Digital Intelligence Analysis System Based on Physical Equipment of Hydropower Station

Abstract

The present invention relates to the technical field of digital and intelligent analysis, specifically a digital and intelligent analysis system based on physical equipment of a hydropower station. The digital and intelligent analysis system based on physical equipment of a hydropower station includes a state feature extraction module, a fault identification and early warning module, a deep analysis module of equipment behavior, and a collaborative learning and prediction module. In the present invention, by calculating the deviation between the current operation data of the equipment and the normal operation benchmark, and automatically identifying and marking the out-of-bounds data points, combined with time series analysis, potential abnormalities in the operation of the equipment can be timely identified in the management of physical equipment of the hydropower station. The behavior patterns of the equipment under different operating conditions are deeply analyzed, which can reveal the change rules of the behavior and the prediction time of periodic faults, reduce unplanned outages, while identifying the performance degradation trend, continuously updating the equipment health status, predicting and preventing potential future risks in advance, optimizing the overall operation efficiency and economy of the hydropower station, and ensuring the safe long-term operation of the equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of digital and intelligent analysis, and particularly to a digital and intelligent analysis system based on physical equipment of a hydropower station. Background Art

[0002] Digital and intelligent analysis technology is the integration of information technology and intelligent technology. Using modern technical means such as big data, cloud computing, artificial intelligence, and the Internet of Things, it deeply analyzes and intelligently processes various types of data. The main purpose is to optimize business processes, improve operation efficiency, and decision-making quality through the efficient management and analysis of data. In the industrial field, digital and intelligent analysis helps enterprises achieve real-time monitoring of equipment status, fault prediction, maintenance optimization, and energy efficiency management, thereby enhancing the reliability and economy of production.

[0003] Among them, the digital and intelligent analysis system based on physical equipment of a hydropower station is a digital and intelligent analysis system specifically for physical equipment of a hydropower station. The purpose is to achieve real-time monitoring and performance analysis of key equipment such as turbines, generators, and sluice gates in a hydropower station by integrating sensor data and operation records. Using advanced data analysis tools, it can predict equipment failures, optimize maintenance plans, and improve energy efficiency. In addition, it can also assist the hydropower station to more accurately manage and dispatch water resources by analyzing the utilization of water resources, enhance the operation efficiency and economic benefits of the hydropower station, and ensure the safety and long-term sustainability of equipment operation.

[0004] In the current technology for the monitoring and maintenance of hydropower station equipment, although real-time monitoring and basic fault prediction are achieved, it often relies on traditional data processing technologies, showing obvious deficiencies in analyzing complex behavior patterns and long-term operation trends of equipment, being unable to effectively identify minor performance degradations or predict the accurate time of periodic failures, resulting in a lag in response, increasing unexpected shutdowns and maintenance costs. In addition, the lack of efficient comprehensive analysis of performance data of physical equipment in hydropower stations leads to the inability of hydropower stations to maximize resource utilization efficiency, affecting economic and environmental sustainability. Summary of the Invention

[0005] The present invention provides a digital and intelligent analysis system based on physical equipment of a hydropower station.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] The digital and intelligent analysis system based on physical equipment of a hydropower station includes:

[0008] The state feature extraction module collects temperature, pressure, and vibration data of the operation of physical equipment in a hydropower station through sensors, filters out background noise, divides data time windows, captures the operation state features of physical equipment in a hydropower station, and outputs a set of equipment state features;

[0009] Based on the device status feature set, the fault identification and early warning module analyzes the operation data of the physical equipment of the hydropower station, calculates the deviation between the current operation data of the physical equipment of the hydropower station and the normal operation reference data, automatically marks the out-of-bounds data points according to the calculation results, and performs time series analysis on the out-of-bounds data points to identify potential abnormal operation of the physical equipment of the hydropower station and generate preliminary fault early warning indicators;

[0010] Based on the preliminary fault early warning indicators, the in-depth analysis module of equipment behavior performs retrospective analysis on the historical behavior of the physical equipment of the hydropower station. By analyzing the behavior patterns of the physical equipment of the hydropower station under different operating conditions, it analyzes and determines the abnormal behavior before the fault of the physical equipment of the hydropower station. According to the analysis and determination results, it performs periodic analysis on the abnormal behavior, reveals the change law of the behavior of the physical equipment of the hydropower station, captures the time frame of periodic fault occurrence, and obtains the equipment behavior analysis results;

[0011] The collaborative learning and prediction module receives the equipment behavior analysis results, initiates the data synchronization process between the cloud and edge computing resources, analyzes and compares the long-term data and short-term activities of the behavior of the physical equipment of the hydropower station, identifies the performance degradation trend of the physical equipment of the hydropower station, continuously monitors the physical equipment of the hydropower station according to the identification results, updates the health status record of the physical equipment of the hydropower station, predicts the potential risks of the physical equipment of the hydropower station in the future, and establishes a health management file for the physical equipment of the hydropower station.

[0012] As a further solution of the present invention, the steps for obtaining the device status feature set are as follows:

[0013] Deploy sensors around the physical equipment of the hydropower station to capture the temperature, pressure and vibration data during the operation of the equipment in real time, and screen out the data points whose frequencies are not within the working frequency range to obtain the pure operation status data of the equipment;

[0014] Based on the pure operation status data of the equipment, perform time series segmentation according to the equipment operation cycle. Each time segment represents a working cycle, and calculate the data mean, standard deviation and range within each cycle to obtain the statistical description of the equipment cycle data;

[0015] Using the statistical description of the equipment cycle data and combining with the equipment operation rules, evaluate the relationship between the data fluctuation within each cycle and the equipment health status, reveal the current state of the equipment, and obtain the device status feature set.

