Waste judgment system and method for in-service black water angle valve based on non-contact monitoring

Abstract

The invention relates to the technical field of industrial valves, in particular to an in-service black water angle valve waste judgment system and method.The system comprises a non-contact information collection platform, a local edge server and a cloud server, the local edge server is used for obtaining target time sequence data in different time scale windows through a sliding window method, and the target time sequence data are sent to the cloud server; data features are extracted from multiple dimensions, and the current operation state of the black water angle valve is judged; the cloud server is used for executing second monitoring analysis, including performing trend prediction on the future operation state of the black water angle valve by adopting a prediction model based on the fusion time sequence data and the analysis result of the local edge server; and based on the prediction result of the future operation state and the current operation state of the black water angle valve, identifying the potential fault type of the black water angle valve through a fault classification model, and calculating the health index and the residual life of the black water angle valve based on a life evaluation model. Therefore, the waste judgment efficiency, accuracy and reliability of the black water angle valve are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial valves, and in particular to a device and method for judging the scrapping of an in-service black water angle valve. Background Art

[0002] Blackwater angle valves are specialized valves used to regulate or control the pressure, level, and flow of blackwater. Widely used in the coal chemical industry, these valves are subject to the harsh, high-temperature, high-pressure conditions of coal chemical and coal gasification processes. These valves are constantly exposed to the combined impact of solid particles (concentrations > 8%), highly corrosive media (Cl⁻ and H₂S concentrations > 500 ppm), and extreme pressure differentials (> 6.5 MPa). This can lead to frequent failures such as valve core erosion wear, valve seat cavitation spalling, barrel corrosion perforation, and valve stem fatigue fracture.

[0003] Existing technologies mainly rely on contact sensors (such as piezoelectric vibration probes and thermocouples) and manual periodic inspections, which have the following technical bottlenecks: Contact sensors (such as piezoelectric vibration probes and thermocouples) have an extremely short lifespan in environments with high temperatures (>300°C), high-velocity particle erosion (>15m / s), and highly corrosive media (Cl⁻ / H2S concentration >500ppm), with an average failure cycle of less than 3 months. Furthermore, due to limitations in valve structure standards, they are unable to directly monitor critical damage areas such as the valve core and valve seat, resulting in the loss of more than 60% of key data.

[0004] Traditional vibration threshold methods have a miss-detection rate exceeding 50% for cavitation microcracks less than 0.3mm in depth and fatigue cracks less than 0.1mm in length. Manual disassembly and inspection require downtime every two to three months, costing over $200,000 per inspection. This still makes it difficult to avoid unplanned downtime caused by sudden leaks (annual losses exceeding 5 million yuan). Consequently, the use of contact sensors (such as piezoelectric vibration probes and thermocouples) and manual periodic disassembly and inspection of blackwater angle valves has a high miss-detection rate, high manual inspection costs, and low reliability and accuracy. Summary of the Invention

[0005] To address the short lifespan, large blind spots, high missed detection rates, high manual inspection costs, and low reliability and accuracy of existing contact sensors, this paper proposes a non-contact monitoring system and method for determining the scrapping of in-service blackwater angle valves. By combining multimodal sensing with intelligent analysis technology, this system accurately assesses the health of valves and determines their retirement.

[0006] In the first aspect, an embodiment of the present invention discloses a scrap judgment system for in-service black water angle valves based on non-contact monitoring, including: the scrap judgment system for in-service black water angle valves based on non-contact monitoring provided by the embodiment of the present invention includes: a non-contact information collection platform, a local edge server and a cloud server; the non-contact information collection platform includes: a circular track, a support structure, a multi-modal non-contact recorder component, a data preprocessing module and a transmission module and a power supply system; the circular track is arranged around the black water angle valve and is fixedly connected to the pipe connected to the black water angle valve through the support structure; the multi-modal non-contact recorder Components, including action frequency monitoring recorder, vibration monitoring recorder, noise monitoring recorder and temperature monitoring recorder, are installed at intervals on the circular track, respectively used to collect the opening and closing action signals, vibration characteristic signals, noise signals and surface temperature signals of the black water angle valve; the data preprocessing module, transmission module and power supply system are integrated and installed in the closed protective box of the circular track, and the data preprocessing module and transmission module are equipped with a micro control unit for multi-channel synchronous sampling, filtering, time alignment, unit normalization and unified time stamp marking of the signals collected by the multimodal non-contact recorder component to form a fusion Time series data, the fused time series data is transmitted to the local edge server in real time via LoRa or Wi-Fi; the local edge server is used to receive the fused time series data, perform feature extraction on it, obtain the target time series data, and perform the first monitoring analysis; the first monitoring analysis includes: using the sliding window method to obtain the target time series data in different time scale windows, and extracting data features from multiple dimensions to determine the current operating status of the blackwater angle valve; after the local edge server completes the first monitoring analysis, it transmits the target time series data and analysis results to the cloud server in real time through the built-in Wi-Fi module or 5G communication module; the cloud server is used to perform the second monitoring analysis, including: based on the fused time series data and the analysis results of the local edge server, using the prediction model to predict the trend of the future operating status of the blackwater angle valve; based on the prediction results of the future operating status and the current operating status of the blackwater angle valve, the potential fault type of the blackwater angle valve is identified through the fault classification model, and the health index and remaining life of the blackwater angle valve are calculated based on the life assessment model, and the potential fault type, health index and remaining life are output through a visual interactive interface and an automatic retirement recommendation is generated.

[0007] In a second aspect, an embodiment of the present invention discloses a method for judging the scrapping of an in-service black water angle valve based on non-contact monitoring, comprising the following steps: The multimodal non-contact recorder component in the non-contact information acquisition platform is used to collect the action signal, vibration characteristic signal, noise signal and surface temperature signal of the blackwater angle valve. The signals collected by the multimodal non-contact recorder component are subjected to multi-channel synchronous sampling, filtering, time alignment, unit normalization and unified timestamp marking through the data preprocessing module to form fused time series data. The fused time series data is transmitted to the local edge server in real time via LoRa or Wi-Fi through the transmission module.

[0008] The fused time series data is received through the local edge server, and features are extracted to obtain target time series data, and a first monitoring and analysis is performed. The first monitoring and analysis includes: using a sliding window method to obtain target time series data in different time scale windows, and extracting data features from multiple dimensions to determine the current operating status of the blackwater angle valve. After the local edge server completes the first monitoring and analysis, the target time series data and analysis results are transmitted to the cloud server in real time through the built-in Wi-Fi module or 5G communication module of the local edge server.

