A method, device and system for rapid detection of valve sealing performance

Abstract

The present application relates to the technical field of sealing detection, and specifically to a method, device and system for rapid detection of valve sealing performance, the method comprising: using a pressure sensor to obtain the pressure at each moment in each acquisition cycle of a standard valve and the pressure at each moment of a valve to be tested; determining the pressure turning point, obtaining the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve; determining the inflation cycle deviation of each acquisition cycle of the standard valve; determining the pressure stabilization cumulative deviation of each acquisition cycle of the standard valve; determining the marginal error parameter of a single-class support vector machine model, and detecting the sealing performance of the valve to be tested according to the pressure at each moment of the valve to be tested. The present application improves the accuracy of valve sealing performance detection and reduces the error of the detection result.

Description

Technical Field

[0001] The present application relates to the technical field of sealing detection, and specifically to a method, device and system for rapid detection of valve sealing performance. Background Art

[0002] A valve is a mechanical device used to control the flow of fluids. It can be used to open or close pipelines, containers, and other fluid delivery systems to regulate the flow of media. When closed, it needs to completely block the gas or liquid transported by the pipeline, so its sealing is crucial. If the sealing of the valve is unqualified, it may cause material leakage or even cause major safety accidents. Therefore, the development of effective valve sealing detection devices and methods is of great significance to ensure the safety and efficiency of fluid delivery systems.

[0003] The traditional direct pressure method fills the valve to be tested with gas at a certain pressure, stabilizes the pressure for a period of time, and then measures the pressure change in the valve to determine whether there is a leakage problem. However, due to factors such as unstable voltage, the power of the inflation equipment will be unstable, which will cause the pressure inside the valve to fluctuate. When using the one-class support vector machine model (One-ClassSVM model) to detect the sealing performance of the valve, its marginal error parameter determines the misclassification ratio that the model can tolerate. The larger the marginal error parameter value, the greater the tolerance for error. Using a fixed value of the marginal error parameter to detect fluctuating valve pressure data will cause a certain error. Summary of the invention

[0004] In order to solve the above technical problems, a method, device and system for rapid detection of valve sealing performance are provided to solve the existing problems.

[0005] The solution to the technical problem of the present application is to provide a method, device and system for rapid detection of valve sealing performance, including the following steps:

[0006] In a first aspect, an embodiment of the present application provides a method for quickly detecting valve sealing performance, the method comprising the following steps:

[0007] Collect the pressure of the standard valve at each moment in each collection cycle and the pressure of the valve to be tested at each moment;

[0008] According to the local difference between the pressures at different times in each acquisition cycle of the standard valve, the pressure turning point is determined, and the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve are obtained;

[0009] Determine the inflation cycle deviation of each acquisition cycle of the standard valve according to the periodic variation characteristics and discreteness of the pressure at all times during the inflation period of each acquisition cycle of the standard valve;

[0010] Determine the cumulative deviation of the steady pressure of each acquisition cycle of the standard valve according to the local deviation of the pressure during the steady pressure period of each acquisition cycle of the standard valve;

[0011] Based on the inflation cycle deviation and the pressure stabilization cumulative deviation, the marginal error parameter of the single-class support vector machine model is determined, and the sealing performance of the valve to be tested is tested according to the pressure of the valve to be tested at each moment.

[0012] Preferably, the step of determining the pressure turning point and obtaining the inflation period and the pressure stabilization period of each acquisition cycle includes:

[0013] Analyze the pressure difference between each moment and its adjacent moments in each acquisition cycle of the standard valve, and obtain the candidate pressure turning point of each acquisition cycle of the standard valve;

[0014] The difference between the pressures at the left and right adjacent moments of each candidate pressure turning point in each acquisition cycle of the standard valve is recorded as the pressure turning difference of each candidate pressure turning point;

[0015] The candidate pressure turning point corresponding to the minimum value of all the pressure turning differences in each acquisition cycle is used as the pressure turning point of each acquisition cycle of the standard valve;

[0016] The pressure at all times before the pressure turning point of each acquisition cycle is recorded as the inflation period, and the pressure at all times after the pressure turning point is recorded as the pressure stabilization period.

[0017] Preferably, the method for obtaining the candidate pressure turning point of each acquisition cycle of the standard valve is:

[0018] The difference between the pressure at each moment in each acquisition cycle of the standard valve and the pressure at the previous moment is recorded as the first difference;

[0019] The difference between the pressure at each moment in each acquisition cycle of the standard valve and the pressure at the next moment is recorded as the second difference;

[0020] The absolute value of the sum of the first difference and the second difference is used as the pressure fluctuation index at each moment in each acquisition cycle;

[0021] In each acquisition cycle, the pressures at the moments corresponding to the previous preset number of pressure fluctuation indices are selected from large to small according to the pressure fluctuation indices as candidate pressure turning points.

