Distributed water leakage detection method and system for hydropower plant powerhouse

Abstract

The application discloses a distributed water leakage detection method and system for a hydropower plant workshop, and belongs to the field of industrial safety, wherein the method comprises the following steps: positioning detection points based on the structure of a workshop pipeline, and setting detection devices at the detection points; setting multi-level detection thresholds according to the position influence of the detection point positioning; performing three-dimensional modeling based on the structure of the workshop pipeline and the position information of the detection point positioning, and synchronously mapping the detection data of each detection device to the three-dimensional model for visual display; judging whether the detection data reaches the multi-level detection thresholds, and determining the threshold level when the detection data reaches the multi-level detection thresholds; matching an early warning mode according to the threshold level, and performing early warning processing according to the early warning mode. The application solves the technical problems of insufficient water leakage detection range and precision for a hydropower plant workshop in the prior art, and achieves the technical effects of realizing comprehensive and accurate water leakage detection and hierarchical early warning for the workshop.

Description

Technical Field

[0001] This invention relates to the field of industrial safety, specifically to a distributed leakage detection method and system for hydropower plant buildings. Background Technology

[0002] Hydropower plants, as a crucial component of the power system, play a vital role in ensuring their safe and stable operation. The powerhouse, as the main building of a hydropower plant, is highly susceptible to damage from leaks. Leaks can not only harm the generating units and equipment but also threaten the structural safety of the powerhouse, potentially leading to serious consequences. Current methods for detecting leaks in hydropower plant buildings primarily rely on point-based or short-segment-arranged detection devices and manual observation. Point-based or short-segment-arranged detection devices have limited coverage, lack comprehensive monitoring, and are prone to missing leaks. Manual observation and judgment are inefficient, prolonging the time required for leak detection and response, and failing to achieve automatic detection and early warning, thus hindering the timely handling of leak incidents. Summary of the Invention

[0003] This application provides a distributed leakage detection method and system for hydropower plant buildings, aiming to solve the technical problems of insufficient detection range and accuracy of leakage in existing hydropower plant buildings.

[0004] In view of the above problems, this application provides a distributed leakage detection method and system for hydropower plant buildings.

[0005] The first aspect disclosed in this application provides a distributed leakage detection method for hydropower plant buildings. This method includes: locating detection points based on the plant's pipeline structure; setting detection devices at these points, wherein the detection points are distributed across multiple locations within the plant; setting multi-level detection thresholds based on the location impact of the detection points; performing three-dimensional modeling based on the plant's pipeline structure and the location information of the detection points; synchronously mapping the detection data from each detection device to the three-dimensional model for visualization; determining whether the detection data reaches the multi-level detection thresholds; and when it does, determining the threshold level; matching an early warning mode according to the threshold level; and performing early warning processing according to the early warning mode, wherein the early warning modes include unit shutdown control, audible and visual alarms, and power line control.

[0006] Another aspect of this application discloses a distributed leakage detection system for a hydropower plant building. This system includes: a detection device setting module for locating detection points based on the plant's pipeline structure and setting detection devices at these points, wherein the detection points are distributed across multiple locations within the plant; a detection threshold setting module for setting multi-level detection thresholds based on the location influence of the detection points; a plant 3D modeling module for performing 3D modeling based on the plant's pipeline structure and the location information of the detection points, and synchronously mapping the detection data from each detection device to the 3D model for visualization; a threshold level determination module for determining whether the detection data reaches the multi-level detection threshold, and when it does, determining the threshold level; and an early warning mode matching module for matching early warning modes according to the threshold levels and performing early warning processing according to the early warning modes, wherein the early warning modes include unit shutdown control, audible and visual alarms, and power line control.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] By employing a multi-point detection system based on the plant's pipeline structure, and deploying detection devices at these points across multiple locations within the plant, comprehensive monitoring of the plant's interior is achieved. Multiple detection thresholds are set based on the location and impact of each detection point, enabling early detection and location of leaks. The detection data is visualized in a 3D model, facilitating rapid leak identification and location by staff. The system determines whether the detection data reaches the preset multi-level detection thresholds, and if so, establishes the threshold level. Based on the threshold level, corresponding early warning modes are matched, triggering unit shutdown, audible and visual alarms, or power line control, thus achieving automated leak monitoring and early warning. This solution addresses the shortcomings of existing hydropower plant leak detection technologies in terms of range and accuracy, achieving comprehensive and accurate leak detection and tiered early warning for the plant.

[0009] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0010] Figure 1 This application provides a possible flowchart of a distributed leakage detection method for a hydropower plant building;

[0011] Figure 2 This application provides a schematic diagram of a possible process for obtaining image recognition information in a distributed leakage detection method for a hydropower plant building.

[0012] Figure 3This application provides a schematic diagram of a possible early warning process in a distributed leakage detection method for a hydropower plant building.

[0013] Figure 4 This application provides a schematic flowchart of an emergency response method for a first water level in a distributed leakage detection method for a hydropower plant building.

[0014] Figure 5 This application provides a schematic flowchart of an emergency response method for a second water level in a distributed leakage detection method for a hydropower plant building.

[0015] Figure 6 This application provides a possible structural diagram of a distributed leakage detection system for a hydropower plant building, which is an embodiment of the present application.

[0016] Explanation of reference numerals in the attached drawings: Detection device setting module 11, detection threshold setting module 12, factory 3D modeling module 13, threshold level determination module 14, early warning mode matching module 15. Detailed Implementation

[0017] The overall concept of the technical solution provided in this application is as follows:

[0018] This application provides a distributed leakage detection method and system for hydropower plant buildings. It adopts a technical solution that combines distributed layout of detection points, fusion analysis of detection data and multi-level early warning, so as to realize comprehensive monitoring and automatic early warning of leakage inside the plant building.

