A method, device, medium and equipment for detecting the state of a hydropower station oil system

Abstract

The application discloses a kind of hydropower station oil system state detection method, device, medium and equipment, belong to artificial intelligence technical field, the present application first utilizes the monitoring trigger of abnormal condition of multi-source monitoring signal to shift to more simple direction, avoid directly to complex flow pattern identification, secondly, the historical flow pattern visual features before and after it are traced back by locating the time point of abnormality, restore the mutation point of flow pattern feature anomaly in real situation, then utilize the spatiotemporal correlation analysis of history and current flow pattern visual features, not only clear monitoring to flow pattern feature situation, more can analyze its evolution development trend, not only realize flow pattern accurate identification, more can the application of identification result, finally generate monitoring early warning information can effectively guide operation and fault early warning, to flow pattern visual features are more valuable use, improve the effect of hydropower station oil system state detection method.

Description

Technical Field

[0001] This invention belongs to the field of artificial intelligence technology, and specifically relates to a method, device, medium and equipment for detecting the status of a hydropower station oil system. Background Technology

[0002] Gas-liquid two-phase flow is widely present in industrial fields, and accurate identification of its flow patterns is crucial for understanding flow mechanisms, optimizing operating parameters, and ensuring system safety and efficiency. In hydropower stations and their auxiliary systems (such as oil systems), turbine oil, insulating oil, hydraulic oil, and other oil media often form complex gas-liquid two-phase flows during circulation, cooling, sealing, and control processes due to factors such as pressure changes, temperature fluctuations, and the introduction of trace amounts of moisture or air. The state of gas-liquid two-phase flow in hydropower station oil systems is a key microscopic characterization reflecting the operational health of main equipment (such as turbine bearings, speed control systems, and transformers). Existing identification and detection methods are limited to flow pattern classification and identification, only completing the perception task. The deficiency in the cognitive layer causes the flow pattern identification results to become information silos, without further utilization. Evaluation, analysis, guidance for operation and maintenance, and fault early warning all lack data support, resulting in poor effectiveness of state monitoring methods for hydropower station oil systems. Summary of the Invention

[0003] In order to overcome the shortcomings of the prior art, the present invention aims to provide a method, device, medium and equipment for detecting the status of a hydropower station oil system.

[0004] To achieve the above objectives, the present invention employs the following technical solution: In a first aspect, embodiments of the present invention provide a method for detecting the status of a hydropower station's oil system, comprising the following steps: In response to anomalies in multi-source monitoring signals, the system obtains the anomaly time point and the target flow pattern visual features; wherein, the target flow pattern visual features are the flow pattern visual features at the time corresponding to the anomaly time point. Based on the visual features of historical flow patterns before and after the anomalous time points, the anomalous mutation points are obtained; Based on anomalous mutation points and multi-source monitoring signals, spatiotemporal correlation analysis is performed on the visual features of historical flow patterns and the visual features of target flow patterns to obtain source tracing results. Based on the source tracing results, monitoring and early warning information is generated.

[0005] In one possible implementation of the first aspect, before generating monitoring and early warning information based on the source tracing results, the method further includes: Based on the output of the visual features of the target flow pattern using the health status association model, a quantitative health score of the equipment status is obtained. Based on the source tracing results, monitoring and early warning information is generated, including: Based on the source tracing results and quantitative health scores, tiered monitoring and early warning information is generated.

[0006] In one possible implementation of the first aspect, before obtaining a quantitative health score of the device status based on the output of the visual features of the target flow pattern from the health status association model, the method further includes: A health status association model is trained based on synchronized manifold visual features and ground truth data of device status under different health states.

[0007] In one possible implementation of the first aspect, before training a health status association model based on synchronized manifold visual features and ground truth data of device status under different health states of the device, the method further includes: Based on the target detection results of the detected images, the device status characteristics are obtained; Based on device status characteristics, flow pattern visual features are obtained.

[0008] In one possible implementation of the first aspect, flow pattern visual features are obtained based on device state characteristics, including: Higher-order feature calculations are performed based on device state characteristics to obtain spatial distribution features and temporal evolution features.

[0009] In one possible implementation of the first aspect, before obtaining the device state features based on the target detection results of the detected image, the method further includes: The target detection network is used to perform target detection on the detection image to obtain the target detection result. The target detection network uses the YOLOv10-s network as the basic recognition network and includes the CSPHet module, ADown module and SPPELAN module.