[0022] Preferably, the step of determining the inflation cycle deviation of each acquisition cycle of the standard valve comprises:

[0023] Analyze the pressure change differences at each moment during the inflation period and determine the inflation pressure deviation of each acquisition cycle of the standard valve;

[0024] The delay time corresponding to the maximum peak value of the autocorrelation function of the pressure at all times during the inflation period, except for zero delay, is selected as the pressure change period during the inflation period of each acquisition cycle of the standard valve;

[0025] Based on the pressure variation cycle, the inflation period of each acquisition cycle of the standard valve is divided into various cycle periods;

[0026] Calculate the mean value of pressure at all times in each cycle period;

[0027] The difference between the inflation pressure deviation and the mean value of each period is recorded as the period pressure difference of each period;

[0028] The degree of dispersion of the cycle pressure differences in all cycle periods is determined as the inflation cycle deviation of each acquisition cycle of the standard valve.

[0029] Preferably, the determining of the inflation pressure deviation of each acquisition cycle of the standard valve comprises:

[0030] Calculate the average of the pressures at all times during the inflation period of each acquisition cycle of the standard valve, and record it as the first average;

[0031] Determine the difference between the pressure at each moment in the inflation period of each acquisition cycle of the standard valve and the first mean value as the relative pressure difference;

[0032] The relative pressure differences at all times during the inflation period are merged to determine the inflation pressure deviation of each acquisition cycle of the standard valve.

[0033] Preferably, the step of determining the cumulative deviation of the pressure regulation of each acquisition cycle of the standard valve includes:

[0034] Starting from the pressure at the first moment in the pressure stabilization period, a sliding window of a preset size is set;

[0035] Analyze the pressure variation differences at each moment in each sliding window during the pressure stabilization period, and determine the relative cumulative deviation of each sliding window during the pressure stabilization period of the standard valve;

[0036] The time interval between the pressure turning point in the acquisition period and the first pressure in each sliding window is recorded as the time weight of each sliding window;

[0037] The product of the relative cumulative deviation of each sliding window in the voltage stabilization period and the time weight of the corresponding sliding window is used as the cumulative deviation coefficient of each sliding window in the voltage stabilization period;

[0038] The average of the cumulative deviation coefficients of all sliding windows within the pressure stabilization period is taken as the cumulative pressure stabilization deviation of each acquisition cycle of the standard valve.

[0039] Preferably, the relative cumulative deviation of each sliding window during the pressure stabilization period of the standard valve includes:

[0040] Calculate the mean of the pressure at all times in each sliding window during the pressure stabilization period, and record it as the window pressure mean;

[0041] Determine the difference between the pressure at each moment in each sliding window during the pressure stabilization period and the window pressure average as a third difference;

[0042] The third differences at all times in each sliding window during the pressure stabilization period are fused to determine the relative cumulative deviation of each sliding window during the pressure stabilization period of the standard valve.

[0043] Preferably, determining the marginal error parameter of the single-class support vector machine model includes:

[0044] When the relative cumulative deviation of the sliding window in the pressure stabilization period of each acquisition cycle of the standard valve is 0, or the sliding window slides to the pressure at the last moment in the pressure stabilization period, the sliding stops, and the time weight of the sliding window corresponding to the stop sliding is used as the pressure recovery time of each acquisition cycle;

[0045] Count the duration of the inflation period of each acquisition cycle of the standard valve;

[0046] Taking the product of the duration and the inflation cycle deviation as the first product;

[0047] The product of the pressure recovery time and the steady pressure cumulative deviation is used as a second product;

[0048] The first product and the second product of each acquisition cycle of the standard valve are merged to determine the valve sealing deviation of each acquisition cycle of the standard valve;

[0049] The normalized result of the mean value of the valve sealing deviation of all acquisition cycles of the standard valve is used as the marginal error parameter of the single-class support vector machine model.

[0050] In a second aspect, the embodiment of the present application further provides a valve sealing performance rapid detection device, the device comprising:

[0051] Valve pressure acquisition module, which collects the pressure of the standard valve at each moment in each acquisition cycle and the pressure of the valve to be tested at each moment;

[0052] A model building module determines the pressure turning point according to the local difference between the pressures at different moments in each acquisition cycle of the standard valve, and obtains the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve; determines the inflation cycle deviation of each acquisition cycle of the standard valve according to the periodic variation characteristics and the degree of discreteness of the pressure at all moments in the inflation period of each acquisition cycle of the standard valve; determines the pressure stabilization cumulative deviation of each acquisition cycle of the standard valve according to the local deviation of the pressure in the pressure stabilization period of each acquisition cycle of the standard valve; determines the marginal error parameter of the single-class support vector machine model based on the inflation cycle deviation and the pressure stabilization cumulative deviation;

[0053] The valve detection module detects the sealing performance of the valve to be tested based on the marginal error parameters of the single-class support vector machine model and the pressure of the valve to be tested at each moment.

