Equipment fault feature recognition and inspection system based on pressure vessel
By using a multi-dimensional fault monitoring system that combines torque and corrosion parameters, the valve status of pressure vessels can be monitored in real time. This solves the problem of unpredictable fastening force decay and corrosion morphology in existing technologies, enabling early warning of pressure vessel faults and improving the safety and reliability of the system.
Patent Information
- Application Number
- CN202511593670.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-03
- Publication Date
- 2026-01-02
AI Technical Summary
Existing technologies cannot monitor the decay trend of the fastening force of pressure vessel valves in real time, nor can they predict the corrosion morphology, resulting in insufficient safety and reliability of pressure vessels when operating under complex conditions.
A multi-dimensional fault monitoring system is adopted, which combines torque and corrosion parameters, collects multi-condition data through sensors, constructs a dynamic threshold comparison and fault early warning module, and uses image recognition technology to provide fault early warning, thereby realizing a comprehensive quantitative evaluation of the valve.
It enables early fault warning for pressure vessel valves, improving system safety and operational reliability.
Smart Images

Figure CN121253149A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of pressure vessel testing, specifically a pressure vessel-based equipment fault feature identification and inspection system. Background Technology
[0002] A pressure vessel is a sealed device that holds gas or liquid under pressure. Pressure vessel valves are mechanical devices installed on pressure vessels, pressurized equipment, and their connecting pipelines to control the flow direction, pressure, and flow rate of the medium or to achieve safety protection. The safe operation of a pressure vessel depends heavily on the reliable sealing of its valve connections.
[0003] Currently, valve tightening mainly relies on operators using torque wrenches and periodic manual inspections after completion. This method has significant drawbacks: it cannot monitor in real time. The initial tightening torque decays due to vibration, thermal cycling, pressure fluctuations, and material creep, and manual inspection cannot capture this dynamic process. Torque and seal are not directly correlated; a simple torque value cannot fully reflect the actual compression state of the sealing surface. Fluctuations in the coefficient of friction, thread damage, and other factors can cause the same torque to produce different sealing forces. Existing technologies cannot predict the decay trend of tightening force and cannot provide early warnings before failures occur. Furthermore, pressure vessel valves operate under complex conditions for extended periods and are susceptible to media corrosion, electrochemical corrosion, and environmental factors, resulting in various corrosion forms, including uniform corrosion, pitting corrosion, and stress corrosion cracking. Traditional methods often rely on manual visual inspection or localized ultrasonic thickness measurement, which cannot comprehensively reflect the overall corrosion status of the valve. This application aims to combine torque parameters, which reflect the mechanical fastening state of valves, with corrosion parameters, which reflect the material degradation state of valves, to perform dual-dimensional fault monitoring of valves. By analyzing the trend of effective fastening torque and integrating corrosion data, a comprehensive quantitative assessment of valves is achieved, enabling early warning of valve failures and improving the overall safety and operational reliability of pressure vessel systems. Summary of the Invention
[0004] The purpose of this invention is to provide a system for identifying and inspecting equipment fault characteristics based on pressure vessels, in order to solve the problems in the prior art.
[0005] To achieve the above objectives, the present invention provides the following technical solution: A pressure vessel-based equipment fault feature identification and inspection system includes a pressure vessel multi-condition identification module, a pressure vessel valve basic fault feature identification and analysis module, a pressure vessel valve dynamic torque online monitoring and analysis module, a pressure vessel valve corrosion state multi-modal detection module, and a pressure vessel valve fault urgency assessment and prediction module. The pressure vessel multi-condition identification module uses several sensors to collect multiple parameters of the pressure vessel's operating conditions, dynamically identifies data under different operating conditions, dynamically adjusts the evaluation threshold according to the current operating conditions, performs dynamic multi-parameter threshold comparison, and triggers corresponding faults for classification, early warning, and alarm based on the comparison and verification results. The pressure vessel valve basic fault feature identification and analysis module collects the operating status and images of the pressure vessel valve. Based on the monitoring data and images collected by several sensors during the valve opening and closing process, it performs preprocessing. Based on the preprocessed monitoring data and images, it detects micro-fault defect features, identifies basic fault defect features, and performs fault assessment and early warning based on the degree of feature deviation. The pressure vessel valve dynamic torque online monitoring and analysis module continuously monitors the dynamic torque of pressure vessel valves, constructs valve operating torque curves, performs quantitative analysis based on the monitored valve torque signals and torque growth rates within different cycles, and simultaneously constructs a dynamic attenuation model of effective valve fastening torque to analyze valve fastening force attenuation and provide proactive fault warning. The multi-modal detection module for corrosion status of pressure vessel valves acquires pre-processed images of valves exhibiting abnormalities, extracts feature parameters from the corrosion images on the valve surface for intelligent evaluation, monitors environmental data around the valve for combined analysis, analyzes the overall corrosion status of the valve, and outputs a corresponding risk index based on the overall corrosion status for fault warning. The pressure vessel valve failure urgency assessment and prediction module summarizes the dynamic torque monitoring and corrosion status monitoring data of the pressure vessel valves, and constructs a valve maintenance urgency classification framework based on the summarized data to provide graded failure early warning.