[0019] First, multiple detection points are established based on the factory's piping structure, and detection devices are deployed at each point. These points, distributed across multiple locations within the factory, enable comprehensive monitoring of the interior. Then, multi-level detection thresholds are set according to the location and impact of each detection point, achieving varying levels of detection sensitivity and enabling more rigorous monitoring of critical areas. Next, the data acquired from each detection point is fused onto a 3D digital model to generate a map showing the distribution of leak locations within the factory, providing a clear visual display of leak locations and facilitating rapid assessment by personnel. Based on this, the system determines whether the detection data reaches preset multi-level thresholds. If so, the system automatically matches the corresponding early warning mode, such as a unit shutdown command, audible and visual alarms, or power line control, achieving automatic detection and early warning. By adopting a tiered response, over-reaction or under-reaction is avoided.

[0020] In summary, by analyzing the pipeline structure of the hydropower plant, distributed detection points were deployed; by utilizing the fusion and modeling of detection data, accurate acquisition and location display of leakage information were achieved; and a multi-level early warning mode triggered based on the leakage data judgment results was implemented to complete the automatic monitoring, location, and early warning of leakage. This systematically and comprehensively solved the technical problems of insufficient leakage detection range and accuracy, and achieved the technical effect of realizing comprehensive and accurate leakage detection and graded early warning of the plant.

[0021] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0022] Example 1

[0023] like Figure 1 As shown in the embodiment of this application, a distributed leakage detection method for a hydropower plant building is provided. This method is applied to a leakage detection system, which includes a set of detection devices. The method includes:

[0024] The distributed leakage detection method for hydropower plant buildings is applicable to systems that detect leaks in pipelines and the environment within hydropower plant buildings. This system comprises a set of detection devices installed at multiple detection points within the plant, enabling distributed monitoring of the plant. Each detection device monitors parameters of the plant's pipelines and environment. The device itself includes a processor connected to a sensor interface, a communication module, and an audible / visual alarm interface. The sensor interface connects to leakage detection sensors via a sensor communication cable to detect leaks. The communication module connects to the power plant's monitoring unit LCU, public LCU, and leakage control cabinet, enabling timely warning processing based on early warning modes. The audible / visual alarm interface connects to an audible / visual alarm via a cable for timely sound and light alarm activation. The distributed leakage detection method includes:

[0025] The detection points are located based on the factory building's pipeline structure, and detection devices are installed at the detection points, which are distributed in multiple locations within the factory building.

[0026] The plant piping structure refers to the structural distribution of the piping system, such as water pipes and power lines, within the plant; the detection point location refers to determining the specific locations for setting up detection devices based on the plant piping structure. These locations are distributed across multiple areas of the plant to achieve full coverage monitoring of the plant; the detection device refers to the device set up at the detection point location to monitor relevant parameters of the piping, such as a radar level gauge or an infrared camera.

[0027] First, detailed information about the plant's piping is obtained, including the location of nodes, pipe diameter, flow rate, and connection relationships, to construct a structural model of the plant's piping. Then, based on the piping structure model, the importance of each node is assessed according to parameters such as the number of nodes, flow rate, and pipe diameter, and key nodes are selected as monitoring points. Simultaneously, considering the monitoring coverage of the plant, monitoring points are selected in different locations within the plant. Finally, matching monitoring devices are installed at the determined monitoring points to complete the distributed monitoring layout of the plant's piping.

[0028] By analyzing the plant's pipeline structure, and based on the locations of pipeline nodes, branches, and critical sections, multiple detection points are identified. Detection devices are then installed at these points to achieve precise location of the detection points and a reasonable layout of the detection devices. This enables distributed monitoring of the plant's pipelines, providing a foundation for subsequent leak monitoring and early warning.

[0029] Based on the locational impact of the detection points, multiple levels of detection thresholds are set.

[0030] The location impact of a detection point refers to the degree to which the location of the detection point affects the importance and sensitivity of the pipeline. The greater the location impact, the more sensitive the monitored data changes, and the more stringent the detection threshold needs to be set to achieve accurate alarms. Multi-level detection thresholds refer to setting multiple detection threshold levels according to different location impacts, such as level 1 threshold, level 2 threshold, level 3 threshold, etc., corresponding to the location impact from small to large.

[0031] First, the locational impact of the detection point is assessed based on parameters such as the importance of the pipeline, flow rate, pipe diameter, and the impact range of the detection point on surrounding pipelines and equipment. Then, based on the locational impact, multiple levels of detection thresholds are set. For example, a three-level detection threshold system can be set: Level 1, Level 2, and Level 3. Level 1 thresholds are suitable for detection points with low locational impact, used to monitor the normal operation of the pipeline. When the detection data exceeds Level 1 thresholds, an alert or record is triggered. Level 2 thresholds are suitable for detection points with moderate locational impact, used to monitor pipeline changes. When the detection data exceeds Level 2 thresholds, a shutdown inspection or localized handling is triggered. Level 3 thresholds are suitable for detection points with high locational impact, used for precise monitoring of pipeline changes. When the detection data exceeds Level 3 thresholds, emergency response is triggered. During system operation, the data from each detection point is monitored in real time to ensure it does not exceed the set thresholds. If it does, the corresponding early warning mechanism is immediately triggered to ensure the safety of the hydropower plant operation.

[0032] By setting matching detection thresholds based on the locational impact of detection points, the greater the locational impact, the more stringent the detection threshold setting. This enables refined monitoring and tiered early warning of changes in plant pipelines, providing a foundation for accurate leak detection and early warning, and ensuring the safe and stable operation of the hydropower plant.

[0033] Based on the plant pipeline structure and the location information of the detection points, a three-dimensional model is created, and the detection data of each detection device is synchronously mapped to the three-dimensional model for visualization.