[0009] The cloud server is used to perform the second monitoring and analysis, including: based on the target time series data and the analysis results of the local edge server, using the prediction model to predict the trend of the future operating status of the blackwater angle valve; based on the prediction results of the future operating status and the current operating status of the blackwater angle valve, the potential fault type of the blackwater angle valve is identified through the fault classification model, and the health index and remaining life of the blackwater angle valve are calculated based on the life assessment model, and the potential fault type, health index and remaining life are output through a visual interactive interface to generate automatic retirement suggestions.

[0010] In this way, the embodiment of the present invention obtains the operating data of the blackwater angle valve through a non-contact acquisition platform, eliminating the need to physically contact or disassemble the blackwater angle valve to install a contact sensor. This reduces the risk of equipment damage that may be introduced by installing the sensor, while also avoiding the problem of low accuracy of the collected data caused by the impact of harsh environments such as high corrosion on the sensor, further improving the accuracy and stability of the blackwater angle valve scrapping judgment. The local edge server directly preprocesses and performs the first monitoring analysis on the collected operating data, reducing the time and network bandwidth resources required to transmit the data to the cloud. The data can be quickly processed and analyzed locally, and the operating status of the blackwater angle valve can be output in real time, enhancing the real-time, reliability, and efficiency of the blackwater angle valve scrapping judgment. In addition, the target time series data and real-time operating status are input into the prediction model to predict the future status of the blackwater angle valve in advance. Based on the prediction results, a more scientific and reasonable maintenance plan can be formulated, changing from passive maintenance to active preventive maintenance, avoiding equipment downtime and production interruption caused by sudden failures, reducing operation and maintenance costs, and extending the service life of the blackwater angle valve. In addition, the embodiment of the present invention uses the above-mentioned automated method to inspect the black water angle valve, which will not be affected by human subjective factors, improves the efficiency, real-time and accuracy of scrap judgment, reduces the missed detection rate, and improves the accuracy and reliability of scrap judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic structural diagram of a scrapping judgment system for in-service black water angle valves based on non-contact monitoring provided by an embodiment of the present invention.

[0012] Figure 2 A schematic structural diagram of a circular track provided in an embodiment of the present invention.

[0013] Figure 3 A schematic diagram of the matching structure of a black water angle valve and a ring track provided in an embodiment of the present invention.

[0014] Figure 4 A schematic flow chart of a method for judging the scrapping of an in-service black water angle valve based on non-contact monitoring is provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0015] To further illustrate the technical means and effectiveness of the present invention in achieving its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effectiveness of a system and method for determining the scrapping of in-service blackwater angle valves based on non-contact monitoring, as proposed by the present invention. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0016] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0017] like Figure 1 As shown, the embodiment of the present invention also provides a system for judging the scrapping of in-service black water angle valves based on non-contact monitoring, such as Figure 1 As shown, Figure 1 This figure shows a schematic diagram of a non-contact monitoring system for scrapping in-service blackwater angle valves, provided in an embodiment of the present invention. The non-contact monitoring system 100 comprises a non-contact information collection platform 101, a local edge server 102, and a cloud server 103. The non-contact information collection platform includes a circular track, a support structure, a multimodal non-contact recorder assembly, a data preprocessing module, a transmission module, and a power supply system.

[0018] The annular track is arranged around the black water angle valve and is fixedly connected to the pipeline connected to the black water angle valve through a supporting structure.

[0019] The multimodal non-contact recorder components, including the action frequency monitoring recorder, vibration monitoring recorder, noise monitoring recorder and temperature monitoring recorder, are installed at intervals on the circular track and are used to collect the opening and closing action signals, vibration characteristic signals, noise signals and surface temperature signals of the black water angle valve respectively.

[0020] The data preprocessing module, transmission module and power supply system are integrated and installed in a closed protective box on the circular track. The data preprocessing module and transmission module are equipped with a microcontroller unit for performing multi-channel synchronous sampling, filtering, time alignment, unit normalization and unified timestamp marking on the signals collected by the multimodal contactless recorder components to form fused time series data. The fused time series data is transmitted to the local edge server in real time via LoRa or Wi-Fi.

[0021] The local edge server is used to receive the fused time series data, perform feature extraction on it, obtain the target time series data, and perform the first monitoring analysis; the first monitoring analysis includes: using the sliding window method to obtain the target time series data in different time scale windows, and extracting data features from multiple dimensions to determine the current operating status of the blackwater angle valve; after the local edge server completes the first monitoring analysis, it transmits the target time series data and analysis results to the cloud server in real time through the built-in Wi-Fi module or 5G communication module.

[0022] The cloud server is used to perform the second monitoring and analysis, including: based on the analysis results of the fused time series data and the local edge server, using the prediction model to predict the trend of the future operating status of the blackwater angle valve; based on the prediction results of the future operating status and the current operating status of the blackwater angle valve, identifying the potential fault type of the blackwater angle valve through the fault classification model, and calculating the health index and remaining life of the blackwater angle valve based on the life assessment model, outputting the potential fault type, health index and remaining life through a visual interactive interface and generating automatic retirement recommendations.

[0023] Specifically, the data preprocessing module is equipped with a low-power microcontroller unit (MCU) with a built-in high-speed analog-to-digital converter. It receives analog signals from laser displacement sensors, laser Doppler vibration sensors, directional sound pickup sensors, and infrared cameras, converting them into digital signals in real time to meet subsequent data analysis and processing requirements. Initial preprocessing is performed on the MCU, including missing value filling (interpolation), denoising filtering (low-pass filtering and wavelet denoising), and timestamping. The transmission module utilizes a combination of low-power wide area network (LoRa) and wireless fidelity (Wi-Fi) wireless communication technologies to efficiently transmit time-series data from the blackwater angle valve's operation to a local edge server. A dual-mode LoRa and Wi-Fi communication solution is employed to address diverse transmission requirements. Over long distances or when Wi-Fi signals are unstable, the time-series data collected by the contactless information collection platform can be transmitted over long distances via LoRa. Within Wi-Fi coverage, it automatically switches to a high-speed Wi-Fi network and transmits data to the design workstation, achieving stable and efficient data transmission in multiple scenarios.

[0024] Furthermore, in the embodiment of the present invention, the motion frequency monitoring recorder, vibration monitoring recorder, noise monitoring recorder and temperature monitoring recorder can respectively adopt a laser displacement sensor, a laser Doppler vibration sensor, a directional sound pickup sensor and an infrared camera.