[0054] In the third aspect, an embodiment of the present application also provides a valve sealing performance rapid detection system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned valve sealing performance rapid detection methods when executing the computer program.

[0055] This application has at least the following beneficial effects:

[0056] The present application determines the pressure turning point by analyzing the local differences between the pressures at each moment of each acquisition cycle of the standard valve, and obtains the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve. The beneficial effect of the application is that the pressures at each moment of the acquisition are distinguished between the inflation period and the pressure stabilization period, so that the pressures of the two periods can be subsequently analyzed, and the valve sealing performance can be further more accurately detected. The inflation cycle deviation of each acquisition cycle of the standard valve is determined according to the periodic change characteristics and discrete degree of the pressure at all moments in the inflation period of each acquisition cycle of the standard valve. The beneficial effect of the application is that the periodic fluctuation characteristics of the pressure during the inflation period are taken into account, and the pressure deviation of each cycle is analyzed to reflect the pressure in the entire The degree of deviation during the inflation period; according to the local deviation of pressure during the pressure stabilization period of each acquisition cycle of the standard valve, the cumulative pressure stabilization deviation of each acquisition cycle of the standard valve is determined, and its beneficial effect is that the pressure deviation in the fluctuation process during the pressure stabilization period is gradually reduced, and the impact on the model detection is measured; according to the inflation cycle deviation and the cumulative pressure stabilization deviation of all acquisition cycles of the standard valve, the marginal error parameters of the single-class support vector machine model are determined to detect the sealing performance of the valve to be tested. Its beneficial effect is to adjust the parameters of the single-class support vector machine model to avoid the model detecting the normal fluctuation of the internal pressure of the valve as abnormal, thereby improving the accuracy of the valve sealing performance detection and reducing the error of the detection result. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] A rapid detection method for valve sealing performance of the present application is further described in detail below in conjunction with the accompanying drawings.

[0058] Figure 1 A flowchart of the steps of a rapid detection method for valve sealing performance provided in this application;

[0059] Figure 2 A flowchart of the steps of the method for obtaining the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve provided in the present application;

[0060] Figure 3 A flowchart of the steps of a method for obtaining the inflation cycle deviation of each acquisition cycle of a standard valve provided in the present application;

[0061] Figure 4 A flowchart of the steps of a valve sealing performance rapid detection device provided in this application. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the following is a further detailed description of a valve sealing performance rapid detection method, device and system proposed in the present application in conjunction with the accompanying drawings and implementation examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0063] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0064] See also Figure 1 , which shows a flowchart of a method for rapid detection of valve sealing performance provided by an embodiment of the present application, the method comprising the following steps:

[0065] Step 1: Collect the pressure of the standard valve at each moment in each collection cycle and the pressure of the valve to be tested at each moment.

[0066] A standard valve, that is, a valve with good sealing performance and no leakage, is installed on the test bench. A pressure sensor is installed on the valve to collect the pressure changes inside the valve. The direct pressure method is used to test the sealing performance of the standard valve. There are three steps: the first step is to fill the valve with gas, which is called the inflation process; the second step is to keep the gas in the valve for a period of time, which is called the pressure stabilization process; the third step is to release the gas, which is called the deflation process.

[0067] Therefore, the direct pressure method is used to collect the pressure changes during the sealing process of the standard valve. The specific process is: after closing the standard valve, use the inflation equipment to fill the standard valve with gas until the internal pressure of the valve reaches the preset pressure, and then end the inflation. After the inflation is completed, the pressure is stabilized for a period of time, and then the valve is opened.

[0068] The pressure inside the valve at each moment from the beginning of inflation to the end of inflation and finally to the pressure stabilization process is collected as a collection cycle. The standard valve is operated N times to collect the pressure inside the valve at each moment in the N collection cycles, where the collection time interval of the pressure sensor is t. The collected pressure inside the valve is preprocessed and the pressure at each moment in each collection cycle of the standard valve is obtained in the order of collection time.

[0069] According to the above method, the pressure of the valve to be tested at each moment in a collection cycle is collected to obtain the pressure of the valve to be tested at each moment.

[0070] Preferably, in this embodiment, the preset pressure is 1.1 times the pressure of the standard valve during normal use; the number of acquisitions N = 100, and the acquisition time interval is 0.1s; secondly, the pressure at each moment in all acquisition cycles is cleaned, outliers are removed, missing values ​​are filled with cubic splines, and the pressure at each moment is standardized. Different valves are subjected to different pressures during use according to different usage scenarios. As other implementation methods, the implementer can set them according to actual conditions. Among them, data cleaning, cubic spline interpolation and data standardization are well-known technologies, and the specific process will not be repeated.

