A safety monitoring method for hydropower station operations

Abstract

The present application discloses a method for safety monitoring of hydropower station operations, including: obtaining historical monitoring status data of hydro-generator unit equipment, preprocessing the obtained data, screening the data in the vibration alarm state, clustering and clustering the two-dimensional array samples in the screened data, constructing an independent vibration zone sub-cluster model, encapsulating the constructed vibration zone sub-cluster model into an independent software module or hardware plug-in, that is, a sub-cluster form plug-in, receiving real-time data from the monitoring system and processing it, and the sub-cluster form plug-in comparing and matching the real-time data with the preset sub-cluster model to determine whether the device is currently in the vibration zone; by making a meticulous division of the vibration characteristics and constructing a series of highly specific key sub-clusters based on this, the accuracy and accuracy of the vibration monitoring system are improved, ensuring that even under complex and variable working conditions, the subtle changes in the vibration of hydro-generator unit equipment can be captured.

Description

Technical Field

[0001] The present invention relates to the technical field of hydro-generating units, and particularly to a method for monitoring the operation safety of a hydropower station. Background Art

[0002] When the hydro-generating unit equipment is operating normally, excessive vibration and swing will cause certain damage to the hydro-generating unit equipment. Therefore, in order to effectively avoid such damage, it is necessary to ensure that the hydro-generating unit equipment can avoid the vibration area during operation. The vibration area refers to a certain range or multiple ranges of active power values that will cause excessive amplitude under a certain water head when the hydro-generating unit equipment is operating normally.

[0003] Comparative document 202011095435.0 discloses a vibration area determination method and device. By selecting target historical monitoring status data in a vibration warning state from a large amount of historical monitoring status data, and then clustering and dividing each two-dimensional array sample in the target historical monitoring status data to obtain the clusters after clustering convergence corresponding to each two-dimensional array sample, and then performing numerical analysis on the two-dimensional array corresponding to each cluster after clustering convergence to obtain the vibration area corresponding to each cluster.

[0004] The above document cannot accurately and quickly determine the vibration area. At the same time, it cannot predict the future vibration trend and cannot avoid the damage during the operation of the hydro-generating unit equipment in time. Summary of the Invention

[0005] The present application provides a method for monitoring the operation safety of a hydropower station. By dividing the vibration characteristics and constructing a sub-cluster form plug-in sub and a series of key sub-clusters with high specificity, the accuracy and precision of the vibration monitoring system are improved.

[0006] The present application provides a method for monitoring the operation safety of a hydropower station, including:

[0007] S110, obtaining the historical monitoring status data of the hydro-generating unit equipment, preprocessing the obtained data, extracting the contour points of the equipment. When the equipment is in a non-vibrating state, using a high-precision sensor to capture the contour points of each key part of the equipment as the static reference; during the operation of the equipment, capturing the contour points of the same key part in real time as the dynamic data;

[0008] S120, screening the data in the vibration warning state, and clustering and dividing the two-dimensional array samples in the screened data;

[0009] S130, constructing an independent vibration area sub-cluster model, and saving the generated clusters as the vibration area sub-cluster model. Each sub-cluster represents a specific vibration area, and its boundary is determined by the distribution of the cluster center and data points;

[0010] Among them, the model contains information about the clustering centers, reflecting the distribution of data points in space;

[0011] S140. Package the constructed vibration zone sub-cluster model into an independent software module or hardware plug-in, namely the sub-cluster form plug-in;

[0012] S150. Receive real-time data from the monitoring system, process it, and compare the real-time data with the positions of the contour points in the non-vibrating state in the sub-cluster form plug-in. If the contour points of the device in the real-time data coincide with the positions of the contour points in the non-vibrating state, it means that the device is not currently in the vibration zone. If the contour points of the device in the real-time data do not coincide with the positions of the contour points in the non-vibrating state, it means that the device is currently in the vibration zone, and trigger corresponding alarms or maintenance measures.

