Method for determining threshold value of pneumatic ball valve failure experiment parameter

Abstract

The application is suitable for the technical field of pneumatic ball valves, and particularly relates to a method for determining a fault experiment parameter threshold value of a pneumatic ball valve, which comprises the following steps: obtaining an environmental working condition of the pneumatic ball valve, an initial leakage amount threshold value, and a fault parameter; improving the calibration accuracy of the leakage amount threshold value through multi-parameter calibration; combining a pre-trained threshold value prediction model to determine a threshold value correction coefficient of the initial leakage amount threshold value under the environmental working condition and the fault parameter, and obtaining a target threshold value, so as to dynamically update the leakage amount threshold value of the pneumatic ball valve. Through the above method, the problem that the leakage amount threshold value is fixed and cannot be adaptively adjusted according to the environmental working condition and the fault parameter in the prior art can be solved.

Description

Technical Field

[0001] This application belongs to the field of pneumatic ball valve technology, and in particular relates to a method for determining the threshold of experimental parameters for pneumatic ball valve failure. Background Technology

[0002] Pneumatic ball valves, as core actuators in industrial fluid control systems, are widely used in power, chemical, aerospace, and other fields. Especially under harsh conditions such as high temperature, high pressure, strong corrosion, and dust, they perform the functions of opening, closing, and regulating fluid media. Their operational reliability directly affects the safety, stability, energy consumption, and production efficiency of the entire industrial system. Leakage is the most common type of failure in pneumatic ball valves. When the system pressure reaches above 4.0 MPa and the medium temperature exceeds 300°C, the sealing rings of traditional ball valves are prone to developing tiny gaps due to thermal expansion and contraction, leading to steam leakage. This not only wastes energy but may also pose safety risks. Therefore, accurately determining the parameter thresholds for leakage fault diagnosis in the laboratory before the pneumatic ball valve leaves the factory is a crucial prerequisite for achieving early warning and timely investigation of leakage faults.

[0003] The main problem with existing methods for determining parameter thresholds in pneumatic ball valve leakage fault diagnosis is that traditional methods often use a fixed threshold mode, where the leakage threshold is fixed and cannot be adaptively adjusted according to environmental conditions and fault parameters. For example, the parameter thresholds of pneumatic ball valves that have been returned for repair need to be reconfirmed.

[0004] To address the shortcomings of the existing technologies, there is an urgent need for a method to determine the threshold values ​​for pneumatic ball valve fault test parameters that can be dynamically updated, linking environmental conditions and fault parameters, combining comprehensive fault levels to achieve accurate threshold correction and adaptation verification, in order to improve the accuracy of pneumatic ball valve leakage fault diagnosis. Summary of the Invention

[0005] This application provides a method for determining the threshold of experimental parameters for pneumatic ball valve failure, which can solve the problem that the leakage threshold is fixed in the prior art and cannot be adaptively adjusted according to environmental conditions and failure parameters.

[0006] In a first aspect, embodiments of this application provide a method for determining the threshold values ​​of experimental parameters for pneumatic ball valve failures, including: Obtain the environmental operating conditions, initial leakage threshold, and fault parameters of the pneumatic ball valve; Based on the environmental conditions, the fault parameters, and the pre-trained threshold prediction model, a threshold correction coefficient for the initial leakage threshold under the environmental conditions and the fault parameters is determined; wherein, the threshold correction coefficient is used to represent the speed of adaptive adjustment of the initial leakage threshold. Based on the threshold correction coefficient of the pneumatic ball valve and the initial leakage threshold, a target threshold is obtained; wherein, the target threshold represents the upper limit of leakage under the environmental conditions and the fault parameters.

[0007] The technical solutions described in this application embodiment have at least the following technical effects: The method for determining the threshold of experimental parameters for pneumatic ball valve faults provided in this application obtains the environmental operating conditions, initial leakage threshold, and fault parameters of the pneumatic ball valve. Based on the environmental operating conditions, fault parameters, and a pre-trained threshold prediction model, a threshold correction coefficient for the initial leakage threshold under the environmental operating conditions and the fault parameters is determined. The threshold correction coefficient represents the rate of adaptive adjustment of the initial leakage threshold. Based on the threshold correction coefficient and the initial leakage threshold of the pneumatic ball valve, a target threshold is obtained. The target threshold represents the upper limit of leakage under the current environmental operating conditions and the fault parameters. This application improves the verification accuracy of the leakage threshold through multi-parameter verification of the pneumatic ball valve's environmental operating conditions, initial leakage threshold, and fault parameters. Combined with a pre-trained threshold prediction model, the threshold correction coefficient for the initial leakage threshold under the environmental operating conditions and the fault parameters is determined, resulting in the target threshold, thereby dynamically updating the leakage threshold of the pneumatic ball valve. This method solves the problem of a fixed leakage threshold that cannot adaptively adjust with environmental operating conditions and fault parameters.

[0008] Secondly, embodiments of this application provide a device for determining the threshold of experimental parameters for pneumatic ball valve failures, applied to electronic devices, for implementing the method for determining the threshold of experimental parameters for pneumatic ball valve failures as described in the first aspect above. The device for determining the threshold of experimental parameters for pneumatic ball valve failures includes: The acquisition unit is used to acquire the environmental operating conditions, initial leakage threshold, and fault parameters of the pneumatic ball valve. The coefficient unit is used to determine the threshold correction coefficient of the initial leakage threshold under the environmental conditions and the fault parameters based on the environmental conditions, the fault parameters and the pre-trained threshold prediction model; wherein the threshold correction coefficient is used to represent the speed of adaptive adjustment of the initial leakage threshold. The target unit is used to obtain a target threshold based on the threshold correction coefficient of the pneumatic ball valve and the initial leakage threshold; wherein the target threshold represents the upper limit of leakage under the environmental conditions and the fault parameters.

[0009] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method as described in any of the foregoing aspects.

[0010] Fourthly, embodiments of this application provide a computer program product that, when run on an electronic device, causes the electronic device to perform the method described in any one of the first aspects above.

[0011] It is understood that the beneficial effects of the second to fourth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart illustrating a method for determining the threshold of experimental parameters for pneumatic ball valve failure according to an embodiment of this application. Figure 2 This is a schematic diagram of the operation of a method for determining the threshold of experimental parameters for pneumatic ball valve failure provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of the device for determining the threshold of experimental parameters for pneumatic ball valve failure provided in the embodiments of this application; Figure 4 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0014] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0015] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0016] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0017] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0018] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0019] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0020] In existing pneumatic ball valve leakage fault diagnosis technologies, the main problem with parameter threshold determination methods is that traditional methods often employ fixed threshold modes. This results in a fixed leakage threshold that cannot be adaptively adjusted according to environmental conditions and fault parameters. For example, it requires secondary confirmation of parameter thresholds for pneumatic ball valves undergoing repair. Unlike traditional manual or electric ball valves, the sealing performance of pneumatic ball valves depends not only on the fit between the valve seat and the ball but also on the output characteristics of the pneumatic actuator. Fluctuations in air source pressure, aging of internal actuator seals, and transmission clearance between the valve stem and actuator all directly affect the valve closing accuracy and sealing specific pressure, thus altering the leakage rate.

[0021] To address the shortcomings of the existing technologies, there is an urgent need for a method to determine the threshold values ​​for pneumatic ball valve fault test parameters that can be dynamically updated, linking environmental conditions and fault parameters, combining comprehensive fault levels to achieve accurate threshold correction and adaptation verification, in order to improve the accuracy of pneumatic ball valve leakage fault diagnosis.

[0022] To address the aforementioned issues, this application provides a method for determining the threshold values ​​of experimental parameters for pneumatic ball valve failures. This method includes: acquiring the environmental operating conditions, initial leakage threshold, and fault parameters of the pneumatic ball valve; determining a threshold correction coefficient for the initial leakage threshold under the environmental operating conditions and the fault parameters based on the environmental operating conditions, fault parameters, and a pre-trained threshold prediction model; wherein the threshold correction coefficient represents the rate of adaptive adjustment of the initial leakage threshold; and obtaining a target threshold based on the threshold correction coefficient and the initial leakage threshold of the pneumatic ball valve; wherein the target threshold represents the upper limit of leakage under the current environmental operating conditions and the fault parameters. This application improves the verification accuracy of the leakage threshold through multi-parameter verification of the pneumatic ball valve's environmental operating conditions, initial leakage threshold, and fault parameters. Combined with a pre-trained threshold prediction model, the threshold correction coefficient for the initial leakage threshold under the environmental operating conditions and the fault parameters is determined to obtain the target threshold, thereby dynamically updating the leakage threshold of the pneumatic ball valve. This method solves the problem of a fixed leakage threshold that cannot adaptively adjust with environmental operating conditions and fault parameters.

[0023] The method for determining the threshold of experimental parameters for pneumatic ball valve failure provided in this application embodiment can be applied to electronic devices. In this case, the electronic device is the executing subject of the method for determining the threshold of experimental parameters for pneumatic ball valve failure provided in this application embodiment. This application embodiment does not impose any restrictions on the specific type of electronic device.

[0024] For example, electronic devices can be programmable logic controllers, embedded industrial computers, edge computing gateways, dedicated valve diagnostic terminals, cloud servers, cloud hosts, commercial desktop computers, laptops, e-commerce dedicated smart terminals, tablet computers, etc. Electronic devices include memory, processors, and computer programs stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements methods as described above.

[0025] To better understand the method for determining the threshold of experimental parameters for pneumatic ball valve failure provided in this application, the specific implementation process of the method for determining the threshold of experimental parameters for pneumatic ball valve failure provided in this application will be described in the following exemplary manner.

[0026] Figure 1 A flowchart illustrating the method for determining the threshold of experimental parameters for pneumatic ball valve failure provided in an embodiment of this application is shown. Figure 2 This illustration shows a schematic diagram of the operation of the method for determining the threshold of experimental parameters for pneumatic ball valve failure provided in an embodiment of this application. The method for determining the threshold of experimental parameters for pneumatic ball valve failure includes: S100 obtains the environmental operating conditions, initial leakage threshold, and fault parameters of the pneumatic ball valve.

