Hydropower plant oil pressure system solenoid valve fault early warning and life prediction method and device

Abstract

This invention discloses a method and device for fault early warning and life prediction of solenoid valves in hydropower plant hydraulic systems, relating to the field of condition monitoring and fault early warning technology for auxiliary equipment in hydropower plants. The method includes: synchronously acquiring dynamic current, mechanical vibration, and temperature signals during the solenoid valve's operation, extracting multi-dimensional electromagnetic, mechanical, and thermodynamic state characteristics; establishing a personalized health baseline based on historical operating data, comparing real-time characteristics with the baseline, and calculating a comprehensive health index; judging the health status based on the index and a dynamic early warning threshold, outputting a fault warning when abnormal, and adaptively updating the health baseline when normal; constructing a performance degradation model based on the temporal changes of multi-dimensional characteristics, predicting the trend of key indicators, and calculating the remaining service life in conjunction with the failure threshold, thereby achieving full-cycle health monitoring and life prediction of the solenoid valve. This invention achieves early warning of solenoid valves through multi-dimensional online monitoring, solving the safety risk of sudden failures easily leading to unit outages.

Description

Technical Field

[0001] This invention relates to the field of condition monitoring and fault early warning technology for auxiliary equipment in hydropower plants, and in particular to a method and device for fault early warning and life prediction of solenoid valves in hydraulic systems of hydropower plants. Background Technology

[0002] Hydropower plants, as crucial power sources in the power system responsible for peak shaving, frequency regulation, and emergency backup, directly impact the safe and reliable supply of the regional power grid. The hydraulic system, a key control actuator of the hydro-generator unit, is widely used in core operating conditions such as governor regulation, unit braking, main shaft seal engagement / disengagement, and emergency gate opening / closing. It is a vital foundation for ensuring normal unit start-up and shutdown, precise load regulation, and rapid response in emergency situations. Solenoid valves, as key components in the hydraulic system that convert electrical signals into hydraulic actions, not only operate at high frequencies and in complex environments, but their core components, such as valve cores, coils, and return springs, are all easily worn structures. Early problems such as valve core jamming, coil overheating and burnout, and mechanical fatigue wear can easily lead to valve refusal to operate, malfunctions, or internal leakage, directly affecting the control accuracy of the speed regulation system. In severe cases, this can even cause unplanned unit shutdowns, equipment damage, and power grid fluctuations, resulting in safety accidents.

[0003] Currently, the operation and maintenance management of solenoid valves in hydraulic systems at domestic hydropower plants still largely relies on traditional manual inspections and periodic maintenance. Routine inspections often depend on simple methods such as manual observation and infrared thermography, which can only provide a rough assessment of the macroscopic condition of the solenoid valve surface temperature and external wiring, making it difficult to detect potential deterioration trends inside the equipment. Periodic maintenance plans are often formulated based on uniform operating hours or number of operations, without considering the actual operating conditions, load intensity, and individual performance differences of different solenoid valves. This can easily lead to over-maintenance, resulting in wasted spare parts and manpower costs, or delayed maintenance leading to equipment operating with defects. Post-failure repairs can only be carried out after the solenoid valve has completely failed. The failures are sudden and the response cycle is long, often accompanied by unit shutdowns and power generation losses, which cannot meet the management requirements for safe and efficient operation of modern hydropower plants.

[0004] With the continuous development of equipment condition monitoring and intelligent diagnostic technologies, some hydropower plants have gradually begun to conduct online condition monitoring of solenoid valves. However, existing monitoring and diagnostic solutions still have significant limitations. Some technologies only monitor electrical parameters such as coil resistance and supply voltage, failing to reflect mechanical faults such as valve core jamming and mechanical wear. Some solutions only collect feedback signals of valve opening and closing positions, making it difficult to identify dynamic performance degradation characteristics during operation. The few systems that integrate current and temperature monitoring mostly remain at the level of simple threshold alarms, without in-depth mining and fusion analysis of sensitive health information such as dynamic current waveforms, vibration and impact characteristics, and temperature change rates.

[0005] Overall, existing solenoid valve condition monitoring technologies generally suffer from problems such as limited monitoring dimensions, insufficient feature extraction, and inadequate intelligence. Most can only provide passive alarms after a fault occurs, making it difficult to establish personalized health assessment benchmarks for individual devices, let alone extrapolate remaining service life based on equipment performance degradation patterns, thus failing to provide maintenance personnel with a basis for proactive maintenance decisions. Therefore, constructing a multi-dimensional, high-precision, and adaptive solenoid valve health status monitoring and life prediction system, and promoting the transformation of the operation and maintenance model from traditional periodic inspections and reactive maintenance to predictive maintenance, has significant engineering application value for improving the operational reliability of hydropower plant hydraulic systems, reducing operation and maintenance costs, and ensuring the safe and stable operation of generating units and the power grid. Summary of the Invention

[0006] The main objective of this invention is to provide a method for early warning of failure and life prediction of solenoid valves in hydraulic systems of hydropower plants.

[0007] Another objective of this invention is to provide a device for early warning and life prediction of solenoid valve failure in a hydroelectric power plant hydraulic system.

[0008] The third objective of this invention is to provide an electronic device.

[0009] A fourth objective of this invention is to provide a non-transitory computer-readable storage medium.