[0071] At this point, the pressure at each moment of each acquisition cycle of the standard valve and the pressure at each moment of the valve to be tested are obtained.

[0072] Step 2: Determine the pressure turning point according to the local difference between the pressures at different times in each acquisition cycle of the standard valve, and obtain the inflation period and pressure stabilization period of each acquisition cycle.

[0073] When the inflation process begins, the pressure inside the valve rises rapidly. When the pressure inside the valve reaches the preset pressure, the inflation process ends and the pressure stabilization process begins. During the pressure stabilization process, the pressure inside the valve no longer increases due to the fact that inflation is no longer in progress, and the fluctuation gradually decreases.

[0074] From the inflation process to the pressure stabilization process, the pressure inside the valve fluctuates, so when the pressure inside the valve reaches the preset pressure, it is not necessarily the real pressure turning point. In addition, the pressure fluctuation changes greatly during the inflation process, and the pressure fluctuation changes little during the pressure stabilization process, so the pressure turning point can be obtained according to the pressure fluctuation degree during the detection process to distinguish the inflation process and the pressure stabilization process.

[0075] Based on the above analysis, the pressure fluctuation index is obtained through the pressure changes at adjacent moments to reflect the pressure fluctuation degree at adjacent moments, specifically:

[0076] The difference between the pressure at each moment in each acquisition cycle of the standard valve and the pressure at the previous moment is recorded as the first difference;

[0077] The difference between the pressure at each moment in each acquisition cycle of the standard valve and the pressure at the next moment is recorded as the second difference;

[0078] The absolute value of the sum of the first difference and the second difference is used as the pressure fluctuation index of the standard valve at each moment in each acquisition cycle;

[0079] Preferably, in this embodiment, the pressure fluctuation index is: B k,t =|(p k,t -p k,t-1 )+(p k,t -p k,t+1 )|, where B k,t is the pressure fluctuation index of the standard valve at the tth moment in the kth acquisition cycle, p k,t is the pressure of the standard valve at the tth moment in the kth acquisition cycle, p k,t-1 is the pressure of the standard valve at the time t-1 in the kth acquisition cycle, p k,t+1 is the pressure of the standard valve at the time t+1 in the kth acquisition cycle, || is the absolute value, p k,t -p k,t-1 is the first difference, p k,t -p k,t+1 The second difference.

[0080] It should be noted that the pressure fluctuation index is not calculated for the first moment and the last moment; secondly, the pressure fluctuation index indicates the degree of pressure fluctuation between adjacent moments. The larger the pressure fluctuation index, the more obvious the pressure fluctuation; the smaller the pressure fluctuation index, the less obvious the pressure fluctuation.

[0081] Furthermore, based on the pressure fluctuation index, the pressure turning point is obtained to distinguish the inflation period and the pressure stabilization period, specifically:

[0082] For each acquisition cycle of the standard valve, the pressure at the time corresponding to the preset number of pressure fluctuation indexes is selected from large to small according to the pressure fluctuation index as the candidate pressure turning point;

[0083] Preferably, in this embodiment, the pressures at the moments corresponding to the first five pressure fluctuation indexes are selected as candidate pressure turning points. As for other implementation modes, implementers can set them according to actual conditions.

[0084] Since the pressure at adjacent moments at the pressure turning point fluctuates and the fluctuations are similar, there may be multiple moments with the same pressure fluctuation index. Therefore, the moment with the largest pressure fluctuation index is not necessarily the pressure turning point. Therefore, it is necessary to further determine the candidate pressure turning point based on the pressure difference at adjacent moments, specifically:

[0085] The difference between the pressures at the left and right adjacent moments of each candidate pressure turning point in each acquisition cycle of the standard valve is recorded as the pressure turning difference of each candidate pressure turning point in each acquisition cycle of the standard valve;

[0086] The candidate pressure turning point corresponding to the minimum value of the pressure turning difference of all candidate pressure turning points is used as the pressure turning point of each acquisition cycle of the standard valve;

[0087] Preferably, in this embodiment, the calculation method of the pressure turning difference is: is the candidate pressure turning point in the kth acquisition cycle of the standard valve The pressure transition difference, is the candidate pressure turning point in the kth acquisition cycle of the standard valve The pressure of the previous moment, is the candidate pressure turning point in the kth acquisition cycle of the standard valve The pressure of the last moment.

[0088] It should be noted that after the inflation process is completed, the pressure inside the valve no longer increases, so the point with the largest pressure fluctuation index is most likely to be the pressure turning point. However, due to the pressure fluctuation, the pressure fluctuation index at the moment before the pressure turning point may also reach the maximum. Therefore, the pressure difference at adjacent moments on the left and right of the candidate pressure change turning point is analyzed. The candidate pressure change turning point with the smallest pressure turning difference has the largest volatility and is most likely to be the pressure change turning point.