[0025] Furthermore, the local edge server uses an embedded AI chip that supports local feature extraction and clustering calculations, receiving pre-processed, fused time series data transmitted by the contactless information collection platform in real time. The local edge server can extract features from the time series operation data, including key parameters such as the mean, variance, peak value, and frequency domain characteristics of each operation data, to obtain the target time series data.

[0026] Furthermore, a sliding window method can be used to divide the time series into short-term, medium-term, and long-term time scales, where the short-term time scale can be 1 hour, the medium-term time scale can be 1 day, and the long-term time scale can be 30 days. Of course, depending on the actual situation, the short-term, medium-term, and long-term time scales can also be other situations, and the embodiment of the present invention is not limited here. To improve data processing efficiency, the embodiment of the present invention standardizes and normalizes the target time series data obtained in different time scale windows using the sliding window method, so that the feature data of each time scale window is analyzed on the same dimension, improving data processing efficiency and accuracy.

[0027] Specifically, if Figure 2 and Figure 3 As shown, Figure 2 A schematic structural diagram of a circular track provided in an embodiment of the present invention, Figure 3 This diagram illustrates the coordinated structure of a blackwater angle valve and a circular track according to an embodiment of the present invention. The motion frequency monitoring recorder 201, vibration monitoring recorder 202, noise monitoring recorder 203, and temperature monitoring recorder 204 are all installed at intervals on the circular track. The circular track surrounds the blackwater angle valve and is connected to the pipe 206 connected to the blackwater angle valve via a support structure 205. The data preprocessing module, transmission module, and power supply system are integrated within a closed protective box 207 on the circular track. Furthermore, as an optional embodiment of the present invention, the diameter D of the annular track satisfies: 200mm≤D−Dv≤400mm, Dv is the maximum outer diameter of the black water angle valve, and the minimum distance between the annular track and the black water angle valve is not less than 50mm.

[0028] Furthermore, as an optional embodiment of the present invention, the support structure 205 includes a support column 2051 and a sleeve installed on the outer wall of the pipe connected to the black water angle valve. The sleeve is fixed to the outer surface of the pipe by a clamp. A flat base plate 2050 is welded above the sleeve. A triangular reinforcing rib is provided between the flat base plate 2050 and the sleeve to enhance the support strength. The lower end of the support column is connected to the flat base plate 2050 through a flange. The upper end of the support column 2051 is sleeved with a movable support sleeve 2052. Support arms 2053 are welded on both sides of the support sleeve 2052. The support arms 2053 are fixedly connected to the bottom of the annular track. A quick locking mechanism is provided on the outside of the support sleeve 2052 to adjust the height of the annular track in the axial direction of the support column 2051 and to perform positioning and locking. The annular track is used to carry a multimodal contactless recorder assembly, a data preprocessing module, a transmission module and a power supply system.

[0029] Furthermore, as an optional embodiment of the present invention, a multimodal non-contact recorder assembly is arranged on a circular track, with the position of the support column on the right side of the circular track as the origin, and each monitoring recorder is arranged in sequence in the counterclockwise direction of the circular track starting from the origin. The inner ring of the circular track is provided with a layout card point for positioning and installation. The layout card point is a plurality of grooves for assisting the installation of the non-contact recorder. Each monitoring recorder is arranged at the following angular positions of the circular track: the action frequency monitoring recorder is arranged at the 90° position of the circular track, and is used to project a laser beam at an incident angle of 45°±5° onto the valve stem of the black water angle valve to non-contactly monitor the number of opening and closing times, opening and closing cycles, and motion trajectory of the valve stem; The vibration monitoring recorders are symmetrically arranged at the 45° and 225° positions of the circular track. They adopt a 10° oblique laser incidence method and cooperate with the reflective lens adsorbed on the valve body surface of the black water angle valve to perform non-contact monitoring of the vibration mode, amplitude and frequency response of the key parts of the black water angle valve; the temperature monitoring recorder is arranged at the 135° position of the circular track, and its pitch angle can be adjusted from 30° to 45°. It is aimed at the entire valve body of the black water angle valve for non-contact monitoring of the surface temperature distribution, local hot spots and temperature gradient of the valve body; the noise monitoring recorders are arranged at the 15° and 255° positions of the circular track respectively, forming a 120° fan-shaped beam, which is used to collect the intensity and spectral characteristics of the cavitation noise and impact noise of the black water angle valve in real time.

[0030] Furthermore, as an optional embodiment of the present invention, the data preprocessing module, the transmission module and the power supply system are integrated and installed in a closed protective box, and the closed protective box is arranged at 170° outside the circular track; the multimodal non-contact recorder component is connected to the data preprocessing module and the transmission module inside the closed protective box through a shielded cable, and is connected to the power supply system through an independent power supply line; the power supply system includes a power supply device arranged in the closed protective box and a solar panel arranged on the outer ring of the circular track, and the power supply device and the solar panel form a power supply path through a photovoltaic charging module, which is used to provide power support for the multimodal non-contact recorder component, the data preprocessing module and the transmission module.

[0031] The embodiments of the present invention acquire the operating data of blackwater angle valves through a non-contact data collection platform, eliminating the need to physically contact or disassemble the valves to install contact sensors. This reduces the risk of equipment damage that may be introduced by sensor installation and avoids the impact of harsh environments such as high corrosion on the sensors, which can lead to lower data accuracy. This further improves the accuracy and stability of blackwater angle valve scrapping judgments. The local edge server directly preprocesses and performs initial monitoring and analysis on the collected operating data, reducing the time and network bandwidth required to transmit data to the cloud. Data can be quickly processed and analyzed locally, enabling real-time output of the blackwater angle valve's operating status, enhancing the real-time, reliability, and efficiency of blackwater angle valve scrapping judgments. Furthermore, by inputting target time series data and real-time operating status into a prediction model, the future state of the blackwater angle valve can be predicted in advance. Based on the prediction results, a more scientific and reasonable maintenance plan can be developed, shifting from reactive maintenance to proactive preventive maintenance, avoiding equipment downtime and production interruptions caused by sudden failures, reducing operation and maintenance costs, and extending the service life of the blackwater angle valve. In addition, the embodiment of the present invention uses the above-mentioned automated method to inspect the black water angle valve, which will not be affected by human subjective factors, improves the efficiency, real-time and accuracy of scrap judgment, reduces the missed detection rate, and improves the accuracy and reliability of scrap judgment.