[0016] As a further solution of the present invention, the steps for calculating the deviation are as follows:

[0017] Based on the device status feature set, extract the real-time operation data of the physical equipment of the hydropower station, capture the current working state of the equipment, and generate a real-time data set;

[0018] Based on the real-time data set, combined with the normal operation reference data of the equipment, use the formula:

[0019]

[0020] Calculate the total deviation value Deviation to generate the total deviation result, where n represents the total number of data points, i represents the data point index, and D r represents the real-time dataset, D b Denotes the benchmark dataset, D ri represents the value of the i-th data point in the real-time data set, D bi represents the value of the i-th data point in the benchmark dataset;

[0021] The total deviation result is compared with a preset abnormal threshold T. If Deviation>T, it is marked as abnormal and a device deviation analysis result is generated.

[0022] As a further solution of the present invention, the steps for obtaining the preliminary fault warning indicator are:

[0023] According to the device deviation analysis result, all data points exceeding the preset abnormal threshold are automatically marked as out-of-bounds data points, the deviation size and timestamp of each out-of-bounds data point are recorded, and a set of out-of-bounds data points is output;

[0024] Performing time series analysis based on the cross-border data point set, identifying periodic fluctuations and irregular changes in the data by examining the time dependency of the data points, revealing potential operating problems, and obtaining time series analysis results;

[0025] Using the time series analysis results, the current analysis results are matched with the fault signatures stored in the database to determine the fault type corresponding to the current out-of-bounds data point and obtain preliminary fault warning indicators.

[0026] As a further solution of the present invention, the steps of analyzing and determining the abnormal behavior before the fault are:

[0027] Calling the preliminary fault warning indicator, integrating the historical operation data of the hydropower station physical equipment, counting the operation time, load conditions and maintenance records, classifying the data according to the time axis, removing abnormal data points, and generating a historical behavior data set of the hydropower station physical equipment;

[0028] Based on the historical behavior data set of the hydropower station physical equipment, the behavior change law of the equipment under different operating conditions is analyzed, and the formula is used:

[0029]

[0030] Calculate the device behavior pattern score S and generate the behavior pattern determination result, where B i is the operating load under different operating conditions, is the average value of the operation load, C i is the maintenance frequency, is the average value of the maintenance frequency, and N is the total number of data points;

[0031] Determine the result according to the described behavior pattern, identify the abnormally elevated part in the pattern score, compare the corresponding positions of the fault warning index and the abnormal behavior of the score, and obtain the abnormal behavior recognition result before the fault.

[0032] As a further solution of the present invention, the steps for obtaining the device behavior analysis result are:

[0033] Utilize the abnormal behavior recognition result before the fault to perform time series analysis on the abnormal behavior, determine the occurrence frequency and periodicity, analyze the time points when the abnormal behavior occurs, and generate a periodic behavior pattern data set;

[0034] Based on the periodic behavior pattern data set, analyze the frequency components of the device behavior, identify the key periodic components, count the frequency peaks in the data set, reveal the key change rules of the device behavior, and obtain the change rule analysis result;

[0035] Based on the change rule analysis result, combine the device maintenance records and the fault history comparison to determine the possible occurrence time frame of the periodic fault, analyze the future potential fault time points of the device, and generate the device behavior analysis result.

[0036] As a further solution of the present invention, the steps for identifying the performance degradation trend are:

[0037] Receive the device behavior analysis result, count the operation data, performance indicators and abnormal event records of the device, perform data integrity verification and timestamp check, confirm the consistency and reliability of the data, and generate the device data verification result;

[0038] By establishing a data channel, synchronize the device data verification result between the cloud and the edge computing resources, confirm that all computing nodes have a consistent data view, and generate a data synchronization completion confirmation result;

[0039] According to the data synchronization completion confirmation result, compare the long-term and short-term device behavior data, identify the performance degradation trend of the device, analyze the device performance degradation mode, and obtain the device performance degradation analysis result.

[0040] As a further solution of the present invention, the steps for establishing the physical device health management file of the hydropower station are:

[0041] According to the device performance degradation analysis result, monitor the device temperature, pressure and load, perform real-time data acquisition to obtain the real-time operation data of the key parameters, and generate an organized real-time monitoring data set;

[0042] Based on the sorted real-time monitoring data set, combined with the historical performance data of the equipment, analyze the changing trend of the current health status of the equipment item by item, and use the formula:

[0043]

[0044] Calculate the change amplitude H of the equipment health status, and generate an updated equipment health status record. Among them, P i and P i-1 respectively represent the key performance parameter values at adjacent time points, t i和 t i-1 are the corresponding time points respectively, and M is the total number of data points;

[0045] According to the updated equipment health status record, combined with the historical fault record, sort out the changing trend of the key parameters of equipment performance degradation, analyze the potential risk probability of the equipment in the future, and generate a health management file for the physical equipment of the hydropower station.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0047] In the present invention, by calculating the deviation between the current operation data of the equipment and the normal operation benchmark, automatically identifying and marking the out-of-bounds data points, and combining time series analysis, potential anomalies in the equipment operation can be timely identified in the management of the physical equipment of the hydropower station, deeply analyzing the behavior patterns of the equipment under different operating conditions, revealing the changing rules of the behavior and the prediction time of periodic faults, reducing unplanned outages, simultaneously identifying the performance degradation trend, continuously updating the equipment health status, predicting and preventing potential future risks in advance, optimizing the overall operation efficiency and economy of the hydropower station, and ensuring the safe long-term operation of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for description in the embodiments. The drawings are only used for the purpose of showing the implementation manner and are not considered as a limitation to the present invention.