[0089] Furthermore, based on the pressure turning point, the inflation period and the pressure stabilization period are distinguished, specifically:

[0090] The pressure at all times before the corresponding time of the pressure turning point is recorded as the inflation period, and the pressure at all times after the corresponding time of the pressure turning point is recorded as the pressure stabilization period.

[0091] Further, the step flow chart of the method for obtaining the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve provided in the embodiment of the present application is as follows: Figure 2 shown.

[0092] At this point, the inflation period and pressure stabilization period of each acquisition cycle of the standard valve are obtained.

[0093] Step 3, determining the inflation cycle deviation of each acquisition cycle of the standard valve according to the periodic variation characteristics and discreteness of the pressure at all times within the inflation period of each acquisition cycle of the standard valve.

[0094] Since the inflation process of the valve is usually completed by a pump or a compressor, the collected pressure data is not stable, especially in reciprocating pumps, which may cause periodic pressure fluctuations. At the same time, when the power of the inflation equipment changes due to voltage instability, periodic fluctuations may also occur. Therefore, the periodic changes in pressure during the inflation process are analyzed through the autocorrelation function, specifically:

[0095] The delay time corresponding to the maximum peak value of the autocorrelation function of the pressure at all times during the inflation period, except for zero delay, is selected as the pressure change period during the inflation period of each acquisition cycle of the standard valve;

[0096] It should be noted that the autocorrelation function is a well-known technology and will not be elaborated on here; secondly, the pressure change period is expressed as an estimated value of the periodic change of pressure during the inflation period. The larger the pressure change period, the greater the periodicity of the pressure change during the inflation period, and the better the pressure stability; the smaller the pressure change period, the smaller the periodicity of the pressure change during the inflation period, the more drastic the pressure change, and the worse the stability.

[0097] Furthermore, based on the pressure change cycle, the pressure periodic deviation during the inflation period of each acquisition cycle of the standard valve is analyzed to reflect the periodic change of pressure, specifically:

[0098] Calculate the average of the pressures at all times during the inflation period of each acquisition cycle of the standard valve, and record it as the first average;

[0099] Determine the difference between the pressure at each moment in the inflation period of each acquisition cycle of the standard valve and the first mean value as the relative pressure difference;

[0100] The relative pressure differences at all times in the inflation period of each acquisition cycle of the standard valve are merged to determine the inflation pressure deviation of each acquisition cycle of the standard valve;

[0101] It is understandable that the fusion can be expressed as a multiplication relationship, an addition relationship, etc., which is determined by the actual application and the present application does not impose any special restrictions on this.

[0102] Preferably, in this embodiment, the calculation method of the inflation pressure deviation is: Among them, C k is the inflation pressure deviation of the kth acquisition cycle of the standard valve, p′ k,t is the pressure at time t during the inflation period of the kth acquisition cycle of the standard valve, μ′ kis the mean value of the pressure at all times during the inflation period of the kth acquisition cycle of the standard valve; t0 is the time corresponding to the pressure turning point, |p′ k,t -μ′ k | represents the relative pressure difference, and || represents the absolute value.

[0103] Based on the pressure variation cycle, the inflation period of each acquisition cycle of the standard valve is divided into various cycle periods;

[0104] Calculate the mean value of pressure at all times in each cycle period;

[0105] The difference between the inflation pressure deviation and the mean value of each period is recorded as the period pressure difference of each period;

[0106] The discrete degree of the cycle pressure difference in all cycle periods of the inflation period is determined as the inflation cycle deviation of each acquisition cycle of the standard valve;

[0107] Preferably, in this embodiment, e k,r =C k -μ′ k,r , where e k,r is the periodic pressure difference of the rth period in the inflation period of the kth acquisition cycle of the standard valve, μ′ k,r is the mean value of the pressure at all times in the rth cycle period during the inflation period of the kth acquisition cycle of the standard valve; C k is the inflation pressure deviation of the kth acquisition cycle of the standard valve; secondly, the inflation cycle deviation of each acquisition cycle of the standard valve can be obtained by calculating the standard deviation of the cycle pressure differences of all cycle time periods.

[0108] It can be understood that in this embodiment, the degree of dispersion is measured by calculating the standard deviation. As other implementation methods, the implementer can adopt other methods of the prior art, such as variance, coefficient of variation or information entropy, for measurement, and this embodiment does not impose any special restrictions on this.

[0109] Further, the step flow chart of the method for obtaining the inflation cycle deviation of each acquisition cycle of the standard valve provided in the embodiment of the present application is as follows: Figure 3 shown.

[0110] It should be noted that the inflation cycle deviation reflects the deviation of pressure during the inflation period, while taking into account the periodicity of pressure changes; the larger the inflation cycle deviation, the greater the pressure deviation within a pressure change cycle, and the greater the impact on subsequent test results.