[0027] It can be understood that environmental operating conditions refer to all external environmental and media operating conditions that directly affect the sealing performance, valve seat wear rate, media flow state, and leakage amount of the pneumatic ball valve during actual operation in the industrial field. Specifically, these include, but are not limited to: ambient temperature (range -40℃ to 200℃, directly affecting the thermal expansion and contraction of the seals); media pressure (range 0.1MPa to 10MPa, affecting media permeability); media viscosity (range 1mPa・s to 100mPa・s, affecting leakage flow rate); air source pressure (range 0.4MPa to 0.8MPa, affecting the actuator's valve closing torque); and pipeline vibration intensity (range 0.1g to 1g, affecting valve seat contact accuracy). Environmental operating conditions can be obtained by real-time data collection using corresponding sensors (temperature sensor, pressure sensor, viscosity sensor, vibration sensor) installed on the ball valve body, pipeline, and surrounding environment. The sampling frequency is 1 time / second, and the collected data is transmitted to electronic equipment via PLC to improve data real-time performance and accuracy. The initial leakage threshold refers to the benchmark critical value used to initially determine whether leakage has occurred under the standard rated operating conditions specified by the factory (ambient temperature 25℃, medium pressure 0.6MPa, air source pressure 0.6MPa, medium viscosity 20mPa・s, no vibration, and no wear on the seals) of the pneumatic ball valve. It is jointly determined by the equipment's factory technical specifications, the industry standard GB / T13927-2008 "Pressure Test of Industrial Valves", and factory test data. The unit is mL / min (small flow scenario) or L / h (large flow scenario). The initial leakage threshold is determined by recording the maximum leakage under standard operating conditions without affecting normal operation through the leakage test at the factory. This value is used as the initial leakage threshold and is preset in the parameter library of the electronic equipment. It can be configured in batches or calibrated individually according to the ball valve model and specifications. Fault parameters refer to characteristic parameters that directly reflect the health status of key components of the pneumatic ball valve, such as the sealing structure, actuator, valve seat, and packing, and are directly related to leakage faults. Specifically, these include: seal wear (reflecting the contact gap between the valve seat and the ball); actuator lag time (reflecting the actuator response speed; excessive lag can easily lead to incomplete valve closure). In particular, for pneumatic ball valves, actuator lag time reflects the overall health status of the pneumatic pipeline, solenoid valve, and cylinder piston seals. Excessive lag time indicates insufficient actuator output torque or slow response, easily leading to incomplete valve closure. Position, thus causing leakage; the valve closing position deviation directly reflects the fitting accuracy between the valve seat and the ball, and its deviation value is affected by both the wear of the seal and the clearance of the actuator transmission mechanism; valve closing position deviation (reflecting the fitting accuracy between the valve seat and the ball); internal leakage flow characteristic value (reflecting the degree of internal leakage of the valve body); external leakage detection value (reflecting the degree of external leakage of parts such as the stuffing box); the fault parameter acquisition method can be to detect in real time by displacement sensors and flow sensors installed on the seal and actuator, and then integrate and transmit to the electronic equipment, so that the fault parameters cover the key leakage-related components of the ball valve.For example, for a certain model of pneumatic ball valve, the ambient temperature (85℃) and medium pressure (2.5MPa) are collected by a temperature sensor, the medium viscosity (35mPa·s) by a viscosity sensor, the air source pressure (0.65MPa) by an air source pressure sensor, and the pipeline vibration intensity (0.3g) by a vibration sensor. The initial leakage threshold of 5mL / min is read from the electronic equipment parameter library. The wear of the seal is indirectly estimated to be 0.28mm by analyzing the stroke-torque curve of the actuator. The stroke-torque curve is a function curve plotted with stroke as the abscissa and actuator output torque as the ordinate during the valve's transition from fully closed to fully open, reflecting the "variation of the required driving torque at different opening degrees". The actuator lag time is 0.3s obtained by a timer, the valve closing deviation is 0.15mm obtained by a displacement sensor, the internal leakage flow characteristic value is 1.2mL / min obtained by a flow sensor, and the external leakage detection value is 0.3mL / min obtained by a real-time acoustic emission sensor installed at the valve stem stuffing box. All data are integrated and imported into the electronic equipment.

[0028] It should be noted that during the opening and closing of a pneumatic ball valve, the valve stem torque is determined by the combined forces of the pneumatic actuator output torque overcoming the friction between the sealing ring and the ball, and the resistance torque generated by the medium pressure. When the sealing ring is in good condition, there is a large interference fit between the valve seat and the ball during the closing process, resulting in a significant steep rise in the torque curve at the end of the closing phase. As the sealing ring wears, the interference fit decreases, the elastic recovery ability of the sealing ring declines, and this steep rise gradually flattens out or even disappears. Simultaneously, the peak starting torque at the moment of opening also changes with the degree of wear. By calibrating the stroke-torque curves of the same model of ball valve under different wear levels in the laboratory beforehand, a mapping table of seal wear amount and torque characteristic values ​​can be established. In actual operation, by collecting real-time stroke-torque curves and extracting torque characteristic values, the seal wear amount can be deduced.

[0029] S200, based on environmental conditions, fault parameters, and a pre-trained threshold prediction model, determines the threshold correction coefficient for the initial leakage threshold under the given environmental conditions and fault parameters. The threshold correction coefficient represents the rate at which the initial leakage threshold is adaptively adjusted.

[0030] It is understandable that a threshold prediction model refers to a prediction model that is pre-trained using machine learning algorithms based on a large amount of pneumatic ball valve operating data under different working conditions and with different fault levels. The core function of the threshold prediction model is to establish a mapping relationship. The training method of the model is as follows: collect at least 1,000 sets of pneumatic ball valve operating data under different working conditions and with different fault levels, use environmental operating parameters and fault parameters as input features, use manually calibrated reasonable threshold correction coefficients as output labels, train the model using a random forest algorithm, verify the model after training, and then embed the model in electronic devices after verification. It is only used for inference and does not involve retraining on site. It can be fine-tuned according to the ball valve model and usage scenario. The threshold correction coefficient is a dimensionless adjustment coefficient ranging from 0.8 to 1.2. It is used to adjust the adaptive adjustment range and response speed of the initial leakage threshold according to the severity of the current environmental conditions and the severity of the fault parameters. The core logic of the threshold correction coefficient is: the more severe the operating conditions (such as high temperature, high pressure, high vibration) and the more severe the fault (such as large seal wear, large valve closing deviation), the larger the correction coefficient, the faster the adjustment speed and the more significant the adjustment range of the initial leakage threshold. Conversely, the smaller the correction coefficient, the smoother the adjustment. The threshold correction coefficient is determined by: using environmental conditions and fault parameters as input, obtaining a preliminary value through threshold prediction model reasoning, and then verifying its rationality to finally determine the usable threshold correction coefficient.

[0031] As an optional embodiment of this application, S200, based on environmental conditions, fault parameters, and a pre-trained threshold prediction model, determines the threshold correction coefficient of the initial leakage threshold under the environmental conditions and fault parameters, including: S210 performs parameter decomposition and normalization on environmental operating conditions and fault parameters to obtain multiple normalized operating condition parameters and normalized fault parameters.

[0032] It can be understood that parameter decomposition refers to breaking down mixed environmental operating condition data and mixed fault parameter data into independent, quantifiable, and individually processable single parameters according to parameter type and physical meaning. This avoids interference between different physical quantities and different types of parameters, and improves the accuracy of subsequent processing. The parameter decomposition is implemented by using the parameter decomposition algorithm built into the electronic device to split the mixed data according to preset parameter classification rules (such as environmental and fault categories). For example, environmental operating condition data can be split into five single operating condition parameters: ambient temperature, medium pressure, medium viscosity, gas source pressure, and pipeline vibration intensity. Fault parameters can be split into five single fault parameters: seal wear, actuator action lag time, valve closing deviation, internal leakage flow characteristic value, and external leakage detection value. After splitting, the data is stored and processed separately. Normalization refers to mapping the individual parameters (with different dimensions and numerical ranges, such as temperature (°C), pressure (MPa), wear (mm), and flow rate (mL / min)) to the standard dimensionless interval [0, 1] using a linear normalization algorithm. This eliminates differences in dimensions and numerical magnitudes, making different types of parameters comparable and facilitating subsequent classification, model input, and weighted calculation. The normalization process is implemented using the linear normalization formula: Normalized value = (Actual parameter value - Minimum parameter value) ÷ (Maximum parameter value - Minimum parameter value). The minimum and maximum values ​​of each parameter are determined by the rated operating range of the pneumatic ball valve, industry standards, and field operating experience (e.g., minimum ambient temperature -40°C, maximum 200°C; minimum medium pressure 0.1MPa, maximum 10MPa; minimum seal wear 0mm, maximum 1mm). The electronic equipment automatically calls the linear normalization formula to calculate each individual parameter and outputs the normalized parameter value. Normalized operating condition parameters refer to dimensionless individual operating condition parameters obtained after parameter decomposition and linear normalization. Their values ​​range from [0, 1]. Each normalized operating condition parameter corresponds to an original individual operating condition parameter, used for subsequent level classification and model input. Normalized operating condition parameters are obtained by performing linear normalization on the decomposed individual operating condition parameters. Normalized fault parameters refer to dimensionless individual fault parameters obtained after parameter decomposition and linear normalization. Their values ​​range from [0, 1]. Each normalized fault parameter corresponds to an original individual fault parameter, used for subsequent level classification and model input. Normalized fault parameters are obtained by performing linear normalization on the decomposed individual fault parameters.For example, the environmental conditions are broken down into ambient temperature (85℃), medium pressure (2.5MPa), medium viscosity (35mPa·s), gas source pressure (0.65MPa), and pipeline vibration intensity (0.3g); the fault parameters are broken down into seal wear (0.28mm), actuator lag time (0.3s), valve closing deviation (0.15mm), internal leakage flow characteristic value (1.2mL / min), and external leakage detection value (0.3mL / min); then normalization is performed, where the minimum value of the ambient temperature in the environmental conditions is considered. -40℃, maximum 200℃, normalized value = (85 - (-40)) ÷ (200 - (-40)) = 125 ÷ 240 ≈ 0.52; minimum medium pressure 0.1MPa, maximum 10MPa, normalized value = (2.5 - 0.1) ÷ (10 - 0.1) = 2.4 ÷ 9.9 ≈ 0.24; medium viscosity 35mPa・s: normalized value = (35 - 1) ÷ (100 - 1) = 34 ÷ 99 ≈ 0.34; gas source pressure 0.65MPa: normalized value = (0.65-0.4)÷(0.8-0.4)=0.25÷0.4=0.635≈0.63; Pipeline vibration intensity 0.3g: Normalized value=(0.3-0.1)÷(1-0.1)=0.2÷0.9≈0.22; The minimum value of seal wear in the fault parameters is 0mm, and the maximum value is 1mm, normalized value=(0.28-0)÷(1-0)=0.28; Actuator action lag time is 0.3s, minimum value is 0s, maximum value is 1s, normalized value=(0.3-0.4)÷(0.8-0.4)≈0.25÷0.4≈0.635≈0.63; Pipeline vibration intensity 0.3g: Normalized value=(0.3-0.1)÷(1-0.1)≈0.22; 3-0)÷(1-0)=0.3; Valve closing deviation 0.15mm: Normalized value=(0.15-0)÷(1-0)=0.15; Internal leakage flow characteristic value minimum 0mL / min, maximum 5mL / min, normalized value=(1.2-0)÷(5-0)=0.24; External leakage detection value 0.3mL / min, external leakage detection value minimum 0mL / min, maximum 1mL / min: normalized value=(0.3-0)÷(1-0)=0.3. Finally, the normalized operating condition parameters (0.52, 0.24, 0.34, 0.63, 0.22) and normalized fault parameters (0.28, 0.3, 0.15, 0.24, 0.3) are obtained.

