A valve operation early warning method and system based on multi-source state cross-validation

Abstract

The application discloses a valve operation early warning method and system based on multi-source state cross-validation, angle sensors, vibration sensors and voiceprint sensors are arranged on the valve; first data feature criteria corresponding to normal authorized operation and second data feature criteria corresponding to expected abnormal events are preset; current feature data of the valve is acquired; the current feature data is matched with the first data feature criteria and the second data feature criteria, according to the matching result, it is determined that the current valve operation is the normal authorized operation or the expected abnormal event, if it is the expected abnormal event, the early warning is triggered. The application can accurately identify the valve state, reduce the valve action misjudgment probability and improve the valve operation convenience.

Description

Technical Field

[0001] This invention relates to the field of valve condition detection technology, and in particular to a valve operation early warning method and system based on multi-source condition cross-verification. Background Technology

[0002] In existing technologies, common methods for monitoring the on / off status of industrial valves and pipelines include angle sensor monitoring and limit switch monitoring. Angle sensor monitoring involves installing an angle sensor on the valve handwheel or valve stem tip to determine the valve's open / closed state and degree of opening by detecting the rotation angle. Mechanical or proximity limit switches are installed at the fully open and fully closed positions of the valve. When the valve reaches its limit position, the limit switch is triggered, and the system receives a "fully open" or "fully closed" signal.

[0003] However, valve angle sensors are prone to angle jumps or drifts due to loose mounting brackets, detached magnets, or poor wiring connections. The system may misinterpret these as valve movement, making it impossible to determine whether the valve actually moved or the sensor malfunctions. This necessitates on-site troubleshooting, increasing manpower and time costs. Furthermore, current technology cannot differentiate between "authorized personnel operating normally," "unauthorized personnel misoperation," or "external impact (such as tool knocks or vandalism)" when sensors detect changes in status or limit switch states. This results in a lack of statistical data to support safety management, making it difficult to trace equipment accidents. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a valve operation early warning method and system based on multi-source state cross-validation, which can accurately identify the valve status, reduce the probability of valve action misjudgment, and improve the convenience of valve operation and maintenance.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A valve operation early warning method based on multi-source state cross-validation includes: The communication connection includes an angle sensor located on the valve rotation shaft, a vibration sensor located on the valve body, and an acoustic sensor located on the valve lock housing. The system pre-defines a first data feature criterion corresponding to normal authorized operations and a second data feature criterion corresponding to anticipated abnormal events. Both the first and second data feature criterions include angle feature criteria, vibration feature criteria, and voiceprint feature criteria. Obtain the current characteristic data of the valve, which includes the angle characteristic data of the angle sensor, the vibration characteristic data of the vibration sensor, and the acoustic characteristic data of the acoustic sensor; The current feature data is matched with the first data feature criterion and the second data feature criterion. Based on the matching result, it is determined whether the current valve operation is a normal authorized operation or a anticipated abnormal event. If it is the anticipated abnormal event, an early warning is triggered.

[0006] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A valve operation early warning system based on multi-source state cross-validation includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the aforementioned valve operation early warning method based on multi-source state cross-validation.

[0007] The beneficial effects of this invention are as follows: It provides a valve operation early warning method and system based on multi-source state cross-verification. Angle sensors, vibration sensors, and acoustic sensors are respectively set at different positions of the valve. Then, a first data feature criterion corresponding to normal authorized operation and a second data feature criterion corresponding to expected abnormal event are preset. Based on this, the relationship between the changes in angle, vibration, and acoustic signature caused by normal authorized operation and expected abnormal event is constructed. The current feature data of the valve is obtained, and the current feature data is matched with the first data feature criterion and the second data feature criterion. Multi-source state cross-verification is achieved through feature data comparison, thereby accurately identifying the state of the valve and whether the current valve operation causing the state is a normal authorized operation or an expected abnormal event. If it is an expected abnormal event, an early warning is triggered, reducing the probability of valve action misjudgment and improving the convenience of valve operation and maintenance. Attached Figure Description

