Hydroelectric generating set health state monitoring method and system based on machine learning

Abstract

The invention relates to the technical field of hydroelectric generating set operation and maintenance management, in particular to a hydroelectric generating set health state monitoring method and system based on machine learning, and realizes more accurate, efficient and comprehensive health management of a hydroelectric generating set. The method comprises the following steps: monitoring operating parameters of a hydroelectric generating set to obtain multi-dimensional operating parameter information; performing abnormal state recognition on the multi-dimensional operation parameter information by using a preset hydroelectric generating set state recognition model to obtain abnormal state information of the hydroelectric generating set; identifying and extracting trend characteristics of the abnormal state information of the hydroelectric generating set obtained at a plurality of continuous preset time nodes, and obtaining abnormal trend information of the hydroelectric generating set; comprehensively considering the multi-dimensional operation parameter information, the abnormal state information of the hydroelectric generating set and the abnormal trend information of the hydroelectric generating set to obtain a health state analysis index of the hydroelectric generating set; and based on a preset health state index threshold, performing state health evaluation on the health state analysis index of the hydroelectric generating set.

Description

Technical Field

[0001] The present invention relates to the technical field of hydropower unit operation and maintenance management, and in particular to a method and system for monitoring the health status of a hydropower unit based on machine learning. Background Art

[0002] As the importance of hydropower energy in the power system increases, the operating stability and reliability of hydropower units, as the core equipment for hydropower energy production, are directly related to the safe and economic operation of the power system. Hydropower units are usually composed of multiple subsystems such as turbines, generators, speed governors, and control systems. During long-term operation, due to mechanical wear, electrical aging, environmental changes, and control logic failure, the unit performance may decline or even fail suddenly. These failures will not only affect the power generation efficiency of the unit, but may also cause serious consequences such as equipment damage and downtime for maintenance, resulting in huge economic losses.

[0003] Existing methods for monitoring the health status of hydropower units often focus on the measurement and analysis of a single parameter. For example, they only focus on a single indicator such as bearing temperature or voltage. This method is difficult to fully reflect the health status of the unit, because the abnormality of a hydropower unit may be the result of the combined effect of multiple factors. A seemingly normal single parameter may be masked when other related parameters have already become abnormal. At the same time, the intrinsic connection between different types of parameters (mechanical, electrical, environmental and control) and their changing trends are not fully considered. For example, only mechanical vibration abnormalities are found without combining the downward trend of electrical power output and changes in environmental water flow for comprehensive judgment, which is likely to lead to misjudgment of the health status of the unit. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method and system for monitoring the health status of a hydropower unit based on machine learning, which realizes more accurate, efficient and comprehensive health management of the hydropower unit.

[0005] In a first aspect, the present invention provides a method for monitoring the health status of a hydropower unit based on machine learning, the method comprising:

[0006] Monitor the operating parameters of hydropower units and obtain multi-dimensional operating parameter information;

[0007] Using a preset hydropower unit state recognition model to identify abnormal states of the multi-dimensional operating parameter information, and obtaining abnormal state information of the hydropower unit;

[0008] Identify and extract trend features of abnormal state information of the hydropower unit obtained at multiple consecutive preset time nodes to obtain abnormal trend information of the hydropower unit;

[0009] Comprehensively considering the multi-dimensional operating parameter information, the abnormal state information of the hydropower unit and the abnormal trend information of the hydropower unit, to obtain a health status analysis index of the hydropower unit;

[0010] Based on a preset health status index threshold, the health status analysis index of the hydropower unit is evaluated, and the evaluation result is used as the health status monitoring result of the hydropower unit.

[0011] Furthermore, the multi-dimensional operating parameter information includes mechanical monitoring data, electrical monitoring data, environmental monitoring data and control monitoring data.

[0012] Furthermore, the method for constructing the hydropower unit state identification model includes:

[0013] Collect multi-dimensional operating parameter information from various monitoring points of the hydropower units;

[0014] Clean the collected data to remove noise and redundant information; normalize the data to eliminate the impact of different dimensions on the data;

[0015] Selecting a machine learning model as the basic architecture of the hydropower unit state recognition model; the machine learning model includes a support vector machine, a neural network, a random forest, and a gradient boosting tree;

[0016] Use the cleaned data to train the model;

[0017] Use an independent validation dataset to validate the trained model and evaluate the accuracy and generalization ability of the model;

[0018] Based on the verification results, the model is further optimized;

[0019] The trained and verified model is deployed to the hydropower unit health status monitoring system.