[0049] Figure 1 is the system flow chart of the present invention;

[0050] Figure 2 is the acquisition flow chart of the equipment status feature set of the present invention;

[0051] Figure 3 is the calculation flow chart of the deviation of the present invention;

[0052] Figure 4 is the acquisition flow chart of the preliminary fault warning index of the present invention;

[0053] Figure 5 Flow chart for analyzing and determining abnormal behavior before a fault in the present invention

[0054] Figure 6 Flow chart for obtaining the analysis result of the device behavior in the present invention

[0055] Figure 7 Flow chart for identifying the performance degradation trend in the present invention

[0056] Figure 8 Flow chart for establishing the health management file of the physical equipment of the hydropower station in the present invention Detailed implementation manners

[0057] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without making creative efforts belong to the scope of protection of the present invention.

[0058] Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs; the terms used in the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The terms "including" and "having" in the specification and claims of the present invention and any variations thereof are intended to cover non-exclusive inclusions.

[0059] In the description of the embodiments of the present invention, technical terms such as "first" and "second" are only used to distinguish different objects, and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity, specific order or primary-secondary relationship of the indicated technical features. In the description of the embodiments of the present invention, the meaning of "a plurality of" is more than two, unless otherwise specifically defined.

[0060] In the description of the embodiments of the present invention, the term "and / or" is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the front and rear associated objects.

[0061] In the description of the embodiments of the present invention, the term "a plurality of" refers to more than two (including two). Similarly, "a plurality of groups" refers to more than two groups (including two groups), and "a plurality of pieces" refers to more than two pieces (including two pieces).

[0062] In the description of the embodiments of the present invention, the orientation or positional relationship indicated by technical terms such as "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the embodiments of the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the embodiments of the present invention.

[0063] In the description of the embodiments of the present invention, unless otherwise clearly specified and limited, technical terms such as "installation", "connection", "connection", "fixation", etc. should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or integrated; it can also be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of the present invention can be understood according to specific situations.

[0064] Embodiment 1. The embodiments of the present invention provide a digital and intelligent analysis system based on the physical equipment of a hydropower station, as Figure 1 shown, including:

[0065] The state feature extraction module collects the temperature, pressure and vibration data of the operation of the physical equipment of the hydropower station through sensors, filters out background noise, divides the data time window, captures the operation state features of the physical equipment of the hydropower station, and outputs the equipment state feature set;

[0066] The fault identification and early warning module analyzes the operation data of the physical equipment of the hydropower station based on the equipment state feature set, calculates the deviation between the current operation data of the physical equipment of the hydropower station and the normal operation reference data, automatically marks the out-of-bounds data points according to the calculation results, and performs time series analysis on the out-of-bounds data points to identify potential abnormal operation of the physical equipment of the hydropower station and generate preliminary fault early warning indicators;

[0067] The equipment behavior in-depth analysis module performs a retrospective analysis of the historical behavior of the physical equipment of the hydropower station according to the preliminary fault early warning indicators. By analyzing the behavior patterns of the physical equipment of the hydropower station under different operating conditions, it analyzes and determines the abnormal behavior before the fault of the physical equipment of the hydropower station, and performs periodic analysis on the abnormal behavior according to the analysis and determination results to reveal the change law of the behavior of the physical equipment of the hydropower station, capture the time frame of the occurrence of periodic faults, and obtain the equipment behavior analysis results;

[0068] The collaborative learning and prediction module receives the analysis results of device behavior, initiates the data synchronization process between the cloud and edge computing resources, analyzes and compares the long-term data and short-term activities of the physical device behavior of the hydropower station, identifies the performance degradation trend of the physical device of the hydropower station, continuously monitors the physical device of the hydropower station according to the identification results, updates the health status record of the physical device of the hydropower station, predicts the potential risks of the physical device of the hydropower station in the future, and establishes a health management file for the physical device of the hydropower station.

[0069] The device status feature set includes temperature features, pressure features, and vibration features; the preliminary fault warning indicators include out-of-bounds data points, time series analysis results, and potential anomaly marker records; the device behavior analysis results include behavior pattern change analysis results, periodicity analysis results, and periodic fault time frames; the health management file for the physical device of the hydropower station includes performance degradation trend identification results, continuous monitoring records, and potential risk prediction records.

[0070] Please refer to Figure 2 , and the steps for obtaining the device status feature set are as follows:

[0071] Deploy sensors around the physical device of the hydropower station to capture the temperature, pressure, and vibration data during the operation of the device in real time, and screen out the data points whose frequencies are not within the working frequency range to obtain the pure operation status data of the device;

[0072] Configure a variety of sensors around the physical device of the hydropower station, including temperature sensors, pressure sensors, and vibration sensors. These sensors are fixed at the key parts of the device to monitor the physical parameters of the device during operation in real time. These sensors continuously collect data and transmit it to the monitoring center through wiring. First, timestamp the collected raw data to ensure that each piece of data can accurately reflect the physical state at the time of collection. Subsequently, through a preset frequency filtering function, eliminate all data that clearly does not conform to the device operation frequency, such as occasional noise or errors. This process ensures the purity and usability of the data, making subsequent data analysis more accurate. In addition, the integrity of the data will also be checked to ensure that each data collection is complete and not interfered by any external factors.