[0010] To achieve the above objectives, a first aspect of the present invention provides a method for early warning and life prediction of solenoid valve failure in a hydroelectric power plant's hydraulic system, comprising:

[0011] The system synchronously acquires dynamic current signals, mechanical vibration signals, and temperature change signals of the solenoid valve during its operation, and extracts multidimensional state characteristic parameters that characterize its electromagnetic, mechanical, and thermodynamic properties. A personalized health baseline is constructed based on the historical operating data of the solenoid valve. Multidimensional state characteristic parameters are compared and analyzed with the personalized health baseline to calculate a comprehensive health index that reflects the current health status of the solenoid valve. The health status is determined based on the comprehensive health index and the preset dynamic warning threshold. When an abnormal health status is detected, a fault warning message is generated. When the health status is determined to be healthy, the health baseline is adaptively updated using the current multi-dimensional status feature parameters. Based on the temporal evolution law of the multidimensional state characteristic parameters, a performance degradation model is constructed to deduce the future change trend of key performance indicators. Combined with the preset failure threshold, the remaining service life of the solenoid valve is calculated, thus completing the full-cycle health monitoring and remaining service life prediction of the solenoid valve.

[0012] Optionally, the dynamic current signal, mechanical vibration signal, and temperature change signal of the solenoid valve during operation are simultaneously acquired, and multi-dimensional state characteristic parameters characterizing the electromagnetic, mechanical, and thermodynamic properties are extracted, including: The instantaneous value of the coil current is collected by a current sensor installed on the coil circuit of the solenoid valve. The characteristics of the peak value of the solenoid valve starting current and the effective value of the steady-state holding current are obtained, which respectively characterize the mechanical jamming of the valve core, the fatigue state of the spring and the abnormal change of the DC resistance of the coil. Vibration acceleration signals during the valve core movement process are collected by vibration sensors attached to the valve body of the solenoid valve. The valve core movement time, vibration impact energy and vibration main frequency energy ratio are extracted to characterize the mechanical movement time of the solenoid valve, the impact intensity between the valve core and the valve seat and the internal mechanical wear and loosening state, respectively. By collecting the coil housing temperature and ambient temperature through a temperature sensor attached to the solenoid valve coil housing, the steady-state temperature rise and temperature rise rate characteristics are calculated to characterize the coil power consumption, heat dissipation conditions and abnormal thermal conditions. After the signals from each sensor are amplified and filtered by a signal amplifier, the controller completes data acquisition and feature fusion to finally obtain multi-dimensional state characteristic parameters of electromagnetic, mechanical and thermodynamic dimensions.

[0013] Optionally, a personalized health baseline is constructed based on the historical operating data of the solenoid valve. Multidimensional state characteristic parameters are compared and analyzed with the personalized health baseline to calculate a comprehensive health index reflecting the current health status of the solenoid valve, including: Based on the historical health operation data of the solenoid valve, the baseline mean and standard deviation of each state characteristic parameter are determined to form a personalized health baseline for each device. The real-time collected state feature parameters of each dimension are normalized and the deviation is calculated with the corresponding health baseline to obtain the independent deviation index of each feature parameter. The weighting coefficients of each feature parameter are determined by using the entropy weighting method or the expert scoring method. Under the constraint that the sum of the weights is 1, the comprehensive health index is obtained by fusion calculation.

[0014] Optionally, the health status is determined based on the comprehensive health index and a preset dynamic early warning threshold. When an abnormal health status is detected, a fault warning message is generated. When the health status is determined to be healthy, the health baseline is adaptively updated using the current multi-dimensional state feature parameters, including: An adaptive healthy baseline update model is constructed using an exponentially weighted moving average algorithm, and a reasonable learning rate is set to control the baseline update speed. When determining that the current multidimensional state characteristic parameters are in a healthy state, the mean and standard deviation of the health baseline are slowly updated using the current characteristic parameters so that the health baseline can adapt to the natural aging of the solenoid valve and seasonal environmental changes. When the overall health index is below the warning threshold or a single feature is significantly abnormal, corresponding fault warning information is generated in real time.

[0015] Optionally, a performance degradation model is constructed based on the temporal evolution of the multidimensional state characteristic parameters to predict the future changing trends of key performance indicators, including: The valve core action time was selected as a key performance indicator for degradation. An exponential degradation trajectory model was fitted based on historical time series data to determine the initial health value and degradation rate coefficient. A zero-mean random noise term is introduced into the degradation model to simulate the fluctuations in operating conditions and random disturbances during actual operation. By using a degradation model to extrapolate the future change path of key performance indicators at each moment, dynamic tracking and trend prediction of the performance degradation process of solenoid valves can be achieved.

[0016] Optionally, the remaining service life of the solenoid valve can be calculated by combining a preset failure threshold, including: Pre-set failure thresholds for key performance indicators, and use the moment when the indicator reaches the failure threshold as the end of its lifespan. Based on the analytical relationship of the exponential degradation model, combined with the measured values ​​of current performance indicators, degradation rate coefficient and failure threshold, the remaining service life of the solenoid valve is calculated. By combining real-time health status assessment, dynamic fault early warning and remaining service life calculation results, the full-cycle health monitoring and life prediction management of solenoid valves from commissioning to failure can be completed.

[0017] To achieve the above objectives, a second aspect of the present invention provides a device for early warning and life prediction of solenoid valve failure in a hydropower plant hydraulic system, comprising: The multi-state acquisition module is used to synchronously acquire dynamic current signals, mechanical vibration signals, and temperature change signals of the solenoid valve during its operation, and extract multi-dimensional state characteristic parameters that characterize its electromagnetic, mechanical, and thermodynamic properties. The health assessment module is used to construct a personalized health baseline based on the historical operating data of the solenoid valve, compare and analyze the multi-dimensional status characteristic parameters with the personalized health baseline, and calculate a comprehensive health index that reflects the current health status of the solenoid valve. The early warning update module is used to make a judgment based on the comprehensive health index and the preset dynamic early warning threshold. When an abnormal health status is detected, it generates a fault early warning information. When the health status is determined to be healthy, it adaptively updates the health baseline using the current multi-dimensional status feature parameters. The life prediction module is used to construct a performance degradation model based on the time-series evolution law of the multidimensional state characteristic parameters, deduce the future change trend of key performance indicators, and calculate the remaining service life of the solenoid valve in combination with the preset failure threshold, so as to complete the full-cycle health monitoring and remaining life prediction of the solenoid valve.