[0033] S220 classifies multiple normalized operating condition parameters and normalized fault parameters into a comprehensive fault level.

[0034] It can be understood that level classification refers to classifying each normalized operating condition parameter and each normalized fault parameter into three distinct abnormality levels: "minor," "moderate," and "severe," based on the magnitude of the normalized value and according to a preset single-parameter level judgment range. This achieves a quantitative mapping from continuous dimensionless values ​​to discrete fault severity levels. The level classification is implemented as follows: the electronic equipment compares the actual value of each normalized parameter with the preset single-parameter level judgment range. Based on the range in which the value falls, the single-parameter level corresponding to each normalized operating condition parameter and normalized fault parameter is determined. Each parameter is classified into a level independently, without affecting each other. The comprehensive fault level refers to a unique level that comprehensively characterizes the health status and leakage risk level of the pneumatic ball valve by considering all normalized operating condition parameters and the single parameter levels corresponding to all normalized fault parameters, and by performing a fusion judgment on all single parameter levels according to the preset logic of "fault parameter priority, operating condition parameter auxiliary, severity level priority, and multi-parameter anomaly superposition to increase the level". It is divided into three levels: minor fault level, moderate fault level, and severe fault level. The comprehensive fault level can be obtained by the electronic equipment performing logical operations and fusion judgments on all single parameter levels according to the preset comprehensive judgment rules, and finally outputting a comprehensive fault level. The comprehensive fault level directly reflects the current overall leakage risk of the ball valve. For example: normalize the operating condition parameters (0.52, 0.24, 0.34, 0.63, 0.22), normalize the fault parameters (0.28, 0.3, 0.15, 0.24, 0.3), first classify each parameter into levels to obtain the single parameter levels: ambient temperature (0.52 for moderate), medium pressure (0.24 for slight), medium viscosity (0.34 for moderate), gas source pressure (0.63 for moderate), pipeline vibration (0.22 for slight); seal wear (0.28 for slight), actuator hysteresis (0.3 for slight), valve closing deviation (0.15 for slight), internal leakage (0.24 for slight), external leakage (0.3 for slight); then, according to the comprehensive judgment rules, all fault parameters are of the slight level, there are 3 moderate levels and 2 slight levels among the operating condition parameters, and no severe level, so the final comprehensive fault level is slight fault level.

[0035] In one possible implementation, S220, multiple normalized operating condition parameters and normalized fault parameters are classified into levels to obtain a comprehensive fault level, including: S221, based on the rated operating standard of pneumatic ball valves, obtains the single parameter level judgment range.

[0036] It is understandable that the rated operating standard of a pneumatic ball valve refers to the normal operating range, warning range, and danger range of each individual parameter of the pneumatic ball valve, jointly determined by the technical specifications of the pneumatic ball valve at the time of its manufacture, industry standards such as GB / T13927-2008 "Industrial Valves Pressure Test" and GB / T24925-2010 "Industrial Valves Terminology and Definitions", and statistics of failure cases. It is the core basis for classifying the level of each single parameter. The rated operating standard of a pneumatic ball valve is determined by the equipment manufacturer and on-site technicians in combination with the ball valve model, specifications, and usage scenario, clarifying the reasonable operating range of each individual parameter, and then combining failure cases to determine the numerical range corresponding to the degree of abnormality of each parameter. This value is fixed in the electronic equipment and can be fine-tuned periodically according to the on-site operating conditions. The single-parameter level judgment interval refers to a pre-defined [0, 1] interval used to determine the degree of abnormality (slight, moderate, severe) of each normalized chemical condition parameter and each normalized fault parameter. Each parameter corresponds to an independent judgment interval. The method for obtaining the single-parameter level judgment interval is as follows: based on the rated operating standard of the pneumatic ball valve, the normal, warning, and danger ranges of each single parameter are converted into a [0, 1] interval through linear normalization to obtain the single-parameter level judgment interval. For example, for normalized fault parameters and normalized chemical condition parameters... The leakage parameters (which directly affect leakage, and the range can be appropriately strict) are: slight level [0, 0.3], moderate level (0.3, 0.7], and severe level (0.7, 1.0]. All ranges are non-overlapping and non-omitted, ensuring that each normalized value uniquely corresponds to a level. For example, for a certain model of pneumatic ball valve, based on the factory rated standard of the pneumatic ball valve, combined with industry standards and field failure cases, the single parameter level judgment range is determined as follows: internal leakage flow characteristic value (after normalization) slight [0, 0.3], moderate (0.3, 0.7], and severe (0.7, 1.0).

[0037] S222, the actual values ​​of each normalized working condition parameter and normalized fault parameter are compared with the single parameter level judgment range to obtain the single parameter level corresponding to each normalized working condition parameter and normalized fault parameter.

[0038] It can be understood that the actual values ​​of each normalized operating condition parameter and normalized fault parameter refer to the independent, dimensionless parameter values ​​obtained after parameter decomposition and normalization. These values ​​are the direct basis for level determination, and each value is within the interval [0, 1]. The single parameter level refers to the degree of abnormality corresponding to a single normalized operating condition parameter or a single normalized fault parameter, based on the single parameter level determination interval into which the actual normalized value of the single parameter level falls. These levels are divided into minor, moderate, and severe levels. Each parameter corresponds to a unique single parameter level, reflecting the individual degree of abnormality of the single normalized operating condition parameter or a single normalized fault parameter, without involving the influence of other parameters. The single parameter level is obtained by comparing the actual value of each normalized parameter with the single parameter level determination interval to determine which interval the value falls into, thus determining the single parameter level of each normalized operating condition parameter and normalized fault parameter. For example, normalized parameters (ambient temperature 0.52, medium pressure 0.24, seal wear 0.28, internal leakage flow characteristic value 0.24) and single parameter level determination intervals are used. The actual values ​​of each parameter are compared with the intervals: ambient temperature 0.52 falls within (0.3, 0.7], corresponding to a moderate level; medium pressure 0.24 falls within [0, 0.3], corresponding to a slight level; seal wear 0.28 falls within [0, 0.3], corresponding to a slight level; internal leakage flow characteristic value 0.24 falls within [0, 0.3], corresponding to a slight level; finally, the single parameter level corresponding to each parameter is obtained.

[0039] S223, comprehensively judges all individual parameter levels to determine the overall fault level.

[0040] It can be understood that all single-parameter levels refer to the single-parameter levels corresponding to all normalized operating condition parameters and all normalized fault parameters (such as the levels of 5 operating condition parameters and 5 fault parameters). These levels are the basic units for comprehensive fault level determination, covering abnormal information of all key parameters of ball valve operation. The comprehensive fault level refers to the level obtained through comprehensive determination, which can comprehensively characterize the leakage risk of the pneumatic ball valve. It is divided into minor fault level, moderate fault level, and severe fault level. The core meaning of the comprehensive fault level is: minor fault level (the ball valve is in good overall condition, with no obvious leakage risk, only a few operating condition parameters are slightly abnormal, and all fault parameters are normal), moderate fault level (the ball valve has a certain leakage risk, some fault parameters are moderately abnormal, or multiple operating condition parameters are moderately abnormal), and severe fault level (the ball valve has a serious leakage risk, some fault parameters are severely abnormal, or operating condition parameters are severely abnormal and accompanied by moderately abnormal fault parameters). The comprehensive fault level is obtained by executing comprehensive determination logic through electronic equipment and outputting a unique comprehensive fault level. For example: Single parameter level (operating condition parameters: moderate ambient temperature, slight medium pressure, moderate medium viscosity, moderate gas source pressure, slight pipeline vibration; fault parameters: slight seal wear, slight actuator hysteresis, slight valve closing deviation, slight internal leakage, slight external leakage), according to the comprehensive judgment rules, all fault parameters are of the slight level, there are 3 moderate levels and 2 slight levels among the operating condition parameters, and no severe level. Therefore, the comprehensive fault level of the current environmental operating conditions and fault parameters is determined to be the slight fault level.

[0041] By adopting the above steps S221 to S223, it is helpful to solve the technical problems of vague fault level judgment standards, one-sided single parameter judgment, and lack of unified and implementable rules in the existing technology, so as to realize the quantification and accurate judgment of the overall fault level from scattered parameters, and improve the accuracy and reliability of threshold correction.

[0042] For example, S223, a comprehensive judgment is made on all individual parameter levels to determine the overall fault level, including: S2231, if any normalized fault parameter is a severe level in the single parameter level, the pneumatic ball valve is determined to be at the severe fault level in the comprehensive fault level.