[0073] Based on the pure operation status data of the device, perform time series segmentation according to the device operation cycle. Each time segment represents a working cycle, and calculate the mean value, standard deviation, and range of the data within each cycle to obtain the statistical description of the device cycle data;

[0074] After data cleaning, the purified data is segmented by time series, splitting the continuous data stream into multiple independent time periods, with each time period corresponding to a complete operating cycle of the device. In this process, first, a time window is defined, usually bounded by a complete cycle of the device's operation. Then, the data is cut according to this time window, segmented into multiple independent data blocks, and each block contains the operating data of a complete cycle. For each data block, basic statistical metrics are calculated, including the maximum, minimum, mean, and standard deviation within the data block. These statistical metrics can describe the operating condition of the device in each cycle. For example, the mean and standard deviation can reflect the average level and fluctuation of the device's operation, while the maximum and minimum values provide intuitive data on the extreme states of the device's operation.

[0075] Using the statistical description of the device cycle data and combining with the device operation rules, evaluate the relationship between the data fluctuation in each cycle and the device health condition, reveal the current state of the device, and obtain the device state feature set;

[0076] By analyzing the statistical data of each time window in detail, further explore the deep features of the device operation state, and evaluate how each statistical metric reflects the operation efficiency and possible functional abnormalities of the device alone and jointly. First, according to the design and function requirements of the device, identify key operating metrics, such as the stability of temperature, the normal fluctuation range of pressure, and the frequency characteristics of vibration. Then, for the data within each time window, extract the specific values and change trends of the key metrics. For example, analyze the frequency distribution in the vibration data to identify whether there are unusual high-frequency components, indicating wear or faults in certain components. Further combine the key metrics to construct a comprehensive device state feature set, covering the data extracted from each time window, reflecting the overall operation quality of the device during the observation period, providing an intuitive understanding of the device's performance in each detection cycle, enhancing the immediate understanding of the device state, and also strengthening the ability of preventive maintenance.

[0077] Please refer to Figure 3 For the calculation steps of the deviation:

[0078] Based on the device state feature set, extract the real-time operation data of the physical devices in the hydropower station, capture the current working state of the devices, and generate a real-time data set;

[0079] Obtaining real-time operating data based on the device status feature set first involves confirming data access permissions to ensure that the operator has sufficient permissions to obtain information from the server. Next, it is necessary to cooperate with the database administrator to ensure data security and integrity. The specific operations include querying the database, writing SQL query statements, and selecting and filtering specific data tables and fields, such as temperature, pressure, and vibration, to ensure that the obtained data set contains all the required data points. In addition, the query operation also needs to consider the timestamp of the data to ensure that the extracted data is in real-time or near-real-time chronological order. Convert any non-standard data formats in the queried data so that the output of the data preparation stage is a structured and standardized data set.

[0080] Based on the real-time data set, combined with the device normal operation benchmark data, use the formula:

[0081]

[0082] Calculate the sum of deviation values Deviation to generate the total deviation result. Among them, n represents the total number of data points, i represents the data point index, D r represents the real-time data set, D b represents the benchmark data set, D ri represents the value of the i-th data point in the real-time data set, D bi represents the value of the i-th data point in the benchmark data set;

[0083] The following temperature data (unit: °C) was collected within a certain period of time:

[0084] Real-time operating temperature D ri =[65,68,67,70,72]

[0085] Benchmark temperature D bi =[60,60,60,60,60]

[0086] The number of data points n = 5;

[0087] Calculate the square of the difference for each data point:

[0088]

[0089] Calculate the sum of the squares of the differences:

[0090] 25 + 64 + 49 + 100 + 144 = 382

[0091] Calculate the square root of the deviation:

[0092]

[0093] The calculated deviation value is approximately 8.74°C. This result represents the average deviation between the actual measured temperature and the normal operating temperature of the device. This relatively large deviation indicates that the device has an overheating problem and further inspection is needed to determine if there are any faults or external factors causing the device's operating temperature to rise.

[0094] Using the total deviation result, compare it with the preset abnormal threshold T. If Deviation > T, then mark it as abnormal and generate the device deviation analysis result.

[0095] After completing the deviation calculation, compare the deviation value with the preset abnormal threshold. For each device type, determine the normal range of the deviation based on historical maintenance data and the manufacturer's instruction manual. Once the threshold is defined, then write a logical judgment statement to compare whether the deviation value of each data point exceeds its corresponding threshold, such as whether the deviation value is greater than the threshold T. The threshold is usually preset during the data monitoring process or adjusted by the operator according to the actual situation. Data points that exceed the threshold will be marked as abnormal for subsequent analysis and investigation to ensure that each anomaly is properly recorded and responded to. The generated result will reflect the operating status of the device and indicate the need for further technical evaluation or direct maintenance intervention.