[0013] Preferably, the steps of the clustering and clustering include:

[0014] S121. Select a preset number of two-dimensional arrays as the initial clustering contour points;

[0015] S122. Calculate the Euclidean distance between the initial clustering contour points of each two-dimensional array sample, and according to the principle of the shortest distance, assign the sample to the first clustering cluster;

[0016] S123. Calculate the first clustering center in the first clustering cluster, and confirm the second clustering cluster according to the relationship between the two-dimensional array sample and the first clustering center;

[0017] S124. Take the second clustering cluster as the new clustering cluster, and send it to step S121, and repeat the above three steps for cyclic iteration;

[0018] S125. When the maximum distance between the new and old clustering centers is less than the set threshold, the iteration stops, and the final clustering cluster of the two-dimensional array sample is output.

[0019] Preferably, classify the sub-cluster form plug-ins:

[0020] S210. Classify the plug-ins in the sub-cluster form according to the vibration type, and divide them into axial vibration plug-ins, radial vibration plug-ins, and torsional vibration plug-ins;

[0021] Among them, the axial vibration plug-in specifically monitors the vibration of the axis of the water turbine runner;

[0022] The radial vibration plug-in monitors the radial vibration of key components such as generator bearings and water turbine guide bearings;

[0023] The torsional vibration plug-in monitors the torsional vibration of the unit during start-up, acceleration, stable operation, and shutdown;

[0024] S220, classify the plug-in connectors in the sub-cluster form according to the severity of vibration, and divide them into slightly vibrating plug-in connectors, moderately vibrating plug-in connectors, and severely vibrating plug-in connectors;

[0025] S230, according to different types of plug-in connectors, integrate the plug-in connectors in the sub-cluster form in the monitoring system;

[0026] S240, receive vibration data from each measuring point of the unit, and distribute it to the corresponding vibration monitoring plug-in connectors for processing. Each plug-in connector analyzes the data independently and matches it with the preset clustering model to determine whether the unit is in the vibration area, the vibration type, and the severity.

[0027] Preferably, the method for integrating the plug-in connectors in the sub-cluster form:

[0028] S231, design the overall architecture of the monitoring system, including a data acquisition layer, a data processing layer, a monitoring and evaluation layer, and an alarm response layer. Reserve a plug-in connector interface in the data processing layer so that different types of plug-in connectors can be integrated into the system;

[0029] S232, select appropriate vibration monitoring plug-in connectors for configuration according to actual needs. When comprehensive vibration monitoring of the unit is required, axial vibration plug-in connectors, radial vibration plug-in connectors, and torsional vibration plug-in connectors can be configured simultaneously.

[0030] Preferably, use the key sub-cluster to perform plug-in comparison with the sub-cluster plug-in connector. The comparison steps are as follows:

[0031] S310, select a group of vibration area sub-clusters with high representativeness and specificity. This vibration area sub-cluster is the key sub-cluster;

[0032] S320, there is a standardized vacant part on the plug-in connector in the sub-cluster form, and this vacant part is used to receive and compare with the key sub-cluster;

[0033] S330, perform plug-in comparison between the key sub-cluster and the plug-in connector in the sub-cluster form.

[0034] Preferably, add a contour point fluctuation block to the key sub-cluster. The specific steps are as follows:

[0035] S410, add a contour point fluctuation block to the key sub-cluster. This contour point fluctuation block is used to record and analyze the change of each sub-cluster contour point over time;

[0036] S420, during the process of constructing the key sub-cluster, calculate the position of each sub-cluster contour point and store it as a time series in the contour point fluctuation block;

[0037] Among them, the contour point fluctuation block will update the data regularly to reflect the latest change of vibration characteristics;

[0038] S430, Plug-in comparison. In real-time plug-in comparison, the sub-cluster plug-in not only compares with the static features of the key sub-cluster, but also analyzes the dynamic changes of the vibration characteristics by using the contour point fluctuation block in the time dimension;

[0039] S440, If the position of the contour point in the real-time data does not coincide with the position of the contour point in the non-vibrating state, the sub-cluster immediately triggers the warning mechanism to prompt the operator to pay attention to the equipment status.

[0040] Preferably, the vibration warning state is divided into a first warning state and a second warning state. The first warning state is the vibration area warning. When the vibration of the unit exceeds the preset safety threshold, the system will trigger the first warning state, and the state value is usually 1, indicating that the unit is in a potential vibration risk area. The second warning state is the non-vibration area warning. When the vibration of the unit is within the normal range, the system is in the second warning state, and the state value is usually 0, indicating that the unit is operating smoothly.