[0020] Furthermore, the method for obtaining abnormal trend information of a hydropower unit includes:

[0021] Preprocess the abnormal status information of hydropower units, including removing outliers and filling missing values;

[0022] Arrange the preprocessed abnormal status information in chronological order to construct time series information;

[0023] Select trend feature extraction algorithm based on the characteristics of abnormal status information of hydropower units and monitoring requirements;

[0024] Apply the selected trend feature extraction algorithm to extract trend features from time series information;

[0025] Select the features that have an impact on the health status monitoring of hydropower units from the extracted trend features, and further optimize the extracted features;

[0026] The screened and optimized features are integrated into abnormal trend information of hydropower units.

[0027] Furthermore, the method for obtaining the health status analysis index of the hydropower unit includes:

[0028] Integrate multi-dimensional operating parameter information, abnormal status information of hydropower units, and abnormal trend information of hydropower units, and clean them up to remove outliers and fill in missing values;

[0029] Through data analysis, the features that have an impact on the unit health status assessment are screened out, and the corresponding weights are set according to the importance and experience of each feature;

[0030] Standardize and normalize all selected features;

[0031] For each characteristic, several health levels are defined according to its normal operating range, and a corresponding score is assigned to each level;

[0032] According to the score of each feature and the weight corresponding to each feature, the health status analysis index of the hydropower unit is calculated.

[0033] Furthermore, the formula for calculating the health status analysis index of the hydropower unit is:

[0034]

[0035] Among them, S is the final calculated health status analysis index, n is the total number of features involved in the scoring, and s i is the score obtained by the i-th feature according to its health level, w i is the weight of the ith feature.

[0036] Furthermore, factors affecting the setting of the preset health status index threshold include unit type, historical operating data, industry standards, operating environment conditions, and a balance between safety and economy.

[0037] On the other hand, the present application also provides a hydropower unit health status monitoring system based on machine learning, the system comprising:

[0038] The operating parameter monitoring module monitors the operating parameters of the hydropower units and obtains multi-dimensional operating parameter information;

[0039] An abnormal state identification module uses a preset hydropower unit state identification model to identify the abnormal state of the multi-dimensional operating parameter information to obtain abnormal state information of the hydropower unit;

[0040] An abnormal trend extraction module is used to identify and extract the trend characteristics of the abnormal state information of the hydropower unit obtained at multiple consecutive preset time nodes to obtain abnormal trend information of the hydropower unit;

[0041] A health status analysis module, which comprehensively considers the multi-dimensional operation parameter information, the abnormal state information of the hydropower unit and the abnormal trend information of the hydropower unit to obtain a health status analysis index of the hydropower unit;

[0042] The health status evaluation module performs health status evaluation on the health status analysis index of the hydropower unit based on a preset health status index threshold, and uses the evaluation result as the health status monitoring result of the hydropower unit.

[0043] In a third aspect, the present application provides an electronic device, comprising a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, and the computer program, when executed by the processor, implements the steps of any one of the above methods.

[0044] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps in any one of the above-mentioned methods when executed by a processor.

[0045] Compared with the prior art, the present invention has the following beneficial effects: the method can comprehensively reflect the health status of the unit by monitoring the operating parameters of the hydropower unit in multiple dimensions; it effectively avoids the limitation of the traditional method of focusing on only a single parameter, so that the possible abnormalities of the unit can be captured more accurately;

[0046] The preset hydropower unit status recognition model can intelligently identify abnormal status information in multi-dimensional operating parameters; it not only improves the accuracy of fault detection, but also greatly shortens the time of fault discovery, which helps operation and maintenance personnel take timely measures to prevent the fault from further deteriorating;

[0047] By identifying and extracting the trend features of abnormal status information at multiple consecutive preset time nodes, this method can predict the development trend of the unit's health status; it provides valuable early warning information for operation and maintenance personnel, allowing them to take preventive measures before failures occur, thereby avoiding or reducing serious consequences such as downtime for maintenance;

[0048] Taking into account multi-dimensional operating parameter information, abnormal status information and abnormal trend information, this method can generate a health status analysis index for hydropower units. It not only provides intuitive unit health status assessment results for operation and maintenance personnel, but also provides strong support for them to make maintenance plans and decisions.

[0049] Since this method can detect and warn of possible abnormalities in the unit in a timely manner, operation and maintenance personnel can arrange maintenance work more efficiently and reduce unnecessary downtime. It not only improves the operational stability and reliability of the hydropower unit, but also significantly reduces the economic losses caused by failures.