[0043] It can be understood that any normalized fault parameter refers to any one of all normalized fault parameters (such as the wear of the seal, the lag time of the actuator, the deviation of the valve closing position, etc.). As long as one fault parameter meets the condition, the judgment can be triggered. The severity level in a single parameter level refers to the actual value of the normalized fault parameter falling within a preset severity level range (e.g., the severity level range of fault parameters [0.7, 1.0]). The corresponding single parameter level indicates that the component corresponding to the normalized fault parameter has severely failed, directly leading to a serious leakage risk. The severity level in a single parameter level is determined by interval comparison. Judgment refers to the electronic device determining whether the condition "any normalized fault parameter is of severity level" is met. If it is met, the overall fault level is directly output as a severe fault level. The judgment is implemented by the electronic device traversing the single parameter levels of all normalized fault parameters and determining whether a severity level exists. If it exists, the severe fault level is immediately output. For example, if the normalized value of one of the normalized fault parameters (e.g., the wear of the seal) is 0.82, and it is determined to be of severity level by comparing it with the preset severity level range of fault parameters [0.7, 1.0], then regardless of the levels of other operating parameters or other fault parameters, the electronic device directly determines the overall fault level as a severe fault level.

[0044] S2232, if there is no normalized fault parameter that is at the severe level in the single parameter level, but any normalized operating condition parameter is at the severe level in the single parameter level, and at least one normalized fault parameter is at the moderate level in the single parameter level, the pneumatic ball valve is determined to be at the severe fault level in the comprehensive fault level.

[0045] It is understandable that "no normalized fault parameter is of a severe level" means that the individual parameter levels of all obtained normalized fault parameters are either mild or moderate, and no single fault parameter is of a severe level. This can be determined by having the electronic device iterate through the individual parameter levels of all normalized fault parameters and confirm that no severe level is found. "Any normalized operating condition parameter is of a severe level" means that any single parameter of any of the obtained normalized operating condition parameters is of a severe level (the normalized value falls within the severe level range of the operating condition parameter, such as [0.7, 1.0]). A severe level indicates that the current environmental conditions are extremely harsh, which will exacerbate the risk of leakage. The method for determining if any normalized operating condition parameter is of a severe level is through interval comparison. "At least one normalized fault parameter is of a moderate level" means that at least one parameter among all obtained normalized fault parameters has an individual parameter level of a moderate level (the normalized value falls within the moderate level range of the fault parameter, such as [0.3, 0.7]). A moderate level indicates that the component corresponding to the normalized fault parameter has malfunctioned and will further deteriorate under extreme operating conditions. A serious leak is caused by at least one normalized fault parameter being classified as moderate. The determination method is as follows: the electronic equipment traverses the individual parameter levels of all normalized fault parameters, confirming at least one moderate level. For example: all normalized fault parameters have individual parameter levels of slight or moderate (no severe level), one normalized operating condition parameter (such as medium pressure) has a normalized value of 0.85, which, according to the severe level range (0.7, 1.0), is determined to be severe. Simultaneously, another normalized fault parameter (such as internal leakage flow characteristic value) has a normalized value of 0.45, which, according to the moderate level range (0.3, 0.7), is determined to be moderate. Since all three conditions are met, the overall fault level is determined to be severe.

[0046] S2233, if any normalized fault parameter is at the medium level in the single parameter level, the pneumatic ball valve is determined to be at the medium fault level in the comprehensive fault level.

[0047] It can be understood that any normalized fault parameter being of a medium level means that among all the obtained normalized fault parameters, the single parameter level of any one parameter is of a medium level (the normalized value falls within the medium level range of fault parameters, such as (0.3, 0.7)). A medium fault level indicates that the component corresponding to the normalized fault parameter has malfunctioned and will have a certain risk of leakage, but has not reached the level of serious failure. The method for determining any normalized fault parameter to be of a medium level is: obtained through interval comparison. For example, if the normalized value of one of the normalized fault parameters (such as valve closing deviation) is 0.42, and it is determined to be of a medium level by comparing it with the medium level range of fault parameters (0.3, 0.7), then regardless of the level of other operating condition parameters, the electronic equipment will determine the overall fault level to be a medium fault level.

[0048] S2234 If there is no normalized fault parameter that is at the medium level in the single parameter level, but at least four normalized operating condition parameters correspond to single parameter levels that are at the medium level in the single parameter level, the pneumatic ball valve is determined to be at the medium fault level in the comprehensive fault level.

[0049] It is understandable that "no normalized fault parameters of moderate severity" means that the single-parameter levels of all obtained normalized fault parameters are of the minor level, with no fault parameter of moderate or severe severity. "At least four normalized operating condition parameters of moderate severity" means that among all obtained normalized operating condition parameters, four or more parameters have a single-parameter level of moderate severity. The situation of at least four normalized operating condition parameters of moderate severity indicates that multiple environmental operating conditions are simultaneously deviating from the normal range. Although the component does not show obvious abnormalities, the superposition of multiple abnormal operating conditions increases the risk of leakage. The determination method can be: the electronic device iterates through the single-parameter levels of all normalized operating condition parameters and counts the number of moderate levels. If the number is ≥4, then the condition is met and a moderate fault level is output. For example: all normalized fault parameters are of the minor level, and the normalized values ​​of the ambient temperature (0.52) and medium viscosity (0.34) parameters fall into the moderate level range of the operating condition parameters [0.3, 0.7], with a moderate level count of 2, which does not meet the condition. Therefore, the next step of judgment is required.

[0050] S2235, if all normalized fault parameters are minor in the single parameter level, and the number of moderate levels in the single parameter level corresponding to the normalized chemical condition parameters is less than four and there is no severe level, the pneumatic ball valve is determined to be at the minor fault level in the comprehensive fault level.

[0051] It is understandable that "all normalized fault parameters are of the minor level" means that the individual parameter levels of all normalized fault parameters are of the minor level. This indicates that all critical components of the ball valve related to leakage are in normal condition with no obvious abnormalities. This can be confirmed by traversing the individual parameter levels of all normalized fault parameters using electronic equipment. "Less than four moderate levels and no severe levels" means that none of the individual parameter levels of all normalized operating conditions are of the severe level, and the number of moderate levels is less than four. This indicates that the environmental operating conditions are basically within the normal range, with only a few operating condition parameters deviating slightly, which will not cause significant leakage risk. This can be confirmed by traversing the individual parameter levels of all normalized operating conditions using electronic equipment to count the number of moderate levels and confirm that there are no severe levels. For example: the individual parameter levels of all normalized fault parameters (seal wear, actuator hysteresis, etc.) are all minor. Among the normalized operating condition parameters, only the ambient temperature (0.52) is at the moderate level, and the number of moderate level parameters is 1 (less than four). Moreover, no operating condition parameter is at the severe level. The conditions are met, so the overall fault level is determined to be minor.

[0052] By adopting the above steps S2231 to S2235, it is helpful to solve the technical problems of chaotic comprehensive fault level determination logic, lack of clear priority, incomplete coverage of scenarios, and poor operability in the existing technology. It constructs a comprehensive fault level determination rule that fits the leakage mechanism of pneumatic ball valves, prioritizes fault parameters, is rigorous, complete, and feasible, so that the comprehensive fault level can be accurately determined for each operating scenario, avoiding omissions and misjudgments. At the same time, it clarifies the priority of the impact of different parameters on leakage risk, provides a reliable basis for the accurate determination of subsequent threshold correction coefficients, and improves the practicality and feasibility of the entire fault diagnosis solution.

[0053] S230, based on the comprehensive fault level and threshold prediction model, determines the threshold correction coefficient of the initial leakage threshold under environmental conditions and fault parameters.

[0054] It can be understood that the comprehensive fault level refers to the minor, moderate, or severe level that comprehensively characterizes the leakage risk of the pneumatic ball valve. It is the core input feature of the threshold prediction model and directly determines the approximate range of the correction coefficient (severe levels correspond to larger correction coefficients, and minor levels correspond to smaller correction coefficients). The comprehensive fault level is obtained through level classification and comprehensive judgment. The threshold prediction model is a pre-trained random forest machine learning model. Its core function is to infer and output suitable threshold correction coefficients based on the input feature parameters. The threshold prediction model is invoked by the electronic device importing the comprehensive fault level and related parameters into the model, triggering model inference. The threshold correction coefficient is a dimensionless coefficient used to adjust the adaptive adjustment speed of the initial leakage threshold, ranging from 0.8 to 1.2. The value of the threshold correction coefficient is positively correlated with the comprehensive fault level. The threshold correction coefficient is obtained by using the comprehensive fault level as the core input, combined with normalized operating condition parameters and normalized fault parameters, inputting them into the threshold prediction model, obtaining preliminary correction coefficients through model inference, and then verifying their rationality before finally determining the usable threshold correction coefficients. For example, the comprehensive fault level (minor fault level) is input together with the normalized operating condition parameters and normalized fault parameters into a pre-trained threshold prediction model. Based on the built-in mapping relationship, the model infers and outputs an initial correction coefficient of 0.87. After verification (0.87 is in the range of 0.8 to 1.2 and matches the minor fault level), the final threshold correction coefficient is determined to be 0.87. The threshold correction coefficient indicates that the initial leakage threshold will be adaptively adjusted at a slower rate to adapt to the current minor fault and basically normal operating conditions.

[0055] By adopting the above steps S210 to S230, it is helpful to solve the technical problems in the prior art where the threshold correction coefficient is not determined based on any basis, does not conform to the actual working conditions and fault states, and has poor operability. It transforms the complex multi-source working conditions and fault information into standardized inputs that the model can recognize, realizes the adaptive and intelligent determination of the correction coefficient, and enables the correction coefficient to accurately match the current leakage risk.

[0056] In one possible implementation, S230, based on a comprehensive fault level and threshold prediction model, determines the threshold correction coefficient for the initial leakage threshold under environmental conditions and fault parameters, including: S231, based on the comprehensive fault level, integrates to obtain the input parameter set of the threshold prediction model.

[0057] It is understandable that the comprehensive fault level is the core feature of the input parameter set. The comprehensive fault level needs to be quantified into a numerical value that the model can recognize (e.g., 0 for minor faults, 1 for moderate faults, and 2 for severe faults) to facilitate model inference. The comprehensive fault level quantification method is as follows: the electronic device converts the comprehensive fault level into a corresponding numerical value according to a preset quantization rule and embeds it in the model. The input parameter set refers to the structured feature vector that the threshold prediction model can recognize and process. It consists of the quantized comprehensive fault level, all normalized operating condition parameters, and all normalized fault parameters. The structure of the input parameter set is "quantized comprehensive fault level value + normalized operating condition parameter 1 + normalized operating condition parameter 2 + ... + normalized fault parameter 1 + normalized fault parameter 2 + ...", ensuring that the input parameter set is complete, standardized, and without redundancy or missing values. The input parameter set is obtained by: the electronic device first quantifies the comprehensive fault level, then integrates the quantized value with all normalized operating condition parameters and normalized fault parameters in a preset order to form a structured input parameter set. Simultaneously, the parameter set is standardized in format to prevent the model from being unable to recognize it. For example, the comprehensive fault level (minor fault level, quantified as 0) and normalized operating condition parameters (0.52, 0.24, 0.34, 0.63, 0.22), and normalized fault parameters (0.28, 0.3, 0.15, 0.24, 0.3) are integrated in the order of "comprehensive fault level quantification value + normalized operating condition parameter + normalized fault parameter" to obtain the input parameter set [0, 0.52, 0.24, 0.34, 0.63, 0.22, 0.28, 0.3, 0.15, 0.24, 0.3]. After format standardization, this is used as the model input.