[0008] Figure 1 This is a schematic diagram illustrating the steps of a valve operation early warning method based on multi-source state cross-validation according to the present invention. Figure 2 This is a valve information display diagram of a valve operation early warning method based on multi-source state cross-validation according to an embodiment of the present invention; Figure 3 This is a flowchart of a three-level cross-validation process for a valve operation early warning method based on multi-source state cross-validation, as described in an embodiment of the present invention. Figure 4 This is a system block diagram of a valve operation early warning system based on multi-source state cross-verification according to the present invention. Detailed Implementation

[0009] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0010] The present invention provides a valve operation early warning method and system based on multi-source state cross-verification, which is applicable to industrial fields related to valve applications, such as the energy and power industry and the petrochemical industry.

[0011] In existing technologies, the commonly used methods for monitoring the on / off status of industrial valves and pipelines include angle sensor monitoring and limit switch monitoring. However, valve angle sensors are prone to angle jumps or drift signals due to loose mounting brackets, detached magnets, poor wiring connections, etc. The system may misinterpret these as valve movement, making it impossible to determine whether the valve has actually moved or the sensor is faulty. This necessitates on-site troubleshooting, increasing manpower and time costs. Furthermore, when sensors detect changes in status or limit switch status, existing technologies cannot differentiate between specific causes such as "normal operation by authorized personnel," "misoperation by unauthorized personnel," or "external impact (such as tool knocks or deliberate damage)." This results in a lack of statistical data support for safety management, making it difficult to trace equipment accidents.

[0012] To address at least the aforementioned issues, this invention distributes angle sensors, vibration sensors, and acoustic sensors on the valve to achieve multi-source data acquisition. It then pre-sets a first data feature criterion corresponding to normal authorized operations and a second data feature criterion corresponding to anticipated abnormal events. Through feature input and cross-validation of the multi-source data, it determines whether the current valve operation is a normal authorized operation or a anticipated abnormal event, and whether an alert should be issued. This invention utilizes multi-source data cross-validation to accurately identify the valve's state and whether the current valve operation causing that state is a normal authorized operation or a anticipated abnormal event, reducing the probability of misjudged valve actions and improving the convenience of valve operation and maintenance.

[0013] Please refer to Figures 1 to 3 This invention discloses a valve operation early warning method 100 based on multi-source state cross-validation, the method 100 including the following steps: In step S102, the angle sensor located on the valve rotation shaft, the vibration sensor located on the valve body, and the acoustic sensor located on the valve lock housing are connected in communication.

[0014] Unlike existing methods that use a single sensor for monitoring, this invention integrates three types of sensors. Specific installation locations and configuration examples are as follows: Angle sensor: Installed on the valve rotating shaft or handwheel, used to collect the valve rotation angle in real time; optionally, the sampling frequency of the angle sensor is ≥10Hz, the angle resolution is ≤1°, and the measurement range is 0°~360°. When dealing with multi-turn valves, a multi-turn absolute encoder can be selected.

[0015] Vibration sensor: Installed on the valve body or lock base, used to collect vibration signals generated during valve operation, with a frequency response range of 10Hz~5kHz and a sensitivity of ≥100mV / g.

[0016] Acoustic sensor: Installed inside the lock housing near the valve, used to collect sound signals during valve operation, with a frequency response range of 20Hz~20kHz and a signal-to-noise ratio of ≥65dB.

[0017] Specifically, common communication methods such as Bluetooth are used to connect the aforementioned sensors, enabling real-time acquisition of angular characteristic data from the angle sensor, vibration characteristic data from the vibration sensor, and acoustic characteristic data from the acoustic sensor. These sensors are battery-powered and normally operate in low-power sleep mode, only reaching full speed when a change in the valve's rotation angle is detected or a wake-up command is received.