[0050] In summary, the health status monitoring method of hydropower units based on machine learning overcomes the limitations of traditional monitoring methods, and by introducing advanced machine learning technology, it achieves more accurate, efficient and comprehensive health management of hydropower units. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a flow chart of a method for monitoring the health status of a hydropower unit based on machine learning in Embodiment 1;

[0052] Figure 2 It is a flow chart of a method for obtaining abnormal trend information of a hydropower unit;

[0053] Figure 3 It is a structural diagram of the hydropower unit health status monitoring system based on machine learning in Example 2. DETAILED DESCRIPTION

[0054] In the description of the present application, those skilled in the art should know that the present application can be implemented as a method, an apparatus, an electronic device, and a computer-readable storage medium. Therefore, the present application can be specifically implemented in the following forms: complete hardware, complete software (including firmware, resident software, microcode, etc.), a combination of hardware and software. In addition, in some embodiments, the present application can also be implemented in the form of a computer program product in one or more computer-readable storage media, and the computer-readable storage medium contains computer program code.

[0055] The above-mentioned computer-readable storage medium may adopt any combination of one or more computer-readable storage media. Computer-readable storage media include: electrical, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or devices, or any combination of the above. More specific examples of computer-readable storage media include: portable computer disks, hard disks, random access memories, read-only memories, erasable programmable read-only memories, flash memories, optical fibers, optical disc read-only memories, optical storage devices, magnetic storage devices, or any combination of the above. In the present application, computer-readable storage media can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or device.

[0056] The acquisition, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of national laws.

[0057] The present application describes the provided methods, devices, and electronic devices through flowcharts and / or block diagrams.

[0058] It should be understood that each box in the flowchart and / or block diagram and the combination of boxes in the flowchart and / or block diagram can be implemented by computer-readable program instructions. These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine, and these computer-readable program instructions are executed by a computer or other programmable data processing device to produce a device that implements the functions / operations specified by the boxes in the flowchart and / or block diagram.

[0059] These computer-readable program instructions may also be stored in a computer-readable storage medium that enables a computer or other programmable data processing device to work in a specific manner. In this way, the instructions stored in the computer-readable storage medium produce an instruction device product including functions / operations specified in the blocks in the flowchart and / or block diagram.

[0060] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device, so that a series of operational steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby enabling the instructions executed on the computer or other programmable data processing apparatus to provide a process for implementing the functions / operations specified in the blocks in the flowchart and / or block diagram.

[0061] The present application is described below in conjunction with the drawings in the present application.

[0062] Embodiment 1: Figure 1 to Figure 2 As shown, the method for monitoring the health status of a hydropower unit based on machine learning of the present invention specifically comprises the following steps:

[0063] S1. Monitor the operating parameters of the hydropower units to obtain multi-dimensional operating parameter information;

[0064] The multi-dimensional operation parameter information includes mechanical monitoring data, electrical monitoring data, environmental monitoring data and control monitoring data;

[0065] The machinery monitoring data includes:

[0066] Vibration signal: Vibration sensors are used to monitor the vibration of the unit during operation, including parameters such as vibration frequency and amplitude. Abnormal vibration is often related to mechanical failures, such as bearing wear and rotor imbalance.

[0067] Bearing temperature: Monitor the temperature changes of the bearing through the temperature sensor; high bearing temperature indicates poor lubrication or bearing damage;

[0068] Pressure changes: monitor pressure changes at various locations within the unit, such as oil pressure, water pressure, etc. Abnormal pressure indicates leakage or blockage, etc.

[0069] The electrical monitoring data includes:

[0070] Voltage: Monitor the generator output voltage to ensure it fluctuates within normal range; voltage anomalies may be related to excitation system or load changes;

[0071] Current: Monitors the generator output current and the balance of currents in each phase; abnormal currents may indicate an electrical fault or excessive load;

[0072] Power output: monitors the active and reactive power output of the generator; a drop in power may indicate a degradation in unit performance or a load mismatch;

[0073] The environmental monitoring data include:

[0074] Water flow: monitor changes in water flow at the turbine inlet and outlet; abnormal water flow may be related to reduced turbine efficiency or leakage;

[0075] Water temperature: monitor the water temperature changes in the unit's cooling system; excessively high water temperature will affect the cooling effect and cause the unit to overheat;

[0076] Ambient temperature: monitor the temperature changes of the unit's operating environment; excessively high or low ambient temperature will affect the unit's heat dissipation performance;

[0077] The control monitoring data includes:

[0078] Governor signal: monitor the governor's control signal and feedback signal to ensure the normal operation of the speed control system; governor abnormalities may cause unstable speed or overload of the unit;

[0079] Hydraulic system pressure: monitor the working pressure and flow changes of the hydraulic system; abnormalities in the hydraulic system may affect the control performance and stability of the unit.