[0041] Preferably, the sub-cluster form plug-in contains the position of the contour point in the non-vibrating state, necessary algorithm logic and data structure for storing the information of the sub-cluster model and performing real-time analysis, and designs a standardized interface so that the plug-in can integrate different systems, and the interface should support common communication protocols.

[0042] Preferably, the formula for the Euclidean distance is

[0043] Preferably, the vibration area is divided according to different parts, and the specific steps are as follows:

[0044] S510, Before screening the vibration warning data, mark the parts of the vibration records in the data according to the equipment structure diagram and operation and maintenance records;

[0045] S520, According to the part marks of the vibration data, group the data into different subsets, and each subset corresponds to a specific part of the equipment;

[0046] S530, Perform clustering on each subset to form a vibration area sub-cluster unique to the corresponding part, and bring it into step S140.

[0047] One or more technical solutions provided in this application have at least the following technical effects or advantages: By making a meticulous division of the vibration characteristics and constructing a series of highly specific key sub-clusters based on this, the accuracy and precision of the vibration monitoring system are improved, ensuring that even under complex and variable working conditions, the subtle changes in the vibration of the hydro-generator unit equipment can be captured.

[0048] To achieve the wide adaptability and flexibility of the system, a sub-cluster plug-in is introduced. According to different types of hydro-generator unit equipment and their specific operating conditions, corresponding monitoring modules can be combined or replaced, simplifying the installation and commissioning processes. It also greatly expands the application scope of the system, ensuring comprehensive coverage and accurate monitoring of various complex equipment.

[0049] In addition, a contour point fluctuation block is introduced, which can accurately capture the subtle change trends of vibration characteristics. Through continuous monitoring and in-depth analysis of these trends, the development trajectory of the equipment vibration state can be predicted in advance, providing valuable forward-looking warning information for operation and maintenance personnel. This enables staff to take necessary preventive measures before equipment failures occur, effectively avoiding the adverse impacts of sudden failures on production operations and ensuring the safe and stable operation of hydro-generator unit equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic flow chart of a method for monitoring the safety of hydropower station operations according to the present invention;

[0051] Figure 2 It is a schematic flow chart of the clustering and clustering method of an embodiment of the present invention;

[0052] Figure 3 It is a schematic flow chart of the plug-in classification of sub-cluster forms in an embodiment of the present invention;

[0053] Figure 4 It is a schematic flow chart of the comparison between the key sub-cluster and the plug-in of the sub-cluster form in an embodiment of the present invention;

[0054] Figure 5 It is a schematic flow chart of the division of the vibration area by part in an embodiment of the present invention;

[0055] Figure 6 It is a schematic diagram of the vibration of the hydro-generator unit parts in an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0056] To facilitate the understanding of the present invention, the present application will be described more comprehensively with reference to the relevant drawings; the preferred embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein; on the contrary, these embodiments are provided to make the disclosure of the present invention more thorough and comprehensive.

[0057] It should be noted that the terms "vertical", "horizontal", "upper", "lower", "left", "right" and similar expressions used herein are for illustrative purposes only and do not represent the only embodiments.

[0058] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this invention belongs; the terms used in the specification of this invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention; the term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0059] Embodiment 1: Figure 1 It is a schematic flow chart of a method for monitoring the safety of hydropower station operations according to an embodiment of the present invention, including the following steps:

[0060] S110. Obtain the historical monitoring status data of the hydro-generator unit equipment, preprocess the obtained data, extract the contour points of the equipment, and use a high-precision sensor to capture the contour points of each key part of the equipment under the non-vibrating state of the equipment as the static reference; during the operation of the equipment, capture the contour points of the same key part in real time as the dynamic data.

[0061] This data includes the basic operating parameters of the equipment, active power value, head value, temperature value, and vibration alarm status. Clean, denoise, and standardize the historical monitoring data;

[0062] The vibration alarm status is divided into a first alarm status and a second alarm status. The first alarm status is the vibration area alarm. When the vibration of the unit exceeds the preset safety threshold, the system will trigger the first alarm status, and the status value is usually 1, indicating that the unit is in a potential vibration risk area. The second alarm status is the non-vibration area alarm. When the vibration of the unit is within the normal range, the system is in the second alarm status, and the status value is usually 0, indicating that the unit is operating smoothly.