[0034] Specifically, 3D modeling refers to establishing a 3D digital model of the plant's pipeline structure and the location information of the testing points based on the plant's pipeline structure and the location information of the testing points; testing data refers to the relevant parameter data obtained by each testing device, such as flow rate, liquid level, temperature, and image data; visualization display refers to mapping the testing data onto the 3D model and realizing an intuitive display of the data through virtual display.

[0035] First, detailed data on the plant's water pipes, electrical wires, and other pipelines are collected, including their spatial coordinates, diameters, and connections. This data is then input into 3D modeling software to construct a 3D model of the plant's pipelines. Second, the specific locations of the monitoring points are marked on the 3D model. Third, monitoring data acquired by various monitoring devices, such as flow rate, temperature, and liquid level, are collected and matched with the corresponding pipelines and monitoring points in the 3D model. Finally, the monitoring data is displayed on the 3D model, for example, using different colors to represent flow rate, achieving a visual representation of the monitoring data.

[0036] By establishing a 3D model and using virtual data mapping to display the data, the detection data is presented intuitively, enabling a clear and comprehensive understanding and monitoring of the hydropower plant system's operating status. This provides a visual and intuitive reference for anomalies, facilitating rapid location and handling, and improving the operational intelligence level of the system's monitoring and early warning capabilities.

[0037] Determine whether the detected data reaches a multi-level detection threshold; if so, determine the threshold level.

[0038] The alarm water levels in the hydropower plant are determined based on experience, categorized into three levels: primary, secondary, and tertiary. Multiple detection thresholds, such as primary, secondary, and tertiary thresholds, are set for each level. During system operation, the detection data acquired by each device is monitored in real time and compared with the set primary, secondary, and tertiary thresholds. If the detection data reaches the tertiary threshold, the threshold level is confirmed as tertiary; if it reaches the secondary threshold but not the tertiary threshold, the threshold level is confirmed as secondary; if it only reaches the primary threshold, the threshold level is confirmed as primary.

[0039] When determining thresholds, a reasonable time window is set for data comparison based on the data collection frequency and latency to avoid false alarms caused by excessively rapid data collection. Simultaneously, the impact of weather conditions and seasonal changes on the detection data is considered, and the detection threshold is adjusted appropriately to reduce misjudgments caused by environmental factors.

[0040] By comparing the detected data with the set thresholds, the specific threshold level reached is identified, providing a basis for further early warning processing. Different threshold levels correspond to different early warning levels, enabling refined early warning and control of abnormal situations.

[0041] The warning mode is matched according to the threshold level, and the warning is processed according to the warning mode, wherein the warning mode includes unit shutdown control, audible and visual alarm, and power line control.

[0042] The early warning mode refers to an early warning processing scheme that matches the detection threshold level, used to match the threshold level when the detected data reaches the threshold; early warning processing refers to activating the emergency measures corresponding to the early warning mode to deal with abnormal situations. Unit shutdown control refers to initiating the shutdown procedure for the corresponding unit; audible and visual alarm refers to activating the audible and visual alarm to provide a warning; power line control refers to activating backup power or switching control measures for the relevant lines.

[0043] When the detection data reaches the first-level threshold, the audible and visual alarm warning mode is activated, alerting the user to an abnormality. When the detection data reaches the second-level threshold, the power line control warning mode is activated, monitoring or assisting in the control of relevant power lines, such as issuing audible and visual alarms, activating maintenance leakage drainage pumps, and simultaneously monitoring the power plant's monitoring unit LCU in real time. When the detection data reaches the third-level threshold, the unit shutdown control warning mode is activated, shutting down or inspecting the unit, such as issuing commands to the public LCU to disconnect the AC circuit breaker for plant power and the DC battery outlet circuit breaker. The warning modes are set according to the plant system structure and pipeline layout, with corresponding handling schemes for pipe sections and equipment. Multiple warning mode schemes are also set according to different seasons and operating environments to adapt to changes under different conditions.

[0044] Upon activation of the early warning mode, relevant operators immediately implement emergency procedures according to the corresponding requirements, investigate the cause of the anomaly, and prevent serious consequences. Simultaneously, the leak detection system continuously monitors data changes at other detection points to determine if the anomaly is worsening. If the data continues to deteriorate, the system immediately switches to a higher-level early warning mode, implementing more stringent emergency measures to ensure system safety. This achieves the technical effect of tiered early warning and ensures the safe and stable operation of the hydropower plant.

[0045] Preferably, a flooding alarm signal is added to the factory building, positioned in the corner most prone to water accumulation, such as in a low-lying area of ​​the factory building, to monitor and alarm for flooding caused by severe leaks. A flooding alarm level is set; when the actual water level exceeds the preset alarm level, the flooding alarm signal outputs an alarm signal, emitting a visual flash and an alarm sound to alert factory management personnel. Simultaneously, early warning processing is performed according to the warning mode to prevent more serious accidents. This achieves rapid response in the event of internal flooding of the factory building, preventing the expansion of accident losses. The flooding alarm signal works in conjunction with a distributed leak detection system to further improve the safety of the hydropower plant. Furthermore, the embodiments of this application also include:

[0046] Information on the location, connection, and parameters of the water and electricity pipelines in the factory building is collected to obtain the factory building pipeline structure, which includes the factory building water pipe structure and the factory building power grid structure.

[0047] Based on the plant's water pipe structure, the number of pipe connections, flow rate, and pipe diameter parameters are analyzed. Evaluation and calculation rules for the number of pipe connections, flow rate, and pipe diameter parameters are set, and the evaluation results for each water pipe node are determined.

[0048] The evaluation results of each water pipe node are screened according to preset screening conditions to obtain the detection points of the water pipe.

[0049] Based on the aforementioned power grid structure, current, connected equipment information, and load are analyzed, and evaluation conversion rules for each current, connected equipment information, and load are set to determine the evaluation results for each power grid node.

[0050] Based on the evaluation results of each power grid node, the power grid pipeline detection points are determined by screening according to preset screening conditions.