[0080] In this step, by comprehensively monitoring mechanical data such as vibration signals, bearing temperature, and pressure changes, mechanical faults such as bearing wear and rotor imbalance can be discovered early, so that maintenance measures can be taken in advance to avoid sudden failures; electrical parameters such as voltage, current, and power output are monitored to ensure that the generator operates within a safe and efficient range; this helps maintain the stability and reliability of the power grid and prevent power outages or power quality degradation due to electrical problems; monitoring of environmental parameters such as water flow, water temperature, and ambient temperature can adjust operating conditions in a timely manner to optimize the working efficiency of the turbine, while ensuring the effective operation of the cooling system to prevent performance degradation or equipment damage due to overheating; monitoring of control parameters such as governor signals and hydraulic system pressure ensures the normal operation of the speed control system and maintains the unit's rotation. It improves the response speed and control accuracy of the entire system and reduces the risk of overload. Comprehensive data collection provides a solid foundation for subsequent abnormal state identification and trend analysis, allowing the operation and maintenance team to formulate scientific and reasonable maintenance plans based on the data analysis results, shifting from passive maintenance to active prevention, reducing maintenance costs and extending equipment life. It reduces unplanned downtime, improves power generation efficiency, reduces unnecessary maintenance costs, and thus increases the overall economic benefits of the hydropower station. At the same time, accurate health status monitoring also helps to reasonably plan maintenance cycles and avoid waste of resources caused by excessive maintenance. The collection and analysis of multi-dimensional operating parameter information provides detailed data support for management, helping them make more informed operational decisions and promote the effective use of water resources and the sustainable development of energy production.

[0081] S2, using a preset hydropower unit state recognition model to identify the abnormal state of the multi-dimensional operating parameter information, and obtain abnormal state information of the hydropower unit; the abnormal state information of the hydropower unit includes mechanical fault signals, electrical fault signals, environmental abnormality signals and control system abnormality signals;

[0082] Input multi-dimensional operating parameter information into the trained state recognition model in real time;

[0083] The model will automatically compare the current input data to identify whether there is any deviation from the normal range; if an abnormality is found, the corresponding abnormal status information will be generated;

[0084] For detected anomalies, the model further classifies them to determine whether they belong to mechanical fault signals, electrical fault signals, environmental abnormality signals, or control system abnormality signals;

[0085] As new operational data accumulates, the model is regularly retrained or fine-tuned to adapt to equipment aging and environmental changes;

[0086] The method for constructing the hydropower unit state identification model comprises:

[0087] Collect multi-dimensional operating parameter information from various monitoring points of the hydropower units;

[0088] Clean the collected data to remove noise and redundant information; normalize the data to eliminate the impact of different dimensions on the data; smooth the time series data to reduce the impact of random fluctuations on the model;

[0089] According to the requirements and data characteristics of hydropower unit state identification, a machine learning model is selected as the basic architecture of the hydropower unit state identification model; the machine learning model includes support vector machine, neural network, random forest and gradient boosting tree;

[0090] Use the cleaned data to train the model; optimize the performance of the model by adjusting the model parameters;

[0091] Use an independent validation dataset to validate the trained model and evaluate the accuracy and generalization ability of the model; quantitatively evaluate the performance of the model through indicators such as confusion matrix, accuracy, recall rate, and F1 score;

[0092] According to the verification results, the model is further optimized; the optimization direction includes adjusting model parameters, increasing or decreasing the number of features, using more advanced models, etc.;

[0093] The trained and verified model is deployed to the hydropower unit health status monitoring system.

[0094] In this step, by inputting multi-dimensional operating parameter information into the trained state recognition model in real time, it is possible to more accurately detect situations that deviate from the normal range; the inherent relationship between multiple related parameters is comprehensively evaluated, thereby improving the accuracy and comprehensiveness of fault detection; the model can promptly detect potential mechanical, electrical, environmental or control system anomalies and classify them; enabling the operation and maintenance team to take preventive measures before the problem worsens, reduce the probability of sudden failures, and avoid economic losses caused by unplanned downtime and equipment damage; abnormal state recognition helps to quickly locate factors that affect unit performance, thereby making maintenance work more targeted; it can not only maintain the optimal performance of the unit, but also Extend its service life and improve power generation efficiency; real-time monitoring and analysis of multi-dimensional operating parameters to ensure that the hydropower units operate within a safe range; any abnormal situation will immediately trigger an alarm to help operation and maintenance personnel respond quickly, prevent small problems from turning into major accidents, and ensure the stability and reliability of the power system; the state recognition model based on machine learning provides scientific data support, enabling management to make more informed operational decisions based on quantitative evaluation results, optimize resource allocation, and promote the effective use of water resources; accurate abnormal state identification reduces unnecessary excessive maintenance and improves the efficiency of maintenance work; by predicting faults in advance and taking preventive measures, it can significantly reduce maintenance costs and time consumption.