[0010] In one possible implementation of the first aspect, based on anomalous mutation points and multi-source monitoring signals, spatiotemporal correlation analysis is performed on the visual features of historical flow patterns and the visual features of the target flow pattern to obtain source tracing results, including: Based on anomalous mutation points and multi-source monitoring signals, a correlation analysis of the spatial distribution and temporal evolution of the visual features of historical flow patterns and the visual features of target flow patterns is conducted to obtain the source tracing results.

[0011] Secondly, embodiments of the present invention provide a hydropower station oil system status detection device, comprising: An anomaly response module is used to respond to anomalies in multi-source monitoring signals and obtain the anomaly time point and target flow pattern visual features; wherein, the target flow pattern visual features are the flow pattern visual features at the time corresponding to the anomaly time point; The mutation detection module is used to obtain abnormal mutation points based on the visual features of historical flow patterns before and after the abnormal time points; The source tracing module analyzes the spatiotemporal correlation of historical flow pattern visual features and target flow pattern visual features based on abnormal mutation points and multi-source monitoring signals to obtain source tracing results. The monitoring and early warning module generates monitoring and early warning information based on the source tracing results.

[0012] Thirdly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when loaded and executed by a processor, implements the hydropower station oil system status detection method provided in any of the first aspects above.

[0013] Fourthly, embodiments of the present invention provide an electronic device, including a processor and a memory, wherein: Memory is used to store computer programs; The processor is used to load and execute computer programs to cause electronic devices to perform the hydroelectric power plant oil system status detection method provided in any of the first aspects above.

[0014] Compared with the prior art, the present invention has the following beneficial effects: This invention proposes a method, apparatus, medium, and equipment for detecting the state of an oil system in a hydropower station. The method includes: responding to an anomaly in a multi-source monitoring signal, obtaining the anomaly time point and the visual features of the target flow pattern; wherein, the visual features of the target flow pattern are the flow pattern visual features at the time corresponding to the anomaly time point; obtaining the anomalous abrupt change point based on the historical flow pattern visual features before and after the anomaly time point; performing spatiotemporal correlation analysis on the historical flow pattern visual features and the target flow pattern visual features based on the anomalous abrupt change point and the multi-source monitoring signal to obtain the source tracing result; and generating monitoring and early warning information based on the source tracing result.

[0015] This invention first utilizes multi-source monitoring signals to shift the monitoring triggering of abnormal situations to a simpler direction, avoiding direct and complex flow pattern identification. Second, by locating the time point of the anomaly, it traces the historical flow pattern visual features before and after it, reconstructing the abrupt change points of the flow pattern anomalies in the real situation. Then, it uses historical and current flow pattern visual features for spatiotemporal correlation analysis, which not only clearly monitors the flow pattern features but also analyzes their evolution and development trends. This not only achieves accurate flow pattern identification but also allows for the reapplication of the identification results, ultimately generating monitoring and early warning information that can effectively guide operation and maintenance and fault early warning. This provides more valuable utilization of flow pattern visual features and improves the effectiveness of the state detection method for hydropower station oil systems. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the electronic device structure for the hardware operating environment involved in this invention; Figure 2 A flowchart illustrating the hydropower station oil system status detection method provided by the present invention; Figure 3 This is a schematic diagram of the module structure of the target detection network in the hydropower station oil system status detection method provided by the present invention; Figure 4 A schematic diagram of the three-layer system integrated for the hydropower station oil system condition detection method provided by the present invention; Figure 5 This is a schematic diagram of the module of the hydropower station oil system status detection device provided by the present invention.

[0017] Wherein: 101-processor, 102-communication bus, 103-network interface, 104-user interface, 105-memory. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0020] See attached document Figure 1 , attached Figure 1This is a schematic diagram of the electronic device structure of the hardware operating environment involved in the embodiments of the present invention. The electronic device may include: a processor 101, such as a central processing unit (CPU), a communication bus 102, a user interface 104, a network interface 103, and a memory 105. The communication bus 102 is used to realize the connection and communication between these components. The user interface 104 may include a display screen and an input unit such as a keyboard. Optionally, the user interface 104 may also include a standard wired interface and a wireless interface. The network interface 103 may include a standard wired interface and a wireless interface (such as a Wi-Fi interface). The memory 105 may be a storage device independent of the aforementioned processor 101. The memory 105 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as at least one disk storage device. The processor 101 may be a general-purpose processor, including a central processing unit, a network processor, etc., or it may be a digital signal processor, an application-specific integrated circuit, a field-programmable gate array or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component.