[0018] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0019] To achieve the above objectives, a third aspect of this application provides an electronic device, including a processor and a memory; wherein the processor reads executable program code stored in the memory to run a program corresponding to the executable program code, for implementing a method for early warning and life prediction of solenoid valve failure in a hydropower plant hydraulic system as described in the first aspect embodiment.

[0020] To achieve the above objectives, the fourth aspect of this application proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for early warning and life prediction of solenoid valve failure in a hydroelectric power plant hydraulic system as described in the first aspect embodiment.

[0021] The embodiments of the present invention have the following beneficial effects: 1. Improve operational safety by enabling early warning of solenoid valve failures, accurately identifying potential hazards such as valve core jamming, coil overheating, and mechanical fatigue, and significantly reducing the risk of unplanned unit shutdowns and safety accidents caused by sudden solenoid valve failures.

[0022] 2. Reduce the economic cost of operation and maintenance by changing the traditional periodic uniform replacement of solenoid valves to on-demand replacement based on health status, reducing the waste of spare parts and manpower, and changing emergency repairs to planned maintenance, effectively reducing the economic losses caused by unit downtime.

[0023] 3. Improve the efficiency and intelligence of operation and maintenance management, support remote automatic batch monitoring of solenoid valve health status, reduce on-site inspection workload, and accumulate equipment life cycle data to provide reliable support for equipment selection optimization and operation strategy improvement. Attached Figure Description

[0024] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 A flowchart illustrating a method for early warning and life prediction of solenoid valve failure in a hydroelectric power plant's hydraulic system, provided by an embodiment of the present invention; Figure 2 This is a flowchart illustrating the execution and feedback process of the intelligent monitoring terminal provided in this embodiment of the invention. Figure 3 A flowchart of a platform-side early warning and predictive analysis system provided in an embodiment of the present invention; Figure 4This is a flowchart of the signal acquisition and processing of an intelligent monitoring terminal provided in an embodiment of the present invention; Figure 5 A flowchart illustrating the implementation of a lifespan warning system based on the number of actions provided in this embodiment of the invention; Figure 6 A flowchart of remaining lifetime prediction based on a performance degradation model provided in an embodiment of the present invention; Figure 7 This is a structural diagram of a solenoid valve fault early warning and life prediction device for a hydroelectric power plant hydraulic system, provided in an embodiment of the present invention. Detailed Implementation

[0025] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0027] The following description, with reference to the accompanying drawings, describes a method and apparatus for early warning and life prediction of solenoid valve failure in a hydroelectric power plant hydraulic system according to an embodiment of the present invention.

[0028] Example 1 This invention provides a method for early warning and life prediction of solenoid valve failure in a hydroelectric power plant's hydraulic system. Figure 1 This is a flowchart illustrating a method for early warning and lifespan prediction of solenoid valve faults in a hydroelectric power plant's hydraulic system, provided by an embodiment of the present invention. Figure 1 As shown, the method includes the following steps: Step S1: Synchronously acquire dynamic current signals, mechanical vibration signals, and temperature change signals of the solenoid valve during its operation, and extract multidimensional state characteristic parameters that characterize its electromagnetic properties, mechanical action characteristics, and thermodynamic properties.

[0029] This step corresponds to the execution and feedback process of the intelligent monitoring terminal, such as... Figure 2As shown, the intelligent monitoring terminal mainly consists of a power supply module, a sensor group, a signal processing module, a microcontroller (MCU), and a communication module. The power supply module provides a stable power supply to the entire terminal. The sensor group comprises three types: a high-precision current sensor, a miniature vibration sensor, and a temperature sensor, corresponding to the monitoring needs of different physical states of the solenoid valve. The high-precision current sensor is mounted on the solenoid valve coil circuit to collect the instantaneous value of the coil current. The miniature vibration sensor is attached to the surface of the solenoid valve body to collect the vibration acceleration signal during the valve core's movement. The temperature sensor is attached to the solenoid valve coil housing to simultaneously collect the coil housing temperature and the ambient temperature. The raw signals collected by the three types of sensors are first sent to the signal processing module, where they are amplified and filtered by a signal amplifier to eliminate electromagnetic interference and environmental noise. Then, they are transmitted to the MCU, where a built-in feature extraction algorithm extracts multi-dimensional state feature parameters. Finally, the processed data is transmitted via the communication module and network to the platform-side early warning and predictive analysis system.

[0030] In this application embodiment, feature extraction for different types of sensors corresponds to specific calculation methods and physical meanings. For current sensors, the core extraction focuses on two types of dynamic current features: one is the peak value of the starting current, calculated using the following formula:

[0031] in, This is the peak starting current. This is the instantaneous value of the coil current, in amperes (A). The moment when the coil is energized. The time when the valve core completes engagement is the time when the valve core completes engagement. The physical meaning of this parameter is to reflect the balance between the electromagnetic attraction and mechanical resistance of the solenoid valve. It can characterize faults such as mechanical jamming of the valve core and spring fatigue. When the peak value of the starting current increases abnormally, it usually indicates that there is a jamming problem in the valve core.

[0032] The second is the effective value of the steady-state holding current, calculated using the following formula:

[0033] in, To maintain the effective value of the current, , T represents the start and end times of the steady-state holding phase, in milliseconds (ms), and T is the duration of the steady-state holding phase, T = - This parameter is used to reflect the change in the DC resistance of the coil. If the effective value of the holding current increases abnormally, it may indicate that there is a risk of inter-turn short circuit or excessive temperature rise in the solenoid valve coil.