[0111] Thus, the inflation cycle deviation of each acquisition cycle of the standard valve is obtained.

[0112] Step 4: Determine the cumulative deviation of the pressure stabilization in each acquisition cycle of the standard valve according to the local deviation of the pressure in the pressure stabilization period of each acquisition cycle of the standard valve.

[0113] During the pressure stabilization period, no more gas is added, and the pressure fluctuation inside the valve begins to decrease until it reaches a steady state, that is, the pressure inside the valve no longer fluctuates. By measuring the degree to which the internal air pressure of the valve reaches a stable state during the pressure stabilization period, the accumulated pressure stabilization deviation is analyzed to reflect the pressure fluctuation during the pressure stabilization period. Specifically:

[0114] Starting from the pressure at the first moment in the pressure stabilization period of each acquisition cycle of the standard valve, a sliding window of a preset size is set;

[0115] Preferably, in this embodiment, the preset size of the sliding window is 1*10, and the step size of the sliding window is 1. As other implementation modes, the implementer can set them according to actual conditions.

[0116] Calculate the mean value of the pressure at all times within each sliding window of the standard valve, and record it as the mean window pressure;

[0117] The difference between the pressure at each moment in each sliding window and the mean pressure of the window is determined as a third difference;

[0118] The third differences at all moments in each sliding window are merged to determine the relative cumulative deviation of each sliding window;

[0119] Preferably, in this embodiment, the absolute value of the difference between the pressure at each moment in each sliding window and the window pressure mean is taken as the third difference; the sum of the third differences at all moments in each sliding window is taken as the relative cumulative deviation of each sliding window.

[0120] The time interval between the pressure turning point in the acquisition period and the first pressure in each sliding window is recorded as the time weight of each sliding window;

[0121] The product of the relative cumulative deviation of each sliding window in the voltage stabilization period and the time weight of the corresponding sliding window is used as the cumulative deviation coefficient of each sliding window in the voltage stabilization period;

[0122] The average value of the cumulative deviation coefficients of all sliding windows in the pressure stabilization period is taken as the cumulative pressure stabilization deviation of each acquisition cycle of the standard valve;

[0123] It should be noted that the pressure stabilization cumulative deviation reflects the pressure deviation of each small period of time during the pressure stabilization period, and introduces a time weight, which shows that the closer the window is to the pressure stabilization period, the greater the proportion of its relative cumulative deviation, and the greater the impact on the overall stability. The larger the relative cumulative deviation, the greater the pressure value deviation during the pressure stabilization period, and the greater the impact on model training.

[0124] At this point, the cumulative deviation of the pressure regulation of each acquisition cycle of the standard valve is obtained.

[0125] Step 5, determining the marginal error parameter of the single-class support vector machine model according to the inflation cycle deviation and the pressure stabilization cumulative deviation of all acquisition cycles of the standard valve, and testing the sealing performance of the valve to be tested according to the pressure of the valve to be tested at each moment.

[0126] Furthermore, the valve sealing deviation is determined by combining the pressure stabilization cumulative deviation and the inflation cycle deviation to reflect the overall deviation of the pressure value during the valve sealing performance test, specifically:

[0127] When the relative cumulative deviation of the sliding window in the pressure stabilization period of each acquisition cycle of the standard valve is 0, or the sliding window slides to the pressure at the last moment in the pressure stabilization period, the sliding stops, and the time weight of the sliding window corresponding to the stop sliding is used as the pressure recovery time of each acquisition cycle;

[0128] Count the duration of the inflation period of each acquisition cycle of the standard valve;

[0129] Taking the product of the duration and the inflation cycle deviation as the first product;

[0130] The product of the pressure recovery time and the steady pressure cumulative deviation is used as a second product;

[0131] The first product and the second product of each acquisition cycle of the standard valve are merged to determine the valve sealing deviation of each acquisition cycle of the standard valve;

[0132] Preferably, in this embodiment, the sum of the first product and the second product is used as the valve sealing deviation of the standard valve in each acquisition cycle.

[0133] It should be noted that the inflation cycle weight and pressure recovery time reflect the proportion of the two time periods in the overall deviation. The greater the valve sealing deviation, the greater the overall deviation of the pressure value during the valve sealing performance test, that is, the greater the volatility of the pressure change, and there are larger changes and noise errors.