[0063] S120. Screen the data in the vibration alarm status, and perform clustering on the two-dimensional array samples in the screened data, as Figure 2 shown.

[0064] S121. Select a preset number of two-dimensional arrays as the initial clustering contour points.

[0065] Specifically, randomly select 3 samples from 100 two-dimensional array samples in the vibration alarm status. The x and y coordinate values of these 3 samples will be used as the contour points of 3 initial clustering clusters respectively.

[0066] S122. Calculate the Euclidean distance between the initial clustering contour points of each two-dimensional array sample, and assign the sample to the first clustering cluster according to the principle of the shortest distance.

[0067] For the remaining 97 samples (the initial contour point samples have been excluded), calculate their Euclidean distances from the 3 initial clustering contour points one by one. The Euclidean distance is a common method for measuring the straight-line distance between two points, and the calculation formula is Each sample will be assigned to the cluster where the nearest clustering contour point is located according to the principle of the shortest distance.

[0068] S123, calculate the first clustering center in the first cluster, and confirm the second cluster according to the relationship between the two-dimensional array samples and the first clustering center.

[0069] S124, take the second cluster as the new cluster, transfer it to step S121, and repeat the above three steps for cyclic iteration.

[0070] After the first assignment, calculate the average value (or contour point) of the x and y coordinates of all samples in the first cluster. This average value will become the first clustering center of this cluster. Subsequently, calculate the second cluster according to the distance between the two-dimensional array samples and the first clustering center. According to the average value (or contour point) of the x and y coordinates of all samples in the second cluster, this new average value will become the second clustering center of this cluster, and prepare for the next iteration.

[0071] S125, when the maximum distance between the new and old clustering centers is less than the set threshold, the iteration stops, and the final clusters of the two-dimensional array samples are output.

[0072] Specifically, take the newly calculated clustering center as the new contour point, and repeat the steps of S122 and S123. In each iteration, the samples will be re-assigned to each cluster according to the position of the current clustering center, and the new clustering center will be calculated again. This process will continue until the convergence condition is met. The convergence condition is usually set as the maximum distance between the new and old clustering centers is less than a preset threshold, or the preset maximum number of iterations is reached.

[0073] When the iteration process converges, the algorithm stops, and the final clustering labels of each sample are output, that is, which cluster each sample is assigned to. At this time, we obtain 3 clusters, and each cluster contains a set of two-dimensional array samples with similar vibration characteristics.

[0074] S130, construct an independent vibration zone sub-cluster model, and save the generated clusters as the vibration zone sub-cluster model. Each sub-cluster represents a specific vibration region, and its boundary is determined by the clustering center and the distribution of data points.

[0075] Based on the above clustering results, an independent sub-cluster model for the vibration area is constructed for each cluster. These models not only contain the information of the clustering center but also reflect the distribution of data points in space, thus defining the boundaries of different vibration regions. Each sub-cluster represents a specific vibration mode or fault type.

[0076] S140, encapsulate the constructed sub-cluster model of the vibration area into an independent software module or hardware plug-in, that is, the sub-cluster form plug-in.

[0077] Among them, the sub-cluster form plug-in contains the positions of the contour points in the non-vibrating state, necessary algorithmic logics and data structures for storing the information of the sub-cluster model and performing real-time analysis, and designs standardized interfaces so that the plug-in can be integrated into different systems. The interfaces should support common communication protocols (such as Modbus, OPC UA, etc.) for data exchange with devices and software from different manufacturers.

[0078] S150, receive the real-time data from the monitoring system, process it, and compare the real-time data with the positions of the contour points in the non-vibrating state in the sub-cluster form plug-in. If the contour points of the device in the real-time data coincide with the positions of the contour points in the non-vibrating state, it means that the device is not currently in the vibration area. If the contour points of the device in the real-time data do not coincide with the positions of the contour points in the non-vibrating state, it means that the device is currently in the vibration area.