[0034] For vibration sensors, the core extraction of three types of mechanical motion features is as follows: The first category is the valve core actuation time, calculated using the following formula:

[0035] in, For valve core actuation time, The peak moment of valve core impact detected by the vibration sensor, in milliseconds. The starting time of coil energization, measured in milliseconds (ms), represents the total time it takes for the solenoid valve to complete its mechanical action from energization. It is a typical indicator of performance degradation, and an extended action time directly reflects the deterioration of the solenoid valve's mechanical performance.

[0036] The second category is vibration and impact energy, calculated using the following formula:

[0037] in, This is a vibration acceleration signal, with units of m / s². Δt represents the moment when the valve core impact occurs, and Δt is the impact duration window, typically ranging from 5 to 10 ms. This parameter reflects the mechanical impact intensity between the valve core and the valve seat, and energy decay can indicate problems such as wear of buffer components or sludge buffering.

[0038] The third category is the proportion of energy at the dominant vibration frequency, calculated using the following formula:

[0039] Where E(f) is the energy of the vibration signal power spectral density at frequency f. The pre-calibrated main frequency under healthy conditions, measured in Hz, can effectively indicate mechanical faults such as loosening and wear inside the solenoid valve by observing changes in the spectral energy distribution.

[0040] For temperature sensors, two core types of thermodynamic features are extracted: One is the steady-state temperature rise, calculated using the following formula:

[0041] in, This refers to the temperature of the coil casing, in degrees Celsius (°C). The ambient temperature is expressed in °C. This parameter reflects the coil power consumption and heat dissipation conditions. Abnormal temperature rise can indicate problems such as short circuits between coil turns or poor ventilation.

[0042] The second is the rate of temperature rise, calculated using the following formula:

[0043] in, , The parameter, obtained from two consecutive sampling points after power-on, can rapidly assess the thermal state of the coil through the initial temperature rise rate, enabling early identification of thermal faults. After the aforementioned feature extraction, multidimensional state characteristic parameters encompassing electromagnetic, mechanical, and thermodynamic properties are ultimately obtained, providing a data foundation for subsequent health assessments and lifespan predictions.

[0044] Step S2: Construct a personalized health baseline based on the historical operating data of the solenoid valve, compare and analyze the multi-dimensional state characteristic parameters with the personalized health baseline, and calculate a comprehensive health index that reflects the current health status of the solenoid valve.

[0045] This step relies on the platform-side early warning and predictive analysis system to achieve a quantitative assessment of health status, such as... Figure 3 As shown, after receiving multi-dimensional status feature data uploaded via the communication network, the platform-side system performs real-time data parsing and verification through the data receiving server, and stores the valid data in the historical database for persistent storage. At the same time, it retrieves the built-in fault feature library and the device-specific health basic model to provide sufficient data support and model foundation for subsequent health baseline comparison and comprehensive evaluation.

[0046] In this embodiment, to avoid problems such as poor adaptability and high misjudgment rate caused by uniform threshold determination, the long-term historical operating data of the solenoid valve under healthy operating conditions are first retrieved, and the baseline mean value corresponding to each state characteristic parameter is determined through statistical analysis. with standard deviation This allows for the construction of a personalized health baseline that adapts to the actual operating conditions, environmental conditions, and individual performance differences of a single solenoid valve. Compared to the traditional fixed threshold method, this personalized baseline can more realistically reflect the normal operating range of the equipment, significantly improving the accuracy and reliability of health status determination.

[0047] After completing the construction of a personalized health baseline, the obtained state feature parameters of each dimension will be collected and extracted in real time. The normalized deviation from the corresponding health baseline is calculated to obtain an independent deviation index for each feature parameter. The specific calculation formula is as follows:

[0048] In the formula Let j be the measured value of the j-th feature parameter at the current time. This is the baseline mean of the characteristic parameter under historical health conditions. This represents the historical baseline standard deviation corresponding to the feature parameter. This deviation index can intuitively reflect the degree of deviation of the real-time feature relative to its health status. In this embodiment, corresponding judgment rules are set; when the deviation index of a certain feature... When the value is greater than 3, it can be determined that the feature has a significant abnormality, and the system will immediately trigger a single-dimensional feature abnormality alarm, thereby realizing rapid and accurate identification of single-type faults such as electrical, mechanical and thermal faults, and providing a direct basis for early hidden danger investigation.

[0049] On the basis of completing the single-dimensional feature deviation analysis, in order to comprehensively reflect the overall health level of the solenoid valve, the entropy weight method or expert scoring method is used to assign weights to each feature parameter to obtain the corresponding weighting coefficients , all weighting coefficients satisfy the constraint condition that the sum of weights is 1, ensuring that the contribution distribution of each feature to the overall health state is reasonable. Then, multi-parameter fusion calculation is performed based on the deviation degree of each feature and the corresponding weighting coefficients to obtain the normalized comprehensive health index HI, and the calculation formula is . The value range of the comprehensive health index is standardized between 0 and 1, which can intuitively quantify the current overall health deterioration degree of the solenoid valve. In the embodiment of the present application, a hierarchical determination standard is set: when HI≥0.85, the solenoid valve is determined to be in a healthy state; when 0.6≤HI<0.85, it is determined to be in an attention state, indicating that the equipment has a slight deterioration trend; when HI<0.6, it is determined to be in an early warning state, indicating that the equipment has obvious performance abnormality. Through the hierarchical determination of the comprehensive health index, the current overall health degree of the solenoid valve can be reflected comprehensively and objectively, providing a key basis for subsequent fault early warning output, adaptive update of health baseline and remaining useful life prediction.

[0050] Step S3: determining according to the comprehensive health index and a preset dynamic early warning threshold, generating fault early warning information when abnormal health status is detected, and adaptively updating the health baseline by using the current multi-dimensional state feature parameters when the health status is determined.