[0095] S3, identifying and extracting trend characteristics of the abnormal state information of the hydropower unit obtained at multiple consecutive preset time nodes to obtain abnormal trend information of the hydropower unit;

[0096] The method for obtaining abnormal trend information of hydropower units includes:

[0097] Preprocess the abnormal status information of hydropower units collected from multiple consecutive preset time nodes, including removing outliers and filling missing values ​​to ensure the integrity and accuracy of the data;

[0098] Arrange the preprocessed abnormal status information in chronological order to construct time series information; each time series represents the change of a specific abnormal status indicator over time;

[0099] According to the characteristics of abnormal status information of hydropower units and monitoring requirements, select appropriate trend feature extraction algorithms; commonly used algorithms include but are not limited to moving average method, exponential smoothing method, time series decomposition, ARIMA model, machine learning model, etc.; these algorithms can capture long-term trends, seasonal changes and irregular fluctuations in data;

[0100] Apply the selected algorithm to extract trend features from time series data; use the moving average method to smooth the data and identify the overall trend; use the ARIMA model to predict future trends and evaluate the significance of trends; use the LSTM neural network to learn complex patterns in the data and predict future abnormal conditions;

[0101] The most critical indicators for monitoring the health status of hydropower units can be selected from the extracted trend features; this can be achieved through methods such as correlation analysis and feature importance assessment; at the same time, the features can be further optimized to improve the efficiency and accuracy of subsequent analysis;

[0102] The screened and optimized trend features are integrated into abnormal trend information of hydropower units.

[0103] In this step, through the time series analysis of multi-dimensional operating parameters, complex patterns and trends that cannot be revealed by a single parameter can be captured, making early detection of potential faults possible; it helps to take preventive measures in advance to avoid the occurrence of sudden faults; the extracted key trend features provide scientific data support for operation and maintenance personnel, helping them to make more informed maintenance and management decisions; based on the predicted trend information, maintenance plans can be reasonably arranged to reduce unnecessary downtime; accurate trend analysis allows more effective allocation of human and material resources to ensure targeted maintenance when it is most needed, rather than blindly conducting regular inspections or replacing parts, saving costs while improving equipment availability; continuous monitoring and analysis of abnormal trends helps to keep the hydropower units operating in their optimal working condition, reduce safety risks caused by equipment aging or operating errors, and ensure the stable supply of the power system; the introduction of advanced algorithms and technologies not only improves the current monitoring level, but also lays the foundation for future technological development; through effective trend analysis and timely intervention measures, direct economic losses caused by failures can be significantly reduced.

[0104] S4, comprehensively considering the multi-dimensional operation parameter information, the abnormal state information of the hydropower unit and the abnormal trend information of the hydropower unit, to obtain a health status analysis index of the hydropower unit;

[0105] The method for obtaining the health status analysis index of a hydropower unit includes:

[0106] Integrate multi-dimensional operating parameter information, abnormal status information of hydropower units, and abnormal trend information of hydropower units, clean up the data, remove outliers, fill in missing values, and ensure the integrity and accuracy of the data;

[0107] Through historical data analysis, the features that affect the unit health status assessment are screened out, and the corresponding weights are set according to the importance and experience value of each feature;

[0108] Standardize and normalize all selected features so that they can be compared on the same scale;

[0109] For each characteristic, several health levels are defined according to its normal operating range, and a corresponding score is assigned to each level;

[0110] According to the score of each feature and the weight corresponding to each feature, the health status analysis index of the hydropower unit is calculated;

[0111] The formula for calculating the health status analysis index of hydropower units is:

[0112]

[0113] Among them, S is the final calculated health status analysis index, n is the total number of features involved in the scoring, and s i is the score obtained by the i-th feature according to its health level, w i is the weight of the ith feature.

[0114] In this step, multi-dimensional operating parameter information, abnormal status information of hydropower units and abnormal trend information are comprehensively considered, so as to more comprehensively reflect the health status of the unit; it is more accurate than monitoring of a single parameter, because the health status of the unit is often determined by multiple factors; before integrating the information, the data is cleaned, outliers are removed and missing values ​​are filled to ensure the integrity and accuracy of the data; this is the basis for subsequent analysis and evaluation, avoiding misjudgment caused by data quality problems; through historical data analysis, the features that affect the health status assessment of the unit are screened out, and the corresponding weights are set according to the importance and experience value of each feature; the method of feature selection and weight setting makes the evaluation results more scientific and reasonable, and can more accurately reflect the actual health status of the unit; all selected features are standardized and normalized so that different features can be compared on the same scale; it eliminates The dimensional differences between different characteristics improve the accuracy and comparability of the assessment; several health levels are defined for each characteristic according to its normal operating range, and corresponding scores are assigned to each level; the health level definition and score allocation method make the assessment results more intuitive and easy to understand, and also provide a clear reference for operation and maintenance personnel; according to the score and weight of each feature, the health status analysis index of the hydropower unit is calculated; the health status analysis index of the hydropower unit is a quantitative indicator that comprehensively reflects the health status of the unit, which can provide intuitive unit health status information for operation and maintenance personnel, and help to timely discover and solve potential problems; the results of this step can provide scientific decision-making support for operation and maintenance personnel, helping them to formulate reasonable operation and maintenance plans and maintenance strategies based on the health status analysis index of the unit; it helps to reduce unplanned downtime, improve the reliability and power generation efficiency of the unit, and reduce operation and maintenance costs.