[0079] The technical solutions in the embodiments of the present application at least have the following technical effects or advantages:

[0080] The sub-cluster form plug-in can pre-load and optimize the storage of the sub-cluster model of the constructed vibration area, enabling the plug-in to immediately start the built-in algorithm for fast matching and analysis after receiving the real-time operation data, thus greatly shortening the response time from data reception to the output of the vibration area judgment result, ensuring the rapid response ability of the monitoring system, and providing a valuable time window for timely taking intervention measures.

[0081] The sub-cluster form plug-in can accurately capture and reflect the unique vibration modes and characteristics of this type of device, making the judgment result closer to the actual situation, effectively avoiding misjudgment or missed judgment problems caused by insufficient model generalization. While improving the accuracy of vibration area judgment, the sub-cluster form plug-in also enhances the adaptability and compatibility of the system to different device types.

[0082] Example 2: To improve the vibration monitoring ability of the hydropower station and ensure the safety and stability of the unit operation, the sub-cluster form plug-ins are further classified to comprehensively and highly accurately monitor the three key vibration types of the axial, radial, and torsional vibrations of the unit equipment, so as to be able to timely and comprehensively detect potential faults, optimize the maintenance strategy, and extend the service life of the unit. The following is a detailed description of the construction process of this monitoring network, as Figure 3 shown.

[0083] S210, classify the plug-ins in the form of sub-clusters according to the vibration type, into axial vibration plug-ins, radial vibration plug-ins, and torsional vibration plug-ins.

[0084] Among them, the axial vibration plug-in specifically monitors the vibration of the axis of the water turbine runner to prevent axial vibration problems caused by shafting imbalance, bearing failure, etc.; the radial vibration plug-in monitors the radial vibration of key components such as the generator bearing and the water turbine guide bearing, and timely discovers abnormal vibrations caused by poor lubrication, component loosening or wear, etc.; the torsional vibration plug-in monitors the torsional vibration of the unit during startup, acceleration, stable operation, and shutdown processes to prevent serious problems caused by rotor imbalance, coupling failure, etc.

[0085] Specifically, when the axial vibration plug-in receives real-time vibration data, it automatically matches with the model. If it is determined to be medium vibration, it sends an alarm of the "attention" level to the monitoring system and recommends further inspection; the radial vibration plug-in monitors that there is abnormal high-frequency vibration at the generator bearing, and after matching with the model, it is determined to be "mild vibration caused by wear", and then issues a maintenance reminder, recommending to change the lubricating oil and check the bearing condition; the torsional vibration plug-in monitors that the unit suddenly has a low-frequency torsional vibration with a high amplitude, and after comparing with the model, it is determined to be "severe vibration caused by coupling failure", immediately triggers an emergency shutdown procedure, and issues a detailed fault report for the maintenance personnel to refer to.

[0086] S220, classify the plug-ins in the form of sub-clusters according to the vibration severity, into mild vibration plug-ins, medium vibration plug-ins, and severe vibration plug-ins.

[0087] Specifically, the preset vibration severity threshold of the mild vibration plug-in is relatively low, and a prompt is triggered when the equipment vibration level approaches but does not exceed this threshold, so that the operator can take measures in advance to prevent the vibration from intensifying; the vibration threshold of the medium vibration plug-in is set between mild and severe. When the vibration level reaches this threshold, in addition to the prompt, more detailed monitoring and recording operations may also be triggered; the severe vibration plug-in sets a higher vibration threshold, and once triggered, it immediately issues an emergency alarm, prompting the operator to immediately stop the machine for inspection to avoid equipment damage or safety accidents.

[0088] Furthermore, according to the specific conditions and monitoring objectives of the hydropower unit, the above-mentioned different types of plug-ins are flexibly combined and configured. Through the collaborative work of the plug-ins, the system can comprehensively cover and accurately analyze various vibration conditions of the unit, including identifying the vibration type, evaluating the severity of vibration, and predicting potential fault risks.

[0089] S230. Integrate the plug-ins in the sub-cluster form in the system according to different types of plug-ins.

[0090] S231. Design the overall architecture of the system, including the data acquisition layer, data processing layer, monitoring and evaluation layer, and alarm response layer. Reserve plug-in interfaces in the data processing layer so that different types of plug-ins can be integrated into the system.