[0051] The detection points of the water pipe and the power grid pipe are analyzed for intersection, and the intersection points are merged. The merged detection points are then used to locate the detection points.

[0052] In one feasible embodiment, to locate the detection point, firstly, the plant's pipeline structure is obtained. This structure includes the plant's water pipe structure and electrical grid structure. The water pipe structure is constructed by consulting the plant's water pipe design drawings, recording the pipe laying locations, connection details, and parameters such as pipe diameter. The electrical grid structure is constructed by consulting the plant's electrical grid design drawings, recording the locations of electrical grid equipment (such as transformers and circuit breakers), power line routes, and current loads. Then, based on the obtained water pipe structure information, evaluation calculation rules are set. For example, the number of connections in each water pipe segment is counted; segments with more connections receive higher evaluation results. The water flow rate is calculated based on the water flow sensor measurements or the water pipe diameter and head; segments with higher flow rates receive higher evaluation results. The pipe diameter of each water pipe segment is also counted; segments with larger diameters receive higher evaluation results. Subsequently, considering three factors—the number of pipe connections, flow rate, and pipe diameter—each water pipe node is comprehensively evaluated. For example, the evaluation result for each water pipe node is calculated based on the ratio of the number of pipe connections to the total number of connections, the ratio of flow rate to the total flow rate, and the ratio of pipe diameter to the maximum pipe diameter. Nodes with higher evaluation results are more likely to leak. Next, the evaluation results of the water pipe nodes are filtered according to preset screening criteria to obtain the detection points for the water pipe system. For example, upper and lower limits are set for the evaluation results, and nodes with evaluation results falling within these limits are selected as detection points.

[0053] Similarly, based on the obtained information related to the plant's power grid structure, such as current, connected equipment information, and load, evaluation conversion rules are set, and the evaluation results for each power grid node are calculated. Then, the evaluation results are filtered according to preset screening conditions to determine the power grid pipeline inspection points. After the water pipe and power grid pipeline inspection points are selected, an intersection analysis is performed on the inspection points of the two systems, merging the intersection points into the final inspection points to facilitate the location of the inspection points. First, the spatial locations of the water pipe and power grid pipeline inspection points are marked in the plant's pipeline structure; considering parallel or intersecting pipe sections of the water pipe and power grid, inspection points near these sections are more likely to become intersection points of the two systems, and these pipe sections and inspection points within a certain range on both sides are marked; then, the distance between the marked inspection points is measured in the plant's pipeline structure, and inspection point pairs with a distance less than a preset threshold (e.g., 1 meter) are selected as preliminary intersection point candidates; then, for the preliminary candidate intersection points, the structure of the water pipe and power grid pipeline within the space of this pipeline intersection is examined. If the structure of the water pipes and power grids corresponding to a certain intersection point is relatively complex, such as multiple parallel pipes or multiple connections, then this intersection point is highly likely and is marked as a high-probability intersection point pair. For high-probability intersection point pairs, on-site inspection or historical testing data is used to determine whether the two systems truly intersect. If so, it is determined as the final intersection point; otherwise, it is removed from the candidates. For the finally confirmed intersection points, they are marked on the plant's pipe structure. These intersection points are simultaneously water pipe testing points and power grid testing points, representing the most important and sensitive testing locations for both water pipes and power grids. The water pipe testing points and power grid testing points at the intersection point are merged to obtain a unique cross-testing point for that intersection point. Testing points other than the intersection points are still retained. The obtained testing points include the final cross-testing point and independent testing points in both structures. Testing point location is then performed based on the finally obtained testing points.

[0054] By accurately determining the location of leak detection points based on the factory's pipeline structure, a foundation is laid for subsequent leak detection and early warning, which can effectively improve the accuracy and efficiency of leak detection.

[0055] Furthermore, embodiments of this application also include:

[0056] Obtain the acquisition range of the detection device;

[0057] Based on the collection range, a preset distance is set for the detection points, and the preset distance for the detection points is used as the preset filtering condition.

[0058] In a preferred embodiment, to further improve the targeting of the detection points, the detection points are selected based on the acquisition range of the detection device.

[0059] The acquisition range refers to the coverage and monitoring range of the area being monitored by the detection device. Different types of detection devices, such as radar level gauges and infrared detectors, have different acquisition ranges due to differences in their working principles and performance parameters. Therefore, it is essential to first obtain the detailed technical parameters of each type of detection device to be used. Based on this, the maximum acquisition distance and range of each device can be determined, thus establishing the acquisition range and providing a foundation for setting the matching relationship between detection points and acquisition ranges. Subsequently, a preset distance for the detection points is set based on the acquisition range of the detection device as a preset screening condition to determine whether a detection point is within the effective acquisition range of a particular detection device. For example, for a detection device with an acquisition range of 20 meters, 15 meters or 18 meters would be used as the preset distance for this device. Then, when selecting detection points, the actual distance from each candidate detection point to the detection device is calculated and compared with the preset distance. Detection points with actual distances less than the preset distances are considered to be within the acquisition range and should be prioritized. Detection points with actual distances greater than the preset distances are considered to be outside the acquisition range and have a lower priority.

[0060] By setting a preset distance for the detection points based on the acquisition range of the detection device during the selection process, and using this as a preset screening condition, each candidate detection point is judged, and the detection points with shorter distances and located within the acquisition range of the detection device are selected. This improves the relevance of the final selected detection points and allows the detection devices set at the detection points to play their maximum role.

[0061] Furthermore, embodiments of this application also include:

[0062] The rate of water level rise is monitored in real time using radar level gauges;

[0063] Using the cross-sectional area of ​​the factory building and the rate of water level rise, the real-time leakage flow rate is calculated according to the formula: Flow rate = Volume / Time.

[0064] The flow difference is calculated based on the real-time leakage flow rate and the flow rate of the preset drainage measures to determine the drainage control information. Based on the drainage control information, the matching drainage measures are executed or adjusted.