[0096] Please refer to Figure 4 , and the steps to obtain the preliminary fault warning indicators are as follows:

[0097] According to the device deviation analysis result, automatically mark all data points that exceed the preset abnormal threshold as out-of-bounds data points, record the deviation magnitude and timestamp of each out-of-bounds data point, and output the set of out-of-bounds data points;

[0098] During the process of marking out-of-bounds data points, first obtain each data point from the real-time data stream, compare each data point with the preset threshold. When the value of the data point exceeds the threshold, that point is marked as out-of-bounds, and the specific value of the data point and its corresponding timestamp are recorded. The recording method is to create a record entry in the database for each out-of-bounds point, including the identifier, timestamp, and value of the data point. The records are then integrated into a set of out-of-bounds data points, and this set will be used for further analysis.

[0099] Conduct time series analysis based on the set of out-of-bounds data points. By examining the time dependence of the data points, identify the periodic fluctuations and irregular variations in the data, reveal potential operating problems, and obtain the time series analysis result;

[0100] First, the out-of-bounds data points are arranged in chronological order. Then, the data is analyzed in detail to identify potential trends and patterns. This is mainly done by observing the distribution of data points on the time axis to identify periodicity and random fluctuations in the data. For example, if multiple consecutive data points show the same out-of-bounds behavior, it may be a sign of a systematic failure. During the analysis process, the time interval and degree of violation between data points are carefully considered to determine whether there are any identifiable patterns or trends. This is done by checking the data point set one by one. Each data point is evaluated to determine its impact on the overall trend, helping to understand the abnormal operating mode of the equipment and providing support for further fault diagnosis.

[0101] Using the time series analysis results, the current analysis results are matched with the fault signatures stored in the database to determine the fault type corresponding to the current out-of-bounds data point and obtain preliminary fault warning indicators;

[0102] In the process of comparing the results of time series analysis with historical fault data, the analysis results are first matched with a database containing historical fault patterns. The matching process involves checking whether the new abnormal pattern is similar to the known patterns in the database. If a match is found, the matched fault type and the corresponding warning level are recorded. This is done by comparing each newly identified pattern with each pattern in the database one by one. The comparison benchmarks include the duration of the abnormality, the affected equipment components, and the conditions under which the fault occurred. Each successfully matched event will be recorded in detail, including the confidence of the match and the potential impact assessment, which can quickly identify possible equipment problems and warn before the problem becomes a larger failure, thereby ensuring the stable operation and safety of the equipment.

[0103] See also Figure 5 , the analysis and determination steps of abnormal behavior before failure are:

[0104] Call the preliminary fault warning indicators, integrate the historical operation data of the hydropower station's physical equipment, count the operation time, load conditions and maintenance records, classify the data according to the time axis, remove abnormal data points, and generate a historical behavior data set of the hydropower station's physical equipment;

[0105] The preliminary fault warning indicators are called up, and the historical operating data of the physical equipment of the hydropower station are integrated. During the data integration process, the required data types are first defined, including operating time, load conditions, maintenance records, etc., and then directly exported from the monitoring data according to the actual operating conditions of the hydropower station. In order to ensure the accuracy and availability of the data, the data is cleaned and verified, and obvious errors or abnormal data are eliminated, such as data points with abnormally long operating time or sudden load changes. The data is manually verified and automatically detected by comparing daily maintenance logs and operation logs to ensure that the data used can truly reflect the historical operating status of the equipment. This process involves data preprocessing techniques, such as null value processing, outlier detection, etc. The processed data is stored in a central database for subsequent behavioral analysis.

[0106] Based on the historical behavior data set of the hydropower station physical equipment, the behavior change law of the equipment under different operating conditions is analyzed, and the formula is used:

[0107]

[0108] Calculate the device behavior pattern score S and generate the behavior pattern determination result, where B i is the operating load under different operating conditions, is the average value of the operating load, C i To maintain the frequency, is the average value of the maintenance frequency, and N is the total number of data points;

[0109] Setting B i is the operating load parameter of the equipment under different operating conditions. The specific data is [100, 105, 98, 103, 107]. The average value is =102.6; set C i is the maintenance frequency parameter under the same operating conditions. The specific data is [3, 3, 2, 3, 4]. The average value is =3.0; Set N to the total number of data points, which is 5.

[0110] First calculate

[0111]

[0112] Then calculate

[0113]

[0114] Then substitute the above results into the formula:

[0115]

[0116] The results show that in the given dataset, the average deviation score of the device behavior pattern is 3.1628. The score reflects the degree of change in the device's behavior under different operating conditions. A higher S value indicates that the device exhibits unusual behavior under certain operating conditions, which is related to potential device problems.

[0117] According to the results determined by the behavior pattern, identify the part where the pattern score abnormally increases, compare the fault warning indicators with the corresponding positions of the abnormal behavior of the score, and obtain the recognition result of the abnormal behavior before the fault;

[0118] Through the results determined by the behavior pattern, identify the abnormal behavior pattern associated with the fault. First, determine the threshold value at which the score abnormally increases in the behavior pattern analysis results. This threshold value is obtained through statistical analysis of the behavior scores of past fault cases. Then, compare the current analysis results with this threshold value to identify the behavior patterns that exceed the threshold value, which are often closely related to potential faults of the device. In this way, potential problems can be warned before the device shows significant faults, providing data support for the technical team to enable targeted maintenance or adjustment, thereby improving the reliability and safety of the device. The final result is the recognition result of the abnormal behavior before the device fault, which is crucial for improving the predictability of device operation and reducing unexpected downtime.