[0058] S232, input the input parameter set into the pre-trained threshold prediction model, and output preliminary correction coefficients that are adapted to the input parameter set. The threshold prediction model is a machine learning model that takes the input parameter set as input and outputs the preliminary correction coefficients.

[0059] The input parameter set refers to a standardized structured feature vector. The format, parameter type, and arrangement of the input parameter set are consistent with those used during model training, allowing the model to directly recognize and process it. The input parameter set is imported by the electronic device through a pre-defined interface, transmitting it to the threshold prediction model and triggering its inference calculations. The pre-trained threshold prediction model is a machine learning model that has been trained and validated using a random forest algorithm with a large number of environmental conditions, fault parameters, and threshold correction coefficients. It is then embedded in the electronic device. The core structure of the threshold prediction model includes an input layer, a hidden layer, and an output layer. The input layer receives the input parameter set, the hidden layer performs feature extraction and computation, and the output layer outputs preliminary correction coefficients. The model does not require retraining in the field; it is only used for inference. Model parameters (such as adjusting feature weights) can be periodically fine-tuned based on the field performance. The preliminary correction coefficient refers to the threshold correction coefficient directly output by the threshold prediction model that is compatible with the current input parameter set. It is an intermediate result that has not undergone reasonableness verification, and its value may exceed the reasonable range of 0.8 to 1.2, or it may not match the current comprehensive fault level. The preliminary correction coefficient is obtained as follows: after receiving the input parameter set, the model uses built-in feature extraction and logical reasoning algorithms to output a preliminary correction coefficient, which is then directly transmitted to the electronic device. For example, given the input parameter set [0, 0.52, 0.24, 0.34, 0.63, 0.22, 0.28, 0.3, 0.15, 0.24, 0.3], inputting this parameter set into a pre-trained random forest threshold prediction model, the model, through feature extraction and logical reasoning, outputs a preliminary correction coefficient of 0.87.

[0060] It should be noted that the threshold prediction model is a regression-type random forest (because the initial correction coefficients output are continuous values, not classification labels), and it has already been trained: Training phase: A large amount of sample data consisting of "input parameter set (11-dimensional features) - actual correction coefficients (labels)" is used to train a random forest composed of N decision trees (N=100 trees in the example, 50-200 trees are commonly used in industrial scenarios); Structure of each decision tree: Based on the feature splitting of the training samples (such as selecting splitting features and splitting thresholds by minimizing the mean squared error), a decision path of "feature judgment - branch - leaf node value" is formed; Model solidification: After training, the splitting rules of each decision tree and the mean value of the correction coefficients of the leaf nodes are fixed. During inference, it is only necessary to follow the rules to follow the path, without the need for retraining.

[0061] Taking the input parameter set X=[x0=0, x1=0.52, x2=0.24, x3=0.34, x4=0.63, x5=0.22, x6=0.28, x7=0.3, x8=0.15, x9=0.24, x10=0.3] (a total of 11 features, corresponding to environmental conditions, fault parameters, etc.) as an example, the model inference calculation steps are as follows: Step 1: Input parameter set preprocessing (built into the model, executed automatically) The 11-dimensional features of the input have been standardized (consistent with those during training). The model does not need to process them separately. It directly assigns X to each decision tree in the random forest (a total of 100 trees), and each tree independently performs inference on X. Step 2: Inference Calculation of a Single Decision Tree (Core: Following the decision path to obtain the output value of a single tree) Each tree in a random forest has a different structure. Here, we take the first decision tree as an example and break down its inference process (the logic of the other 99 trees is the same, only the splitting features / thresholds differ): Root Node Splitting (First Layer Judgment): The root node splitting feature of the first decision tree is x4 (the 5th dimension, value 0.63), and the splitting threshold is 0.60 (the optimal splitting value determined during training); Judgment rule: If x4 ≥ 0.60, take the right branch; otherwise, take the left branch; Calculation: 0.63 ≥ 0.60, then take the right branch. Second Layer Node Splitting: The child node splitting feature of the right branch is x1 (the 2nd dimension, value 0.52), and the splitting threshold is 0.50; Judgment rule: If x1 ≥ 0.50, take the right branch; otherwise, take the left branch; Calculation: 0.52 ≥ 0.50, then take the right branch. The third-level node split: The third-level node split feature is x6 (7th dimension, value 0.28), and the split threshold is 0.30; the judgment rule is: if x6 ≥ 0.30, take the right branch; otherwise, take the left branch; calculation: if 0.28 < 0.30, take the left branch. Reaching the leaf node (terminating inference): The left branch directly points to the leaf node, and the "mean of correction coefficient" stored in the leaf node is 0.86 (the mean of the labels of all samples falling into the leaf node during training); therefore, the output value of the first decision tree for input X is 0.86. Step 3: Summarize the inference results of all decision trees. Perform the inference of step 2 on each of the 100 decision trees one by one. Each tree will output a correction coefficient value of a leaf node. The output value distribution of the 100 trees in the example (partial) is as follows: 20 trees output: 0.86, 30 trees output: 0.87, 25 trees output: 0.88, 15 trees output: 0.85, 10 trees output: 0.89. Step 4: Final aggregation calculation of the random forest (resulting in 0.87). The aggregation rule for the regression-type random forest is the arithmetic mean of the output values ​​of all decision trees. The initial correction coefficient is calculated using the following formula: ,in, Indicates the first The output value of the decision tree, This represents the total number of decision trees. Substitute the example data to calculate: =0.8695≈0.87.

[0062] S233, verify the preliminary correction coefficients to obtain the threshold correction coefficients under environmental conditions and fault parameters.

[0063] It is understandable that the preliminary correction coefficient refers to the intermediate correction coefficient directly output by the threshold prediction model. It may have problems such as exceeding the reasonable range or not matching the current comprehensive fault level. It is necessary to remove outliers through verification to improve reliability. Verification refers to the process by which electronic equipment checks the rationality and matching of preliminary correction coefficients according to preset verification rules, ensuring that the preliminary correction coefficients meet actual operational requirements. The verification rules include: Range verification: The preliminary correction coefficient must be within a reasonable range of 0.8 to 1.2; exceeding this range is considered abnormal. Matching verification: The preliminary correction coefficient must match the overall fault level (0.8 to 0.9 for minor faults, 0.9 to 1.1 for moderate faults, and 1.1 to 1.2 for severe faults); mismatch is considered abnormal. Stability verification: The difference between the preliminary correction coefficient and the previously output correction coefficient should be ≤0.1 to avoid excessive coefficient jumps that could affect the stability of threshold adjustment. The verification process is as follows: The electronic equipment sequentially performs the above three verifications. If the preliminary correction coefficient meets all verification rules, it is considered valid. If not, interpolation (based on the reasonable range of the current overall fault level, interpolation is used to obtain a coefficient that meets the requirements) for correction until all verification rules are met. The threshold correction coefficient refers to the final correction coefficient obtained after verification, which is stable, reliable, and adapted to the current environmental conditions and fault parameters. Its value is within the range of 0.8 to 1.2 and matches the overall fault level, allowing it to be directly used for subsequent adjustment of the initial leakage threshold. The threshold correction coefficient is obtained by verifying and correcting the initial correction coefficient to obtain the final usable threshold correction coefficient. For example: the initial correction coefficient is 0.87. Verification is performed as follows: Range verification: 0.87 is within the range of 0.8 to 1.2, which is acceptable; Matching verification: the current overall fault level is minor, and 0.87 is within the range of 0.8 to 0.9, which is acceptable; Stability verification: the previously output correction coefficient was 0.86, and the difference is 0.01 ≤ 0.1, which is acceptable; all verification rules are met, therefore the threshold correction coefficient is determined to be 0.87.

[0064] By adopting the above steps S231 to S233, it is helpful to solve the technical problems in the existing technology where the correction coefficients of the model output have outliers, do not match the actual working conditions, and have poor stability. This improves the accuracy, stability and reliability of the threshold correction coefficients, avoids subsequent threshold calculation errors caused by abnormal threshold correction coefficients, and thus improves the accuracy and feasibility of the entire fault diagnosis solution.

[0065] S300, based on the threshold correction coefficient of the pneumatic ball valve and the initial leakage threshold, yields the target threshold. The target threshold represents the upper limit of leakage under environmental conditions and fault parameters.

[0066] The threshold correction coefficient (e.g., 0.87) is used to adjust the initial leakage threshold to match the current environmental conditions and fault state. It is obtained through preprocessing, level determination, model inference, and verification. The initial leakage threshold (e.g., 5 mL / min) is the baseline leakage threshold under standard operating conditions. It is obtained by reading from the electronic equipment parameter library and is determined by factory and industry standards. The target threshold (or target threshold) is the final upper limit of leakage obtained after adjusting the threshold correction coefficient, matching the current actual environmental conditions and fault state. Its unit is the same as the initial leakage threshold (mL / min or L / h). It is the core criterion for diagnosing pneumatic ball valve faults; exceeding the target threshold indicates a leakage anomaly. The target threshold is obtained by dividing the threshold correction coefficient by the initial leakage threshold to obtain the initial correction threshold, then performing adaptation verification to obtain the target threshold. The target threshold accurately reflects the maximum allowable leakage of the ball valve under current environmental conditions and fault parameters, avoiding the problem of mismatch between fixed thresholds and environmental conditions. For example, with a threshold correction factor of 0.87 and an initial leakage threshold of 5 mL / min, the initial correction threshold of 5 ÷ 0.87 = 5.75 mL / min is obtained by division. Then, the initial correction threshold is adapted and verified. After the verification is successful, the target threshold of 6.34 mL / min is obtained. The target threshold indicates that the maximum allowable leakage of the ball valve under the current environmental conditions and minor fault conditions is 6.34 mL / min. If the target threshold is exceeded, it is judged as an abnormal leakage.