[0021] Those skilled in the art will understand that the appendix Figure 1 The structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0022] As attached Figure 1 As shown, the memory 105, which serves as a storage medium, may include an operating system, a network communication module, a user interface module, and a hydroelectric power station oil system status detection device.

[0023] In the appendix Figure 1 In the electronic device shown, the network interface 103 is mainly used for data communication with the network server; the user interface 104 is mainly used for data interaction with the user; the processor 101 and the memory 105 in this invention can be set in the electronic device. The electronic device calls the hydropower station oil system status detection device stored in the memory 105 through the processor 101 and executes the hydropower station oil system status detection method provided in the embodiment of this invention.

[0024] See attached document Figure 2 Based on the hardware device described in the foregoing embodiments, embodiments of the present invention provide a method for detecting the status of a hydropower station oil system, comprising the following steps: S10: In response to an anomaly in the multi-source monitoring signal, obtain the anomaly time point and the target flow pattern visual features; wherein, the target flow pattern visual features are the flow pattern visual features at the time corresponding to the anomaly time point.

[0025] In practical implementation, multi-source monitoring signals such as pressure, vibration, and temperature signals can be collected by setting up corresponding sensors. These monitoring signals are commonly used in monitoring equipment within the field. Compared to identifying flow patterns and then making anomaly judgments, these direct and simple monitoring signals can respond more quickly to abnormal situations. When anomalies occur in multiple monitoring signals, accompanied by flow pattern anomalies, locating the time point when the anomaly occurs and obtaining the corresponding flow pattern visual characteristics can assist in subsequent identification and judgment of flow pattern characteristics.

[0026] S20: Based on the visual features of historical flow patterns before and after the anomalous time point, obtain the anomalous mutation point.

[0027] In the specific implementation process, since the anomaly of the flow pattern may not occur simultaneously with the anomaly of the monitoring signal, the abnormal time of the multi-source monitoring signal, which is a visual representation, cannot be directly used as the time of the flow pattern anomaly. Therefore, the abnormal time point of the multi-source monitoring signal is used to help locate the anomaly of the flow pattern characteristics. The historical flow pattern visual features before and after the abnormal time point are extracted. On the one hand, it can trace the abrupt change point of the real flow pattern characteristics. On the other hand, it can provide a reliable data foundation for subsequent spatiotemporal correlation analysis.

[0028] S30: Based on anomalous mutation points and multi-source monitoring signals, spatiotemporal correlation analysis is performed on the visual features of historical flow patterns and the visual features of target flow patterns to obtain source tracing results.

[0029] In the specific implementation process, after tracing the anomaly point of the flow pattern mutation, the starting time of the anomaly in the flow pattern was obtained. From this time, spatiotemporal correlation analysis was performed by combining the visual characteristics of historical flow patterns and the visual characteristics of the current target flow pattern to trace its development and evolution trend from the starting time to the present time, providing a basis for subsequent monitoring and early warning. The spatiotemporal correlation analysis includes temporal and spatial analyses, specifically spatiotemporal distribution and temporal evolution. Spatiotemporal distribution characterizes the static features of the flow pattern, while temporal evolution introduces time changes into the static features, thereby obtaining how the visual characteristics of the flow pattern gradually develop from the initial state to the current state. That is, based on the anomalous mutation point and multi-source monitoring signals, spatiotemporal correlation analysis is performed on the visual characteristics of historical flow patterns and the visual characteristics of the target flow pattern to obtain the source tracing results, including: Based on anomalous mutation points and multi-source monitoring signals, a correlation analysis of the spatial distribution and temporal evolution of the visual features of historical flow patterns and the visual features of target flow patterns is conducted to obtain the source tracing results.

[0030] S40: Generate monitoring and early warning information based on the source tracing results.

[0031] In the specific implementation process, after obtaining the source tracing results, they are utilized more valuablely, enabling manifold detection to leap from identification to diagnosis. This not only solves the problem of perceiving what a manifold is, but also, through analysis and evaluation, addresses the problem of understanding what the manifold means. Specifically, before generating monitoring and early warning information based on the source tracing results, the method also includes: Based on the output of the visual features of the target flow pattern using the health status association model, a quantitative health score of the equipment status is obtained. Based on the source tracing results, monitoring and early warning information is generated, including: Based on the source tracing results and quantitative health scores, tiered monitoring and early warning information is generated.