[0119] Please refer to Figure 6 , the steps to obtain the device behavior analysis results are as follows:

[0120] Using the recognition result of the abnormal behavior before the fault, conduct time series analysis on the abnormal behavior to determine the occurrence frequency and periodicity, analyze the time points when the abnormal behavior occurs, and generate a dataset of periodic behavior patterns;

[0121] Using the recognition result of the abnormal behavior before the fault, conduct time series analysis to determine the periodicity of the abnormal behavior. First, collect the detailed logs of the device operation, including the specific date and time of each operation. The log data is filtered to exclude any records during non-operation times, such as during maintenance or downtime. Next, the filtered data is sorted and organized in chronological order to ensure the continuity of the analysis. By examining the changes in operation frequency and duration, possible periodic patterns are identified. The identification process includes checking each time point of the log data, comparing the data points before and after, and finding common patterns in the time intervals. For example, if it is found that the time intervals remain consistent in consecutive multiple operation cycles, this will be marked as a potential periodic behavior. Finally, these marked periodic behaviors will be further analyzed to determine whether they are related to known device problems. The generated dataset of periodic behaviors provides a basis for subsequent in-depth analysis.

[0122] Based on the periodic behavior pattern dataset, analyze the frequency components of the device behavior, identify the key periodic components, count the frequency peaks in the dataset, reveal the key change rules of the device behavior, and obtain the analysis results of the change rules;

[0123] Through frequency analysis of the labeled periodic behavior data, further reveal the main change rules of the device behavior. The analysis first involves converting time data into frequency data, that is, examining the speed and pattern of the device behavior changing over time, calculating the average time interval between each periodic behavior, and monitoring whether these intervals remain consistent over time or show a certain trend. By plotting a scatter plot of the time intervals and observing whether the data points cluster around a certain average line, it helps to identify frequently occurring behavior patterns. For example, if most of the time intervals cluster around a few specific values, this indicates that the device behavior has strong periodicity. In addition, perform statistical analysis on the frequency data, such as calculating its standard deviation and variance, to evaluate the consistency and predictability of the periodicity, and help determine which periodic behaviors are associated with device performance degradation or potential failures, so as to provide specific operation suggestions for the maintenance team.

[0124] Based on the analysis results of the change rules, combined with the comparison of the device maintenance records and the fault history, determine the possible time frame for the occurrence of periodic faults, analyze the potential future fault time points of the device, and generate the analysis results of the device behavior;

[0125] After confirming the periodic behavior pattern of the device, compare the information with the device's maintenance records and historical fault data. First, collect the device's maintenance and fault records, and detail the time of each fault occurrence and the maintenance measures taken. Classify and code the records for convenient comparison and analysis. For each recorded fault, analyze the corresponding relationship between its occurrence time and the periodic behavior pattern, and check whether the fault appears after certain specific periodic behaviors. This requires a detailed comparison of the behavior patterns before and after each fault to find any possible similarities or patterns. Through detailed comparative analysis, it can be revealed which periodic behaviors may lead to device performance problems or faults. In addition, by calculating the time difference between the periodic behavior and the fault, evaluate the warning time of these behavior patterns, so as to provide a basis for the preventive maintenance of the device.

[0126] Please refer to Figure 7 , the steps for identifying the performance degradation trend are as follows:

[0127] Receive the analysis results of the device behavior, count the operation data, performance indicators and abnormal event records of the device, perform data integrity verification and timestamp checking, confirm the consistency and reliability of the data, and generate the device data verification results;

[0128] Receive the analysis results of device behavior. First, check the format and structure of the incoming data stream to ensure that the received information is consistent with the data reception standard. Then, by reading the data records line by line, check the timestamps and data integrity of each record, which is completed by comparing the data length of the record with the preset data structure. If it is found that the data is damaged or missing during transmission, mark this batch of data as incomplete and request retransmission. In addition, the received data will be initially sorted in chronological order to establish a basis for subsequent data processing and analysis.

[0129] By establishing a data channel, synchronize the device data verification results between the cloud and edge computing resources, confirm that all computing nodes have a consistent data view, and generate a data synchronization completion confirmation result;

[0130] Once the data is received and initially processed, start the data synchronization process with the cloud to ensure that all operation data is updated to the cloud server in real time. By establishing a scheduled task, new data is automatically pushed from the local server to the cloud at regular intervals. Before data is pushed, an integrity check will be performed to compare the differences between the latest records of local data and the data after the last synchronization to ensure that all new data is included. The data transmission uses an encrypted communication protocol to protect the security of data transmission over the Internet. Once the data is successfully uploaded to the cloud, automatically record a synchronization log locally, including the synchronization time and data volume, for future auditing and troubleshooting.

[0131] According to the data synchronization completion confirmation result, compare the long-term and short-term device behavior data, identify the degradation trend of device performance, analyze the device performance degradation mode, and obtain the device performance degradation analysis result;

[0132] After completing data synchronization, conduct a detailed comparative analysis of long-term and short-term data, aiming to identify and quantify the device performance degradation trend. First, extract the historical performance data of the device from the database, including past operating efficiency, maintenance records, and failure rates, etc., and compare it with the current latest data. By calculating the percentage change in device performance metrics, such as the reduction in operating efficiency or the growth trend of failure rates, the degradation speed of the device can be quantitatively evaluated. Through detailed calculations and comparisons, a detailed report on device performance degradation can be obtained to provide decision support for the maintenance team.