[0067] As an optional embodiment of this application, in step S300, a target threshold is obtained based on the threshold correction coefficient of the pneumatic ball valve and the initial leakage threshold, including: S310, divide the threshold correction coefficient of the pneumatic ball valve by the initial leakage threshold to obtain the initial correction threshold.

[0068] The threshold correction coefficient refers to the final correction coefficient (e.g., 0.87), which is dimensionless and used for coarse adjustment of the initial leakage threshold to initially adapt it to the current operating conditions and fault status. The initial leakage threshold refers to the baseline leakage threshold under standard operating conditions (e.g., 5 mL / min); it is obtained by reading it from the electronic equipment parameter library. The initial correction threshold is an intermediate threshold obtained through division, initially adapting to the current operating conditions. Its unit is the same as the initial leakage threshold, and it is an intermediate result that has not undergone adaptation verification, potentially leading to a mismatch with the current safe leakage range. It is obtained by dividing the initial leakage threshold by the threshold correction coefficient. For example, with a threshold correction coefficient of 0.87 and an initial leakage threshold of 5 mL / min, the electronic equipment performs a division operation: 5 × 0.87 = 5.75 mL / min, resulting in an initial correction threshold of 5.75 mL / min. This initial correction threshold is an intermediate value after preliminary coarse adjustment.

[0069] S320, the initial modified threshold is adapted and verified to obtain the target threshold.

[0070] It is understandable that adaptation verification refers to the process by which electronic devices perform rationality, safety, and matching verification on the initial correction threshold according to preset verification rules, so that the initial correction threshold is adapted to the current environmental conditions, comprehensive fault level, and preset safe leakage range. The verification rules include: safety range verification: the initial correction threshold must be within the preset safe leakage threshold range; if it exceeds the range, fine-tuning is required; operating condition matching verification: the initial correction threshold must match the current environmental conditions; fault level matching verification: the initial correction threshold must match the comprehensive fault level. The adaptation verification is implemented as follows: the electronic device performs the above verifications sequentially, and if the initial correction threshold meets all the rules, it is directly used as the target threshold.

[0071] By adopting the above steps S310 to S320, it is helpful to solve the technical problems in the existing technology, such as the single threshold adjustment, the lack of coarse and fine adjustment process, and the poor adaptability of the threshold to the actual working conditions. It can avoid misjudgment and missed judgment of faults caused by threshold deviation, and improve the accuracy and feasibility of fault diagnosis.

[0072] In one possible implementation, S320 involves adapting and verifying the initial modified threshold to obtain the target threshold, including: S321, calculate the threshold deviation rate between the initial correction threshold and the midpoint of the preset safe leakage threshold range. The threshold deviation rate reflects the degree of deviation from the initial correction threshold.

[0073] It can be understood that the initial correction threshold refers to the intermediate threshold obtained by multiplying the initial leakage threshold by the threshold correction coefficient. The preset safe leakage threshold range refers to the normal leakage allowable range that is pre-set based on the factory rated operating standard of pneumatic ball valves, the industry standard GB / T13927-2008 "Pressure Test of Industrial Valves", long-term field operating experience and failure case statistics, and is adapted to the current environmental conditions and comprehensive failure level. It is the core basis for judging whether the initial correction threshold is reasonable. The preset safe leakage threshold range can be determined by electronic equipment calling the corresponding preset range from the parameter library according to the current comprehensive failure level (minor, moderate, severe) and environmental conditions (temperature, pressure, viscosity, etc.). The upper and lower limits of the range can be dynamically adjusted (for example: under the minor failure level and medium temperature and pressure conditions, the preset safe leakage threshold range is 5 to 7.6 mL / min; under the moderate failure level, the preset safe leakage threshold range is 7.6 to 10 mL / min; under the severe failure level, the preset safe leakage threshold range is 10 to 15 mL / min), so that the range matches the actual operating state. The median value refers to the middle value within the preset safe leakage threshold range. It serves as a benchmark for measuring the deviation of the initial correction threshold. The median value is calculated as follows: "Median = (Upper limit of preset safe leakage threshold range + Lower limit of preset safe leakage threshold range) ÷ 2". The calculation process is automatically executed by the electronic equipment, ensuring accuracy. The threshold deviation rate is the percentage deviation of the initial correction threshold from the median value of the preset safe leakage threshold range. It is dimensionless and ranges from 0% to 100%. Its core function is to quantify the degree of deviation of the initial correction threshold. A larger deviation rate indicates that the initial correction threshold deviates further from the reasonable range; a smaller deviation rate indicates that the initial correction threshold is closer to the reasonable range, requiring less fine-tuning. The threshold deviation rate is calculated as follows: "Threshold deviation rate = |Initial correction threshold - Median| ÷ Median × 100%". To avoid negative deviation rates, the absolute value is used. The calculation process is automatically executed by the electronic equipment, retaining two decimal places to ensure precision. For example: the initial correction threshold is 5.75 mL / min, the current comprehensive fault level is minor fault, medium temperature and medium pressure condition, and the preset safe leakage threshold range is 5 to 7.6 mL / min; first calculate the median: (5+7.6)÷2=6.3 mL / min; then calculate the threshold deviation rate: |5.75-6.3|÷6.3×100%=8.75%, that is, the deviation of the current initial correction threshold from the median is 8.75%, which is a slight deviation and requires minor fine-tuning.

[0074] S322, a weighted verification value is obtained based on the first deviation weight of the threshold deviation rate and the second deviation weight of the fault deviation rate corresponding to the comprehensive fault level. The first deviation weight is a pre-set weight corresponding to the threshold deviation rate. The second deviation weight is a pre-set weight corresponding to the comprehensive fault level.

[0075] The threshold deviation rate, as understood, reflects the percentage deviation from the initial correction threshold and is one of the core inputs for weighted verification. The first deviation weight is a pre-set weight coefficient corresponding to the threshold deviation rate. It is dimensionless and ranges from 0.3 to 0.5. Its core function is to measure the influence of the threshold deviation rate on the weighted verification value. The larger the first deviation weight, the greater the impact of the threshold deviation rate on the final fine-tuning effect. The first deviation weight is determined by pre-setting corresponding weights according to the range of the threshold deviation rate (e.g., 0% to 5% deviation rate, first deviation weight 0.3; 5% to 10% deviation rate, first deviation weight 0.4; deviation rate above 10% deviation rate, first deviation weight 0.5). This preset weight is stored in the electronic equipment parameter library and can be fine-tuned according to the ball valve's usage scenario to match the weight with the degree of deviation. The comprehensive fault level characterizes the overall leakage risk level of the ball valve and is the core basis for determining the fault deviation rate and the second deviation weight. The fault deviation rate is a pre-set quantitative value that corresponds to the overall fault level and reflects the severity of the ball valve fault. It is dimensionless and ranges from 0% to 100%. Its core function is to quantify the impact of the fault level on the weighted verification value, and it is positively correlated with the overall fault level. The fault deviation rate is determined by: based on the severity of the overall fault level, a corresponding fault deviation rate is preset (for example: a fault deviation rate of 15% for a minor fault level, 40% for a moderate fault level, and 70% for a severe fault level), and then the preset rate is fixed in the electronic equipment. The second deviation weight refers to a pre-set weight coefficient corresponding to the comprehensive fault level. It is dimensionless and ranges from 0.5 to 0.7. Its core function is to measure the degree of influence of the fault deviation rate on the weighted verification value. The larger the second deviation weight, the greater the impact of the fault level on the final fine-tuning effect, which is in line with the core logic of "fault parameters first". The second deviation weight is determined as follows: based on the priority of the comprehensive fault level, a corresponding weight is preset (for example: minor fault level, second deviation weight is 0.5; moderate fault level, second deviation weight is 0.6; severe fault level, second deviation weight is 0.7). After being preset, it is fixed in the electronic equipment parameter library and can be fine-tuned according to the field operation effect. The weighted verification value is a comprehensive calibration coefficient obtained by integrating the threshold deviation rate, the fault deviation rate, and the corresponding weights. It is dimensionless and ranges from 0.9 to 1.1. Its core function is to fine-tune the initial correction threshold so that the fine-tuned target threshold is both suitable for the threshold deviation level and matches the current fault level. The weighted verification value is obtained by weighted summation of "threshold deviation rate × first deviation weight + fault deviation rate × second deviation weight".

[0076] For example, in S322, a weighted verification value is obtained based on the first deviation weight of the threshold deviation rate and the second deviation weight of the fault deviation rate corresponding to the comprehensive fault level, including: S3221, based on the threshold deviation rate and the fault deviation rate preset by the comprehensive fault level, determine the deviation adaptation coefficient.

[0077] The threshold deviation rate is understood to be a percentage reflecting the degree of deviation from the initial correction threshold. The fault deviation rate preset for the comprehensive fault level refers to a pre-set quantitative deviation value that corresponds one-to-one with the current comprehensive fault level. The deviation adaptation coefficient is a compensation coefficient dynamically determined based on the magnitude of the threshold deviation rate and the fault deviation rate. It is dimensionless and ranges from 0.9 to 1.1. Its core function is to dynamically compensate for the first and second deviation weights, making the weighted verification value more realistic. When the difference between the threshold deviation rate and the fault deviation rate is large, the deviation adaptation coefficient balances their influence; when the difference is small, the deviation adaptation coefficient is 1.0 (no compensation). The specific formula for calculating the deviation adaptation coefficient is "deviation adaptation coefficient = 1.0 + (threshold deviation rate - fault deviation rate) × 0.01", where "(threshold deviation rate - fault deviation rate) × 0.01" is the compensation amount, ensuring a reasonable compensation range and avoiding over-compensation. The calculation process is automatically executed by the electronic equipment and is rounded to three decimal places. For example: the threshold deviation rate is 8.75%, and the preset fault deviation rate corresponding to the minor fault level is 15%; substituting into the deviation adaptation coefficient calculation formula, the deviation adaptation coefficient is calculated as follows: 1.0 + (8.75% - 15%) × 0.01 = 1.0 + (-6.25%) × 0.01 = 1.0 - 0.0625 = 0.9375, which is rounded to 3 decimal places as 0.938. That is, the current deviation adaptation coefficient is 0.938, which is used for subsequent weight compensation.