[0006] Further configuration: The pressure vessel multi-condition identification module includes a pressure vessel multi-parameter synchronous acquisition submodule and a multi-parameter threshold identification and early warning model construction submodule. The pressure vessel multi-parameter synchronous acquisition submodule includes an adaptive sensor management and monitoring unit and a data acquisition and aggregation unit. The adaptive sensor management and monitoring unit uses several sensors to monitor and collect pressure vessel operating parameters. The sensor types are selected and set by the operator. The collected parameters include the internal transmission medium of the pressure vessel, pressure, ambient temperature and humidity, ambient gas concentration, and valve wall thickness. The data acquisition and aggregation unit preprocesses the collected raw data, aligns all data streams on the time axis to ensure consistency of data at the same time, and classifies and summarizes the different detection data according to different detection categories. The multi-parameter threshold identification and early warning model construction submodule obtains the parameter thresholds corresponding to different detection categories, obtains the detection data under different monitoring operating condition categories of the pressure vessel, and obtains the peak value and mean value of the data for each category. The peak value of the data detected for a certain category is set as... The mean is When both the peak and average values of the detection data for a given category exceed the corresponding parameter threshold, a fault alarm for that category is triggered. When both the peak and average values of the detection data for a given category are less than the corresponding parameter threshold, it indicates that the detection data for that category is within the safety standard. When the average value of the detection data for a given category is less than the corresponding parameter threshold, but the peak value is greater than the corresponding parameter threshold, the corresponding parameter threshold is set to [value missing]. Introduce a safety weight coefficient for this category. Set the threshold for the security controllable parameters corresponding to this category as follows: The safety weight coefficient is set manually, and the dynamic safety peak range for the detection data of this category is defined, with the safety peak range set as follows: Screen the test data for this category, according to the formula: When the extreme values within the dynamic safety peak range of the detection data for this category do not meet the above formula, a fault alarm for the corresponding category is triggered. When the extreme values within the dynamic safety peak range of the detection data for this category all meet the above formula, the number of times the detection data for this category exceeds the corresponding parameter threshold within the dynamic safety peak range is analyzed. When the number exceeds the set threshold, a warning is uploaded for the corresponding category fault. When the number is less than the set threshold, it indicates that the detection data for this category is within the safety standard. The set thresholds are all preset by the operator.
[0007] Further configuration: The pressure vessel valve basic fault feature identification and analysis module includes a valve multi-dimensional feature acquisition submodule and a basic fault feature classification, identification, and evaluation submodule. The valve multi-dimensional feature acquisition submodule includes a sensor signal acquisition unit and a valve image acquisition unit. The sensor signal acquisition unit includes a torque sensor, a temperature and humidity sensor, and an ultrasonic sensor, which respectively acquire valve status parameters, including valve temperature and humidity, operating torque, and valve stem displacement data. The valve image acquisition unit includes several industrial cameras, which are used to acquire overall valve appearance images and partial feature appearance images, including valve stem, valve disc, handwheel, and flange connection. The acquired raw detection data and images are preprocessed to match the detected valve parameters with the valve's overall appearance. The door images undergo spatial alignment and temporal matching. Valve parameter data and image data are aggregated according to different valve numbers and sent to the basic fault feature classification and evaluation submodule. This submodule acquires the fluctuation characteristics of different valve category detection parameters and different valve images acquired by the valve image acquisition unit. Based on the different valve category detection parameters, it obtains the positive and negative peak values of the detection parameters and compares them with the corresponding set threshold range. If the positive and negative peak values are not within the set threshold range, fault identification and early warning are performed based on the data detection for that category. If both the positive and negative peak values are within the set threshold range, the fluctuation amplitude of the detection parameter is identified, and the detection parameter is set as follows: The positive and negative peak values of this detection parameter were screened out as follows: and After removing the positive and negative peak values from the detection equipment, the remaining data are summarized, and the truncated mean is calculated using the remaining data. , ,in The data for valves is collected in real time, and the standard deviation is calculated using the remaining data. Analyze the fluctuation and dispersion coefficient of the detection data for this category, and set the fluctuation and dispersion coefficient of the detection data for this category as . : Based on the fluctuation and dispersion coefficient of the detection data for this category, the stability of the real-time detection data for this category is analyzed, and the stability coefficient of the detection data is set as follows: According to the formula: when A value of 1 indicates that the real-time detection data for this category is highly stable. A smaller value indicates that the real-time monitoring of that category is less stable. When the stability coefficient of the monitoring data is less than a set threshold, the real-time monitoring data for that valve category is marked and uploaded for fault warning. Factors affecting data fluctuations; Based on the different valve images acquired by the valve image acquisition unit, each acquired image is compared with the fault map. According to the fault map uploaded by the administrator, each acquired image is marked with valve color features, defect areas, and structural anomalies. By analyzing the valve color features, areas with obvious rust spots are identified and marked as having valve color anomalies. By monitoring the defect areas, images with valve defects are marked as having abnormal contours or deformations. By detecting structural anomalies, images with abnormal bolts, aging seals, or abnormal connections are marked as having valve anomalies. The valve numbers corresponding to the images with identified and marked faults are analyzed and warned of basic valve faults.
[0008] Further configuration: The pressure vessel valve dynamic torque online monitoring and analysis module includes a dynamic torque signal parameter monitoring submodule and a valve effective tightening torque dynamic attenuation analysis submodule. The dynamic torque signal parameter monitoring submodule includes several torque sensors, which synchronously collect the valve's torque value, capture the valve's dynamic torque data in real time, extract peak torque and mean torque data from the dynamic torque signal, construct the valve's torque curve, obtain the reference torque value of the valve's rated operating condition, perform dynamic torque signal detection on the valve, obtain the peak torque collected and monitored for a specific valve within the current operating cycle, calculate the average value and standard deviation of the peak torque in the data of different operating cycles based on the monitored peak torque signals within different cycles, and set the peak torque monitored in the current operating cycle as... The average peak torque value is set as the average value in the data from different operating cycles. and standard deviation Calculate the rate of increase of valve torque within the current cycle, and set the rate of increase of torque within the current cycle as... According to the formula: In the above formula, The initial torque at the start of the current operating cycle. This refers to the time elapsed from the start of the current operating cycle to reaching peak torque. The start time of the current operation cycle is marked as the baseline time point. Quantitative analysis is performed based on the valve torque data collected within different cycles, according to the formula: in, This is the standard deviation factor, which is set manually. The threshold for the torque growth rate within the current cycle is defined as follows: if the peak torque and torque growth rate monitored in the current operating cycle satisfy the above formula, it indicates that the valve torque data is normal. The detected valve number and the corresponding torque data are then sent to the valve effective tightening torque dynamic decay analysis submodule. If the above formula is not satisfied, the data is sent to the administrator to trigger an alarm indicating that there is an abnormal fault in the valve torque data within the current cycle.