[0018] like Figure 2 As shown, during routine operation, the authorized computer key approaches the lock of the target valve via Bluetooth to activate the valve status monitoring device, actively collecting and uploading data from three sensors and other data; such as Figure 2 As shown, the displayed data includes: current valve angle, vibration noise baseline, battery level, and sensor self-test status.

[0019] In step S104, a first data feature criterion corresponding to normal authorized operation and a second data feature criterion corresponding to expected abnormal event are preset. Both the first data feature criterion and the second data feature criterion include angle feature criterion, vibration feature criterion and voiceprint feature criterion.

[0020] In this embodiment, in order to preset the first data feature criterion corresponding to normal authorized operation, the authorized personnel use a standard computer key to perform 10 or more normal authorized operations on the valve according to the normal operating procedure, such as standard open / close operation; the feature data of each operation process is collected, as shown in Table 1: Table 1. Valve characteristic data under normal authorized operation

[0021] In creating the first data feature criterion, a special library can be created based on the feature category. The specific parameters corresponding to the features under the feature category are compared with the standard data to establish the first data feature criterion. For example, the angular velocity curve of the current feature data is compared with the standard curve in the standard data. Other items are also compared one by one. When the similarity between the current feature data and the standard data is greater than or equal to 0.85, it is considered to meet the standard.

[0022] In some embodiments, anticipated abnormal events include external impact events and other events suspected of being due to misoperation. Taking external impact events as an example, a "external impact" scenario is simulated, such as striking a valve handwheel with a rubber hammer or bumping a tool against the valve body. Different types of feature data are collected to establish a "external force damage" feature library. For the data features of external impact events, corresponding preset features are used: vibration time-domain features to reflect impact intensity, vibration frequency-domain features to reflect impact source characteristics, acoustic fingerprint time-domain features to reflect impact sound characteristics, acoustic fingerprint frequency-domain features to reflect the impact sound spectrum, and angle correlation features to reflect the impact impact. The sensor sources and specific feature parameters of each data category in the "external force damage" feature library are shown in Table 2. Table 2. Data Composition of the "External Damage" Feature Database

[0023] Therefore, based on vibration time-domain features, vibration frequency-domain features, acoustic signature time-domain features, acoustic signature frequency-domain features, and angle correlation features, a second data feature criterion corresponding to an external force impact event is created. For example, if the vibration peak value in the vibration time-domain feature is >8m / s² and is an impact pulse with a small angle change (specifically, the angle change can be set to <2°), this can be combined with other data criteria to determine whether the valve has been impacted by an external force.

[0024] Furthermore, changes in valve data may also be caused by environmental background noise, equipment installation abnormalities, or sensor malfunctions. For example, background vibrations from fluid flow in pipelines or the operation of adjacent equipment, sensor malfunctions, loose installations, or external electromagnetic interference can cause slight fluctuations in vibration peak values ​​and acoustic signature energy. These fluctuations do not reach the energy threshold corresponding to valid operation, nor do they match the characteristic patterns of normal authorized operation or anticipated abnormal events, thus indicating no valid operation. Therefore, a third data feature criterion is preset to indicate no valve operation. This third data feature criterion includes angle feature criteria, vibration feature criteria, and acoustic signature criteria. The current feature data is matched against the third data feature criterion. If the match is successful, it indicates no valid operation, meaning the valve is in an inactive state. For example, if the rotation angle of the valve's rotating shaft does not change, the vibration peak value is less than the background noise threshold, and the acoustic signature energy does not change significantly, it is determined as "no valid operation," meaning no operation was performed or the operation was invalid. In this case, a "sensor malfunction" or "loose installation" report is submitted. If the sound energy during the operation period is less than 1.5 times the background energy and the sound pressure level increases by no more than 2dB, it is determined that there is no significant change in the sound signature energy.

