Valve detection system and detection method thereof

Abstract

The invention provides a valve detection system and a detection method thereof. The valve detection system comprises a valve data module, an associated valve module, a first prediction module, an associated verification module and a second prediction module, the valve data module comprises associated features and fault types of valves in the system; the associated valve module can obtain an associated valve having fault propagation relevance with the to-be-detected valve according to the associated characteristics of the valve; the first prediction module can judge whether the valve has a fault risk or not according to the deviation between actual detection data and theoretical data of the valve; when the judgment result is yes, the association verification module can verify the judgment result of the first prediction module through the obtained detection data of the associated valve; and when the verification result is that the first prediction module is misjudged, the fault risk of the valve is judged again through the second prediction module. The detection accuracy is improved, and false alarms are reduced.

Description

Technical Field

[0001] The present application relates to the field of valve safety detection, and in particular to a valve detection system and a detection method thereof. Background Art

[0002] Valves are control components in fluid delivery systems, and have functions such as shutting off and unblocking fluids, adjusting flow rates, guiding, checking, stabilizing pressure, dispersing fluids, and overflow and pressure relief. Valves come in a wide variety of types and specifications, including simple-structured stop valves and ball valves, as well as complex valves suitable for automatic control systems. With the expansion of production demand and the improvement of industrial automation technology, efficient and automatic valve detection methods are gaining more and more attention in order to ensure smooth production and prevent accidents, especially in the fields of aerospace, nuclear power, thermal power, nuclear energy, high temperature and high pressure. In order to ensure the normal operation of pressure pipelines, valves in pressure pipelines need to be inspected regularly to promptly detect problems with the valves. Current valve detection methods include using instruments to detect changes in pressure and other parameters in pressure pipes, and using photographic instruments or infrared, laser and other instruments to periodically collect data on the overall working status of the valve, obtain its appearance data and working status data, and use qualitative judgment models to collect these detection data to achieve fault warning. However, the use of such methods is often rigid in design and can only provide warnings after a fault occurs, but cannot effectively prevent the fault. There is also a lack of means to verify the fault, so the judgment result is either an existing fault or a large number of false alarms, resulting in the detection system being unable to make accurate judgments. Summary of the invention

[0003] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art can think of other obvious variations. The basic principles of the present invention defined in the following description can be applied to other embodiments, variations, improvements, equivalents, and other technical solutions that do not deviate from the spirit and scope of the present invention.

[0004] In order to solve the technical problem, the present application provides a valve detection system, including a valve data module, an associated valve module, a first prediction module, an associated verification module and a second prediction module; the valve data module includes associated features and fault types of valves in the system; the associated valve module can obtain associated valves with fault propagation correlation with the valve to be inspected according to the associated features of the valves; The first prediction module can detect the valve, and determine whether there is a risk of failure of the valve based on the deviation between the actual detection data and the theoretical data of the valve; when the judgment is yes, the associated verification module can detect the associated valves of the valve to be detected, and verify the judgment result of the first prediction module by obtaining the detection data of the associated valves; when the verification result shows that the first prediction module makes a misjudgment, the second prediction module can re-judge the risk of failure of the valve.

[0005] Among them, preferably, the associated features include valve position, material, function, medium, flow rate, pressure, vibration frequency, and operation sequence.

[0006] Among them, preferably, the failure types include jamming, blockage, adhesion, leakage, and valve body fracture.

[0007] Among them, preferably, when the first prediction module and the second prediction module detect, they judge the risk of failure by obtaining the deviation of several detection indicators corresponding to the failure type.

[0008] Among them, preferably, the judgment result of the first prediction module is verified by obtaining the deviation of the risk-associated indicators with risks among the detection indicators.