[0009] Further settings: The valve effective tightening torque dynamic attenuation analysis submodule acquires the torque data sent by the dynamic torque signal parameter monitoring submodule. It also acquires the valve number and its original torque data that are free of anomalies, and obtains the peak torque of different valves within different acquisition periods. The system collects the current pressure value of the pressure vessel, monitors the friction torque of the valve and the pressure torque generated by the reaction force of the valve working pressure on the sealing surface. The higher the current pressure value of the pressure vessel, the greater the pressure torque generated. It analyzes the maximum effective tightening torque data of different valves within different acquisition cycles and sets the maximum effective tightening torque of different valves within different acquisition cycles as... Set the current valve's friction torque as The pressure torque is According to the formula: In the above formula, This is the torque value generated when the valve is initially started. The torque value generated during valve operation is used, where the pressure torque is set according to the current pressure value of the pressure vessel. The maximum effective tightening torque data of the valve in different acquisition cycles is obtained. The maximum effective tightening torque data of the valve in different acquisition cycles is compared with the set standard torque value. When the maximum effective tightening torque data of the valve in several consecutive acquisition cycles is less than the set standard torque value and is arranged in descending order, it indicates that the valve tightening force is weakening, and an alert is sent to the administrator.
[0010] Further configuration: The multi-modal detection module for pressure vessel valve corrosion status includes a valve corrosion image feature fusion analysis submodule and a comprehensive corrosion status risk coefficient analysis submodule. The valve corrosion image feature fusion analysis submodule acquires images of marked valves exhibiting anomalies, while simultaneously acquiring environmental data such as temperature, humidity, gas type and concentration, and the actual valve wall thickness. It then acquires images containing color anomalies within the images of valves exhibiting anomalies for valve corrosion status monitoring. Based on image recognition technology, it analyzes whether the color-abnormal areas within the images of valves exhibiting color anomalies constitute valve corrosion. If so, it collects the total area of the color-abnormal regions in the image, obtains the inspected surface area of the valve in the image, and calculates the total area of the color-abnormal regions. With the inspected surface area of the valve The ratio analysis of the surface corrosion area ratio of the valve is used to determine the surface corrosion area ratio of the valve. , Obtain the valve design wall thickness The actual wall thickness of the valve being tested Comparison and analysis were conducted to determine the actual wall thickness reduction rate of the valves. , The process involves acquiring ambient temperature, humidity, and gas type and concentration data to analyze the corrosivity coefficient of the environment surrounding the valve. Different corrosive gas concentrations are monitored, and each gas concentration is analyzed to determine if it reaches a set critical corrosion threshold. Corrosive gases that reach the critical corrosion threshold are then selected, and their concentrations are determined. The corrosion coefficient of the valve environment is set as follows: According to the formula: in, These are the weighting coefficients for the impact of corrosive gases on valves. As environmental temperature and humidity are influencing factors, the greater the exceedance of critical values by these factors, the higher the influencing factor becomes. The percentage of surface corrosion area, the actual wall thickness reduction rate, and the environmental corrosion coefficient of the valve are sent to the comprehensive corrosion state risk coefficient analysis submodule for comprehensive evaluation. This submodule retrieves the corresponding values, assigns different weights to different values, and analyzes the comprehensive corrosion state risk coefficient of different valves. According to the formula: The comprehensive corrosion risk coefficient of different valves was calculated. These are different weighting coefficients, and Based on the comprehensive corrosion risk coefficient of different valves, multiple risk levels are divided into high risk, medium risk, and low risk, with the values of the multiple risk levels set manually.
[0011] Further configuration: The pressure vessel valve failure urgency assessment and prediction module includes a dynamic valve status data parameter statistics submodule and a valve failure maintenance urgency classification and early warning submodule. The dynamic valve status data parameter statistics submodule acquires the maximum effective tightening torque datasets for different valves in the pressure vessel and the comprehensive corrosion status risk coefficients for different valves. The valve failure maintenance urgency classification and early warning submodule sets a reference range for valve dynamic torque values and compares the maximum effective tightening torque datasets for different valves with the reference range. If the maximum effective tightening torque data for a valve within different acquisition periods is all within the reference range and the comprehensive corrosion status risk coefficient is low, the valve is marked as low-maintenance. A low-maintenance warning is issued when the ratio of the number of maximum effective tightening torque data for a valve within different acquisition periods that is outside the reference range is greater than a certain threshold. If the valve's overall corrosion risk coefficient is medium risk, the valve will be marked as a general emergency maintenance warning if either condition occurs. This warning is issued when the ratio of the number of valves whose maximum effective tightening torque data within different data collection periods is outside the valve's dynamic torque reference range is greater than a certain threshold. If the overall corrosion risk coefficient of the valve is high, the valve will be marked as having a high emergency maintenance warning if either condition occurs. , The specific ratio is set manually.
[0012] Compared with the prior art, the beneficial effects of the present invention are: it aims to combine the torque parameter reflecting the mechanical fastening state of the valve with the corrosion parameter reflecting the degradation state of the valve material, and to perform dual-dimensional fault monitoring of the valve. By analyzing the trend of the effective fastening torque of the valve and fusion monitoring of corrosion data, the valve is comprehensively and quantitatively evaluated, achieving early warning of the valve and improving the overall safety and operational reliability of the pressure vessel system. Attached Figure Description
[0013] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.