[0115] S5. Based on a preset health status index threshold, the health status analysis index of the hydropower unit is evaluated, and the evaluation result is used as the health status monitoring result of the hydropower unit;

[0116] Compare the health status analysis index of the hydropower unit with the preset health status index threshold to determine the current health status level of the unit; in addition to numerical comparison, a comprehensive assessment should be conducted based on the historical operation records, maintenance records, current operating environment and other factors of the unit to gain a more comprehensive understanding of the health status of the unit;

[0117] Once the unit health index reaches or exceeds the preset health index threshold, the early warning mechanism should be triggered immediately to notify the operation and maintenance personnel to pay attention to the unit's operating status and prepare to take corresponding preventive measures;

[0118] For units with serious abnormalities, a detailed maintenance plan should be formulated, including maintenance content, required materials, personnel arrangements, and estimated downtime, to ensure that the unit can resume normal operation as soon as possible; based on the results of the unit health status assessment, the operation and maintenance strategy should be optimized and adjusted regularly, such as adjusting the monitoring frequency, optimizing the maintenance process, and strengthening personnel training, so as to improve the operation and maintenance efficiency and unit reliability;

[0119] Record the results of each health status assessment, the operation and maintenance measures taken, and the subsequent operation status of the unit in detail for subsequent analysis and summary; establish a feedback mechanism to encourage operation and maintenance personnel to promptly report problems and improvement suggestions found during the assessment process, so as to continuously improve and optimize the health status monitoring and assessment system;

[0120] The factors affecting the setting of the preset health status index threshold include:

[0121] Unit type: Hydropower units of different types and specifications differ in structure, performance, operating environment, etc., so their health status index thresholds should also be different; large hydropower units may have relatively stricter health status index thresholds due to their complex structures and large operating loads;

[0122] Historical operation data: The historical operation data of the unit is an important reference for setting thresholds. By analyzing the unit's operating parameters, fault records, maintenance history and other data over a period of time, we can understand the unit's performance change trends, common fault types and occurrence frequencies, and other information, thereby providing a scientific basis for setting thresholds.

[0123] Industry standards: The operation and maintenance management of hydropower units should comply with relevant industry standards and specifications; these standards and specifications include relevant requirements and recommendations for unit health status monitoring and evaluation, including the threshold setting range of the health status index, etc. Therefore, when setting the threshold, the requirements of these standards and specifications should be fully considered;

[0124] Operating environment conditions: The operating environment conditions of the unit, such as water temperature, water flow, ambient temperature, etc., have an important impact on its health status; therefore, when setting the threshold, full consideration should be given to the current operating environment conditions of the unit and the possible impact of these conditions on the health status of the unit;

[0125] Balance between safety and economy: Weigh the safety and economy of the unit; on the one hand, a threshold that is too low may lead to frequent false alarms and unnecessary downtime for maintenance, increasing operation and maintenance costs; on the other hand, a threshold that is too high may not be able to detect the potential failure risks of the unit in time, threatening the safe operation of the unit; therefore, the threshold should be reasonably set according to the actual situation and operation requirements of the unit to achieve a balance between safety and economy.