[0032] In practical implementation, this invention, through source tracing results, can pinpoint the source of the anomaly and reconstruct the fault development process at or before the fault occurs. For example, the source tracing result, "Two hours before the vibration alarm, the relevant oil circuit showed a continuous increase in the aggregation of small bubbles, indicating that the root cause of the vibration may be early wear of the bearing, and it is recommended to prioritize checking bearing number X," provides an intuitive fault development chain for operation and maintenance. By introducing a quantitative health score for equipment status, with quantifiable values, monitoring and early warning information for monitored equipment such as bearings and transformers can be generated more efficiently and accurately. For example, graded early warning includes three levels: attention, warning, and alarm, with corresponding specific maintenance recommendations including enhanced monitoring, arranging oil filtration, and recommending shutdown for inspection.

[0033] Furthermore, before obtaining a quantitative health score for the device status based on the output of the visual features of the target flow pattern using the health status association model, the method also includes: A health status association model is trained based on synchronized manifold visual features and ground truth data of device status under different health states.

[0034] In the implementation process, a pre-trained health status association model is used to correlate equipment health status with flow pattern visual features, providing a foundation for analysis and early warning applications. Multiple dedicated models can be set up for different equipment. Equipment status ground values, such as vibration values, temperature values, oiling data, and fault records, along with different health states such as normal, wear, and overheating, are used to establish a matching model between flow patterns and equipment states under multiple health states, avoiding incomplete detection and identification problems caused by assessments under a single state. Dedicated models for different health states can be stored in a model library for easy retrieval. Examples include a bearing wear assessment model, which takes the bearing return oil pipeline fingerprint as input and outputs the probability of wear stage, such as normal, slight, and severe; a transformer fault identification model, which takes the insulating oil pipeline features as input and outputs the probability of fault type, such as partial discharge and overheating; and a sealing status assessment model, used to assess the degree of air infiltration. The results of these assessments are converted into specific quantitative health scores, such as 0-100 points, for analysis together with the source tracing results to achieve graded early warning.

[0035] In one embodiment, before training a health status association model based on synchronized manifold visual features and ground truth data of device status under different health states of the device, the method further includes: Based on the target detection results of the detected images, the device status characteristics are obtained; Based on device status characteristics, flow pattern visual features are obtained.

[0036] In the specific implementation process, the extraction of manifold visual features and monitoring and early warning are performed separately to simplify the operation. By deeply processing the target detection results of the detection image, the original detection boxes are transformed into features representing the device status. Specifically, based on the device status features, manifold visual features are obtained, including: Higher-order feature calculations are performed based on device state characteristics to obtain spatial distribution features and temporal evolution features.

[0037] High-order feature calculations are performed on the continuous frame detection results output from the acquisition end to generate feature vectors, i.e., flow pattern visual features. These feature vectors include spatial distribution features such as bubble size distribution histograms, spatial clustering index, and shape irregularity, as well as temporal evolution features such as bubble generation / annihilation rates, average rise rates, and the frequency of occurrence of dominant flow patterns (e.g., slug flow). The acquisition end acquires video streams from the pipeline observation window using a high-speed camera and extracts the target detection results through a target detection network. Before obtaining device status features based on target detection results from the detected images, the method also includes: The target detection network is used to perform target detection on the detection image to obtain the target detection result. The target detection network is based on the YOLOv10-s network as the basic recognition network, and the C2f module is replaced by the CSPHet module, the SCDown module is replaced by the ADown module, and the SPPF module is replaced by the SPPELAN module.

[0038] In practical implementation, a target detection network is used on edge devices to achieve real-time and accurate bubble and flow pattern target detection, outputting data such as target category, location, bounding box, and confidence score. For scenarios involving turbid, highly reflective, and multi-scale bubbles in hydropower station oil, this invention provides an improved YOLOv10-s network, as shown in the attached diagram. Figure 3 As shown, it uses the YOLOv10-s network as the basic identification network, replacing the original C2f module with the CSPHet module, the original SCDown module with the ADown module, and the original SPPF module with the SPPELAN module.