[0014] Figure 1 This is a schematic diagram of the structural process of a pressure vessel-based equipment fault feature identification and inspection system according to the present invention; Figure 2 This is a specific module of the equipment fault feature identification and inspection system based on pressure vessels according to the present invention. Figure 1 ; Figure 3 This is a specific module of the equipment fault feature identification and inspection system based on pressure vessels according to the present invention. Figure 2 ; Figure 4This is a specific module of the equipment fault feature identification and inspection system based on pressure vessels according to the present invention. Figure 3 ; Figure 5 This is a specific module of the equipment fault feature identification and inspection system based on pressure vessels according to the present invention. Figure 4 ; Figure 6 This is a specific module of the equipment fault feature identification and inspection system based on pressure vessels according to the present invention. Figure 5 . Detailed Implementation
[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.
[0016] Please see Figures 1-6 In this embodiment of the invention, a pressure vessel-based equipment fault feature identification and inspection system includes a pressure vessel multi-condition identification module, a pressure vessel valve basic fault feature identification and analysis module, a pressure vessel valve dynamic torque online monitoring and analysis module, a pressure vessel valve corrosion state multi-modal detection module, and a pressure vessel valve fault urgency assessment and prediction module. The pressure vessel multi-condition identification module uses several sensors to collect multiple parameters of the pressure vessel's operating conditions, dynamically identifies data under different operating conditions, dynamically adjusts the evaluation threshold according to the current operating conditions, performs dynamic multi-parameter threshold comparison, and triggers corresponding faults for classification, early warning, and alarm based on the comparison and verification results. like Figure 2Further explanation is needed. The pressure vessel multi-condition identification module includes a pressure vessel multi-parameter synchronous acquisition submodule and a multi-parameter threshold identification and early warning model construction submodule. The pressure vessel multi-parameter synchronous acquisition submodule includes an adaptive sensor management and monitoring unit and a data acquisition and aggregation unit. The adaptive sensor management and monitoring unit uses several sensors to monitor and acquire pressure vessel operating parameters. The sensor types are selected and set by the operator. The acquired parameters include the internal transmission medium of the pressure vessel, pressure, ambient temperature and humidity, ambient gas concentration, and valve wall thickness. The data acquisition and aggregation unit preprocesses the acquired raw data, aligns all data streams on the time axis to ensure consistency of data at the same time, and classifies and summarizes the different detection data according to different detection categories. The multi-parameter threshold identification and early warning model construction submodule obtains the parameter thresholds corresponding to different detection categories, acquires the detection data under different monitoring operating condition categories of the pressure vessel, and obtains the peak value and mean value of the data for each category. The peak value of the data detected for a certain category is set as... The mean is When both the peak and average values of the detection data for a given category exceed the corresponding parameter threshold, a fault alarm for that category is triggered. When both the peak and average values of the detection data for a given category are less than the corresponding parameter threshold, it indicates that the detection data for that category is within the safety standard. When the average value of the detection data for a given category is less than the corresponding parameter threshold, but the peak value is greater than the corresponding parameter threshold, the corresponding parameter threshold is set to [value missing]. Introduce a safety weight coefficient for this category. Set the threshold for the security controllable parameters corresponding to this category as follows: The safety weight coefficient is set manually, and the dynamic safety peak range for the detection data of this category is defined, with the safety peak range set as follows: Screen the test data for this category, according to the formula: When the extreme values within the dynamic safety peak range of the detection data for this category do not meet the above formula, a fault alarm for the corresponding category is triggered. When the extreme values within the dynamic safety peak range of the detection data for this category all meet the above formula, the number of times the detection data for this category exceeds the corresponding parameter threshold within the dynamic safety peak range is analyzed. When the number exceeds the set threshold, a warning is uploaded for the corresponding category fault. When the number is less than the set threshold, it indicates that the detection data for this category is within the safety standard. The set thresholds are all preset by the operator.
[0017] The pressure vessel valve basic fault feature identification and analysis module collects the operating status and images of the pressure vessel valve. Based on the monitoring data and images collected by several sensors during the valve opening and closing process, it performs preprocessing. Based on the preprocessed monitoring data and images, it detects micro-fault defect features, identifies basic fault defect features, and performs fault assessment and early warning based on the degree of feature deviation. like Figure 3 Further explanation is needed: the pressure vessel valve basic fault feature identification and analysis module includes a valve multi-dimensional feature acquisition submodule and a basic fault feature classification, identification, and evaluation submodule. The valve multi-dimensional feature acquisition submodule includes a sensor signal acquisition unit and a valve image acquisition unit. The sensor signal acquisition unit includes a torque sensor, a temperature and humidity sensor, and an ultrasonic sensor, which respectively acquire valve status parameters, including valve temperature and humidity, operating torque, and valve stem displacement data. The valve image acquisition unit includes several industrial cameras, which are used to acquire overall valve appearance images and partial feature appearance images, including valve stem, valve disc, handwheel, and flange connection. The acquired raw detection data and images are preprocessed to obtain the detected valve parameters and... Valve images undergo spatial alignment and temporal matching. Valve parameter data and image data are aggregated according to different valve numbers and sent to the basic fault feature classification and evaluation submodule. This submodule acquires the fluctuation characteristics of different categories of valve detection parameters and different valve images acquired by the valve image acquisition unit. Based on the different categories of valve detection parameters, it obtains the positive and negative peak values of the detection parameters and compares them with the corresponding set threshold ranges. If the positive and negative peak values are not within the set threshold range for that category, a fault identification and warning are issued based on the data detection for that category. If both the positive and negative peak values are within the set threshold range, the fluctuation amplitude of the detection parameter is identified, and the detection parameter is set as follows: The positive and negative peak values of this detection parameter were screened out as follows: and After removing the positive and negative peak values from the detection equipment, the remaining data are summarized, and the truncated mean is calculated using the remaining data. , ,in The data for valves is collected in real time, and the standard deviation is calculated using the remaining data. Analyze the fluctuation and dispersion coefficient of the detection data for this category, and set the fluctuation and dispersion coefficient of the detection data for this category as . : Based on the fluctuation and dispersion coefficient of the detection data for this category, the stability of the real-time detection data for this category is analyzed, and the stability coefficient of the detection data is set as follows: According to the formula: when A value of 1 indicates that the real-time detection data for this category is highly stable. A smaller value indicates that the real-time monitoring of that category is less stable. When the stability coefficient of the monitoring data is less than a set threshold, the real-time monitoring data for that valve category is marked and uploaded for fault warning. Factors affecting data fluctuations; Based on the different valve images acquired by the valve image acquisition unit, each acquired image is compared with the fault map. According to the fault map uploaded by the administrator, each acquired image is marked with valve color features, defect areas, and structural anomalies. By analyzing the valve color features, areas with obvious rust spots are identified and marked as having valve color anomalies. By monitoring the defect areas, images with valve defects are marked as having abnormal contours or deformations. By detecting structural anomalies, images with abnormal bolts, aging seals, or abnormal connections are marked as having valve anomalies. The valve numbers corresponding to the images with identified and marked faults are analyzed and warned of basic valve faults.