[0032] Based on the same inventive concept, the embodiment of the present invention also provides a method for judging the scrapping of an in-service black water angle valve based on non-contact monitoring, such as Figure 4 As shown, Figure 4 A flowchart of a method for determining scrapping of an in-service black water angle valve based on non-contact monitoring is provided in an embodiment of the present invention, comprising the following steps: In step S401, the action signal, vibration characteristic signal, noise signal, and surface temperature signal of the blackwater angle valve are collected through the multimodal non-contact recorder component in the non-contact information collection platform. The signals collected by the multimodal non-contact recorder component are subjected to multi-channel synchronous sampling, filtering, time alignment, unit normalization, and unified timestamp marking through the data preprocessing module to form fused time series data. The fused time series data is transmitted to the local edge server in real time via LoRa or Wi-Fi through the transmission module.

[0033] Specifically, the frequency monitoring recorder, vibration monitoring recorder, noise monitoring recorder, and temperature monitoring recorder in the embodiments of the present invention can respectively employ a laser displacement sensor, a laser Doppler vibration sensor, a directional sound pickup sensor, and an infrared camera. The laser displacement sensor emits a laser aimed at the valve stem of the blackwater valve to collect the valve opening and closing operating signals, such as the number of openings and closings, the opening and closing cycles, and the movement amplitude of the valve stem. The laser Doppler vibration sensor emits a laser aimed at the valve surface of the blackwater valve to collect the vibration characteristic signals of the valve surface, such as the vibration amplitude, vibration frequency, and vibration displacement. The laser Doppler vibration sensor emits a laser aimed at the valve surface of the blackwater valve, receives the reflected light, and calculates the Doppler frequency shift, thereby accurately measuring the vibration amplitude, vibration frequency, and displacement changes of the valve surface. The laser emitting unit of the laser Doppler vibration sensor is fixed to a real-time non-contact data acquisition platform, emitting a stable laser beam toward the valve surface. A small, highly reflective mirror is attached to the target valve measurement point to enhance signal recovery. The photoelectric detection unit of the laser Doppler vibration sensor receives the reflected light signal at a specific angle (typically 5-20 degrees) to the laser emitting unit. A directional sound pickup sensor is used to directionally pick up noise signals generated during the operation of the blackwater valve, such as sound amplitude intensity, sound pressure level, and sound spectrum distribution data. The infrared camera's detection range covers the entire surface of the blackwater angle valve to collect the overall surface temperature distribution, local hotspot temperatures, and temperature gradients. In this way, multiple non-contact measurement devices are used to monitor the blackwater angle valve in real time, achieving multi-channel data collection and enhancing the comprehensiveness and accuracy of monitoring. By integrating features from multiple data sources, the robustness of blackwater angle valve anomaly detection can be improved and the false positive rate can be reduced.

[0034] In step S402, the fused time series data is received through the local edge server, features are extracted to obtain target time series data, and a first monitoring analysis is performed. The first monitoring analysis includes: using a sliding window method to obtain target time series data in different time scale windows, and extracting data features from multiple dimensions to determine the current operating status of the blackwater angle valve. After the local edge server completes the first monitoring analysis, the target time series data and analysis results are transmitted to the cloud server in real time through the built-in Wi-Fi module or 5G communication module of the local edge server.

[0035] Specifically, a sliding window method can be used to divide time series into short-term, medium-term, and long-term time scales, where the short-term time scale can be 1 hour, the medium-term time scale can be 1 day, and the long-term time scale can be 30 days. Of course, depending on the actual situation, the short-term, medium-term, and long-term time scales can also be other situations, and the embodiments of the present invention are not limited here. To improve data processing efficiency, the embodiments of the present invention standardize and normalize the target time series data obtained in different time scale windows using the sliding window method, so that the feature data of each time scale window is analyzed on the same dimension, thereby improving data processing efficiency and accuracy.

[0036] Furthermore, as an optional embodiment of the present invention, a sliding window method is used to obtain target time series data in different time scale windows, and data features are extracted from multiple dimensions to determine the current operating status of the black water angle valve, including: dividing the time series into a short-term time scale window, a medium-term time scale window, and a long-term time scale window; starting from the current moment of the black water angle valve, the first target time series data in the short-term time scale window, the second target time series data in the medium-term time scale window, and the third target time series data in the long-term time scale window are sequentially taken from the current moment and the historical time series before the current moment; and the first target time series data, the second target time series data, and the third target time series data are used to determine the current operating status of the black water angle valve.

[0037] Specifically, the embodiment of the present invention starts from the current moment of the blackwater angle valve and sequentially obtains characteristic data from multiple short-term time scale windows, medium-term time scale windows, and long-term time scale windows from the current moment and the historical time series before the current moment. Among them, the use of data from different time scale windows can capture the dynamic change characteristics of the blackwater angle valve operation data from multiple dimensions. The short time scale window can detect instantaneous faults in a timely manner, and the long time scale window is helpful for analyzing equipment operation trends, etc., to achieve a comprehensive and accurate assessment of the real-time operating status of the blackwater angle valve, effectively avoid the problems of missed judgment or misjudgment caused by single-scale analysis, and improve the accuracy of the blackwater angle valve scrapping judgment.

[0038] Furthermore, as an optional embodiment of the present invention, the current operating state of the black water angle valve is determined by using the first target time series data, the second target time series data and the third target time series data, including: using a dynamic time warping algorithm to respectively calculate the first similarity between the first target time series data and the second target time series data in the short-term time scale window at the current moment, and the second similarity between the first target time series data and the third target time series data in the short-term time scale window at the current moment; when the average of the first similarity and the second similarity is less than the first threshold value, determining that the real-time operating state of the black water angle valve is an abnormal state. Or, The K-means clustering method is used to cluster the first target time series data, the second target time series data, and the third target time series data to obtain cluster clusters of short-term time scale windows, cluster clusters of medium-term time scale windows, and cluster clusters of long-term time scales. The cluster clusters include sub-clusters of different operating modes of the blackwater angle valve; the Euclidean distance between the data point at the current moment and the cluster center of the sub-cluster of the time scale window corresponding to the data point is calculated; when the Euclidean distance is greater than a second threshold, the real-time operating status of the blackwater angle valve is determined to be an abnormal state.

[0039] Alternatively, key parameters in the first target time series data, the second target time series data, and the third target time series data are obtained; if the key parameters in the first target time series data, the second target time series data, and the third target time series data are abnormal at the same time, the real-time operating status of the black water angle valve is determined to be an abnormal state.