[0078] S3222, multiply the first deviation weight of the threshold deviation rate and the second deviation weight of the fault deviation rate corresponding to the comprehensive fault level by the deviation adaptation coefficient respectively to obtain the final deviation weights of the threshold deviation rate and the fault deviation rate.

[0079] It can be understood that the first deviation weight of the threshold deviation rate refers to the weight corresponding to the current threshold deviation rate. The second deviation weight of the fault deviation rate corresponding to the comprehensive fault level refers to the weight corresponding to the current comprehensive fault level. The deviation adaptation coefficient is a coefficient used to compensate for the weights, used to dynamically adjust the magnitude of the first and second deviation weights. The final deviation weight refers to the final weight obtained after compensation by the deviation adaptation coefficient, used for subsequent weighted summation, including the first final deviation weight (corresponding to the threshold deviation rate) and the second final deviation weight (corresponding to the fault deviation rate). It is dimensionless, and the sum of the two is still close to 1.0 (because the deviation adaptation coefficient is 0.9 to 1.1, small deviations do not affect the overall logic). The final deviation weight is obtained by multiplying the first and second deviation weights by the deviation adaptation coefficient respectively. Its core function is to make the weights more consistent with the current deviation situation and improve the accuracy of the weighted verification value. For example: the first deviation weight is 0.4, the second deviation weight is 0.5, and the deviation adaptation coefficient is 0.938; perform multiplication operations respectively: the first final deviation weight = 0.4 × 0.938 = 0.3752, rounded to 3 decimal places as 0.375; the second final deviation weight = 0.5 × 0.938 = 0.469, rounded to 3 decimal places as 0.469; finally, the first final deviation weight is 0.375, the second final deviation weight is 0.469, and the sum of the two is 0.844 (due to the compensation of the deviation adaptation coefficient, the sum is slightly less than 1.0, which does not affect subsequent calculations).

[0080] S3223, the final deviation weights of the threshold deviation rate and the fault deviation rate are summed with the threshold deviation rate and the fault deviation rate to obtain the weighted verification value.

[0081] It can be understood that the threshold deviation rate refers to the calculated percentage deviation, which needs to be converted to decimal form for calculation. The fault deviation rate refers to the preset percentage deviation corresponding to the comprehensive fault level, which also needs to be converted to decimal form for calculation. The final deviation weight refers to the first final deviation weight (0.375) and the second final deviation weight (0.469) after compensation, corresponding to the threshold deviation rate and the fault deviation rate, respectively. The weighted summation refers to the electronic device performing the mathematical operation of "threshold deviation rate × first final deviation weight + fault deviation rate × second final deviation weight". The core logic is to fuse the two types of deviations according to the compensated weights to obtain a calibration coefficient that can comprehensively reflect the degree of deviation and the fault level. The calculation process is executed automatically, retaining 4 decimal places to improve accuracy. The weighted check value refers to the comprehensive calibration coefficient obtained after the weighted summation operation. It is dimensionless and ranges from 0.9 to 1.1. The weighted check value is obtained through the above weighted summation operation. Its core function is to accurately adjust the initial correction threshold so that the adjusted target threshold is both suitable for the degree of threshold deviation and matches the current fault level, avoiding the influence of a single deviation on the fine-tuning effect. For example: the threshold deviation rate (decimal) is 0.0875, and the first final deviation weight is 0.375; the fault deviation rate (decimal) is 0.15, and the second final deviation weight is 0.469; perform a weighted summation operation: 0.0875×0.375+0.15×0.469=0.0328125+0.07035=0.1031625, keep 4 decimal places as 0.1032, convert to coefficient form as 1.1032, that is, the weighted check value is 1.1032.

[0082] By adopting the above steps S3221 to S3223, it is helpful to solve the technical problems in the existing technology of weighted verification without dynamic compensation, fixed weights, low fine-tuning accuracy, and failure to fit the actual working conditions and fault states. It can achieve dual accurate fusion of threshold deviation rate and fault deviation rate, so that the weighted verification value can dynamically adapt to the current deviation and fault level, avoid fine-tuning deviation caused by a single factor, and improve the feasibility and fine-tuning accuracy of the solution.

[0083] S323, multiply the initial correction threshold by the weighted check value to obtain the target threshold.

[0084] The initial correction threshold, as understood, refers to the intermediate threshold after coarse adjustment and forms the basis for fine-tuning. The weighted check value is a comprehensive calibration coefficient obtained through dynamic compensation and weighted fusion, used for precise fine-tuning of the initial correction threshold. The correction magnitude is positively correlated with the degree of deviation and the fault level. The target threshold is the final upper limit of leakage obtained after fine-tuning the initial correction threshold. Its unit is consistent with the initial leakage threshold (mL / min or L / h). It is the core criterion for diagnosing pneumatic ball valve faults. The core significance of the target threshold is to accurately adapt to the current environmental conditions and fault state, clearly defining the maximum allowable leakage of the ball valve under the current conditions. Exceeding the target threshold is considered an abnormal leakage, avoiding misjudgments and missed judgments caused by the mismatch between fixed thresholds and actual operating conditions. The target threshold can be obtained by multiplying the initial correction threshold by the weighted check value, and must meet the requirement of "preset safe leakage threshold range". For example: the initial correction threshold is 5.75 mL / min, and the weighted check value is 1.1032; perform the multiplication operation: 5.75 × 1.1032 ≈ 6.3434, rounded to two decimal places, which is 6.34 mL / min. The preset safe leakage threshold range for minor fault level is 5 to 7.6 mL / min. The final target threshold is determined to be 6.34 mL / min. The target threshold of 6.34 mL / min means that under the current environmental conditions and minor fault level, the maximum allowable leakage of the ball valve is 6.34 mL / min. Exceeding the target threshold of 6.34 mL / min is judged as an abnormal leakage.

[0085] By adopting the above steps S321 to S323, it is helpful to solve the technical problems of simple threshold fine-tuning logic, lack of dynamic compensation, low fine-tuning accuracy, and poor adaptability to actual working conditions in the prior art. It realizes multi-step verification of target threshold deviation quantification, dynamic compensation, and weighted fusion, which not only avoids the deviation problem of the initial correction threshold, but also takes into account the impact of fault level, and significantly improves the accuracy and reliability of the target threshold.

[0086] S400: If the target threshold is greater than or equal to the mechanical threshold of the pneumatic ball valve, a ball valve fault signal is output. The mechanical threshold reflects the leakage amount at which the pneumatic ball valve is rendered unusable.

[0087] The target threshold, as understood, refers to the final, finely adjusted upper limit of leakage. It is obtained through coarse and fine adjustments and represents the maximum permissible leakage of the ball valve under current operating conditions, serving as the core basis for fault diagnosis. "Greater than or equal to" refers to a numerical comparison operation performed by the electronic equipment, comparing the target threshold with the mechanical threshold. If the target threshold is greater than or equal to the mechanical threshold, a fault signal is triggered; if the target threshold is less than the mechanical threshold, the ball valve is considered to be operating normally, and no fault signal is output. The comparison process is automated, requiring no manual intervention, making the determination efficient and accurate, and avoiding human error. The mechanical threshold of a pneumatic ball valve refers to the critical leakage limit at which the valve's sealing structure (seat, seals) completely fails, rendering it unsafe to continue operating and reaching the scrap standard. The unit is the same as the target threshold (mL / min or L / h). It is the core basis for determining whether a ball valve needs to be scrapped and shut down for maintenance. Detailed definition of the mechanical threshold: The mechanical threshold is determined based on the mechanical structural strength of the pneumatic ball valve, the lifespan of the sealing materials, the factory scrap standard, and industry safety regulations. It is the "ultimate upper limit" of the ball valve's leakage. Once the target threshold is reached or exceeded, it indicates that the ball valve's sealing performance has been completely lost, and continued operation will lead to serious leakage and safety accidents (such as media leakage). (Causing fires, explosions, environmental pollution, etc.); The mechanical threshold can be determined by the equipment manufacturer based on the ball valve model, sealing material, mechanical structure, combined with factory test data (such as leakage after the ultimate wear of the seal) and the GB / T13927-2008 industry standard, to preset the mechanical threshold, which is then fixed in the electronic equipment parameter library. Each model of ball valve corresponds to a unique mechanical threshold, which cannot be arbitrarily modified. It can be matched with the actual scrapping state of the ball valve through periodic testing and calibration (for example, for a certain model of pneumatic ball valve, the mechanical threshold is preset to 5.5mL / min, that is, when the leakage reaches 5.5mL / min, the ball valve seal is completely ineffective and needs to be scrapped and replaced). The ball valve fault signal is an alarm signal used to indicate that the pneumatic ball valve has reached the scrapping standard, there is a serious leakage risk, and it must be stopped immediately for maintenance or replacement. The signal contains specific fault information, which makes it easy for on-site technicians to quickly understand the degree of fault and the direction of handling. After the fault signal is issued, the control system can automatically trigger a shutdown command to avoid safety accidents caused by continued operation, and at the same time record fault information (time, operating conditions, threshold data) for subsequent fault analysis and tracing. For example: the target threshold is 6.34 mL / min, and the current mechanical threshold of the ball valve is preset to 5.5 mL / min; the electronic device performs a comparison calculation: 6.34 > 5.5, so it outputs a fault signal, determining that the severe wear of the seal has caused the threshold after fine adjustment to exceed the mechanical threshold, and immediately outputs a ball valve fault signal, the audible and visual alarm is activated, the control system records the fault information, and automatically triggers a shutdown command, prompting on-site technicians to stop the experiment and replace the scrapped ball valve.

[0088] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0089] Corresponding to the method for determining the threshold of experimental parameters for pneumatic ball valve failure described in the above embodiments, this application also provides a device for determining the threshold of experimental parameters for pneumatic ball valve failure. Each unit of this device can implement each step of the method for determining the threshold of experimental parameters for pneumatic ball valve failure. Figure 3 The diagram shows a structural block diagram of the device for determining the threshold of experimental parameters for pneumatic ball valve failure provided in an embodiment of this application. For ease of explanation, only the parts related to the embodiment of this application are shown.

[0090] Reference Figure 3 The device includes: The acquisition unit is used to acquire the environmental operating conditions, initial leakage threshold, and fault parameters of the pneumatic ball valve.