[0018] The pressure vessel valve dynamic torque online monitoring and analysis module continuously monitors the dynamic torque of pressure vessel valves, constructs valve operating torque curves, performs quantitative analysis based on the monitored valve torque signals and torque growth rates within different cycles, and simultaneously constructs a dynamic attenuation model of effective valve fastening torque to analyze valve fastening force attenuation and provide proactive fault warning. like Figure 4 Furthermore, it needs to be explained that the pressure vessel valve dynamic torque online monitoring and analysis module includes a dynamic torque signal parameter monitoring submodule and a valve effective tightening torque dynamic attenuation analysis submodule. The dynamic torque signal parameter monitoring submodule includes several torque sensors, which synchronously collect the valve's torque value, capture the valve's dynamic torque data in real time, extract peak torque and average torque data from the dynamic torque signal, construct the valve's torque curve, obtain the reference torque value of the valve's rated operating condition, perform dynamic torque signal detection on the valve, obtain the peak torque collected and monitored for a specific valve within the current operating cycle, calculate the average value and standard deviation of the peak torque in the data of different operating cycles based on the monitored peak torque signals within different cycles, and set the peak torque monitored in the current operating cycle as... The average peak torque value is set as the average value in the data from different operating cycles. and standard deviation Calculate the rate of increase of valve torque within the current cycle, and set the rate of increase of torque within the current cycle as... According to the formula: In the above formula, The initial torque at the start of the current operating cycle. This refers to the time elapsed from the start of the current operating cycle to reaching peak torque. The start time of the current operation cycle is marked as the baseline time point. Quantitative analysis is performed based on the valve torque data collected within different cycles, according to the formula: in, This is the standard deviation factor, which is set manually. The threshold for the torque growth rate within the current cycle is defined as follows: if the peak torque and torque growth rate monitored in the current operating cycle satisfy the above formula, it indicates that the valve torque data is normal. The detected valve number and the corresponding torque data are then sent to the valve effective tightening torque dynamic decay analysis submodule. If the above formula is not satisfied, the data is sent to the administrator to trigger an alarm indicating that there is an abnormal fault in the valve torque data within the current cycle.
[0019] It should also be noted that the valve effective tightening torque dynamic decay analysis submodule acquires the torque data sent by the dynamic torque signal parameter monitoring submodule for valves with no abnormalities and their original torque data, and obtains the peak torque of different valves in different acquisition periods. The system collects the current pressure value of the pressure vessel, monitors the friction torque of the valve and the pressure torque generated by the reaction force of the valve working pressure on the sealing surface. The higher the current pressure value of the pressure vessel, the greater the pressure torque generated. It analyzes the maximum effective tightening torque data of different valves within different acquisition cycles and sets the maximum effective tightening torque of different valves within different acquisition cycles as... Set the current valve's friction torque as The pressure torque is According to the formula: In the above formula, This is the torque value generated when the valve is initially started. The torque value generated during valve operation is used, where the pressure torque is set according to the current pressure value of the pressure vessel. The maximum effective tightening torque data of the valve in different acquisition cycles is obtained. The maximum effective tightening torque data of the valve in different acquisition cycles is compared with the set standard torque value. When the maximum effective tightening torque data of the valve in several consecutive acquisition cycles is less than the set standard torque value and is arranged in descending order, it indicates that the valve tightening force is weakening, and an alert is sent to the administrator.