[0133] Please refer to Figure 8 , the steps for establishing the physical device health management file of the hydropower station are as follows:

[0134] According to the device performance degradation analysis result, monitor the device temperature, pressure, and load, collect real-time data to obtain the real-time operation data of key parameters, and generate a sorted real-time monitoring data set;

[0135] Before starting data monitoring, first set the key parameters for monitoring based on the historical performance analysis results, such as temperature, pressure, and load. These parameters are crucial for evaluating the health status of the device. Next, continuously collect the operating data. After the data is collected, perform preliminary screening and processing, including verifying the timestamps of the data, checking the integrity and consistency of the data, to confirm that all data conforms to the preset standard format. For incomplete or abnormal data points, remove them from the dataset. This process is implemented through an automated script to ensure that only valid and complete data is used for subsequent analysis, generating a sorted real-time monitoring dataset, providing an accurate data basis for the continuous monitoring of the device status and future analysis work.

[0136] Based on the sorted real-time monitoring dataset, combined with the device historical performance data, analyze the changing trend of the current health status of the device item by item, using the formula:

[0137]

[0138] Calculate the change amplitude H of the device health status, generating an updated record of the device health status. Among them, P i and P i-1 represent the key performance parameter values at adjacent time points respectively, t i和 t i-1 are the corresponding time points respectively, and M is the total number of data points;

[0139] The times corresponding to the time points t1, t2, and t3 are the 1st day, the 2nd day, and the 3rd day.

[0140] The values of the performance parameter P at t1, t2, and t3 are 100, 105, and 95 respectively.

[0141] Calculate the change rate of the performance parameter between every two consecutive time points and square it:

[0142]

[0143] Accumulate the squared values:

[0144] 25 + 100 = 125

[0145] Calculate the value of H:

[0146]

[0147] The calculated value of H is 11.18, reflecting the overall change amplitude of the performance parameter of the device during the monitoring period. A higher value of H indicates that the performance parameter of the device has changed significantly during the observation period, indicating that the device performance is degrading or the device status is unstable, and further inspection and possible maintenance are required.

[0148] Based on the updated device health status records, combined with historical failure records, organize the change trends of key parameters of device performance degradation, analyze the potential risk probability of the device in the future, and generate a physical device health management file for hydropower stations;

[0149] After updating the device health status records, further aggregate and analyze the data to identify any early signs of possible performance degradation or potential failures. In this process, data analysis does not rely on complex external tools but is completed through built-in data processing functions, including trend analysis and periodic checks of the data, which helps to reveal any atypical patterns in the device operation. Then the analysis results are used to update the device's health records, including not only the update of numerical data but also the supplementation of the device operation logs, such as changes in operating conditions, addition of maintenance history, and records of replacement of any key components. Finally, all this information is integrated into the device health management file, which is dynamically updated, can reflect the latest status of the device at any time, and provides data support for formulating future maintenance and monitoring strategies.

[0150] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered by the scope of the claims and the specification of the present invention. In particular, as long as there is no structural conflict, the technical features mentioned in each embodiment can be combined in any way. The present invention is not limited to the specific embodiments disclosed in the text but includes all technical solutions falling within the scope of the claims.

Claims

1. A digital intelligence analysis system based on the physical equipment of a hydropower station, characterized in that The system includes: The state feature extraction module collects the temperature, pressure, and vibration data of the operation of the physical equipment of the hydropower station through sensors, filters out the background noise, divides the data time window, captures the operation state features of the physical equipment of the hydropower station, and outputs the equipment state feature set; Based on the equipment state feature set, the fault identification and early warning module analyzes the operation data of the physical equipment of the hydropower station, calculates the deviation between the current operation data of the physical equipment of the hydropower station and the normal operation reference data, automatically marks the out-of-bounds data points according to the calculation results, and performs time series analysis on the out-of-bounds data points to identify potential abnormal operation of the physical equipment of the hydropower station and generate preliminary fault early warning indicators; According to the preliminary fault early warning indicators, the equipment behavior in-depth analysis module conducts a retrospective analysis of the historical behavior of the physical equipment of the hydropower station. By analyzing the behavior patterns of the physical equipment of the hydropower station under different operating conditions, it analyzes and determines the abnormal behavior before the fault of the physical equipment of the hydropower station. According to the analysis and determination results, it conducts a periodic analysis of the abnormal behavior, reveals the change law of the behavior of the physical equipment of the hydropower station, captures the time frame of the occurrence of periodic faults, and obtains the equipment behavior analysis result; The collaborative learning and prediction module receives the equipment behavior analysis result, starts the data synchronization process between the cloud and edge computing resources, analyzes and compares the long-term data and short-term activities of the behavior of the physical equipment of the hydropower station, identifies the performance degradation trend of the physical equipment of the hydropower station, continuously monitors the physical equipment of the hydropower station according to the identification result, updates the health status record of the physical equipment of the hydropower station, predicts the potential risks of the physical equipment of the hydropower station in the future, and establishes a health management file for the physical equipment of the hydropower station.

2. The digital intelligent analysis system based on the physical equipment of the hydropower station according to claim 1, wherein The steps for obtaining the equipment state feature set are as follows: Deploy sensors around the physical equipment of the hydropower station to capture the temperature, pressure, and vibration data during the equipment operation process in real time, and screen out the data points whose frequencies are not within the working frequency range to obtain the pure operation state data of the equipment; Based on the pure operation state data of the equipment, perform time series segmentation according to the equipment operation cycle. Each time segment represents a working cycle, and calculate the data mean, standard deviation, and range within each cycle to obtain the statistical description of the equipment cycle data; Using the statistical description of the equipment cycle data and combining with the equipment operation rules, evaluate the relationship between the data fluctuation within each cycle and the equipment health status, reveal the current state of the equipment, and obtain the equipment state feature set.