[0040] Specifically, the embodiment of the present invention uses a dynamic time warping (DTW) algorithm to align the first target time series data, the second target time series data, and the third target time series data and calculates the first similarity between the first target time series data and the second target time series data, and the second similarity between the first target time series data and the third target time series data in the short-term time scale window at the current moment. Among them, the principle of using the DTW algorithm to calculate the similarity can refer to the known technology, and the embodiment of the present invention will not be repeated here. After calculating the first similarity and the second similarity, the average of the first similarity and the second similarity is calculated. The first threshold can be determined according to the actual situation. In the embodiment of the present invention, the value is 0.7. If the average of the first similarity and the second similarity is less than 0.7, it can be considered that the real-time operating status of the black water angle valve is abnormal.

[0041] Furthermore, the embodiment of the present invention adopts the K-means clustering method to cluster the first target time series data, the second target time series data and the third target time series data respectively, and classify the operation modes of the target series data at different time scales. Among them, the operation mode includes normal operation mode, mild fluctuation operation mode, high frequency opening and closing operation mode, cavitation tendency operation mode, structural looseness or abnormal wear mode, sealing failure / leakage mode, etc. The embodiment of the present invention classifies the target time series data under each time scale window according to the operation mode, and then calculates the Euclidean distance between the data point at the current moment and the cluster center of the sub-cluster of the time scale window corresponding to the data point, that is, first determine the time scale window where the data point at the current moment is located, then determine the operation mode of the black water angle valve corresponding to the data point at the current moment, and determine the Euclidean distance between the operation mode of the black water angle valve corresponding to the data point at the current moment and the cluster center of the corresponding sub-cluster.

[0042] Furthermore, in this embodiment of the present invention, if multiple key parameters are simultaneously abnormal within the short-term (one hour), medium-term (one day), and long-term (30 days) time scales, the real-time operating status of the blackwater angle valve is determined to be abnormal, triggering an abnormality warning. Key parameters include, but are not limited to, characteristic data collected by various non-contact sensors, such as the frequency of opening and closing movements, vibration amplitude and frequency, noise intensity and spectrum, and temperature distribution and gradient.

[0043] Furthermore, when the real-time operating status of the black water angle valve is abnormal, in response to the user's first input, the first target time series data, the second target time series data and the third target time series data are sent to the cloud, and the cloud uses the first target time series data, the second target time series data and the third target time series data to train the prediction model.

[0044] Specifically, users can view anomaly details in the visualization interface and select: "Normal": If the user confirms that the data is normal, the first, second, and third target time series data will be added to the prediction model update pool, with a weight of 0.5 during training to prevent new data from excessively influencing the model. If "Abnormal": If the user confirms an anomaly, the first, second, and third target time series data will be uploaded to the cloud analysis center for further training and model optimization.

[0045] Furthermore, when the real-time operating status of the black water angle valve is abnormal, the lengths of the short-term time scale, the medium-term time scale and the long-term time scale are reduced; the first monitoring analysis and the second monitoring analysis are performed again on the black water angle valve using the reduced short-term time scale, the medium-term time scale and the long-term time scale.

[0046] Specifically, embodiments of the present invention utilize a longer timescale window when the device is operating stably to reduce computational burden and enhance long-term trend analysis capabilities. When blackwater angle valve anomalies occur frequently, window shortening logic is triggered, adjusting the window sizes for the short-term, medium-term, and long-term timescales. Computational parameters are updated in the local edge server to more quickly capture short-term anomalies in the blackwater angle valve and improve fault response speed. If no anomalies are triggered for N consecutive timescale windows, the default timescale window is restored. The computational parameters include the number of clusters, first threshold, second threshold, and timescale window size for the operating mode described in the above embodiments. Thus, in embodiments of the present invention, when the blackwater angle valve is operating stably, a longer timescale window is utilized to reduce the frequency of anomaly warning triggering and false alarms. If blackwater angle valve anomalies occur frequently, the timescale window is shortened (e.g., from 1 hour to 30 minutes) to improve detection accuracy.

[0047] Step S403, using the cloud server to perform a second monitoring analysis, including: based on the target time series data and the analysis results of the local edge server, using a prediction model to perform trend prediction on the future operating status of the blackwater angle valve; based on the prediction results of the future operating status and the current operating status of the blackwater angle valve, identifying the potential fault type of the blackwater angle valve through a fault classification model, and calculating the health index and remaining life of the blackwater angle valve based on the life assessment model, outputting the potential fault type, health index and remaining life through a visual interactive interface and generating automatic retirement recommendations.

[0048] Specifically, in embodiments of the present invention, the future state of a blackwater angle valve includes, but is not limited to, its deterioration trend, health index, and remaining lifespan. The prediction model can employ a long-short-term memory (LSTM) network model. The fault classification model can be a combination of a clustering model and a classification model. The clustering model can be a K-means clustering model or a density-based spatial clustering of applications with noise (DBSCAN) clustering model. The classification model can be a support vector machine (SVM) or a random forest model. The lifespan prediction model can be a Markov degradation model.

[0049] Furthermore, as an optional embodiment of the present invention, based on the analysis results of the target time series data and the local edge server, a prediction model is used to perform trend prediction on the future operating status of the blackwater angle valve, including: inputting the target time series data of the blackwater angle valve for the past multiple days into the prediction model, using the prediction model to predict the key parameter values ​​of the blackwater angle valve at future moments based on the target time series data of the blackwater angle valve for the past multiple days, and when the absolute value of the difference between the key parameter value at the future moment and the actual key parameter value at the current moment is greater than a third threshold, determining that the blackwater angle valve has a deterioration trend; based on the prediction result of the future operating status and the current operating status of the blackwater angle valve, identifying the potential fault type of the blackwater angle valve through a fault classification model The method includes: inputting the target time series data of the blackwater angle valve in the past few days, the current operating status of the local edge server, and the prediction results of the future operating status into the fault classification model, and using the fault classification model to predict the fault type of the blackwater angle valve through the target time series data of the blackwater angle valve in the past few days, the current operating status, and the prediction results of the future operating status, and the fault type includes valve body wear failure and / or sealing failure failure; calculating the health index and remaining life of the blackwater angle valve based on the life assessment model includes: constructing a life assessment module based on the improved Weibull-particle filtering method, and using the life assessment model to dynamically calculate the health index and confidence interval of the blackwater angle valve, and generating the remaining life of the blackwater angle valve through Monte Carlo simulation.