[0134] Based on the above analysis, the larger the valve sealing deviation is, the larger the marginal error parameter of the one-class support vector machine model (One-Class SVM model) should be, which can increase the fault tolerance of the model and avoid misjudging normal fluctuations as abnormalities; conversely, a smaller marginal error parameter needs to be selected to avoid the model being insensitive to the detection of abnormal values. Therefore, the method for determining the marginal error parameter of the One-Class SVM model is:

[0135] The normalized result of the mean value of the valve sealing deviation of all acquisition cycles of the standard valve is used as the marginal error parameter of the One-Class SVM model;

[0136] The pressure at each moment of all acquisition cycles of the standard valve is trained using the One-Class SVM model;

[0137] Furthermore, the size of the marginal error parameter of the One-Class SVM model determines the degree of penalty imposed by regularization on the complexity of the model. Therefore, based on the marginal error parameter of the One-Class SVM model, the sealing performance of the valve to be tested is tested.

[0138] The trained One-Class SVM model is used to detect the pressure of the valve to be tested at all times to obtain the test results of the sealing performance of the valve to be tested.

[0139] The present application also provides a valve sealing performance rapid detection device, the device comprising:

[0140] The valve pressure acquisition module uses a pressure sensor to obtain the pressure of the standard valve at each moment in each acquisition cycle and the pressure of the valve to be tested at each moment;

[0141] A model building module determines the pressure turning point according to the local difference between the pressures at each moment of each acquisition cycle of the standard valve, and obtains the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve; determines the inflation cycle deviation of each acquisition cycle of the standard valve according to the periodic variation characteristics and the degree of discreteness of the pressure at all moments in the inflation period of each acquisition cycle of the standard valve; determines the pressure stabilization cumulative deviation of each acquisition cycle of the standard valve according to the local deviation of the pressure in the pressure stabilization period of each acquisition cycle of the standard valve; determines the marginal error parameter of the single-class support vector machine model based on the inflation cycle deviation and the pressure stabilization cumulative deviation;

[0142] The valve detection module detects the sealing performance of the valve to be tested based on the marginal error parameters of the single-class support vector machine model and the pressure of the valve to be tested at each moment.

[0143] Further, the flow chart of the steps of a valve sealing performance rapid detection device provided in the embodiment of the present application is as follows: Figure 4 shown.

[0144] Based on the same inventive concept as the above method, an embodiment of the present application also provides a valve sealing performance rapid detection system, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of any one of the above-mentioned valve sealing performance rapid detection methods are implemented.

[0145] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0146] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0147] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the present application. It should be pointed out that, for those of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application. Therefore, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application without departing from the content of the technical solution of the present application, shall fall within the protection scope of the technical solution of the present application.

Claims

1. A method for rapid detection of valve sealing performance, characterized in that: The method comprises the following steps: Collect the pressure of the standard valve at each moment in each collection cycle and the pressure of the valve to be tested at each moment; According to the local difference between the pressures at different times in each acquisition cycle of the standard valve, the pressure turning point is determined, and the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve are obtained; Determine the inflation cycle deviation of each acquisition cycle of the standard valve according to the periodic variation characteristics and discreteness of the pressure at all times during the inflation period of each acquisition cycle of the standard valve; Determine the cumulative deviation of the steady pressure of each acquisition cycle of the standard valve according to the local deviation of the pressure during the steady pressure period of each acquisition cycle of the standard valve; Based on the inflation cycle deviation and the pressure stabilization cumulative deviation, a marginal error parameter of a single-class support vector machine model is determined, and the sealing performance of the valve to be tested is tested according to the pressure of the valve to be tested at each moment; The step of determining the inflation cycle deviation of each acquisition cycle of the standard valve comprises: Analyze the pressure change differences at each moment during the inflation period and determine the inflation pressure deviation of each acquisition cycle of the standard valve; The delay time corresponding to the maximum peak value of the autocorrelation function of the pressure at all times during the inflation period, except for zero delay, is selected as the pressure change period during the inflation period of each acquisition cycle of the standard valve; Based on the pressure variation cycle, the inflation period of each acquisition cycle of the standard valve is divided into various cycle periods; Calculate the mean value of pressure at all times in each cycle period; The difference between the inflation pressure deviation and the mean value of each period is recorded as the period pressure difference of each period; The degree of dispersion of the cycle pressure differences in all cycle periods is determined as the inflation cycle deviation of each acquisition cycle of the standard valve; Determining the cumulative pressure stabilization deviation of each acquisition cycle of the standard valve includes: Starting from the pressure at the first moment in the pressure stabilization period, a sliding window of a preset size is set; Analyze the pressure variation differences at each moment in each sliding window during the pressure stabilization period, and determine the relative cumulative deviation of each sliding window during the pressure stabilization period of the standard valve; The time interval between the pressure turning point in the acquisition period and the first pressure in each sliding window is recorded as the time weight of each sliding window; The product of the relative cumulative deviation of each sliding window in the voltage stabilization period and the time weight of the corresponding sliding window is used as the cumulative deviation coefficient of each sliding window in the voltage stabilization period; The average value of the cumulative deviation coefficients of all sliding windows in the pressure stabilization period is taken as the cumulative pressure stabilization deviation of each acquisition cycle of the standard valve; The step of determining the marginal error parameter of the single-class support vector machine model comprises: When the relative cumulative deviation of the sliding window in the pressure stabilization period of each acquisition cycle of the standard valve is 0, or the sliding window slides to the pressure at the last moment in the pressure stabilization period, the sliding stops, and the time weight of the sliding window corresponding to the stop sliding is used as the pressure recovery time of each acquisition cycle; Count the duration of the inflation period of each acquisition cycle of the standard valve; Taking the product of the duration and the inflation cycle deviation as the first product; The product of the pressure recovery time and the steady pressure cumulative deviation is used as a second product; The first product and the second product of each acquisition cycle of the standard valve are merged to determine the valve sealing deviation of each acquisition cycle of the standard valve; The normalized result of the mean value of the valve sealing deviation of all acquisition cycles of the standard valve is used as the marginal error parameter of the single-class support vector machine model.