[0051] This step corresponds to the core link of intelligent diagnosis and early warning generation of the platform-side early warning and predictive analysis system, such as Figure 3 as shown in the figure, the intelligent diagnosis engine built in the platform-side system will integrate the comprehensive health index calculated in step S2 and the independent deviation index of each feature parameter and other multi-source information, and complete the accurate determination of the current health state of the solenoid valve through multi-dimensional fusion analysis. Specifically, the comprehensive health index is compared in real time with the dynamic early warning threshold preset by the system, and at the same time, combined with the determination result of the single feature deviation degree (i.e. whether it is greater than 3) to comprehensively judge whether to trigger a fault early warning, so as to ensure the comprehensiveness and accuracy of early warning generation and avoid false alarms and missed alarms caused by a single determination standard.

[0052] In this embodiment, considering that solenoid valves will naturally age during long-term operation, and that seasonal changes will cause changes in external conditions such as ambient temperature, if a fixed health baseline is used, baseline deviation will occur over time, thus affecting the accuracy of health assessment and early warning. Therefore, an exponentially weighted moving average algorithm is used to construct an adaptive health baseline update model. The specific calculation formula is as follows:

[0053] Here, α is the learning rate, typically set between 0.05 and 0.2. The core function of this learning rate is to control the update speed of the health baseline, ensuring that the baseline can adapt to changes in device status in a timely manner, while avoiding excessive fluctuations in the baseline due to too rapid updates, thus ensuring the stability and reliability of the baseline.

[0054] When the intelligent diagnostic engine determines that the current multidimensional state feature parameters are in a healthy state (i.e., the comprehensive health index HI ≥ 0.85, and the deviation index of all feature parameters is within a healthy range), When the value is ≤3), the system will automatically use the currently collected feature parameters from each dimension to calculate the mean of the personalized health baseline. with standard deviation By performing slow updates and using the aforementioned adaptive update model, the health baseline can dynamically adapt to the natural aging process of the solenoid valve, while also taking into account the impact of seasonal environmental changes. This effectively avoids baseline deviations caused by long-term equipment operation and external environmental fluctuations, further improving the long-term accuracy and adaptability of health assessments.

[0055] Conversely, when the comprehensive health index is lower than the preset dynamic warning threshold (i.e., HI < 0.6), or when there is a single feature parameter that is significantly abnormal and meets the following conditions... When the condition >3 is met, the platform-side system's early warning generation module will immediately activate the early warning mechanism, generating corresponding fault early warning information in real time. This early warning information will include key information such as the type of abnormality, the degree of abnormality, the current health index, and potential fault hazards. On the one hand, it will be visualized through the platform-side human-machine interface, pushing alarm information to operators through pop-ups, audio-visual prompts, etc., so that operators can quickly know the abnormality of the equipment. On the other hand, the early warning information will be simultaneously transmitted to the operation and maintenance support system, providing accurate data support for maintenance personnel to formulate targeted troubleshooting and repair plans. Ultimately, this achieves dynamic and adaptive health status monitoring and fault early warning of the solenoid valve, ensuring the safe and stable operation of the equipment.

[0056] Step S4: Based on the temporal evolution law of the multidimensional state characteristic parameters, construct a performance degradation model, deduce the future change trend of key performance indicators, and calculate the remaining service life of the solenoid valve in combination with the preset failure threshold, thus completing the full-cycle health monitoring and remaining service life prediction of the solenoid valve.

[0057] This step corresponds to the lifespan prediction module of the platform-side system, such as... Figure 3 As shown, the life prediction module completes performance degradation modeling and remaining service life calculation based on historical time-series data and real-time monitoring data. In this embodiment, the valve core actuation time is first selected. Key performance indicators are used as degradation characteristics. An exponential degradation trajectory model is fitted based on historical time-series data, and the calculation formula is as follows:

[0058] in, As a key performance indicator, The initial health value is given by λ, which is the degradation rate coefficient obtained by fitting historical data. The zero-mean random noise term is used to simulate operating condition fluctuations and random disturbances during actual operation, improving the accuracy of performance trend prediction. This exponential degradation model conforms to the evolution law of most mechanical wear and can accurately characterize the performance degradation process of solenoid valves.

[0059] By using a degradation model to extrapolate the future change paths of key performance indicators (KPIs) time-by-time, dynamic tracking and trend prediction of the solenoid valve's performance degradation process can be achieved. Based on this, the remaining service life (RUL) of the solenoid valve is calculated using a preset failure threshold. First, the failure thresholds for key performance indicators are pre-set. For example, setting a failure threshold when the valve core's operating time exceeds 1.5 times the initial value, and using the moment the indicator reaches the failure threshold as the end of the service life. The formula for calculating the remaining service life is... Under the exponential degradation model, an analytical solution can be derived. ,in The current performance index is the measured value, and λ is the degradation rate coefficient. This is a preset failure threshold.

[0060] After calculating the remaining useful life, the operation and maintenance support system combines real-time health status assessment, dynamic fault early warning, and the remaining useful life calculation results to complete full-cycle health monitoring and life prediction management, such as... Figure 3 As shown, the operation and maintenance support system is connected to the spare parts management system and the maintenance plan system respectively. Based on the remaining service life and health status, it generates accurate spare parts procurement plans and maintenance plans, which ultimately guide on-site operation and maintenance. At the same time, operators can visualize the health status, early warning information and life prediction results of the solenoid valve through the human-machine interface, realize the transformation from "periodic maintenance" to "post-maintenance" to "predictive maintenance", and complete the full-cycle health monitoring and life prediction management of the solenoid valve from commissioning to failure.