3. The digital intelligence analysis system based on the physical equipment of the hydropower station according to claim 2, wherein, The steps for calculating the deviation are as follows: Based on the equipment state feature set, extract the real-time operation data of the physical equipment of the hydropower station, capture the current working state of the equipment, and generate a real-time data set; Based on the real-time data set, combine with the normal operation reference data of the equipment, and use the formula: Calculate the sum of deviation values Deviation to generate the total deviation result, where n represents the total number of data points, i represents the data point index, D r represents the real-time data set, D b represents the reference data set, D ri represents the value of the i-th data point in the real-time data set, D bi represents the value of the i-th data point in the reference data set; Using the total deviation result, compare it with the preset abnormal threshold T. If Deviation > T, then mark it as abnormal and generate the equipment deviation analysis result.

4. The digital intelligent analysis system based on the physical equipment of the hydropower station according to claim 3, characterized in that, The steps for obtaining the preliminary fault early warning indicators are as follows: According to the equipment deviation analysis result, automatically mark all the data points that exceed the preset abnormal threshold as out-of-bounds data points, record the deviation magnitude and timestamp of each out-of-bounds data point, and output the out-of-bounds data point set; Performing time series analysis based on the cross-border data point set, identifying periodic fluctuations and irregular changes in the data by examining the time dependency of the data points, revealing potential operating problems, and obtaining time series analysis results; Using the time series analysis results, the current analysis results are matched with the fault signatures stored in the database to determine the fault type corresponding to the current out-of-bounds data point and obtain preliminary fault warning indicators.

5. The digital intelligence analysis system based on the physical equipment of a hydropower station according to claim 4, wherein, The analysis and determination steps of the abnormal behavior before the fault are as follows: Calling the preliminary fault warning indicator, integrating the historical operation data of the hydropower station physical equipment, counting the operation time, load conditions and maintenance records, classifying the data according to the time axis, removing abnormal data points, and generating a historical behavior data set of the hydropower station physical equipment; Based on the historical behavior data set of the hydropower station physical equipment, the behavior change law of the equipment under different operating conditions is analyzed, and the formula is used: Calculate the behavior pattern score S of the computing device and generate a behavior pattern determination result, where B i is the operation load under different operation conditions, is the average value of the operation load, C i is the maintenance frequency, is the average value of the maintenance frequency, and N is the total number of data points; According to the behavior pattern determination result, the abnormally increased part in the pattern score is identified, and the corresponding position of the fault warning indicator and the scored abnormal behavior is compared to obtain the abnormal behavior identification result before the fault.

6. The digital intelligence analysis system based on the physical equipment of the hydropower station according to claim 5, characterized in that The steps for obtaining the device behavior analysis result are as follows: Using the abnormal behavior recognition results before the fault, perform time series analysis on the abnormal behavior, determine the frequency and periodicity of occurrence, analyze the time point when the abnormal behavior occurs, and generate a periodic behavior pattern data set; Based on the periodic behavior pattern data set, the frequency components of the device behavior are analyzed, key periodic components are identified, frequency peaks in the statistical data set are counted, key change patterns of the device behavior are revealed, and change pattern analysis results are obtained; Based on the analysis results of the change rules, combined with the comparison of equipment maintenance records and fault history, the possible time frame for the occurrence of periodic faults is determined, the potential time points for the equipment to fail in the future are analyzed, and the equipment behavior analysis results are generated.

7. The digital intelligence analysis system based on the physical equipment of the hydropower station according to claim 6, wherein The steps for identifying the performance degradation trend are: Receive the equipment behavior analysis results, collect statistics on equipment operation data, performance indicators and abnormal event records, perform data integrity verification and timestamp verification, confirm data consistency and reliability, and generate equipment data verification results; By establishing a data channel, synchronizing the device data verification results between the cloud and edge computing resources, confirming that all computing nodes have a consistent data view, and generating a data synchronization completion confirmation result; The confirmation result is completed synchronously according to the data, the long-term and short-term equipment behavior data are compared, the degradation trend of equipment performance is identified, the equipment performance degradation mode is analyzed, and the equipment performance degradation analysis result is obtained.

8. The digital intelligent analysis system based on the physical equipment of a hydropower station according to claim 7, characterized in that, The steps for establishing the health management file of the hydropower station physical equipment are as follows: According to the equipment performance degradation analysis results, monitor the equipment temperature, pressure and load, perform real-time data collection to obtain real-time operating data of key parameters, and generate a collated real-time monitoring data set; Based on the real-time monitoring data set, combined with the historical performance data of the equipment, the changing trend of the current health status of the equipment is analyzed item by item, using the formula: Calculate the change amplitude H of the device health state and generate an updated device health state record, where P i and P i-1 represent the key performance parameter values at adjacent time points respectively, and t i和 t i-1 are the corresponding time points respectively, and M is the total number of data points; Based on the updated device health status record, combined with the historical failure records, sort out the change trend of the key parameters of the device performance degradation, analyze the potential risk probability of the device in the future, and generate a physical device health management file for the hydropower station.

Images