[0065] In a preferred embodiment, one of the devices installed at the detection point for monitoring water leakage is a radar level gauge. The radar level gauge at the detection point monitors changes in the water level height and obtains the rate of water level rise; a higher rate of rise indicates a larger leakage flow. Based on the plant's cross-sectional area and the rate of water level rise, the real-time leakage flow rate is calculated using the formula: Flow Rate = Volume / Time. Here, volume is the water level rise height multiplied by the plant's cross-sectional area, and time is the time taken for the water level to rise.

[0066] Subsequently, the maximum drainage flow rate of the factory drainage system is calculated based on parameters such as pump capacity and pipe diameter, and this flow rate is used as the preset drainage measure flow rate. The calculated real-time leakage flow rate is then compared with the preset drainage measure flow rate. If the real-time leakage flow rate is greater than the preset drainage measure flow rate, the flow difference is positive, indicating that the current drainage capacity is insufficient and the drainage volume needs to be increased. If the real-time leakage flow rate is less than or equal to the preset drainage measure flow rate, the flow difference is zero or negative, indicating that the current drainage capacity is sufficient and no adjustment is required.

[0067] When the flow difference is positive, the corresponding drainage control information is determined based on the magnitude of the flow difference. For example, a flow difference of 0-50m³... 3 Between / h, the drainage control information is "Increase drainage pump 1 speed"; the flow rate difference is 50-100m³ / h. 3 Between / h, the drainage control information is "Start standby drainage pump 2"; the flow difference is greater than 100m³ / h. 3 The drainage control information is "Increase the speed of drainage pump 1 and drainage pump 2, and start the standby drainage pump 3". Subsequently, the determined drainage control information is sent to the leakage drainage control cabinet, which then executes corresponding control measures on the plant's drainage system based on the drainage control information. For example, it sends a control signal to increase the speed of a drainage pump or start the standby drainage pump to increase the drainage volume and ensure that the real-time drainage flow meets the requirements.

[0068] By setting up radar level gauges to monitor water level changes at detection points in real time, and calculating the corresponding leakage flow rate based on the water level rise rate, and comparing it with the drainage system flow rate, it can determine whether the drainage system needs to be adjusted to avoid flooding, thereby achieving automated control of the drainage system and ensuring the normal operation of the plant.

[0069] Furthermore, embodiments of this application also include:

[0070] The monitoring area is monitored by an infrared camera to obtain monitoring video information;

[0071] The monitoring video information is segmented using a semantic segmentation model to determine whether there is water leakage in the monitoring area, and the leakage point is located based on the water leakage monitoring results.

[0072] The semantic segmentation model includes a first channel and a second channel. The first channel has a first step length, and the second channel has a second step length, with the first step length being smaller than the second step length. The first channel is used to identify the leakage status, and the second channel is used to confirm the leakage point.

[0073] In a preferred embodiment, another detection device installed at the detection point is an infrared camera. The infrared camera continuously monitors the monitoring area surrounding the detection point to acquire monitoring video information.

[0074] Subsequently, the monitoring video information acquired by the infrared camera is input into the first channel of the semantic segmentation model. The first channel performs semantic segmentation on the video image using a small stride to determine whether there is a leak in each segmented region. If a leak is detected, the first channel determines that there is a leak in the monitored area, which is taken as the leak detection result. If no leak is detected in any segmented region, it is determined that there is no leak in the monitored area. When the first channel determines that there is a leak in the monitored area, the monitoring video information is input into the second channel of the semantic segmentation model. The second channel performs semantic segmentation on the video image using a larger stride to determine which segmented regions correspond to leak points or leaking areas. Based on the segmentation results of the second channel, the specific area of ​​the leak point in the monitoring video image is located. Then, the location information of the leak point located by the second channel is matched with the monitoring video image to specifically determine the actual location of the leak point in the monitored area, realizing the leak point location and providing a location reference for on-site leak handling.

[0075] The monitoring video information obtained by the infrared camera is input into the semantic segmentation model for processing, judgment and localization. The dual-channel design in the semantic segmentation model can accurately and efficiently determine whether there is water leakage in the monitoring area, and locate the location of the water leakage point when necessary. This realizes the automatic monitoring and location of water leakage at the detection point, provides information for water leakage early warning and improves the efficiency of water leakage repair.

[0076] Furthermore, such as Figure 2 As shown, embodiments of this application also include:

[0077] The first monitoring image is acquired using the first channel;

[0078] The first monitoring image is semantically segmented to determine the target monitoring object, and a grayscale region of interest is set based on the grayscale region of the target monitoring object;

[0079] The target recognition information is obtained by using the grayscale region of interest to perform target recognition on the first monitoring image.

[0080] In a preferred embodiment, the monitoring video information is processed using a first channel to obtain an image of the monitored area at a certain point in time, which is then used as the first monitoring image. Next, semantic segmentation is performed on the acquired first monitoring image to identify target monitoring objects, such as water pipes or valves, within the image. Subsequently, based on information such as the color intensity of the target monitoring object in the image, a corresponding grayscale range is defined as the region of interest (GI). Then, using the determined GI, target recognition is performed on the corresponding area in the first monitoring image to identify whether any abnormalities, such as cracks or leaks, exist within the area, and the recognition result is used as image recognition information.

[0081] The first channel of the semantic segmentation model processes the monitoring video information of the detection point to obtain a static image at a certain time point. Then, semantic segmentation and grayscale analysis are performed on the image to determine the target area of ​​interest. Target recognition is performed on the target area of ​​interest to determine whether there is any anomaly. It can accurately determine whether the target object in the first monitoring image has a fault such as water leakage.

[0082] Furthermore, such as Figure 3 As shown, embodiments of this application also include:

[0083] Determine whether the difference between the detected data and the threshold level meets the preset adjustment range;

[0084] When the conditions are met, the matching warning mode is determined by matching the threshold level with the preset warning mode.