[0020] The multi-modal detection module for corrosion status of pressure vessel valves acquires pre-processed images of valves exhibiting abnormalities, extracts feature parameters from the corrosion images on the valve surface for intelligent evaluation, monitors environmental data around the valve for combined analysis, analyzes the overall corrosion status of the valve, and outputs a corresponding risk index based on the overall corrosion status for fault warning. like Figure 4 Furthermore, it needs to be explained that the multi-modal detection module for pressure vessel valve corrosion status includes a valve corrosion image feature fusion analysis submodule and a comprehensive corrosion status risk coefficient analysis submodule. The valve corrosion image feature fusion analysis submodule acquires images marked with valve anomalies, and simultaneously acquires the ambient temperature, humidity, gas type and concentration, and actual valve wall thickness. It then acquires images containing valve color anomalies within the images of existing valve anomalies for valve corrosion status monitoring. Based on image recognition technology, it analyzes whether the color-abnormal areas within the images of existing valve color anomalies constitute valve corrosion. If so, it collects the total area of the color-abnormal areas in the image, obtains the inspected surface area of the valve in the image, and calculates the corrosion status based on the total area of the color-abnormal areas. With the inspected surface area of the valve The ratio analysis of the surface corrosion area ratio of the valve is used to determine the surface corrosion area ratio of the valve. , Obtain the valve design wall thickness The actual wall thickness of the valve being tested Comparison and analysis were conducted to determine the actual wall thickness reduction rate of the valves. , The process involves acquiring ambient temperature, humidity, and gas type and concentration data to analyze the corrosivity coefficient of the environment surrounding the valve. Different corrosive gas concentrations are monitored, and each gas concentration is analyzed to determine if it reaches a set critical corrosion threshold. Corrosive gases that reach the critical corrosion threshold are then selected, and their concentrations are determined. The corrosion coefficient of the valve environment is set as follows: According to the formula: in, These are the weighting coefficients for the impact of corrosive gases on valves. As environmental temperature and humidity are influencing factors, the greater the exceedance of critical values by these factors, the higher the influencing factor becomes. The percentage of surface corrosion area, the actual wall thickness reduction rate, and the environmental corrosion coefficient of the valve are sent to the comprehensive corrosion state risk coefficient analysis submodule for comprehensive evaluation. This submodule retrieves the corresponding values, assigns different weights to different values, and analyzes the comprehensive corrosion state risk coefficient of different valves. According to the formula: The comprehensive corrosion risk coefficient of different valves was calculated. These are different weighting coefficients, and Based on the comprehensive corrosion risk coefficient of different valves, multiple risk levels are divided into high risk, medium risk, and low risk, with the values of the multiple risk levels set manually.
[0021] The pressure vessel valve failure urgency assessment and prediction module summarizes the dynamic torque monitoring and corrosion status monitoring data of the pressure vessel valves, and constructs a valve maintenance urgency classification framework based on the summarized data to provide graded failure early warning.
[0022] like Figure 5 Further explanation is needed: the pressure vessel valve failure urgency assessment and prediction module includes a dynamic valve status data parameter statistics submodule and a valve failure maintenance urgency classification and early warning submodule. The dynamic valve status data parameter statistics submodule acquires the maximum effective tightening torque datasets for different valves in the pressure vessel and the comprehensive corrosion risk coefficients for different valves. The valve failure maintenance urgency classification and early warning submodule sets a reference range for valve dynamic torque values and compares the maximum effective tightening torque datasets for different valves with this reference range. If the maximum effective tightening torque data for a valve within different acquisition periods is consistently within the reference range, and the valve's comprehensive corrosion risk coefficient is low, the valve is marked as low-maintenance. A low-maintenance warning is issued when the ratio of the number of maximum effective tightening torque data for a valve within different acquisition periods that is outside the reference range is greater than [a certain threshold]. If the valve's overall corrosion risk coefficient is medium risk, the valve will be marked as a general emergency maintenance warning if either condition occurs. This warning is issued when the ratio of the number of valves whose maximum effective tightening torque data within different data collection periods is outside the valve's dynamic torque reference range is greater than a certain threshold. If the overall corrosion risk coefficient of the valve is high, the valve will be marked as having a high emergency maintenance warning if either condition occurs. , The specific ratio is set manually.
[0023] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.
Claims
1. A fault feature identification and inspection system for pressure vessels, characterized in that: The system includes a pressure vessel multi-condition identification module, a pressure vessel valve basic fault feature identification and analysis module, a pressure vessel valve dynamic torque online monitoring and analysis module, a pressure vessel valve corrosion state multi-modal detection module, and a pressure vessel valve fault urgency assessment and prediction module. The pressure vessel multi-condition identification module uses several sensors to collect multiple parameters of the pressure vessel's operating conditions, dynamically identifies data under different operating conditions, dynamically adjusts the evaluation threshold according to the current operating conditions, performs dynamic multi-parameter threshold comparison, and triggers corresponding faults for classification, early warning, and alarm based on the comparison and verification results. The pressure vessel valve basic fault feature identification and analysis module collects the operating status and images of the pressure vessel valve. Based on the monitoring data and images collected by several sensors during the valve opening and closing process, it performs preprocessing. Based on the preprocessed monitoring data and images, it detects micro-fault defect features, identifies basic fault defect features, and performs fault assessment and early warning based on the degree of feature deviation. The pressure vessel valve dynamic torque online monitoring and analysis module continuously monitors the dynamic torque of pressure vessel valves, constructs valve operating torque curves, performs quantitative analysis based on the monitored valve torque signals and torque growth rates within different cycles, and simultaneously constructs a dynamic attenuation model of effective valve fastening torque to analyze valve fastening force attenuation and provide proactive fault warning. The multi-modal detection module for corrosion status of pressure vessel valves acquires pre-processed images of valves exhibiting abnormalities, extracts feature parameters from the corrosion images on the valve surface for intelligent evaluation, monitors environmental data around the valve for combined analysis, analyzes the overall corrosion status of the valve, and outputs a corresponding risk index based on the overall corrosion status for fault warning. The pressure vessel valve failure urgency assessment and prediction module summarizes the dynamic torque monitoring and corrosion status monitoring data of the pressure vessel valves, and constructs a valve maintenance urgency classification framework based on the summarized data to provide graded failure early warning.