2. A method for rapid detection of valve sealing performance according to claim 1, characterized in that: The step of determining the pressure turning point and obtaining the inflation period and the pressure stabilization period of each acquisition cycle includes: Analyze the pressure difference between each moment and its adjacent moments in each acquisition cycle of the standard valve, and obtain the candidate pressure turning point of each acquisition cycle of the standard valve; The difference between the pressures at the left and right adjacent moments of each candidate pressure turning point in each acquisition cycle of the standard valve is recorded as the pressure turning difference of each candidate pressure turning point; The candidate pressure turning point corresponding to the minimum value of all the pressure turning differences in each acquisition cycle is used as the pressure turning point of each acquisition cycle of the standard valve; The pressure at all times before the pressure turning point of each acquisition cycle is recorded as the inflation period, and the pressure at all times after the pressure turning point is recorded as the pressure stabilization period.

3. A rapid valve sealing performance detection method as claimed in claim 2, characterized in that: The method for obtaining the candidate pressure turning point of each acquisition cycle of the standard valve is as follows: The difference between the pressure at each moment in each acquisition cycle of the standard valve and the pressure at the previous moment is recorded as the first difference; The difference between the pressure at each moment in each acquisition cycle of the standard valve and the pressure at the next moment is recorded as the second difference; The absolute value of the sum of the first difference and the second difference is used as the pressure fluctuation index at each moment in each acquisition cycle; In each acquisition cycle, the pressures at the moments corresponding to the previous preset number of pressure fluctuation indices are selected from large to small according to the pressure fluctuation indices as candidate pressure turning points.

4. A rapid valve sealing performance detection method as claimed in claim 1, characterized in that: The step of determining the inflation pressure deviation of each acquisition cycle of the standard valve comprises: Calculate the average of the pressures at all times during the inflation period of each acquisition cycle of the standard valve, and record it as the first average; Determine the difference between the pressure at each moment in the inflation period of each acquisition cycle of the standard valve and the first mean value as the relative pressure difference; The relative pressure differences at all times during the inflation period are merged to determine the inflation pressure deviation of each acquisition cycle of the standard valve.

5. A method for rapid detection of valve sealing performance as claimed in claim 1, characterized in that: The relative cumulative deviation of each sliding window during the pressure stabilization period of the standard valve includes: Calculate the mean of the pressure at all times in each sliding window during the pressure stabilization period, and record it as the window pressure mean; Determine the difference between the pressure at each moment in each sliding window during the pressure stabilization period and the window pressure average as a third difference; The third differences at all times in each sliding window during the pressure stabilization period are fused to determine the relative cumulative deviation of each sliding window during the pressure stabilization period of the standard valve.

6. A valve sealing performance rapid detection device, which implements the method as claimed in claim 1, characterized in that: The device comprises: Valve pressure acquisition module, which collects the pressure of the standard valve at each moment in each acquisition cycle and the pressure of the valve to be tested at each moment; A model building module determines the pressure turning point according to the local difference between the pressures at different moments in each acquisition cycle of the standard valve, and obtains the inflation period and the pressure stabilization period of each acquisition cycle of the standard valve; determines the inflation cycle deviation of each acquisition cycle of the standard valve according to the periodic variation characteristics and the degree of discreteness of the pressure at all moments in the inflation period of each acquisition cycle of the standard valve; determines the pressure stabilization cumulative deviation of each acquisition cycle of the standard valve according to the local deviation of the pressure in the pressure stabilization period of each acquisition cycle of the standard valve; determines the marginal error parameter of the single-class support vector machine model based on the inflation cycle deviation and the pressure stabilization cumulative deviation; The valve detection module detects the sealing performance of the valve to be tested based on the marginal error parameters of the single-class support vector machine model and the pressure of the valve to be tested at each moment.

7. A valve sealing performance rapid detection system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of a method for quickly detecting valve sealing performance as described in any one of claims 1 to 5 are implemented.

Images