[0085] The warning loss value is calculated based on the matching warning mode to obtain the warning loss value;

[0086] The threshold level is adjusted by lowering the adjacent values, and the adjusted threshold level is matched with the preset early warning mode to determine the adjusted early warning loss value;

[0087] The warning loss value is compared with the adjusted warning loss value, and the matching warning mode with the smaller warning loss value is selected for warning processing.

[0088] In one feasible embodiment, when the detection data at a certain detection point reaches a threshold level, continuous detection of the detection point continues to determine whether the detection data has changed significantly and whether the difference between the data and the corresponding threshold level meets a preset adjustment range. The preset adjustment range refers to the allowable range of detection data corresponding to the threshold level. When the detection data exceeds this adjustment range, it indicates that the threshold level corresponding to the detection data needs to be adjusted. If the difference is within the preset adjustment range, it indicates that the threshold level corresponding to the detection data does not need to be adjusted, and the warning mode does not need to be adjusted.

[0089] If the numerical difference is determined to be within the preset adjustment range, meaning the difference between the detected data and the threshold level meets the preset adjustment range, it indicates that the detected data is stable and not a false alarm. At this point, based on the currently determined threshold level, a preset early warning mode is matched. This preset early warning mode is pre-set according to the magnitude of the detected data change and corresponds to the corresponding threshold level. Subsequently, based on the determined matched early warning mode, the corresponding early warning loss value is calculated, reflecting the factory's economic or safety losses under this early warning mode. The threshold level corresponds to the early warning mode and the early warning loss value. The loss data corresponding to each early warning mode is pre-set according to the factory's actual situation and stored in the database. When calculating the early warning loss value, the loss data corresponding to the matched early warning mode is retrieved.

[0090] To select the optimal early warning mode, the current threshold level is lowered by one level and then raised by one level. Preset early warning modes are then matched against the lowered and raised threshold levels respectively to determine the two adjusted early warning modes. The corresponding early warning loss values ​​for these two adjusted early warning modes are then calculated and used as the adjusted early warning loss values. Finally, the calculated early warning loss values ​​are compared with the two adjusted early warning loss values, and the early warning mode with the lower early warning loss value is selected as the final matched early warning mode. The early warning operation is then executed, improving the economic efficiency of early warning processing.

[0091] Furthermore, embodiments of this application also include:

[0092] The detection device includes a flow monitoring device;

[0093] Set a traffic threshold;

[0094] The characteristic flow rate of the ball valve layer is monitored using flow monitoring equipment;

[0095] Flow is calculated based on the characteristic values ​​of the flow obtained from monitoring, and the calculated monitoring flow results are used to make early warning judgments based on the flow threshold.

[0096] In one feasible implementation, the detection device includes a flow monitoring device, which refers to various sensor devices installed in the pipeline for measuring fluid flow, such as electromagnetic flow meters, ultrasonic flow meters, turbine flow meters, differential pressure flow meters, etc.

[0097] First, a flow threshold is set based on pipeline parameters and normal flow range, and compared with the flow rate obtained from real-time monitoring. When the real-time flow rate exceeds the flow threshold, a leak is identified. Then, the characteristic flow rate of the butterfly valve layer or ball valve layer is monitored in real-time using a flow meter, which corresponds to the flow rate value under the combined open and closed states of the butterfly or ball valves. The flow monitoring device is installed at the inlet of the ball valve layer or butterfly valve layer. Next, based on the real-time open and closed state of the butterfly or ball valves and combined with pipeline parameters, the monitored flow rate under that state is calculated. Subsequently, the monitored flow rate result is compared with the set flow threshold. If the monitored flow rate result is significantly higher than the flow threshold, a leak is identified. This achieves intelligent monitoring and early warning analysis of the flow rate in the ball valve layer, quickly locating the leak location when an anomaly occurs, improving the response speed of detection and handling, and ensuring the safe and reliable operation of the hydropower plant. In summary, the distributed leak detection method for hydropower plant buildings provided in this application has the following technical effects:

[0098] Based on the plant's piping structure, detection points are located and detection devices are installed at these points. These detection points are distributed across multiple locations within the plant to achieve comprehensive monitoring coverage and prevent missed leaks. Multiple detection thresholds are set according to the location impact of each detection point, allowing for different levels of sensitivity and enabling early detection and location of leaks. A 3D model is created based on the plant's piping structure and the location information of the detection points. The detection data from each device is synchronously mapped onto the 3D model for visualization, clearly showing the leak location and facilitating quick identification and location by staff. The system determines whether the detection data reaches the multiple detection thresholds. When a threshold is reached, the threshold level is determined, and the system automatically assesses the presence and severity of leaks based on the detection data. An early warning mode is matched to the threshold level, and warning actions are taken according to the warning mode. These warning modes include unit shutdown control, audible and visual alarms, and power line control. Different levels of response are adopted based on the severity of the leak, achieving automatic detection and early warning, and enabling comprehensive and accurate leak detection and tiered early warning for the entire plant.

[0099] Furthermore, to prevent false alarms from leak monitoring that could affect the normal operation of the hydropower plant, a dual-signal triggering mechanism is adopted in the emergency response plan for each alarm water level in the hydropower plant building. The monitored water level is first divided into a first water level and a second water level. For example... Figure 4As shown, the first water level includes two leakage monitoring signal actions, continuously monitoring for leakage in the hydropower plant. If the first water level leakage monitoring signal 1 is triggered, it is determined whether the first water level leakage monitoring signal 2 is also triggered. If not, no emergency response is initiated for the first water level. If the first water level leakage monitoring signal 2 is also triggered, emergency response is initiated for the first water level, including a plant-wide audible and visual alarm, activation of the turbine protection circuit "emergency shutdown" for each unit's LCU, startup of the maintenance drainage pump, startup of the leakage drainage pump, and simultaneous control of the leakage monitoring camera to patrol and check whether each unit's LCU is shut down. If it is shut down, it is confirmed that the leakage point has sent a signal to the unit's LCU, at which point the emergency response for the first water level ends.