[0091] S232. Select appropriate vibration monitoring plug-ins for configuration according to actual needs. When comprehensive vibration monitoring of the unit is required, axial vibration plug-ins, radial vibration plug-ins, and torsional vibration plug-ins can be configured simultaneously. Set the parameters and thresholds of the plug-ins in the monitoring system, such as the classification criteria of vibration levels and alarm trigger conditions.

[0092] S240. Receive vibration data from each measuring point of the unit and distribute it to the corresponding vibration monitoring plug-ins for processing. Each plug-in independently analyzes the data and matches it with the preset clustering and sub-clustering model to determine whether the unit is in the vibration area, the vibration type, and the severity.

[0093] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0094] Different types of plug-ins are designed for specific vibration types and severities, which can more accurately identify and evaluate the vibration conditions of hydropower unit equipment, making the plug-ins in the sub-cluster form more targeted. At the same time, the plug-ins cooperate with each other to form a multi-level vibration monitoring network, providing a full range of coverage of the vibration state of the equipment.

[0095] Embodiment 3: To improve the adaptability of vibration monitoring of hydropower unit equipment, a real-time plug-in comparison scheme using key sub-clusters and sub-cluster plug-ins is proposed. This scheme realizes rapid and accurate monitoring and evaluation of the vibration conditions of hydropower unit equipment by finely dividing the vibration characteristics of the equipment, constructing key sub-clusters with strong specificity, and designing flexible and pluggable sub-cluster plug-ins.

[0096] S310. Select a group of vibration area sub-clusters with high representativeness and specificity, and this vibration area sub-cluster is the key sub-cluster.

[0097] Specifically, the key sub-cluster is a set of vibration zone sub-clusters with high specificity and accuracy formed by a clustering algorithm based on historical monitoring data of specific types, models, and operating conditions of hydro-generator unit equipment. With the continuous accumulation of historical monitoring data, the key sub-cluster model is continuously updated based on historical data, and the key sub-cluster contains the real-time profile points of the equipment.

[0098] S320. On the plug-in of the sub-cluster form, there is a standardized vacant part for receiving and comparing the key sub-cluster.

[0099] Among them, the vacant part adopts a standardized data interface protocol to ensure seamless access and comparison analysis of different key sub-clusters. The vacant part is built-in with an intelligent algorithm that can automatically adjust the internal algorithm logic and comparison strategy according to the received key sub-cluster type and characteristics to achieve precise vibration monitoring.

[0100] S330. The key sub-cluster is inserted and compared with the plug-in of the sub-cluster form.

[0101] Specifically, when the hydro-generator unit equipment is running, the key sub-cluster will collect the operating parameters of the equipment in real time and conduct comparison analysis through the sub-cluster plug-in. The key sub-cluster is inserted into the corresponding vacant part of the sub-cluster plug-in, and the intelligent algorithm is used to compare the real-time data with the sub-cluster form plug-in to determine whether the real-time profile points of the equipment in the key sub-cluster coincide with the profile points in the non-vibration state of the sub-cluster plug-in. If they coincide, the equipment is not in the vibration zone; if not, the equipment is in the vibration zone. The sub-cluster plug-in will determine the vibration state of the equipment and trigger the corresponding prompt or alarm mechanism. If the equipment is in the vibration zone, the system will immediately issue an alarm to prompt the operator to take measures for adjustment or shutdown inspection. At the same time, the system can record the vibration event information to provide a basis for subsequent analysis and optimization.

[0102] The technical solutions in the embodiments of the present application described above have at least the following technical effects or advantages:

[0103] By finely dividing the vibration characteristics and constructing key sub-clusters with strong specificity, the accuracy and precision of vibration monitoring are improved, and the pluggable design of the sub-cluster plug-in enables the system to flexibly adapt to hydro-generator unit equipment of different types and different operating conditions.

[0104] Embodiment 4: In the above embodiments, the key sub-cluster master cannot effectively monitor the real-time evolution of the equipment vibration state, nor can it predict the future vibration trend.

[0105] Therefore, the embodiments of the present application are optimized on the basis of the above embodiments, as Figure 4 shown.

[0106] S410. Add a contour point fluctuation block to the key sub-cluster. This contour point fluctuation block is used to record and analyze the change of each sub-cluster contour point over time.

[0107] S420. During the process of constructing the key sub-cluster, calculate the position of each sub-cluster contour point and store it as a time series in the contour point fluctuation block.