[0039] Replacing the C2f module with the CSPHet module effectively avoids excessive complexity and redundancy in the model structure, maintaining excellent performance while reducing the computational complexity of the feature integration stage, enabling the model to more efficiently parse multi-scale and multi-view features of the input data. Replacing the SCDown module with the ADown module ensures that the model maintains high accuracy in object detection tasks. In addition, the ADown module has adaptive learning capabilities, which can dynamically adjust according to different data scenarios, thereby further optimizing the overall performance. Replacing the SPPF module with the SPPELAN module improves the accuracy of feature aggregation and multi-scale perception capabilities by introducing a local attention mechanism, enhancing the model's feature representation capabilities.

[0040] This invention integrates a three-tiered system of perception, cognition, and decision-making, as shown in the appendix. Figure 4 As shown, the perception layer uses an improved YOLOv10-s network for target detection to improve detection performance. The cognition layer establishes a correlation model between flow pattern visual features and unit status, mapping the flow pattern visual fingerprint to equipment health score or failure mode probability. The decision layer analyzes the historical mutation and evolution patterns of the flow pattern visual fingerprint before and after anomalies, and combines the correlation model to generate a source tracing analysis report pointing to the root cause and location of potential faults, realizing a fully automated processing flow from flow pattern image to equipment health assessment, fault tracing and maintenance recommendation output.

[0041] This invention first utilizes multi-source monitoring signals to shift the monitoring triggering of abnormal situations to a simpler direction, avoiding direct and complex flow pattern identification. Secondly, by locating the time point of the anomaly, it traces the historical flow pattern visual features before and after it, reconstructing the abrupt change point of the flow pattern feature anomaly in the real situation. Then, it uses historical and current flow pattern visual features for spatiotemporal correlation analysis, which not only clearly monitors the flow pattern features but also analyzes its evolution and development trend. This not only achieves accurate flow pattern identification but also allows for the reapplication of the identification results. Finally, the generated monitoring and early warning information can effectively guide operation and maintenance and fault early warning, making more valuable use of flow pattern visual features and improving the effectiveness of the state detection method for hydropower station oil systems.

[0042] See attached document Figure 5 Based on the same inventive concept as in the foregoing embodiments, this embodiment of the invention also provides a hydropower station oil system status detection device, comprising: An anomaly response module is used to respond to anomalies in multi-source monitoring signals and obtain the anomaly time point and target flow pattern visual features; wherein, the target flow pattern visual features are the flow pattern visual features at the time corresponding to the anomaly time point; The mutation detection module is used to obtain abnormal mutation points based on the visual features of historical flow patterns before and after the abnormal time points; The source tracing module analyzes the spatiotemporal correlation of historical flow pattern visual features and target flow pattern visual features based on abnormal mutation points and multi-source monitoring signals to obtain source tracing results. The monitoring and early warning module generates monitoring and early warning information based on the source tracing results.

[0043] Those skilled in the art should understand that the division of the various modules in the embodiments is merely a logical functional division. In actual applications, they can be fully or partially integrated onto one or more actual carriers. These modules can be implemented entirely in software through processing unit calls, entirely in hardware, or a combination of software and hardware. It should be noted that each module in the hydropower station oil system status detection device in this embodiment corresponds one-to-one with each step in the hydropower station oil system status detection method in the aforementioned embodiments. Therefore, the specific implementation of this embodiment can refer to the implementation of the aforementioned hydropower station oil system status detection method, which will not be repeated here.

[0044] Based on the same inventive concept as in the foregoing embodiments, embodiments of the present invention also provide a computer-readable storage medium storing a computer program, which, when loaded and executed by a processor, implements the hydropower station oil system status detection method provided in the embodiments of the present invention.

[0045] Based on the same inventive concept as in the foregoing embodiments, embodiments of the present invention also provide an electronic device, including a processor and a memory, wherein: Memory is used to store computer programs; The processor is used to load and execute computer programs to enable electronic devices to perform the hydroelectric power station oil system status detection method provided in the embodiments of the present invention.

[0046] In some embodiments, the computer-readable storage medium may be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; or it may be a device including one or any combination of the above-mentioned memories. The computer may be a variety of computing devices, including smart terminals and servers.

[0047] In some embodiments, executable instructions may take the form of a program, software, software module, script, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a standalone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0048] As an example, executable instructions may, but do not necessarily, correspond to files in the file system. They may be stored as part of a file that holds other programs or data, for example, in one or more scripts in a Hyper Text Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple collaborative files (e.g., a file that stores one or more modules, subroutines, or code sections).