[0050] Specifically, the embodiment of the present invention first performs data fusion on the target time series data of multiple data sources, specifically using interpolation and DTW to align the time series data, eliminate outliers, and standardize the data to eliminate the time differences between different data sources and improve data consistency. And perform more detailed feature parameter extraction on the data. Among them, the first prediction model can be trained based on the historical operation data of the blackwater angle valve, such as using the historical operation data of the blackwater angle valve collected by various non-contact devices in the past N days to train the first prediction model. Then, the operation data of the past period of time is used to predict some key parameter values ​​of the blackwater angle valve in the future. If the deviation between the key parameter value at the future moment and the actual key parameter value collected by each non-contact device at the current moment exceeds the third threshold, it indicates that the blackwater angle valve has a deterioration trend and triggers an early warning. Among them, the third threshold can be determined according to the actual situation, and the embodiment of the present invention is not limited here.

[0051] Furthermore, the embodiment of the present invention combines regression analysis to calculate the predicted rate of change of the key parameter value at the future moment. If the rate of change of the key parameter value at the future moment is greater than the fourth threshold value and the absolute value of the difference between the key parameter value at the future moment and the actual key parameter value at the current moment is greater than the third threshold value, it is determined that the black water angle valve has a deterioration trend. Among them, the fourth threshold value can be determined according to the actual scenario, and the embodiment of the present invention does not limit it here. In this way, judging solely based on the rate of change of the key parameter value or the difference with the actual value at the current moment is prone to misjudgment due to data fluctuations or abnormal values. Setting the third threshold and the fourth threshold at the same time requires that the rate of change and the absolute value of the difference meet the conditions at the same time before determining that there is a deterioration trend, thereby significantly reducing the probability of misjudgment.

[0052] Furthermore, the embodiment of the present invention can adopt a method combining clustering and classification when diagnosing the fault of the black water angle valve. First, the K-means or DBSCAN clustering method is used to calculate the cluster centers of the sub-cluster clusters of the operating modes in different time scale windows based on the collected and pre-processed target time series data. The current data points are classified according to the operating status by comparing the Euclidean distance between the current data points and the cluster centers, that is, the real-time operating status of the black water angle valve corresponding to the current data points is determined. If the real-time operating status of the black water angle valve is abnormal, the SVM classification model or the random forest model is used for fault classification. For example, the vibration amplitude, frequency, noise spectrum distribution, temperature change and other characteristic parameters of the current data points are used to accurately determine the current fault type of the black water angle valve, wherein the fault type includes but is not limited to specific fault types such as sealing failure, cavitation, internal leakage, etc., so as to achieve comprehensive and accurate diagnosis of black water angle valve faults.

[0053] Furthermore, the Markov degradation model calculates the health index of the blackwater angle valve based on the future key parameter values ​​predicted by the prediction model (LSTM neural network model). It also combines the target time series data of the blackwater angle valve from various data sources over the past several days to predict the valve's life decay curve and generate the valve's remaining life. Specifically, the LSTM neural network model processes the target time series data of the blackwater angle valve over the past several days. It inputs multi-source target time series data, such as vibration, temperature, and noise, for training. By learning from historical data, the LSTM neural network model captures long-term dependencies and complex patterns in the target time series data, and thus predicts the future key parameter values ​​of the blackwater angle valve, providing a basis for subsequent remaining life prediction of the blackwater angle valve. The Markov degradation model calculates the health index based on the future key parameter values ​​predicted by the LSTM neural network model. It analyzes multiple data sources, such as vibration amplitude and temperature rise rate, quantifying these factors and incorporating them into the Markov degradation model's calculations. These factors reflect the changes in the operating status of the blackwater angle valve. Through the operation of the Markov degradation model, a numerical value that can comprehensively evaluate the current health of the blackwater angle valve is obtained, namely the health index. After calculating the health index, the Markov degradation model predicts the life decay curve of the blackwater angle valve based on the changes in the health index. The Markov degradation model takes into account that the health status of the blackwater angle valve will gradually deteriorate over time. In combination with the weight of the impact of various factors on the remaining life of the blackwater angle valve, it simulates the changes in the health status of the blackwater angle valve at different time points in the future. When the health index drops to a certain level, or reaches a specific threshold according to the life decay curve, the remaining life of the blackwater angle valve can be determined.

[0054] Furthermore, after obtaining the current operating status and future status of the black water angle valve, the black water angle valve can be judged as scrapped. As an optional embodiment of the present invention, the method also includes: if any one of the following conditions is met, an automatic retirement recommendation for the black water angle valve is generated and displayed; the conditions include: the average of the first similarity and the second similarity is less than the first threshold; the Euclidean distance between the data point at the current moment and the cluster center of the cluster in the time scale window corresponding to the data point is greater than the second threshold; the health index is less than the third threshold; the remaining life is less than or equal to the fourth threshold.

[0055] Specifically, if any of the above conditions are met, the Blackwater Angle Valve's decommissioning assessment process begins. The first threshold can be 0.7, and the second threshold can be 1.5 standard deviations. The third threshold can be 30% of the initial health index, and the fourth threshold can be 30 days. Furthermore, the Blackwater Angle Valve's decommissioning assessment process can also include triggering ≥3 anomalies within a short-term timeframe (1 hour), ≥5 anomalies within a medium-term timeframe (1 day), ≥2 recurrences of the same fault within the past 3 months, or multiple confirmations of equipment failure by operations and maintenance personnel.

[0056] In addition, if the black water angle valve meets the condition that the remaining life is less than or equal to the fourth threshold and the warning is triggered ≥3 times in the past 7 days, the remaining life of the black water angle valve will be corrected, that is, it is recommended to retire the black water angle valve early.

[0057] Furthermore, this embodiment of the present invention generates a recommendation for the blackwater angle valve's scrapping and decommissioning, displaying it on a visualization page for users to review. The visualization page displays data such as the valve's health index, remaining lifespan (estimated decommissioning time), and abnormality frequency, as well as past abnormalities. Maintenance recommendations are also displayed, offering possible repair solutions, such as replacing components or adjusting operating parameters.

[0058] Furthermore, if the blackwater angle valve continues to malfunction after maintenance, the confidence level in the remaining life of the blackwater angle valve will be lowered. If the remaining life of the blackwater angle valve is less than 30 days and the frequency of malfunctions exceeds a set threshold (e.g., three warnings triggered in the past seven days), an early retirement recommendation will be automatically generated.