[0126] In this step, by comparing the health status analysis index of the hydropower unit with the preset health status index threshold, and combining the historical operation record, maintenance record and current operating environment of the unit for comprehensive evaluation, we can have a more comprehensive understanding of the health status of the unit and improve the accuracy of the judgment; once the health status index of the unit reaches or exceeds the preset threshold, the early warning mechanism is immediately triggered to notify the operation and maintenance personnel to pay attention in time and take preventive measures to effectively avoid the occurrence or expansion of potential faults; for units with serious abnormalities, a detailed maintenance plan can be quickly formulated, including maintenance content, required materials, personnel arrangements and expected downtime, etc., to ensure that the unit resumes normal operation as soon as possible and reduce downtime losses; according to the results of the unit health status judgment, the operation and maintenance strategy is regularly optimized and adjusted to improve the operation and maintenance efficiency and unit reliability and reduce the operation and maintenance cost; considering factors such as unit type, specification and operating environment conditions, the health status index threshold is reasonably set. value to ensure the scientific nature and applicability of the threshold; when setting the threshold, weigh the safety and economy of the unit to avoid too low a threshold that leads to frequent false alarms and unnecessary shutdowns for maintenance, and too high a threshold that fails to promptly detect potential failure risks, so as to achieve a balance between safety and economy; record the results of each health status assessment, the operation and maintenance measures taken, and the subsequent operation status of the unit in detail to provide data support for subsequent analysis and summary; establish a feedback mechanism to encourage operation and maintenance personnel to promptly report problems and improvement suggestions found during the assessment process, continuously improve and optimize the health status monitoring and assessment system, and improve the accuracy and effectiveness of monitoring and assessment; this step provides strong support for the health status monitoring of hydropower units through accurate health status assessment, efficient operation and maintenance management, scientific threshold setting, and continuous improvement and feedback mechanism, effectively improving the operating stability and reliability of the units and reducing operation and maintenance costs.

[0127] Embodiment 2: Figure 3 As shown, the hydropower unit health status monitoring system based on machine learning of the present invention specifically includes the following modules:

[0128] The operating parameter monitoring module monitors the operating parameters of the hydropower units and obtains multi-dimensional operating parameter information;

[0129] An abnormal state identification module uses a preset hydropower unit state identification model to identify the abnormal state of the multi-dimensional operating parameter information to obtain abnormal state information of the hydropower unit;

[0130] An abnormal trend extraction module is used to identify and extract the trend characteristics of the abnormal state information of the hydropower unit obtained at multiple consecutive preset time nodes to obtain abnormal trend information of the hydropower unit;

[0131] A health status analysis module, which comprehensively considers the multi-dimensional operation parameter information, the abnormal state information of the hydropower unit and the abnormal trend information of the hydropower unit to obtain a health status analysis index of the hydropower unit;

[0132] The health status evaluation module performs health status evaluation on the health status analysis index of the hydropower unit based on a preset health status index threshold, and uses the evaluation result as the health status monitoring result of the hydropower unit.

[0133] The system uses a multi-dimensional operating parameter monitoring module that can simultaneously collect different types of operating data such as mechanical, electrical, environmental and control data; this comprehensive data collection method helps to more accurately reflect the actual working status of the unit;

[0134] The combination of the abnormal state recognition module and the abnormal trend extraction module can not only detect the abnormal changes of a single parameter, but also capture the changing trends of these parameters over time and their mutual influence; this enables the system to have a deeper understanding of the causes of the fault and avoids the situation where the abnormality of other related parameters is covered up by the normality of a single parameter;

[0135] The health status analysis index calculated by the health status analysis module can predict potential problems in advance, so as to take preventive measures to reduce the probability of sudden failures and improve system reliability;

[0136] Using machine learning algorithms for data analysis can automatically learn patterns from a large amount of historical data and apply these patterns to predict future states, thus improving the accuracy of abnormal state identification.

[0137] The evaluation results provided by the health status evaluation module can help operation and maintenance personnel better plan maintenance plans and achieve the transition from regular maintenance to condition-based maintenance, thereby reducing maintenance costs and extending equipment life;

[0138] Reduce unplanned downtime and maintenance times, improve power generation efficiency, and thus increase economic benefits; at the same time, through effective management of the unit health status, it can also avoid huge economic losses caused by equipment damage; more reliable unit operation status monitoring helps prevent safety accidents caused by faults and ensure the safety of staff and facilities;

[0139] In summary, by introducing advanced machine learning technology, this monitoring system has achieved more accurate, comprehensive and forward-looking health management of hydropower units, effectively solving the limitations of traditional methods.

[0140] The various variations and specific embodiments of the method for monitoring the health status of a hydropower unit based on machine learning in the aforementioned embodiment 1 are also applicable to the system for monitoring the health status of a hydropower unit based on machine learning in the present embodiment. Through the aforementioned detailed description of the method for monitoring the health status of a hydropower unit based on machine learning, those skilled in the art can clearly understand the implementation method of the system for monitoring the health status of a hydropower unit based on machine learning in the present embodiment. Therefore, for the sake of brevity of the specification, it will not be described in detail here.