[0061] In the application of one embodiment of the present invention, the implementation process is as follows: This application addresses the actual operating scenarios of solenoid valves in hydropower plant hydraulic systems. Combining online monitoring, feature extraction, health assessment, and lifespan prediction processes, it provides a complete engineering application example. This example implements early jamming fault warning, lifespan alerts based on rated number of operations, and remaining lifespan prediction based on a performance degradation model. The overall monitoring, diagnosis, early warning, and handling processes are as follows: Figure 4 , Figure 5 , Figure 6 As shown.

[0062] In practical applications, the system described in this application uses an intelligent monitoring terminal to collect dynamic current, mechanical vibration, and temperature signals during the operation of the solenoid valve in real time, and performs multi-dimensional feature extraction according to corresponding formulas. For the current signal, Formula 1 is used to calculate the peak starting current. ,in This is the instantaneous value of the coil current. The moment when power is applied. The moment the valve core engages reflects the balance between electromagnetic attraction and mechanical resistance; the effective value of the holding current is calculated using Formula 3. This is used to determine if the coil resistance and thermal condition are abnormal. For the vibration signal, the action time is calculated using Formula 2. The speed of mechanical action is characterized by the speed of mechanical action; the vibration and impact energy is calculated using Formula 4. This reflects the impact intensity of the valve core; the proportion of vibration dominant frequency energy is calculated using Formula 5. This is used to identify internal wear and loosening. For the temperature signal, the steady-state temperature rise is calculated using Formula 6. The rate of temperature rise is calculated using Formula 7. This enables rapid assessment of the coil's thermal state.

[0063] After feature extraction is completed, the platform-side early warning and predictive analysis system calculates the deviation of each feature based on Formula 9. ,when A score greater than 3 indicates a single-feature abnormality; simultaneously, Formula 8 is used to calculate the comprehensive health index. The system classifies data according to a HI ≥ 0.85 for healthy individuals, 0.6 ≤ HI < 0.85 for those requiring attention, and HI < 0.6 for those in a warning state. Based on this, the system uses the exponentially weighted moving average model from Formula 10. The system adaptively updates the health baseline with a learning rate α ranging from 0.05 to 0.2 to adapt the baseline to equipment aging and environmental changes. When lifetime prediction is required, the system constructs an exponential degradation trajectory based on Equation 10. And combined with the analytical expression of Formula 11 Calculate the remaining reliable service life.

[0064] In a practical application at a hydropower plant, the first step was to implement early warning of solenoid valve jamming faults, such as... Figure 4 As shown, the complete process from data acquisition and feature extraction to early warning push is clearly presented. Specifically, the current sensor, vibration sensor, and temperature sensor of the intelligent monitoring terminal are activated simultaneously to collect the instantaneous value of the solenoid valve coil current, the vibration acceleration signal of the valve core action, the coil shell temperature, and the ambient temperature, respectively. After the acquisition is completed, the signal is amplified and filtered before being transmitted to the controller. The controller uses the built-in algorithm to calculate the action time using formula 2 and the steady-state temperature rise using formula 6. After feature extraction is completed, the data is uploaded to the platform side via the communication module. After receiving the data, the platform side calls formula 8 to calculate the comprehensive health index and formula 9 to calculate the feature deviation. It compares the data with the fault feature database to determine the fault type and generate early warning information. Finally, the early warning information is pushed to the human-machine interface to notify the operation and maintenance personnel to handle the situation.

[0065] In this application, the monitoring system installed on the main pressure regulating valve of the speed governor is in accordance with... Figure 4 The process described above collects action data in real time. It was found that the average action time of the solenoid valve core abruptly changed from 52ms to 70ms within a day, and the coil temperature rose from 34℃ to 37℃. The system, through deviation calculation and health index assessment, determined that the characteristic changes highly matched the early fault modes of valve core sludge adhesion and slight jamming. It then automatically generated a level-two warning, indicating a risk of mechanical action lag. Maintenance personnel promptly inspected the unit during peak and off-peak hours. Upon disassembly, they found dense oily sludge adhering to the valve core surface. After cleaning and replacing the valve core, the action time and temperature returned to normal, effectively preventing accidental downtime caused by jamming and fully reproducing the fault. Figure 4 The entire execution process is shown.

[0066] Secondly, the system implements a lifespan warning function based on the rated number of actions, corresponding to Figure 5 The diagram clearly illustrates the process from parameter input and action count statistics to early warning generation and maintenance handling. Specifically, maintenance personnel input the rated effective action count of the solenoid valve into the life prediction database through the platform-side user interface. The intelligent monitoring terminal collects the solenoid valve action trigger signals in real time, accumulates the actual action count, and uploads it to the platform side. The platform-side storage server records the action count data. The life prediction module continuously compares the actual action count with the rated threshold. When the actual count approaches the rated threshold, the early warning generation module is triggered to generate a life warning information. The warning information is pushed to the human-machine interface and transmitted to the maintenance support system to generate a spare parts procurement and maintenance plan, guiding maintenance personnel to replace the solenoid valve.

[0067] In this application, for the governor locking solenoid valve, the maintenance personnel entered its rated effective number of 2000 operations into the life prediction database according to the product manual. The system then proceeded accordingly. Figure 5 The process shown continuously accumulates the actual number of solenoid valve operations. When the number of operations approaches the set threshold of 2000, a reliability degradation alarm is automatically issued, reminding maintenance personnel to arrange for timely replacement. This achieves preventative maintenance based on the design lifespan, fully complying with... Figure 5 The process logic ensures the timeliness and accuracy of lifespan warnings.