[0025] Combination Figure 3As shown, the first data feature criterion, the second data feature criterion, and the third data feature criterion are combined to form a three-level cross-validation, which can identify different states of the valve. The preconditions for the three-level cross-validation are: the angle feature data is greater than a preset angle, which can be selected from 1° to 3°, preferably 2°.

[0026] Combining Tables 1 and 2, the contents of the three-level cross-validation are shown in Table 3: Table 3. Explanation of Three-Level Cross-Validation

[0027] In step S106, the current characteristic data of the valve is obtained. The current characteristic data includes the angle characteristic data of the angle sensor, the vibration characteristic data of the vibration sensor, and the acoustic characteristic data of the acoustic sensor.

[0028] In existing valve operation-related tasks, the operation ticket is a standardized operating document formulated according to industry safety regulations. It clearly stipulates the tasks, steps, personnel, and safety measures for valve operation. During routine monitoring, it checks whether an operation ticket for the corresponding valve has been received. If so, it indicates that the operator is about to perform valve operation according to the instructions on the operation ticket. Before the valve operation corresponding to the operation ticket is executed, the current characteristic data of the valve is acquired. Based on the current characteristic data, the current valve status is determined and compared with the target valve status required by the operation ticket. If they match, an alert indicating no operation is needed is triggered; otherwise, operation is permitted. In this way, the current valve status can be automatically verified to match the target status on the operation ticket before operation, effectively avoiding duplicate operations and reducing unnecessary work.

[0029] During valve operation, the valve opening degree is fed back in real time. For multi-turn valves, the number of turns is calculated by integrating the angular velocity. If the operator stops operation but the target position has not been reached, the computer key will issue a voice prompt: "Current opening degree is 65%, target is 100%, please continue operation."

[0030] After the valve operation is completed, the angle feature data from the angle sensor is acquired; based on the angle feature data, the valve operation is determined as follows: If the angle sensor's angle characteristics show that the angle change has stopped and remains unchanged for 2 seconds, it indicates that the valve operation is complete. The system then continuously monitors the valve angle for 10-30 seconds. If the valve's rotation shaft angle stabilizes at the target position (i.e., the deviation from the target value is ≤ ±2°), it is determined that "the operation was successful and the valve is in position." If the valve's rotation shaft angle rebounds, for example, from 0° back to 5°, it is determined that "the operation was not in place, and the valve may be stuck or not locked," and a maintenance prompt is generated. If the angle drifts slowly, such as changing by more than 0.5° every 10 seconds, it is determined that "the valve is leaking internally," and an early warning is issued. The identification results, along with the data from each sensor, are stored in the operation log and uploaded to the management platform.

[0031] In step S108, the current feature data is matched with the first data feature criterion and the second data feature criterion. Based on the matching result, it is determined whether the current valve operation is a normal authorized operation or a anticipated abnormal event. If it is a anticipated abnormal event, an early warning is triggered.

[0032] Furthermore, complete data from each operation is stored long-term to obtain historical characteristic data of the valve. Based on this historical characteristic data, a valve state change curve is constructed. The valve state change curve is then used to determine if any abnormal changes occur in the valve. At least two abnormal state characteristics and corresponding abnormal modes are preset. The valve state change curve is matched against these abnormal state characteristics, and the matching result determines whether the valve is in an abnormal mode. If it is in an abnormal mode, a corresponding warning action is provided.

[0033] During the matching process, time series analysis algorithms, such as moving average and exponential smoothing, are used to analyze and compare the data. The abnormal status characteristics and their corresponding abnormal patterns and warning actions are shown in Table 4. Table 4. Characteristics of State Anomalies and Their Corresponding Anomaly Patterns

[0034] Reference Figure 4 Another embodiment of the present invention proposes a valve operation early warning system 200 based on multi-source state cross-validation, including a memory 202, a processor 204, and a computer program stored in the memory 202 and executable on the processor 204. When the processor 204 executes the computer program, it implements the above-mentioned valve operation early warning method based on multi-source state cross-validation.