[0059] The present invention employs a sliding window approach to acquire target time series data within windows of varying time scales, capturing the dynamic characteristics of blackwater angle valve operational data from multiple dimensions. Short-time scale windows can promptly detect transient faults, while long-time scale windows facilitate analysis of equipment operational trends. This enables a comprehensive and accurate assessment of the real-time operational status of blackwater angle valves, effectively avoiding missed or misjudgment errors caused by single-scale analysis and improving the accuracy of blackwater angle valve scrapping. Acquiring blackwater angle valve operational data through a non-contact acquisition platform eliminates the need to physically access or disassemble the valves to install contact sensors, reducing the risk of equipment damage associated with sensor installation and avoiding the potential for lower data accuracy due to the effects of harsh environments such as high corrosion on the sensors. This further improves the accuracy and stability of blackwater angle valve scrapping. The local edge server directly pre-processes and performs primary monitoring and analysis on the collected operational data, reducing the time and network bandwidth required to transmit data to the cloud. Data can be rapidly processed and analyzed locally, enabling real-time output of the blackwater angle valve's operational status, enhancing the real-time, reliability, and efficiency of blackwater angle valve scrapping. Furthermore, by inputting target time series data and real-time operating status into a predictive model, the future status of the blackwater angle valve can be predicted in advance. Based on these predictions, a more scientific and rational maintenance plan can be developed, shifting from reactive maintenance to proactive preventative maintenance. This avoids equipment downtime and production interruptions caused by sudden failures, reduces operational costs, and extends the service life of the blackwater angle valve. Furthermore, the embodiments of the present invention utilize this automated approach to inspect blackwater angle valves, eliminating the influence of subjective human factors and improving the efficiency and accuracy of scrap determination.

Claims

1. A scrapping system for in-service black water angle valves based on non-contact monitoring, characterized in that: include: Contactless information collection platform, local edge server and cloud server; The non-contact information collection platform includes: a ring track, a support structure, a multi-modal non-contact recorder component, a data pre-processing module and a transmission module and a power supply system; The annular track is arranged around the black water angle valve and is fixedly connected to the pipe connected to the black water angle valve through a supporting structure; The multimodal non-contact recorder assembly includes an action frequency monitoring recorder, a vibration monitoring recorder, a noise monitoring recorder, and a temperature monitoring recorder, which are installed at intervals on the annular track and are used to collect the opening and closing action signals, vibration characteristic signals, noise signals, and surface temperature signals of the black water angle valve respectively; The data preprocessing module, the transmission module, and the power supply system are integrated and installed in a closed protective box of the circular track. The data preprocessing module and the transmission module are equipped with a microcontroller unit for performing multi-channel synchronous sampling, filtering, time alignment, unit normalization, and unified time stamp marking on the signals collected by the multimodal contactless recorder component to form fused time series data. The fused time series data is transmitted to the local edge server in real time via LoRa or Wi-Fi; The local edge server is used to receive the fused time series data, perform feature extraction on the fused time series data, obtain target time series data, and perform a first monitoring analysis; the first monitoring analysis includes: using a sliding window method to obtain target time series data in different time scale windows, and extracting data features from multiple dimensions to determine the current operating status of the blackwater angle valve; after the local edge server completes the first monitoring analysis, the target time series data and analysis results are transmitted to the cloud server in real time via a built-in Wi-Fi module or a 5G communication module; The cloud server is used to perform a second monitoring analysis, including: based on the fusion of time series data and the analysis results of the local edge server, using a prediction model to predict the trend of the future operating status of the blackwater angle valve; based on the prediction results of the future operating status and the current operating status of the blackwater angle valve, identifying the potential fault type of the blackwater angle valve through a fault classification model, and calculating the health index and remaining life of the blackwater angle valve based on the life assessment model, outputting the potential fault type, the health index and the remaining life through a visual interactive interface and generating an automatic retirement recommendation.

2. The scrap judgment system for in-service black water angle valves based on non-contact monitoring according to claim 1 is characterized in that: The diameter D of the annular track satisfies: 200 mm ≤ D − Dv ≤ 400 mm, Dv is the maximum outer diameter of the black water angle valve, and the minimum distance between the annular track and the black water angle valve is not less than 50 mm.

3. The scrap judgment system for in-service black water angle valves based on non-contact monitoring according to claim 1 is characterized in that: The support structure includes a support column and a sleeve installed on the outer wall of the pipe connected to the black water angle valve. The sleeve is fixed to the outer surface of the pipe by a clamp. A flat base plate is welded above the sleeve. A triangular reinforcement plate is provided between the flat base plate and the sleeve to enhance the support strength. The lower end of the support column is connected to the planar substrate through a flange, and a movable support sleeve is sleeved on the upper end of the support column. Support arms are welded on both sides of the support sleeve, and the support arms are fixedly connected to the bottom of the annular track. A quick locking mechanism is provided on the outside of the support sleeve to adjust the height of the annular track in the axial direction of the support column and to lock the position. The annular track is used to carry the multimodal contactless recorder assembly, the data preprocessing module, the transmission module and the power supply system.

4. The scrap judgment system for in-service black water angle valves based on non-contact monitoring according to claim 1 is characterized in that: The multimodal non-contact recorder assembly is arranged on a circular track. The position of the support column on the right side of the circular track is used as the origin. Starting from the origin, each monitoring recorder is arranged in sequence in the counterclockwise direction of the circular track. The inner circle of the circular track is provided with layout card points for positioning and installation. The layout card points are a plurality of grooves for assisting the installation of the non-contact recorder. Each monitoring recorder is arranged at the following angular positions on the circular track: The action frequency monitoring recorder is arranged at the 90° position of the circular track, and is used to project a laser beam at an incident angle of 45°±5° onto the valve stem of the black water angle valve, and monitor the opening and closing times, opening and closing cycles and movement trajectory of the valve stem in a non-contact manner; The vibration monitoring recorders are symmetrically arranged at the 45° and 225° positions of the circular track, and adopt a 10° oblique laser incidence method, in conjunction with a reflective lens adsorbed on the valve body surface of the black water angle valve, to non-contact monitor the vibration mode, amplitude and frequency response of the key parts of the black water angle valve; The temperature monitoring recorder is arranged at the 135° position of the circular track, with an adjustable pitch angle range of 30° to 45°, aimed at the entire valve body of the black water angle valve, and is used for non-contact monitoring of the surface temperature distribution, local hot spots and temperature gradient of the valve body; Noise monitoring recorders are respectively arranged at 15° and 255° positions of the circular track to form a 120° fan-shaped beam for real-time collection of the intensity and spectral characteristics of the cavitation noise and impact noise of the black water angle valve.