[0100] Among them, such as Figure 5 As shown, the second water level also includes two leakage monitoring signal actions to prevent false alarms in leakage monitoring. Continuous leakage monitoring is performed on the hydropower plant building. When the leakage in the building exceeds the first water level, if the second water level leakage monitoring signal 1 is triggered, the second water level emergency response is not initiated. Instead, it is first determined whether the second water level leakage monitoring signal 2 is also triggered. If not, the second water level emergency response is not initiated. If the second water level leakage monitoring signal 2 is also triggered, the second water level emergency response is initiated. At this time, the public LCU blocks the automatic transfer switch of the plant auxiliary power supply. After a 2-second delay, the public LCU trips the circuit breakers on the incoming side of each bus of the plant auxiliary power supply. After a 2-second delay, the public LCU trips the 220V DC battery outlet circuit breaker, thus realizing the second water level leakage emergency response.

[0101] Example 2

[0102] Based on the same inventive concept as the distributed leakage detection method for hydropower plant buildings in the aforementioned embodiments, such as Figure 6 As shown in the embodiment of this application, a distributed leakage detection system for a hydropower plant building is provided. The system includes:

[0103] The detection device setting module 11 is used to locate detection points based on the factory pipeline structure and set detection devices at the detection points, wherein the detection points are distributed in multiple locations in the factory.

[0104] The detection threshold setting module 12 is used to set multiple detection thresholds based on the positional influence of the detection point location.

[0105] The 3D modeling module 13 for the factory is used to perform 3D modeling based on the factory pipeline structure and the location information of the detection points, and to synchronously map the detection data of each detection device to the 3D model for visualization display.

[0106] The threshold level determination module 14 is used to determine whether the detection data reaches a multi-level detection threshold, and when it does, to determine the threshold level.

[0107] The early warning mode matching module 15 is used to match an early warning mode according to the threshold level and perform early warning processing according to the early warning mode, wherein the early warning mode includes unit shutdown control, audible and visual alarm, and power line control.

[0108] Furthermore, the detection device setting module 11 includes the following execution steps:

[0109] Information on the location, connection, and parameters of the water and electricity pipelines in the factory building is collected to obtain the factory building pipeline structure, which includes the factory building water pipe structure and the factory building power grid structure.

[0110] Based on the plant's water pipe structure, the number of pipe connections, flow rate, and pipe diameter parameters are analyzed. Evaluation and calculation rules for the number of pipe connections, flow rate, and pipe diameter parameters are set, and the evaluation results for each water pipe node are determined.

[0111] The evaluation results of each water pipe node are screened according to preset screening conditions to obtain the detection points of the water pipe.

[0112] Based on the aforementioned power grid structure, current, connected equipment information, and load are analyzed, and evaluation conversion rules for each current, connected equipment information, and load are set to determine the evaluation results for each power grid node.

[0113] Based on the evaluation results of each power grid node, the power grid pipeline detection points are determined by screening according to preset screening conditions.

[0114] The detection points of the water pipe and the power grid pipe are analyzed for intersection, and the intersection points are merged. The merged detection points are then used to locate the detection points.

[0115] Furthermore, the detection device setting module 11 also includes the following execution steps:

[0116] Obtain the acquisition range of the detection device;

[0117] Based on the collection range, a preset distance is set for the detection points, and the preset distance for the detection points is used as the preset filtering condition.

[0118] Furthermore, the detection device setting module 11 also includes the following execution steps:

[0119] The rate of water level rise is monitored in real time using radar level gauges;

[0120] Using the cross-sectional area of ​​the factory building and the rate of water level rise, the real-time leakage flow rate is calculated according to the formula: Flow rate = Volume / Time.

[0121] The flow difference is calculated based on the real-time leakage flow rate and the flow rate of the preset drainage measures to determine the drainage control information. Based on the drainage control information, the matching drainage measures are executed or adjusted.

[0122] Furthermore, the detection device setting module 11 also includes the following execution steps:

[0123] The monitoring area is monitored by an infrared camera to obtain monitoring video information;

[0124] The monitoring video information is segmented using a semantic segmentation model to determine whether there is water leakage in the monitoring area, and the leakage point is located based on the water leakage monitoring results.

[0125] The semantic segmentation model includes a first channel and a second channel. The first channel has a first step length, and the second channel has a second step length, with the first step length being smaller than the second step length. The first channel is used to identify the leakage status, and the second channel is used to confirm the leakage point.

[0126] Furthermore, the detection device setting module 11 also includes the following execution steps:

[0127] The first monitoring image is acquired using the first channel;

[0128] The first monitoring image is semantically segmented to determine the target monitoring object, and a grayscale region of interest is set based on the grayscale region of the target monitoring object;

[0129] The target recognition information is obtained by using the grayscale region of interest to perform target recognition on the first monitoring image.

[0130] Furthermore, the early warning pattern matching module 15 includes the following execution steps:

[0131] Determine whether the difference between the detected data and the threshold level meets the preset adjustment range;

[0132] When the conditions are met, the matching warning mode is determined by matching the threshold level with the preset warning mode.

[0133] The warning loss value is calculated based on the matching warning mode to obtain the warning loss value;

[0134] The threshold level is adjusted by lowering the adjacent values, and the adjusted threshold level is matched with the preset early warning mode to determine the adjusted early warning loss value;

[0135] The warning loss value is compared with the adjusted warning loss value, and the matching warning mode with the smaller warning loss value is selected for warning processing.