[0035] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A valve operation early warning method based on multi-source state cross-validation, characterized in that, include: The communication connection includes an angle sensor located on the valve rotation shaft, a vibration sensor located on the valve body, and an acoustic sensor located on the valve lock housing. The system pre-defines a first data feature criterion corresponding to normal authorized operations and a second data feature criterion corresponding to anticipated abnormal events. Both the first and second data feature criterions include angle feature criteria, vibration feature criteria, and voiceprint feature criteria. Obtain the current characteristic data of the valve, which includes the angle characteristic data of the angle sensor, the vibration characteristic data of the vibration sensor, and the acoustic characteristic data of the acoustic sensor; The current feature data is matched with the first data feature criterion and the second data feature criterion. Based on the matching result, it is determined whether the current valve operation is a normal authorized operation or a anticipated abnormal event. If it is the anticipated abnormal event, an early warning is triggered.

2. The valve operation early warning method based on multi-source state cross-validation according to claim 1, characterized in that, The anticipated abnormal events include external impact events; The system includes preset vibration time-domain features to reflect impact intensity, vibration frequency-domain features to reflect impact source characteristics, acoustic text time-domain features to reflect impact sound characteristics, acoustic text frequency-domain features to reflect impact sound spectrum, and angle correlation features to reflect impact impact effects. Based on the vibration time-domain features, the vibration frequency-domain features, the acoustic signature time-domain features, the acoustic signature frequency-domain features, and the angle correlation features, a second data feature criterion corresponding to the external force impact event is created.

3. The valve operation early warning method based on multi-source state cross-validation according to claim 1, characterized in that, Also includes: Determine whether an operation ticket for the corresponding valve has been received. If so, obtain the current characteristic data of the valve before the valve operation corresponding to the operation ticket is executed. The current valve status is determined based on the current feature data. The current valve status is compared with the target valve status required by the operation ticket. If they match, an alert that no operation is required is triggered; otherwise, operation is allowed.

4. The valve operation early warning method based on multi-source state cross-validation according to claim 3, characterized in that, Also includes: After the valve operation is completed, the angle feature data of the angle sensor is acquired; Based on the angular feature data, it is determined whether the valve operation is in place.

5. A valve operation early warning method based on multi-source state cross-validation according to claim 3, characterized in that, Also includes: During the execution of the valve operation, the valve opening degree is fed back in real time.

6. The valve operation early warning method based on multi-source state cross-validation according to claim 1, characterized in that, Also includes: Save the historical characteristic data of the valve, and construct the valve state change curve based on the historical characteristic data; Based on the valve status change curve, determine whether there are any abnormal changes in the valve.

7. A valve operation early warning method based on multi-source state cross-validation according to claim 6, characterized in that, The specific steps for determining whether the valve exhibits abnormal changes are as follows: At least two state anomaly features and anomaly patterns that correspond one-to-one with the state anomaly features are preset; The valve state change curve is matched with the state anomaly characteristics, and the valve is determined to be in the anomaly mode based on the matching result.

8. The valve operation early warning method based on multi-source state cross-validation according to claim 1, characterized in that, Also includes: A third data feature criterion is preset for when no valve operation is performed. The third data feature criterion includes angle feature criterion, vibration feature criterion, and voiceprint feature criterion. The current feature data is matched with the third data feature criterion. If the match is successful, it indicates that the valve is in an inactive state.

9. A valve operation early warning method based on multi-source state cross-validation according to claim 1, characterized in that, Also includes: Set a precondition for matching the current feature data with the first data feature criterion and the second data feature criterion; The prerequisite is that the angle feature data is greater than a preset angle.

10. A valve operation early warning system based on multi-source state cross-validation, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a valve operation early warning method based on multi-source state cross-validation as described in any one of claims 1 to 9.

Images