[0068] Simultaneously, the system can also achieve more refined lifetime prediction based on performance degradation trends, corresponding to Figure 6 The diagram fully illustrates the entire process from feature monitoring and degradation modeling to lifespan calculation and work order generation. Specifically, the intelligent monitoring terminal continuously collects signals such as vibration and current from the solenoid valve, extracts feature parameters such as action time and the proportion of vibration main frequency energy, and uploads them. After receiving the data, the platform's intelligent diagnostic engine calculates the comprehensive health index, the lifespan prediction module uses formula 10 to construct an exponential degradation trajectory model, fits the degradation rate coefficient with historical time-series data, and then calculates the remaining lifespan using formula 11. Finally, the early warning generation module generates a lifespan early warning work order and pushes it to the maintenance support system to guide repair and replacement. In this application, during the monitoring of the emergency pressure regulating valve solenoid valve, the system follows... Figure 6 The process shown indicates that the standard deviation of the valve core's full-stroke action time increased from ±1.2ms to ±4.5ms, and the vibration signal characteristics became significantly discretized. Based on health index calculations, the valve's HI value dropped to 72. Combined with the exponential degradation model, the system predicts that its remaining reliable operation count is only 8-15 times, and automatically generates a lifespan warning work order, indicating that wear of mechanical components has led to decreased operational consistency, and recommending replacement during the next maintenance. This method transforms traditional periodic replacement into condition-based precise replacement, ensuring operational safety while fully utilizing spare parts' lifespan, achieving a balance between safety and economy, and fully validating [the system's capabilities]. Figure 6 The practicality and reliability of the lifetime prediction process shown.

[0069] The above embodiments fully combine Figure 4 , Figure 5 , Figure 6 The specific process content, combined with the practical application of each formula, fully verifies the practicality and reliability of this application in early warning of solenoid valve failure, adaptive health assessment and remaining life prediction. It can effectively support the transformation of solenoid valves in hydropower plant hydraulic systems from post-maintenance and periodic inspection to predictive maintenance. The process steps in each figure have been fully implemented in practical applications, ensuring the feasibility and repeatability of the technical solution of this application.

[0070] Example 2 This invention provides a device for early warning and life prediction of solenoid valve failure in a hydroelectric power plant's hydraulic system. Figure 7This is a schematic flowchart illustrating a solenoid valve fault early warning and lifespan prediction device for a hydroelectric power plant's hydraulic system, provided in an embodiment of the present invention. Figure 7 As shown, the device includes: The multi-state acquisition module 100 is used to synchronously acquire the dynamic current signal, mechanical vibration signal and temperature change signal of the solenoid valve during the operation process, and extract multi-dimensional state characteristic parameters that characterize the electromagnetic properties, mechanical action properties and thermodynamic properties. The health assessment module 200 is used to construct a personalized health baseline based on the historical operating data of the solenoid valve, compare and analyze the multi-dimensional status characteristic parameters with the personalized health baseline, and calculate a comprehensive health index that reflects the current health status of the solenoid valve. The early warning update module 300 is used to make a judgment based on the comprehensive health index and the preset dynamic early warning threshold, generate fault early warning information when an abnormal health status is detected, and adaptively update the health baseline using the current multi-dimensional status feature parameters when the health status is determined to be healthy. The life prediction module 400 is used to construct a performance degradation model based on the time-series evolution law of the multidimensional state characteristic parameters, deduce the future change trend of key performance indicators, and calculate the remaining service life of the solenoid valve in combination with the preset failure threshold, thereby completing the full-cycle health monitoring and remaining life prediction of the solenoid valve.

[0071] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0072] Example 3 To implement the methods of the above embodiments, the present invention also provides an electronic device, which includes a memory and a processor; wherein the processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement the various steps of the methods described above.

[0073] Example 4 To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method described in the foregoing embodiments.

[0074] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0075] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0076] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

Claims

1. A method for early warning and life prediction of solenoid valve failure in a hydropower plant's hydraulic system, characterized in that, include: The system synchronously acquires dynamic current signals, mechanical vibration signals, and temperature change signals of the solenoid valve during its operation, and extracts multidimensional state characteristic parameters that characterize its electromagnetic, mechanical, and thermodynamic properties. A personalized health baseline is constructed based on the historical operating data of the solenoid valve. Multidimensional state characteristic parameters are compared and analyzed with the personalized health baseline to calculate a comprehensive health index that reflects the current health status of the solenoid valve. The health status is determined based on the comprehensive health index and the preset dynamic warning threshold. When an abnormal health status is detected, a fault warning message is generated. When the health status is determined to be healthy, the health baseline is adaptively updated using the current multi-dimensional status feature parameters. Based on the temporal evolution law of the multidimensional state characteristic parameters, a performance degradation model is constructed to deduce the future change trend of key performance indicators. Combined with the preset failure threshold, the remaining service life of the solenoid valve is calculated, thus completing the full-cycle health monitoring and remaining service life prediction of the solenoid valve.

2. The method according to claim 1, characterized in that, The system synchronously acquires dynamic current signals, mechanical vibration signals, and temperature change signals of the solenoid valve during its operation, and extracts multidimensional state characteristic parameters characterizing its electromagnetic, mechanical, and thermodynamic properties, including: The instantaneous value of the coil current is collected by a current sensor installed on the coil circuit of the solenoid valve. The characteristics of the peak value of the solenoid valve starting current and the effective value of the steady-state holding current are obtained, which respectively characterize the mechanical jamming of the valve core, the fatigue state of the spring and the abnormal change of the DC resistance of the coil. Vibration acceleration signals during the valve core movement process are collected by vibration sensors attached to the valve body of the solenoid valve. The valve core movement time, vibration impact energy and vibration main frequency energy ratio are extracted to characterize the mechanical movement time of the solenoid valve, the impact intensity between the valve core and the valve seat and the internal mechanical wear and loosening state, respectively. By collecting the coil housing temperature and ambient temperature through a temperature sensor attached to the solenoid valve coil housing, the steady-state temperature rise and temperature rise rate characteristics are calculated to characterize the coil power consumption, heat dissipation conditions and abnormal thermal conditions. After the signals from each sensor are amplified and filtered by a signal amplifier, the controller completes data acquisition and feature fusion to finally obtain multi-dimensional state characteristic parameters of electromagnetic, mechanical and thermodynamic dimensions.