[0141] In addition, the present application also provides an electronic device, including a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor. The transceiver, the memory, and the processor are respectively connected via a bus. When the computer program is executed by the processor, each process of the above-mentioned method for controlling output data is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0142] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for monitoring the health status of a hydropower unit based on machine learning, characterized in that: The method comprises: Monitor the operating parameters of hydropower units and obtain multi-dimensional operating parameter information; Using a preset hydropower unit state recognition model to identify abnormal states of the multi-dimensional operating parameter information, and obtaining abnormal state information of the hydropower unit; Identify and extract trend features of abnormal state information of the hydropower unit obtained at multiple consecutive preset time nodes to obtain abnormal trend information of the hydropower unit; Comprehensively considering the multi-dimensional operating parameter information, the abnormal state information of the hydropower unit and the abnormal trend information of the hydropower unit, to obtain a health status analysis index of the hydropower unit; Based on a preset health status index threshold, the health status analysis index of the hydropower unit is evaluated, and the evaluation result is used as the health status monitoring result of the hydropower unit.

2. The method for monitoring the health status of a hydropower unit based on machine learning according to claim 1, characterized in that: The multi-dimensional operation parameter information includes mechanical monitoring data, electrical monitoring data, environmental monitoring data and control monitoring data.

3. The method for monitoring the health status of a hydropower unit based on machine learning according to claim 1, characterized in that: The method for constructing the hydropower unit state identification model comprises: Collect multi-dimensional operating parameter information from various monitoring points of the hydropower units; Clean the collected data to remove noise and redundant information; normalize the data to eliminate the impact of different dimensions on the data; Selecting a machine learning model as the basic architecture of the hydropower unit state recognition model; the machine learning model includes a support vector machine, a neural network, a random forest, and a gradient boosting tree; Use the cleaned data to train the model; Use an independent validation dataset to validate the trained model and evaluate the accuracy and generalization ability of the model; Based on the verification results, the model is further optimized; The trained and verified model is deployed to the hydropower unit health status monitoring system.

4. The method for monitoring the health status of a hydropower unit based on machine learning according to claim 1, characterized in that: The method for obtaining abnormal trend information of hydropower units includes: Preprocess the abnormal status information of hydropower units, including removing outliers and filling missing values; Arrange the preprocessed abnormal status information in chronological order to construct time series information; Select trend feature extraction algorithm based on the characteristics of abnormal status information of hydropower units and monitoring requirements; Apply the selected trend feature extraction algorithm to extract trend features from time series information; Select the features that have an impact on the health status monitoring of hydropower units from the extracted trend features, and further optimize the extracted features; The screened and optimized features are integrated into abnormal trend information of hydropower units.

5. The method for monitoring the health status of a hydropower unit based on machine learning according to claim 1, characterized in that: The method for obtaining the health status analysis index of a hydropower unit includes: Integrate multi-dimensional operating parameter information, abnormal status information of hydropower units, and abnormal trend information of hydropower units, and clean them up to remove outliers and fill in missing values; Through data analysis, the features that have an impact on the unit health status assessment are screened out, and the corresponding weights are set according to the importance and experience of each feature; Standardize and normalize all selected features; For each characteristic, several health levels are defined according to its normal operating range, and a corresponding score is assigned to each level; According to the score of each feature and the weight corresponding to each feature, the health status analysis index of the hydropower unit is calculated.

6. The method for monitoring the health status of a hydropower unit based on machine learning according to claim 5, characterized in that: The formula for calculating the health status analysis index of hydropower units is: Among them, S is the final calculated health status analysis index, n is the total number of features involved in the scoring, and s i is the score obtained by the i-th feature according to its health level, w i is the weight of the ith feature.

7. The method for monitoring the health status of a hydropower unit based on machine learning according to claim 1, characterized in that: The factors affecting the setting of the preset health status index threshold include unit type, historical operating data, industry standards, operating environment conditions, and a balance between safety and economy.

8. A hydropower unit health status monitoring system based on machine learning, characterized in that: The system comprises: The operating parameter monitoring module monitors the operating parameters of the hydropower units and obtains multi-dimensional operating parameter information; An abnormal state identification module uses a preset hydropower unit state identification model to identify the abnormal state of the multi-dimensional operating parameter information to obtain abnormal state information of the hydropower unit; An abnormal trend extraction module is used to identify and extract the trend characteristics of the abnormal state information of the hydropower unit obtained at multiple consecutive preset time nodes to obtain abnormal trend information of the hydropower unit; A health status analysis module, which comprehensively considers the multi-dimensional operation parameter information, the abnormal state information of the hydropower unit and the abnormal trend information of the hydropower unit to obtain a health status analysis index of the hydropower unit; The health status evaluation module performs health status evaluation on the health status analysis index of the hydropower unit based on a preset health status index threshold, and uses the evaluation result as the health status monitoring result of the hydropower unit.

9. An electronic device, comprising a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the transceiver, the memory, and the processor are connected via the bus, wherein: When the computer program is executed by the processor, the steps in the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps in the method according to any one of claims 1 to 7 are implemented.

Images