2. The equipment fault feature identification and inspection system based on pressure vessels according to claim 1, characterized in that... The pressure vessel multi-condition identification module includes a pressure vessel multi-parameter synchronous acquisition submodule and a multi-parameter threshold identification and early warning model construction submodule. The pressure vessel multi-parameter synchronous acquisition submodule includes an adaptive sensor management and monitoring unit and a data acquisition and aggregation unit. The adaptive sensor management and monitoring unit uses several sensors to monitor and acquire pressure vessel operating parameters. The sensor types are selected and set by the operator. The acquired parameters include the internal transmission medium of the pressure vessel, pressure, ambient temperature and humidity, ambient gas concentration, and valve wall thickness. The data acquisition and aggregation unit preprocesses the acquired raw data, aligns all data streams on the time axis to ensure consistency of data at the same time, and classifies and summarizes the different detection data according to different detection categories. The multi-parameter threshold identification and early warning model construction submodule obtains the parameter thresholds corresponding to different detection categories, acquires the detection data under different monitoring operating condition categories of the pressure vessel, and obtains the peak and average values of the data for each category. The peak value of the data detected for a certain category is set as... The mean is When both the peak and average values of the detection data for a given category exceed the corresponding parameter threshold, a fault alarm for that category is triggered. When both the peak and average values of the detection data for a given category are less than the corresponding parameter threshold, it indicates that the detection data for that category is within the safety standard. When the average value of the detection data for a given category is less than the corresponding parameter threshold, but the peak value is greater than the corresponding parameter threshold, the corresponding parameter threshold is set to [value missing]. Introduce a safety weight coefficient for this category. Set the threshold for the security controllable parameters corresponding to this category as follows: The safety weight coefficient is set manually, and the dynamic safety peak range for the detection data of this category is defined, with the safety peak range set as follows: Screen the test data for this category, according to the formula: When the extreme values within the dynamic safety peak range of the detection data for this category do not meet the above formula, a fault alarm for the corresponding category is triggered. When the extreme values within the dynamic safety peak range of the detection data for this category all meet the above formula, the number of times the detection data for this category exceeds the corresponding parameter threshold within the dynamic safety peak range is analyzed. When the number exceeds the set threshold, a warning is uploaded for the corresponding category fault. When the number is less than the set threshold, it indicates that the detection data for this category is within the safety standard. The set thresholds are all preset by the operator.
3. The equipment fault feature identification and inspection system based on pressure vessels according to claim 1, characterized in that... The pressure vessel valve basic fault feature identification and analysis module includes a valve multi-dimensional feature acquisition submodule and a basic fault feature classification, identification, and evaluation submodule. The valve multi-dimensional feature acquisition submodule includes a sensor signal acquisition unit and a valve image acquisition unit. The sensor signal acquisition unit includes a torque sensor, a temperature and humidity sensor, and an ultrasonic sensor, which respectively acquire valve state parameters, including valve temperature and humidity, operating torque, and valve stem displacement data. The valve image acquisition unit includes several industrial cameras, which are used to acquire overall valve appearance images and partial feature appearance images, including valve stem, valve disc, handwheel, and flange connection. The acquired raw detection data and images are preprocessed, and the detected valve parameters and the valve image are compared. For example, spatial alignment and temporal matching are performed. Different valve parameter data and image data are aggregated according to different valve numbers and sent to the basic fault feature classification and identification evaluation submodule. This submodule acquires the fluctuation characteristics of different categories of valve detection parameters and different valve images acquired by the valve image acquisition unit. Based on the different categories of valve detection parameters, it obtains the positive and negative peak values of the detection parameters and compares them with the set threshold range corresponding to that category. If the positive and negative peak values are not within the set threshold range corresponding to that category, fault identification and early warning are performed based on the data detection for that category. If both the positive and negative peak values are within the set threshold range, the fluctuation amplitude of the detection parameter is identified, and the detection parameter is set as follows: The positive and negative peak values of this detection parameter were screened out as follows: and After removing the positive and negative peak values from the detection equipment, the remaining data are summarized, and the truncated mean is calculated using the remaining data. , ,in The data for valves is collected in real time, and the standard deviation is calculated using the remaining data. Analyze the fluctuation and dispersion coefficient of the detection data for this category, and set the fluctuation and dispersion coefficient of the detection data for this category as . : Based on the fluctuation and dispersion coefficient of the detection data for this category, the stability of the real-time detection data for this category is analyzed, and the stability coefficient of the detection data is set as follows: According to the formula: when A value of 1 indicates that the real-time detection data for this category is highly stable. A smaller value indicates that the real-time monitoring of that category is less stable. When the stability coefficient of the monitoring data is less than a set threshold, the real-time monitoring data for that valve category is marked and uploaded for fault warning. Factors affecting data fluctuations; Based on the different valve images acquired by the valve image acquisition unit, each acquired image is compared with the fault map. According to the fault map uploaded by the administrator, each acquired image is marked with valve color features, defect areas, and structural anomalies. By analyzing the valve color features, areas with obvious rust spots are identified and marked as having valve color anomalies. By monitoring the defect areas, images with valve defects are marked as having abnormal contours or deformations. By detecting structural anomalies, images with abnormal bolts, aging seals, or abnormal connections are marked as having valve anomalies. The valve numbers corresponding to the images with identified and marked faults are analyzed and warned of basic valve faults.
4. The equipment fault feature identification and inspection system based on pressure vessels according to claim 1, characterized in that... The pressure vessel valve dynamic torque online monitoring and analysis module includes a dynamic torque signal parameter monitoring submodule and a valve effective tightening torque dynamic attenuation analysis submodule. The dynamic torque signal parameter monitoring submodule includes several torque sensors that synchronously collect valve torque values, capturing dynamic valve torque data in real time. It extracts peak torque and average torque data from the dynamic torque signal, constructs the valve's torque curve, obtains the reference torque value for the valve's rated operating condition, performs dynamic torque signal detection on the valve, and acquires the peak torque of a specific valve within the current operating cycle. Based on the peak torque signals from different cycles, it calculates the average and standard deviation of the peak torque in the data from different operating cycles, setting the peak torque monitored in the current operating cycle as... The average peak torque value is set as the average value in the data from different operating cycles. and standard deviation Calculate the rate of increase of valve torque within the current cycle, and set the rate of increase of torque within the current cycle as... According to the formula: In the above formula, The initial torque at the start of the current operating cycle. This refers to the time elapsed from the start of the current operating cycle to reaching peak torque. The start time of the current operation cycle is marked as the baseline time point. Quantitative analysis is performed based on the valve torque data collected within different cycles, according to the formula: in, This is the standard deviation factor, which is set manually. The threshold for the torque growth rate within the current cycle is defined as follows: if the peak torque and torque growth rate monitored in the current operating cycle satisfy the above formula, it indicates that the valve torque data is normal. The detected valve number and the corresponding torque data are then sent to the valve effective tightening torque dynamic decay analysis submodule. If the above formula is not satisfied, the data is sent to the administrator to trigger an alarm indicating that there is an abnormal fault in the valve torque data within the current cycle.