[0108] Among them, the contour point fluctuation block will update the data regularly to reflect the latest change of vibration characteristics.

[0109] S430. Plug-in comparison. In the real-time plug-in comparison, the sub-cluster plug-in not only compares with the static characteristics of the key sub-cluster, but also uses the contour point fluctuation block in the time dimension to analyze the dynamic change of vibration characteristics.

[0110] S440. If the position of the contour point in the real-time data does not coincide with the position of the contour point in the non-vibrating state, the sub-cluster immediately triggers an early warning mechanism to prompt the operator to pay attention to the equipment status.

[0111] In some embodiments, the vibration area is divided according to different parts, such as Figure 5 and Figure 6 as shown. The specific steps are as follows:

[0112] S510. Before screening the vibration warning data, according to the equipment structure diagram and operation and maintenance records, mark the vibration records in the data. The marking should be specific to each key component of the equipment, such as bearings, blades, and runners.

[0113] S520. According to the part marking of the vibration data, group the data into different subsets, and each subset corresponds to a specific part of the equipment.

[0114] S530. Perform clustering on each subset to form vibration area sub-clusters unique to the corresponding part, and bring them into step S140.

[0115] Among them, according to the number of times each part falls into the frequently vibrating sub-cluster or area within a period of time, calculate the corresponding frequency or percentage. When the vibration data of a certain part frequently exceeds the set threshold, the system will mark it as a frequently vibrating part and issue corresponding warnings or prompts.

[0116] The technical solutions in the above embodiments of the present application at least have the following technical effects or advantages:

[0117] The contour point fluctuation block can analyze the change trend of vibration characteristics, predict the development of the equipment vibration state, and provide forward-looking warning information for the operation and maintenance personnel.

[0118] Through part marking and screening, vibration data can be accurately mapped to specific components of the device, such as bearings, blades, runners, etc. This precise positioning makes the monitoring more specific and effective, enabling timely detection and focusing on vibration problems of specific components.

[0119] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for monitoring the safety of hydropower station operations, characterized in that: include: S110, obtaining historical monitoring status data of the hydropower unit equipment, preprocessing the obtained data, extracting contour points of the equipment, and using high-precision sensors to capture contour points of key parts of the equipment when the equipment is not vibrating, as a static reference; During the operation of the equipment, the contour points of the same key parts are captured in real time as dynamic data; S120, filtering data in a vibration alarm state, and clustering the two-dimensional array samples in the filtered data; S130, constructing an independent vibration zone sub-cluster model, and saving the generated cluster as a vibration zone sub-cluster model, each sub-cluster represents a specific vibration area, and its boundary is determined by the distribution of the cluster center and the data points; Among them, the model contains the information of cluster centers, which reflects the distribution of data points in space; S140, encapsulating the constructed vibration zone sub-cluster model into an independent software module or hardware plug-in, i.e., a sub-cluster form plug-in; S150, receiving real-time data from the monitoring system and processing it, comparing the real-time data with the positions of the contour points in the sub-cluster morphology plug-in in the non-vibration state, if the positions of the contour points of the device in the real-time data coincide with the positions of the contour points in the non-vibration state, it means that the device is not currently in the vibration zone, if the positions of the contour points of the device in the real-time data do not coincide with the positions of the contour points in the non-vibration state, it means that the device is currently in the vibration zone, triggering corresponding alarms or maintenance measures; Classify the sub-cluster morphology plug-ins: S210, classifying the plugs in the sub-cluster form according to the vibration type into axial vibration plugs, radial vibration plugs, and torsional vibration plugs; Among them, the axial vibration plug-in is specifically used to monitor the vibration of the turbine runner axis; The radial vibration plug-in monitors the radial vibration of key components such as generator bearings and turbine guide bearings; The torsional vibration plug-in monitors the torsional vibration of the unit during startup, acceleration, stable operation and shutdown; S220, classifying the plugs in the sub-cluster form according to the severity of vibration, into light vibration plugs, moderate vibration plugs, and heavy vibration plugs; S230, integrating the plug-ins in the form of sub-clusters in the monitoring system according to different types of plug-ins; S240, receiving vibration data from each measuring point of the unit and distributing it to the corresponding vibration monitoring plug-in for processing, each plug-in independently analyzes the data and matches it with the preset clustering model to determine whether the unit is in the vibration zone and the vibration type and severity; The method for integrating the plug-in sub-cluster form: S231, design the overall architecture of the monitoring system, including the data acquisition layer, data processing layer, monitoring and evaluation layer, and alarm response layer, and reserve a plug-in interface in the data processing layer so that different types of plug-ins can be integrated into the system; S232, selecting a suitable vibration monitoring plug-in for configuration according to actual needs. When comprehensive vibration monitoring of the unit is required, an axial vibration plug-in, a radial vibration plug-in, and a torsional vibration plug-in can be configured at the same time; Use the key sub-cluster and the sub-cluster plug-in to perform plug-in comparison. The comparison steps are as follows: S310, selecting a group of vibration region subclusters with high representativeness and specificity, the vibration region subclusters being key subclusters; S320, a standardized vacant portion is provided on the sub-cluster-shaped plug-in, and the vacant portion is used to receive and compare the key sub-cluster; S330, the key sub-cluster is plugged and compared with the sub-cluster-shaped plug-in; Add a contour point fluctuation block to the key sub-cluster. The specific steps are as follows: 410, adding a contour point fluctuation block to the key sub-cluster, the contour point fluctuation block is used to record and analyze the change of contour points of each sub-cluster over time; S420, in the process of constructing the key sub-clusters, the contour point position of each sub-cluster is calculated and stored as a time series in the contour point fluctuation block; Among them, the contour point fluctuation block will update data regularly to reflect the latest changes in vibration characteristics; S430, plug-in comparison, in the real-time plug-in comparison, the sub-cluster plug-in will not only be compared with the static characteristics of the key sub-cluster, but also the dynamic changes of the vibration characteristics will be analyzed using the contour point fluctuation block in the time dimension; S440, the position of the contour point in the real-time data does not coincide with the position of the contour point in the non-vibration state, and the sub-cluster immediately triggers the early warning mechanism to prompt the operator to pay attention to the equipment status.

