and obtaining the sensor abnormality data of the hydropower unit operation in short intervals based on the set intervals; S402: classifying the sensor type of the short-interval operation sensor abnormality data of the hydropower unit to obtain the short-interval operation sensor type abnormality data of the hydropower unit; S403: Perform frequency statistics according to the short-interval operation sensor type abnormal data, and obtain high-frequency short-interval sensor abnormal data based on the set frequency; S404: performing sensor anomaly clustering on the hydropower unit operation sensor anomaly data according to the high-frequency short-interval sensor anomaly data to obtain the hydropower unit operation sensor high-frequency anomaly data.

8. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 1 is characterized in that: In step S5, the operating performance index data of the hydropower unit is obtained, and the specific process is as follows: S501: extracting flow meter sensor data and water dragon measuring instrument sensor data according to the operating sensor data of the hydropower unit; S502: Calculate the thermal efficiency of the water motor according to the flow meter sensor data and the water dragon measuring instrument sensor data; S503: Obtaining turbine output power; S504: Calculating the hydraulic efficiency of the hydroelectric machine according to the turbine output power, flow meter sensor data, and water dragon measuring instrument sensor data; S5 05: Perform operation performance index analysis based on the thermal efficiency data of the hydropower unit and the hydraulic efficiency data of the hydropower unit to obtain the operation performance index data of the hydropower unit.

9. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 8 is characterized in that: The specific process of the operation performance index analysis is as follows: S5051: Calculate the average efficiency and fluctuation of the hydropower unit based on the thermal efficiency data and hydraulic efficiency data of the hydropower unit; S5052: Obtaining rated efficiency data of the hydropower unit and rated efficiency fluctuation data of the hydropower unit; S5053: performing efficiency performance index comparison with average efficiency data of the hydropower unit according to rated efficiency data of the hydropower unit to obtain efficiency performance index of the hydropower unit; S5054: performing a fluctuation index comparison on the efficiency fluctuation data of the hydropower unit according to the rated efficiency fluctuation data of the hydropower unit to obtain the fluctuation index data of the hydropower unit; S5055: Integrate the index data according to the efficiency performance index of the hydropower unit and the fluctuation index data of the hydropower unit to obtain the operation performance index data of the hydropower unit.

10. The method for evaluating the health status of a hydropower unit based on multidimensional data according to claim 1, characterized in that: In step S6, the operating health status data of the hydropower unit is obtained, and the specific process is as follows: S601: Performing operation fault statistics on the operation performance index data of the hydropower unit and the abnormal frequency data of the operation sensor of the hydropower unit to obtain operation fault data; S602: Calculate the efficiency of the hydropower unit operation performance index data to obtain performance efficiency data; S603: Calculate the change rate according to the flow meter sensor data to obtain flow change rate data; S604: Scoring the fault data, the performance efficiency data, and the flow change data to obtain corresponding scoring data; S605: Calculate the comprehensive health status score according to the fault score data, the performance efficiency score data and the flow rate change rate score data to obtain the operation health status data of the hydropower unit.

Images