3. The method according to claim 2, characterized in that, A personalized health baseline is constructed based on the historical operating data of the solenoid valve. Multidimensional state characteristic parameters are compared and analyzed with the personalized health baseline to calculate a comprehensive health index reflecting the current health status of the solenoid valve, including: Based on the historical health operation data of the solenoid valve, the baseline mean and standard deviation of each state characteristic parameter are determined to form a personalized health baseline for each device. The real-time collected state feature parameters of each dimension are normalized and the deviation is calculated with the corresponding health baseline to obtain the independent deviation index of each feature parameter. The weighting coefficients of each feature parameter are determined by using the entropy weighting method or the expert scoring method. Under the constraint that the sum of the weights is 1, the comprehensive health index is obtained by fusion calculation.

4. The method according to claim 3, characterized in that, The system determines the health status based on the comprehensive health index and a preset dynamic early warning threshold. When an abnormal health status is detected, a fault warning is generated. When the status is determined to be healthy, the health baseline is adaptively updated using the current multi-dimensional status feature parameters, including: An adaptive healthy baseline update model is constructed using an exponentially weighted moving average algorithm, and a reasonable learning rate is set to control the baseline update speed. When determining that the current multidimensional state characteristic parameters are in a healthy state, the mean and standard deviation of the health baseline are slowly updated using the current characteristic parameters so that the health baseline can adapt to the natural aging of the solenoid valve and seasonal environmental changes. When the overall health index is below the warning threshold or a single feature is significantly abnormal, corresponding fault warning information is generated in real time.

5. The method according to claim 4, characterized in that, A performance degradation model is constructed based on the temporal evolution law of the multidimensional state characteristic parameters to deduce the future changing trend of key performance indicators, including: The valve core action time was selected as a key performance indicator for degradation. An exponential degradation trajectory model was fitted based on historical time series data to determine the initial health value and degradation rate coefficient. A zero-mean random noise term is introduced into the degradation model to simulate the fluctuations in operating conditions and random disturbances during actual operation. By using a degradation model to extrapolate the future change path of key performance indicators at each moment, dynamic tracking and trend prediction of the performance degradation process of solenoid valves can be achieved.

6. The method according to claim 5, characterized in that, The remaining service life of the solenoid valve is calculated based on a preset failure threshold, including: Pre-set failure thresholds for key performance indicators, and use the moment when the indicator reaches the failure threshold as the end of its lifespan. Based on the analytical relationship of the exponential degradation model, combined with the measured values ​​of current performance indicators, degradation rate coefficient and failure threshold, the remaining service life of the solenoid valve is calculated. By combining real-time health status assessment, dynamic fault early warning and remaining service life calculation results, the full-cycle health monitoring and life prediction management of solenoid valves from commissioning to failure can be completed.

7. A device for early warning and life prediction of solenoid valve failure in a hydropower plant hydraulic system, characterized in that, include: The multi-state acquisition module is used to synchronously acquire dynamic current signals, mechanical vibration signals, and temperature change signals of the solenoid valve during its operation, and extract multi-dimensional state characteristic parameters that characterize its electromagnetic, mechanical, and thermodynamic properties. The health assessment module is used to construct a personalized health baseline based on the historical operating data of the solenoid valve, compare and analyze the multi-dimensional status characteristic parameters with the personalized health baseline, and calculate a comprehensive health index that reflects the current health status of the solenoid valve. The early warning update module is used to make a judgment based on the comprehensive health index and the preset dynamic early warning threshold. When an abnormal health status is detected, it generates a fault early warning information. When the health status is determined to be healthy, it adaptively updates the health baseline using the current multi-dimensional status feature parameters. The life prediction module is used to construct a performance degradation model based on the time-series evolution law of the multidimensional state characteristic parameters, deduce the future change trend of key performance indicators, and calculate the remaining service life of the solenoid valve in combination with the preset failure threshold, so as to complete the full-cycle health monitoring and remaining life prediction of the solenoid valve.

8. The apparatus according to claim 7, characterized in that, The multi-state acquisition module is also used for: The instantaneous value of the coil current is collected by a current sensor installed on the coil circuit of the solenoid valve. The characteristics of the peak value of the solenoid valve starting current and the effective value of the steady-state holding current are obtained, which respectively characterize the mechanical jamming of the valve core, the fatigue state of the spring and the abnormal change of the DC resistance of the coil. Vibration acceleration signals during the valve core movement process are collected by vibration sensors attached to the valve body of the solenoid valve. The valve core movement time, vibration impact energy and vibration main frequency energy ratio are extracted to characterize the mechanical movement time of the solenoid valve, the impact intensity between the valve core and the valve seat and the internal mechanical wear and loosening state, respectively. By collecting the coil housing temperature and ambient temperature through a temperature sensor attached to the solenoid valve coil housing, the steady-state temperature rise and temperature rise rate characteristics are calculated to characterize the coil power consumption, heat dissipation conditions and abnormal thermal conditions. After the sensor-acquired signals are amplified and filtered by a signal amplifier, the controller completes data acquisition and feature extraction to obtain multi-dimensional state characteristic parameters covering electromagnetic properties, mechanical motion properties, and thermodynamic properties.

9. An electronic device, characterized in that, Including processor and memory; The processor reads executable program code stored in the memory to run a program corresponding to the executable program code, so as to implement the method as described in any one of claims 1-6.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-6.

Images