[0136] Furthermore, embodiments of this application include a traffic monitoring module, which includes the following execution steps:

[0137] The detection device includes a flow monitoring device;

[0138] Set a traffic threshold;

[0139] The characteristic flow rate of the ball valve layer is monitored using flow monitoring equipment;

[0140] Flow is calculated based on the characteristic values ​​of the flow obtained from monitoring, and the calculated monitoring flow results are used to make early warning judgments based on the flow threshold.

[0141] In summary, any step of the method described above can be stored as a computer instruction or program in an unrestricted computer memory, and can be called and identified by an unrestricted computer processor to implement any method in the embodiments of this application, without any additional restrictions.

[0142] Furthermore, the "first" or "second" mentioned above may not only represent a sequential relationship, but may also represent a specific concept, and / or refer to the individual or collective selection of multiple elements. Clearly, those skilled in the art can make various modifications and variations to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A distributed leakage detection method for hydropower plant buildings, characterized in that, The method is applied to a leak detection system, the system comprising a set of detection devices, including: The detection points are located based on the factory building's pipeline structure, and detection devices are installed at the detection points, which are distributed in multiple locations within the factory building. Based on the locational impact of the detection points, multiple levels of detection thresholds are set. Based on the plant pipeline structure and the location information of the detection points, a three-dimensional model is created, and the detection data of each detection device is synchronously mapped to the three-dimensional model for visualization. Determine whether the detected data reaches a multi-level detection threshold; if so, determine the threshold level. The warning mode is matched according to the threshold level, and the warning is processed according to the warning mode, wherein the warning mode includes unit shutdown control, audible and visual alarm, and power line control. The detection device includes an infrared camera, and the method includes: The monitoring area is monitored by an infrared camera to obtain monitoring video information; The monitoring video information is segmented using a semantic segmentation model to determine whether there is water leakage in the monitoring area, and the leakage point is located based on the water leakage monitoring results. The semantic segmentation model includes a first channel and a second channel. The first channel has a first step length and the second channel has a second step length, with the first step length being smaller than the second step length. The first channel is used to identify the leakage status, and the second channel is used to confirm the leakage point. The first monitoring image is acquired using the first channel; The first monitoring image is semantically segmented to determine the target monitoring object, and a grayscale region of interest is set based on the grayscale region of the target monitoring object; The target recognition information is obtained by using the grayscale region of interest to perform target recognition on the first monitoring image. The method of locating detection points based on the factory building's pipeline structure includes: Information on the location, connection, and parameters of the water and electricity pipelines in the factory building is collected to obtain the factory building pipeline structure, which includes the factory building water pipe structure and the factory building power grid structure. Based on the plant's water pipe structure, the number of pipe connections, flow rate, and pipe diameter parameters are analyzed. Evaluation and calculation rules for the number of pipe connections, flow rate, and pipe diameter parameters are set, and the evaluation results for each water pipe node are determined. The evaluation results of each water pipe node are screened according to preset screening conditions to obtain the detection points of the water pipe. Based on the aforementioned power grid structure, current, connected equipment information, and load are analyzed, and evaluation conversion rules for each current, connected equipment information, and load are set to determine the evaluation results for each power grid node. Based on the evaluation results of each power grid node, the power grid pipeline detection points are determined by screening according to preset screening conditions. The detection points of the water pipe and the power grid pipe are analyzed for intersection, and the intersection points are merged. The merged detection points are then used to locate the detection points. Match the warning mode according to the threshold level, and perform warning processing according to the warning mode, including: Determine whether the difference between the detected data and the threshold level meets the preset adjustment range; When the conditions are met, the matching warning mode is determined by matching the threshold level with the preset warning mode. The warning loss value is calculated based on the matching warning mode to obtain the warning loss value; The threshold level is adjusted by lowering the adjacent values, and the adjusted threshold level is matched with the preset early warning mode to determine the adjusted early warning loss value; The warning loss value is compared with the adjusted warning loss value, and the matching warning mode with the smaller warning loss value is selected for warning processing.

2. The method as described in claim 1, characterized in that, Also includes: Obtain the acquisition range of the detection device; Based on the collection range, a preset distance is set for the detection points, and the preset distance for the detection points is used as the preset filtering condition.

3. The method as described in claim 1, characterized in that, The detection device includes a radar level gauge, and the method includes: The rate of water level rise is monitored in real time using radar level gauges; Using the cross-sectional area of ​​the factory building and the rate of water level rise, the real-time leakage flow rate is calculated according to the formula: Flow rate = Volume / Time. The flow difference is calculated based on the real-time leakage flow rate and the flow rate of the preset drainage measures to determine the drainage control information. Based on the drainage control information, the matching drainage measures are executed or adjusted.

4. The method as described in claim 1, characterized in that, The detection device includes a flow monitoring device, and the method includes: Set a traffic threshold; The characteristic flow rate of the ball valve layer is monitored using flow monitoring equipment; Flow is calculated based on the characteristic values ​​of the flow obtained from monitoring, and the calculated monitoring flow results are used to make early warning judgments based on the flow threshold.

5. A distributed leakage detection system for hydropower plant buildings, characterized in that, A distributed leakage detection method for implementing the hydropower plant building according to any one of claims 1-4, the system comprising a set of detection devices, the system comprising: The detection device setting module is used to locate detection points based on the factory pipeline structure and set detection devices at the detection points, wherein the detection points are distributed in multiple locations in the factory. A detection threshold setting module is used to set multiple levels of detection thresholds based on the positional influence of the detection point. The factory building 3D modeling module is used to perform 3D modeling based on the factory building pipeline structure and the location information of the detection points, and synchronously map the detection data of each detection device to the 3D model for visualization display. A threshold level determination module is used to determine whether the detection data reaches a multi-level detection threshold, and when it does, to determine the threshold level. The early warning mode matching module is used to match an early warning mode according to the threshold level and perform early warning processing according to the early warning mode. The early warning mode includes unit shutdown control, audible and visual alarm, and power line control.

Images