[0049] As an example, executable instructions can be deployed to execute on a single computing device, or on multiple computing devices located in one location, or on multiple computing devices distributed across multiple locations and interconnected via a communication network.

[0050] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory / random access memory, magnetic disk, optical disk) and includes several instructions to cause a multimedia terminal device (which may be a mobile phone, computer, television receiver, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0051] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A method for detecting the condition of an oil system in a hydropower station, characterized in that, Includes the following steps: In response to an anomaly in the multi-source monitoring signal, the abnormal time point and the target flow pattern visual features are obtained; wherein, the target flow pattern visual features are the flow pattern visual features at the time corresponding to the abnormal time point; Based on the historical flow pattern visual features before and after the anomalous time point, the anomalous mutation point is obtained; Based on the abnormal mutation points and the multi-source monitoring signals, spatiotemporal correlation analysis is performed on the historical flow pattern visual features and the target flow pattern visual features to obtain the source tracing results. Based on the source tracing results, monitoring and early warning information is generated.

2. The method for detecting the state of a hydropower station oil system according to claim 1, characterized in that, Before generating monitoring and early warning information based on the source tracing results, the method further includes: Based on the output results of the visual features of the target flow pattern using the health status association model, a quantitative health score of the device status is obtained. The generation of monitoring and early warning information based on the source tracing results includes: Based on the source tracing results and the quantitative health score, tiered monitoring and early warning information is generated.

3. The method for detecting the state of a hydropower station's oil system according to claim 2, characterized in that, Before obtaining the quantitative health score of the device status from the output results of the visual features of the target flow pattern based on the health status association model, the method further includes: The health status association model is trained based on the synchronized manifold visual features and ground truth data of the device under different health states.

4. The method for detecting the condition of a hydropower station oil system according to claim 3, characterized in that, Before training the health status association model based on synchronized manifold visual features and ground truth data of device status under different health states of the device, the method further includes: Based on the target detection results of the detected images, the device status characteristics are obtained; The flow pattern visual features are obtained based on the device status characteristics.

5. The method for detecting the condition of a hydropower station oil system according to claim 4, characterized in that, The process of obtaining the flow pattern visual features based on the device state features includes: Based on the device state characteristics, higher-order feature calculations are performed to obtain spatial distribution features and temporal evolution features.

6. The method for detecting the state of a hydropower station oil system according to claim 4, characterized in that, Before obtaining device status features based on the target detection results of the detected image, the method further includes: The target detection result is obtained by performing target detection on the detection image based on the target detection network; wherein, the target detection network uses the YOLOv10-s network as the basic recognition network and includes the CSPHet module, the ADown module and the SPPELAN module.

7. The method for detecting the state of a hydropower station oil system according to claim 1, characterized in that, The process involves performing spatiotemporal correlation analysis on the historical flow pattern visual features and the target flow pattern visual features based on the anomalous mutation points and the multi-source monitoring signals to obtain source tracing results, including: Based on the anomalous mutation points and the multi-source monitoring signals, a correlation analysis of the spatial distribution and temporal evolution of the historical flow pattern visual features and the target flow pattern visual features is performed to obtain the source tracing results.

8. A device for detecting the status of a hydroelectric power station oil system used in printing and packaging, characterized in that, include: An anomaly response module is used to respond to anomalies in multi-source monitoring signals and obtain the anomaly time point and target flow pattern visual features; wherein, the target flow pattern visual features are the flow pattern visual features at the time corresponding to the anomaly time point; The mutation detection module is used to obtain abnormal mutation points based on the historical flow pattern visual features before and after the abnormal time point; The analysis and tracing module performs spatiotemporal correlation analysis on the historical flow pattern visual features and the target flow pattern visual features based on the abnormal mutation points and the multi-source monitoring signals to obtain the tracing results; The monitoring and early warning module generates monitoring and early warning information based on the source tracing results.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is loaded and executed by the processor, it implements the hydropower station oil system status detection method as described in any one of claims 1-7.

10. An electronic device, characterized in that, Including processor and memory, of which: The memory is used to store computer programs; The processor is used to load and execute the computer program so that the electronic device performs the hydropower station oil system status detection method as described in any one of claims 1-7.

Images