[0091] The coefficient unit is used to determine the threshold correction coefficient of the initial leakage threshold under the environmental conditions and fault parameters, based on the environmental operating conditions, fault parameters, and a pre-trained threshold prediction model. The threshold correction coefficient represents the rate at which the initial leakage threshold is adaptively adjusted.

[0092] The target unit is used to obtain the target threshold based on the threshold correction coefficient and the initial leakage threshold of the pneumatic ball valve. The target threshold represents the upper limit of leakage under environmental conditions and fault parameters.

[0093] It should be noted that the information interaction and execution process between the above-mentioned units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0094] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units is used as an example. In practical applications, the above functions can be assigned to different functional units as needed, that is, the internal structure of the device can be divided into different functional units to complete all or part of the functions described above. The functional units in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units in the above device can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0095] This application also provides an electronic device. Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 4 As shown, the electronic device 6 of this embodiment includes: at least one processor 60 ( Figure 4 Only one is shown in the image), at least one memory 61 ( Figure 4 (Only one is shown in the image) and a computer program 62 stored in the at least one memory 61 and executable on the at least one processor 60, wherein when the processor 60 executes the computer program 62, it causes the electronic device 6 to perform the steps in any of the above embodiments of the method for determining the threshold of experimental parameters for pneumatic ball valve failure, or causes the electronic device 6 to perform the functions of the units in the above embodiments of the apparatus.

[0096] For example, the computer program 62 may be divided into one or more units, which are stored in the memory 61 and executed by the processor 60 to complete this application. The one or more units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program 62 in the electronic device 6.

[0097] The electronic device 6 may be a programmable logic controller, an embedded industrial control computer, an edge computing gateway, a dedicated valve diagnostic terminal, a cloud server, a cloud host, a commercial desktop computer, a laptop computer, an e-commerce dedicated smart terminal, a tablet computer, etc. The electronic 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 the method described in any of the foregoing aspects. The electronic device 6 may include, but is not limited to, a processor 60 and a memory 61. Those skilled in the art will understand that... Figure 4 This is merely an example of electronic device 6 and does not constitute a limitation on electronic device 6. It may include more or fewer components than shown, or combine certain components, or different components. For example, it may also include input / output devices, network access devices, buses, etc.

[0098] The processor 60 can be a Central Processing Unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0099] In some embodiments, the memory 61 may be an internal storage unit of the electronic device 6, such as a hard disk or memory of the electronic device 6. In other embodiments, the memory 61 may be an external storage device of the electronic device 6, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device 6. Furthermore, the memory 61 may include both internal and external storage units of the electronic device 6. The memory 61 is used to store the operating system, applications, bootloader, data, and other programs, such as the program code of the computer program. The memory 61 can also be used to temporarily store data that has been output or will be output.

[0100] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in any of the above method embodiments.

[0101] This application provides a computer program product that, when run on an electronic device, causes the electronic device to perform the steps in any of the above-described method embodiments.

[0102] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or system capable of carrying computer program code to an electronic device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0103] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0104] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0105] In the embodiments provided in this application, it should be understood that the disclosed method, device, and electronic equipment for determining the threshold of experimental parameters for pneumatic ball valve failures can be implemented in other ways. For example, the embodiments of the device and electronic equipment for determining the threshold of experimental parameters for pneumatic ball valve failures described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units may be electrical, mechanical, or other forms.

[0106] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0107] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for determining the threshold values ​​of experimental parameters for pneumatic ball valve failure, characterized in that, include: Obtain the environmental operating conditions, initial leakage threshold, and fault parameters of the pneumatic ball valve; Based on the environmental conditions, the fault parameters, and the pre-trained threshold prediction model, a threshold correction coefficient for the initial leakage threshold under the environmental conditions and the fault parameters is determined; wherein, the threshold correction coefficient is used to represent the speed of adaptive adjustment of the initial leakage threshold. Based on the threshold correction coefficient of the pneumatic ball valve and the initial leakage threshold, a target threshold is obtained; wherein, the target threshold represents the upper limit of leakage under the environmental conditions and the fault parameters.

2. The method for determining the threshold of experimental parameters for pneumatic ball valve failure according to claim 1, characterized in that, The step of determining the threshold correction coefficient for the initial leakage threshold under the environmental conditions and the fault parameters based on the environmental conditions, the fault parameters, and the pre-trained threshold prediction model includes: The environmental conditions and fault parameters are decomposed and normalized to obtain multiple normalized operating condition parameters and normalized fault parameters. The normalized operating condition parameters and the normalized fault parameters are classified into levels to obtain a comprehensive fault level. Based on the comprehensive fault level and the threshold prediction model, the threshold correction coefficient of the initial leakage threshold is determined under the environmental conditions and the fault parameters.

3. The method for determining the threshold of experimental parameters for pneumatic ball valve failure according to claim 2, characterized in that, The process of classifying multiple normalized operating condition parameters and normalized fault parameters to obtain a comprehensive fault level includes: Based on the rated operating standard of pneumatic ball valves, the single parameter level determination range is obtained; The actual values ​​of each normalized operating condition parameter and each normalized fault parameter are compared with the single parameter level determination interval to obtain the single parameter level corresponding to each normalized operating condition parameter and each normalized fault parameter. A comprehensive assessment of all the individual parameter levels is performed to determine the overall fault level.

4. The method for determining the threshold values ​​of experimental parameters for pneumatic ball valve failure according to claim 3, characterized in that, The single parameter levels include minor, moderate, and severe levels. The step of comprehensively judging all the single parameter levels to determine the overall fault level includes: If any of the normalized fault parameters is the severity level in the single parameter level, the pneumatic ball valve is determined to be at the severity level in the comprehensive fault level. If no normalized fault parameter is of the severe level in the single parameter level, but any normalized operating condition parameter is of the severe level in the single parameter level, and at least one normalized fault parameter is of the moderate level in the single parameter level, the pneumatic ball valve is determined to be at the severe fault level in the comprehensive fault level. If any of the normalized fault parameters is at the medium level in the single parameter level, the pneumatic ball valve is determined to be at the medium fault level in the comprehensive fault level. If none of the normalized fault parameters are at the medium level in the single parameter level, but at least four of the normalized operating condition parameters are at the medium level in the single parameter level, the pneumatic ball valve is determined to be at the medium fault level in the comprehensive fault level. If all the normalized fault parameters are of the minor level in the single parameter level, and the number of moderate levels in the single parameter level corresponding to the normalized condition parameters is less than four and there is no severe level, the pneumatic ball valve is determined to be of the minor fault level in the comprehensive fault level.

5. The method for determining the threshold of experimental parameters for pneumatic ball valve failure according to claim 2, characterized in that, The step of determining the threshold correction coefficient for the initial leakage threshold under the environmental conditions and the fault parameters based on the comprehensive fault level and the threshold prediction model includes: Based on the comprehensive fault level, the input parameter set of the threshold prediction model is obtained by integration; The input parameter set is input into the pre-trained threshold prediction model, and the output is an initial correction coefficient that is adapted to the input parameter set; the threshold prediction model is a machine learning model that takes the input parameter set as input and the initial correction coefficient as output. The initial correction coefficients are verified to obtain the threshold correction coefficients under the environmental conditions and the fault parameters.

6. The method for determining the threshold values ​​of experimental parameters for pneumatic ball valve failure according to claim 1, characterized in that, The process of obtaining the target threshold based on the threshold correction coefficient of the pneumatic ball valve and the initial leakage threshold includes: The initial correction threshold is obtained by dividing the threshold correction coefficient of the pneumatic ball valve by the initial leakage threshold. The initial modified threshold is adapted and verified to obtain the target threshold.

7. The method for determining the threshold values ​​of experimental parameters for pneumatic ball valve failure according to claim 6, characterized in that, The process of adapting and verifying the initial corrected threshold to obtain the target threshold includes: Calculate the threshold deviation rate between the initial correction threshold and the midpoint of the preset safe leakage threshold range; wherein, the threshold deviation rate reflects the degree of deviation of the initial correction threshold; A weighted verification value is obtained based on the first deviation weight of the threshold deviation rate and the second deviation weight of the fault deviation rate corresponding to the comprehensive fault level; wherein, the first deviation weight is the weight corresponding to the threshold deviation rate that is preset; and the second deviation weight is the weight corresponding to the comprehensive fault level that is preset. The target threshold is obtained by multiplying the initial correction threshold by the weighted check value.

8. The method for determining the threshold values ​​of experimental parameters for pneumatic ball valve failure according to claim 7, characterized in that, The weighted verification value is obtained based on the first deviation weight of the threshold deviation rate and the second deviation weight of the fault deviation rate corresponding to the comprehensive fault level, including: Based on the threshold deviation rate and the fault deviation rate preset by the comprehensive fault level, the deviation adaptation coefficient is determined; The first deviation weight of the threshold deviation rate and the second deviation weight of the fault deviation rate corresponding to the comprehensive fault level are multiplied by the deviation adaptation coefficient to obtain the final deviation weights of the threshold deviation rate and the fault deviation rate. The final deviation weights of the threshold deviation rate and the fault deviation rate are weighted and summed with the threshold deviation rate and the fault deviation rate to obtain the weighted verification value.

9. The method for determining the threshold values ​​of experimental parameters for pneumatic ball valve failure according to claim 1, characterized in that, The method further includes: If the target threshold is greater than or equal to the mechanical threshold of the pneumatic ball valve, a ball valve fault signal is output; wherein, the mechanical threshold is used to reflect the leakage amount of the pneumatic ball valve before it is scrapped.

10. A device for determining the threshold of experimental parameters for a pneumatic ball valve failure, characterized in that, include: The acquisition unit is used to acquire the environmental operating conditions, initial leakage threshold, and fault parameters of the pneumatic ball valve. The coefficient unit is used to determine the threshold correction coefficient of the initial leakage threshold under the environmental conditions and the fault parameters based on the environmental conditions, the fault parameters and the pre-trained threshold prediction model; wherein the threshold correction coefficient is used to represent the speed of adaptive adjustment of the initial leakage threshold. The target unit is used to obtain a target threshold based on the threshold correction coefficient of the pneumatic ball valve and the initial leakage threshold; wherein the target threshold represents the upper limit of leakage under the environmental conditions and the fault parameters.

Images