5. The equipment fault feature identification and inspection system based on pressure vessels according to claim 4, characterized in that... The valve effective tightening torque dynamic attenuation analysis submodule acquires the valve number and its original torque data without abnormality from the torque data sent by the dynamic torque signal parameter monitoring submodule, and obtains the peak torque of different valves in different acquisition periods. The system collects the current pressure value of the pressure vessel, monitors the friction torque of the valve and the pressure torque generated by the reaction force of the valve working pressure on the sealing surface. The higher the current pressure value of the pressure vessel, the greater the pressure torque generated. It analyzes the maximum effective tightening torque data of different valves within different acquisition cycles and sets the maximum effective tightening torque of different valves within different acquisition cycles as... Set the current valve's friction torque as The pressure torque is According to the formula: In the above formula, This is the torque value generated when the valve is initially started. The torque value generated during valve operation is used, where the pressure torque is set according to the current pressure value of the pressure vessel. The maximum effective tightening torque data of the valve in different acquisition cycles is obtained. The maximum effective tightening torque data of the valve in different acquisition cycles is compared with the set standard torque value. When the maximum effective tightening torque data of the valve in several consecutive acquisition cycles is less than the set standard torque value and is arranged in descending order, it indicates that the valve tightening force is weakening, and an alert is sent to the administrator.
6. The equipment fault feature identification and inspection system based on pressure vessels according to claim 1, characterized in that... The multimodal detection module for corrosion status of pressure vessel valves includes a valve corrosion image feature fusion analysis submodule and a comprehensive corrosion status risk coefficient analysis submodule. The valve corrosion image feature fusion analysis submodule acquires images of marked valves exhibiting abnormalities, and simultaneously acquires environmental data such as temperature, humidity, gas type and concentration, and the actual valve wall thickness. It also acquires images containing color anomalies within the images of valves exhibiting abnormalities for valve corrosion status monitoring. Based on image recognition technology, it analyzes whether the color anomalies within the images indicate valve corrosion. If so, it collects the total area of the color-abnormal regions in the image, obtains the inspected surface area of the valve in the image, and calculates the corrosion status based on the total area of the color-abnormal regions. With the inspected surface area of the valve The ratio analysis of the surface corrosion area ratio of the valve is used to determine the surface corrosion area ratio of the valve. , Obtain the valve design wall thickness The actual wall thickness of the valve being tested Comparison and analysis were conducted to determine the actual wall thickness reduction rate of the valves. , The process involves acquiring ambient temperature, humidity, and gas type and concentration data to analyze the corrosivity coefficient of the environment surrounding the valve. Different corrosive gas concentrations are monitored, and each gas concentration is analyzed to determine if it reaches a set critical corrosion threshold. Corrosive gases that reach the critical corrosion threshold are then selected, and their concentrations are determined. The corrosion coefficient of the valve environment is set as follows: According to the formula: in, These are the weighting coefficients for the impact of corrosive gases on valves. As environmental temperature and humidity are influencing factors, the greater the exceedance of critical values by these factors, the higher the influencing factor becomes. The percentage of surface corrosion area, the actual wall thickness reduction rate, and the environmental corrosion coefficient of the valve are sent to the comprehensive corrosion state risk coefficient analysis submodule for comprehensive evaluation. This submodule retrieves the corresponding values, assigns different weights to different values, and analyzes the comprehensive corrosion state risk coefficient of different valves. According to the formula: The comprehensive corrosion risk coefficient of different valves was calculated. These are different weighting coefficients, and Based on the comprehensive corrosion risk coefficient of different valves, multiple risk levels are divided into high risk, medium risk, and low risk, with the values of the multiple risk levels set manually.
7. The equipment fault feature identification and inspection system based on pressure vessels according to claim 1, characterized in that... The pressure vessel valve failure urgency assessment and prediction module includes a dynamic valve status data parameter statistics submodule and a valve failure maintenance urgency classification and early warning submodule. The dynamic valve status data parameter statistics submodule acquires the maximum effective tightening torque datasets for different valves in the pressure vessel and the comprehensive corrosion risk coefficients for different valves. The valve failure maintenance urgency classification and early warning submodule sets a reference range for valve dynamic torque values and compares the maximum effective tightening torque datasets for different valves with this reference range. If the maximum effective tightening torque data for a valve within different acquisition periods is all within the reference range and the valve's comprehensive corrosion risk coefficient is low, the valve is marked as low-maintenance. A low-maintenance warning is issued when the ratio of the number of maximum effective tightening torque data for a valve within different acquisition periods that is not within the reference range is greater than a certain threshold. If the valve's overall corrosion risk coefficient is medium risk, the valve will be marked as a general emergency maintenance warning if either condition occurs. This warning is issued when the ratio of the number of valves whose maximum effective tightening torque data within different data collection periods is outside the valve's dynamic torque reference range is greater than a certain threshold. If the overall corrosion risk coefficient of the valve is high, the valve will be marked as having a high emergency maintenance warning if either condition occurs. , The specific ratio is set manually.