5. The scrap judgment system for in-service black water angle valves based on non-contact monitoring according to claim 1 is characterized in that: The data preprocessing module, the transmission module and the power supply system are integrated and installed in a closed protective box, and the closed protective box is arranged at a position 170° outside the circular track; the multimodal non-contact recorder component is connected to the data preprocessing module and the transmission module inside the closed protective box through a shielded cable, and is connected to the power supply system through an independent power supply line; the power supply system includes a power supply device arranged in the closed protective box and a solar panel arranged on the outer ring of the circular track. The power supply device and the solar panel form a power supply path through a photovoltaic charging module, which is used to provide power support for the multimodal non-contact recorder component, the data preprocessing module and the transmission module.

6. A method for judging the scrapping of in-service black water angle valves based on non-contact monitoring, characterized in that: The scrapping judgment system for in-service black water angle valves based on non-contact monitoring according to any one of claims 1 to 5 comprises the following steps: The multimodal non-contact recorder component in the non-contact information collection platform is used to collect the action signal, vibration characteristic signal, noise signal and surface temperature signal of the black water angle valve. The signals collected by the multimodal non-contact recorder component are subjected to multi-channel synchronous sampling, filtering, time alignment, unit normalization and unified time stamp marking by the data preprocessing module to form fused time series data. The fused time series data is transmitted to the local edge server in real time via LoRa or Wi-Fi by the transmission module; The fused time series data is received by a local edge server, features are extracted from the data to obtain target time series data, and a first monitoring analysis is performed; the first monitoring analysis includes: using a sliding window method to obtain target time series data in different time scale windows, and extracting data features from multiple dimensions to determine the current operating status of the blackwater angle valve; after the local edge server completes the first monitoring analysis, the target time series data and analysis results are transmitted to the cloud server in real time via a built-in Wi-Fi module or 5G communication module of the local edge server; The cloud server is used to perform a second monitoring and analysis, including: based on the target time series data and the analysis results of the local edge server, using a prediction model to predict the trend of the future operating status of the blackwater angle valve; based on the prediction results of the future operating status and the current operating status of the blackwater angle valve, identifying the potential fault type of the blackwater angle valve through a fault classification model, and calculating the health index and remaining life of the blackwater angle valve based on the life assessment model, outputting the potential fault type, the health index and the remaining life through a visual interactive interface and generating an automatic retirement recommendation.

7. The method for judging the scrapping of an in-service black water angle valve according to claim 6 is characterized in that: The sliding window method is used to obtain target time series data in different time scale windows, and data features are extracted from multiple dimensions to determine the current operating status of the blackwater angle valve. Divide the time series into short-term time scale windows, medium-term time scale windows and long-term time scale windows; Starting from the current moment of the black water angle valve, sequentially obtain the first target time series data in the short-term time scale window, the second target time series data in the medium-term time scale window, and the third target time series data in the long-term time scale window from the current moment and the historical time series before the current moment; The current operating state of the black water angle valve is determined using the first target time series data, the second target time series data, and the third target time series data.

8. The method for judging the scrapping of an in-service black water angle valve according to claim 7 is characterized in that: Determining the current operating state of the black water angle valve by using the first target time series data, the second target time series data, and the third target time series data includes: A dynamic time warping algorithm is used to calculate a first similarity between the first target time series data and the second target time series data in the short-term time scale window at the current moment, and a second similarity between the first target time series data and the third target time series data in the short-term time scale window at the current moment; When the average of the first similarity and the second similarity is less than a first threshold, determining that the real-time operating state of the black water angle valve is an abnormal state; or, The first target time series data, the second target time series data, and the third target time series data are clustered using a K-means clustering method to obtain clusters of a short-term time scale window, clusters of a medium-term time scale window, and clusters of a long-term time scale, wherein the clusters include sub-clusters of different operating modes of the blackwater angle valve; Calculating the Euclidean distance between the data point at the current moment and the cluster center of the sub-cluster of the time scale window corresponding to the data point; When the Euclidean distance is greater than a second threshold, determining that the real-time operating state of the black water angle valve is an abnormal state; Alternatively, obtaining key parameters in the first target time series data, the second target time series data, and the third target time series data; If the key parameters in the first target time series data, the second target time series data, and the third target time series data are all abnormal at the same time, it is determined that the real-time operating state of the black water angle valve is an abnormal state.

9. The method for judging the scrapping of an in-service black water angle valve according to claim 6 is characterized in that: The method of using a prediction model to perform trend prediction on the future operating status of the blackwater angle valve based on the target time series data and the analysis results of the local edge server includes: inputting target time series data of the black water angle valve over the past multiple days into a prediction model, using the prediction model to predict key parameter values ​​of the black water angle valve at future moments based on the target time series data of the black water angle valve over the past multiple days, and determining that the black water angle valve has a deterioration trend when the absolute value of the difference between the key parameter value at the future moment and the actual key parameter value at the current moment is greater than a third threshold; The prediction result based on the future operating state and the current operating state of the black water angle valve is used to identify the potential fault types of the black water angle valve through the fault classification model, including: Inputting target time series data of the black water angle valve over the past several days, the current operating state of the local edge server, and the prediction result of the future operating state into a fault classification model, and using the fault classification model to predict the fault type of the black water angle valve through the target time series data of the black water angle valve over the past several days, the current operating state, and the prediction result of the future operating state, the fault type including valve body wear fault and / or seal failure fault; The calculation of the health index and remaining life of the black water angle valve based on the life assessment model includes: The life assessment module is constructed based on the improved Weibull-particle filtering method, and the health index and confidence interval of the black water angle valve are dynamically calculated using the life assessment model, and the remaining life of the black water angle valve is generated through Monte Carlo simulation.

10. The method for judging the scrapping of an in-service black water angle valve according to claim 6, characterized in that: The method further includes: if any one of the following conditions is met, generating and displaying an automatic decommissioning recommendation for the black water angle valve; The conditions include: The average of the first similarity and the second similarity is less than a first threshold; The Euclidean distance between the data point at the current moment and the cluster center of the cluster in the time scale window corresponding to the data point is greater than a second threshold; The health index is less than the third threshold; The remaining lifespan is less than or equal to a fourth threshold.

Images