2. A method for monitoring the safety of hydropower station operations as claimed in claim 1, characterized in that: The clustering step comprises: S121, selecting a preset number of two-dimensional arrays as initial clustering contour points; S122, calculating the Euclidean distance between the initial clustering contour points of each two-dimensional array sample, and assigning the sample to the first cluster according to the shortest distance principle; S123, calculating the first cluster center in the first cluster, and confirming the second cluster according to the relationship between the two-dimensional array samples and the first cluster center; S124, taking the second cluster as a new cluster and sending it to step S121, repeating the above three steps for loop iteration; S125, when the maximum distance between the new and old cluster centers is less than the set threshold, the iteration stops and the final clustering cluster of the two-dimensional array samples is output.

3. A method for monitoring the safety of hydropower station operations as claimed in claim 1, characterized in that: The vibration alarm state is divided into a first alarm state and a second alarm state. The first alarm state is a vibration zone alarm. When the vibration of the unit exceeds a preset safety threshold, the system will trigger the first alarm state. The state value is usually 1, indicating that the unit is in a potential vibration risk area. The second alarm state is a non-vibration zone alarm. When the vibration of the unit is within a normal range, the system is in the second alarm state. The state value is usually 0, indicating that the unit is running smoothly.

4. A method for monitoring the safety of hydropower station operations as claimed in claim 1, characterized in that: The sub-cluster morphology plug-in contains the location of the contour points in the non-vibration state, the necessary algorithm logic and data structure, which are used to store the information of the sub-cluster model and perform real-time analysis. A standardized interface is designed so that the plug-in can integrate different systems. The interface should support common communication protocols.

5. A method for monitoring the safety of hydropower station operations as claimed in claim 2, characterized in that: The formula for the Euclidean distance is 6. A method for monitoring the safety of hydropower station operations as claimed in claim 1, characterized in that: The vibration zone is divided into different parts. The specific steps are as follows: S510, before filtering the vibration warning data, mark the vibration records in the data according to the equipment structure diagram and operation and maintenance records; S520, grouping the data into different subsets according to the location mark of the vibration data, each subset corresponding to a specific location of the device; S530, clustering each subset to form a vibration region subcluster specific to the corresponding part, and bringing it into step S140.

Images