[0009] This application also provides a valve detection method using the valve detection system as described above, including the following steps: S1, the system provides the detection of n valves FM, FM = [FM1, FM2, FM3,..., FM n , where the i-th valve is FM i ; set that the valve FM i can have m types of failures IQ, IQ = [IQ1, IQ2, IQ3,..., IQ m , where the j-th type of failure is IQ j ; Execute the first prediction step for the valve FM i : Set that the judgment of the failure type IQ j requires obtaining k detection indicators JB of the valve, JB = [JB1, JB2, JB3,..., JB k , where the z-th detection indicator is JB z ; Set that the actual detection value of the detection indicator JB z obtained at the first detection time IT1 is Zs, and according to the judgment threshold Zc of the set detection indicator JB z , the indicator deviation θ z of the detection indicator JB z in this detection can be obtained=Zs - Zc; At the first detection time IT1, according to the obtained detection index JB z of the index deviation θ z , the first prediction score of the failure type IQ i of the valve FM j can be obtained; ; According to the set risk threshold IW0 of the failure type IQ j ; When IW1 > IW0, transfer to the associated verification step S2: S2. Obtain the r risk - associated indicators IJB = [IJB1, IJB2, IJB3,..., IJB i of the valve FM and its associated valves according to the index deviation obtained at the detection time IT1, where r ≤ k, and the p - th risk - associated indicator is IJB r , and the index deviation of IJB p is θ p ; p ; Obtain the associated risk threshold GW0; Obtain the c associated valves IG = [IG1, IG2, IG3,..., IG i of the valve FM, where the u - th associated valve is IG c , and the correlation coefficient between FM u and IG i is ω u ; Among them, the associated valves of IQ u are related to the valve FM j , and are the valves that will be affected by the failure propagation themselves when the valve FM i has a failure risk; i At the first association time GT1, GT1 = IT1+△T1, obtain the actual detection value of the risk - associated indicator IJB u of the u - th associated valve IG p , and then according to the judgment threshold of IJB p , the index deviation θ u of the risk - associated indicator IJB p of the associated valve IG u can be obtained; Thus, obtain the associated index deviation p of IJB ; Obtain the associated prediction scores of the r risk - associated indicators IJB ; When GW1≥GW0, give a system warning for the failure type IQ j ;​ When GW1 < GW0, enter the second prediction step S3: S3. At the second detection time IT2, where IT2 = GT1 + △T2, obtain k detection indexes JB of the valve FM i According to the obtained detection indexes JB z and the index deviation θ z ' in this detection, the second prediction score of the failure type IQ i of the valve FM j can be obtained; ; When IW2 ≥ IW1, confirm that there is a risk. At this time, give a system warning for the failure type IQ j ; When IW2 ≤ IW0, confirm that the risk is lifted, end this step, and transfer to the first prediction step S1; When IW0 < IW2 < IW1, the risk situation cannot be confirmed, and transfer to the associated detection step S2.

[0010] In the associated detection step S2, the method for obtaining the risk associated index is as follows: Set the maximum deviation θ z of the detection index JB max ; When the index deviation θ z of the detection index JB obtained at the detection time IT1 z > θ max , use JB z as the risk associated index of the valve FM i and its associated valve.

[0011] In the associated detection step S2, the method for obtaining the associated risk threshold is as follows: Obtain the risk associated distribution λ = r / k; Obtain the associated risk threshold .

[0012] In the associated detection step S2, the method for obtaining the associated valve of the valve FM i is as follows: By establishing a neural network model, use the associated features and the degree of association as the input and output data samples, and train the neural network model.

[0013] The beneficial effects achieved by this application are as follows: This application can overcome the problem of frequent false alarms in the linear judgment of the qualitative model. First, after obtaining the determination of the risk of a fault; then, through the associated influence relationship of the associated valves, the determination result is mapped and verified. When both double determinations represent the occurrence of a risk, an alarm is sent to the system. In this way, the accuracy of the alarm will be greatly improved. At the same time, the threshold of the first determination can also be lowered so that data collection and recording reminders can be carried out when the fault has not fully occurred. The triggering speed is fast, the cost is low, and the startup and verification of the early warning are very rapid, with a wide range of applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments recorded in the present application. For those skilled in the art, other drawings can also be obtained according to these drawings.

[0015] Figure 1 It is a flowchart of the steps of the valve detection method of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts fall within the scope of protection of the present application. It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0017] Set the system of the present application to provide automatic detection of n valves FM, FM = [FM1, FM2, FM3,..., FM n , where the i-th valve is FM i . It is known that the valve FM i can have m types of faults IQ, IQ = [IQ1, IQ2, IQ3,..., IQ m, where the j-th type of fault is IQ j .

[0018] For example, for globe valves and regulating valves in hydraulic pipeline applications, due to long-term contact with liquid media, the common types of faults that are likely to occur generally include jamming, blockage, valve body fracture, etc.; while for boiler safety valves, since they need to bounce according to pressure, the specific types of faults include valve jamming, adhesion, leakage, etc.

[0019] One of the design purposes of this application is to predict and alarm in advance the possible faults of valves before they cause harm. Therefore, first, for valve FM i Execute the first prediction step: The first prediction step includes: setting that to judge the fault type IQ j it is necessary to obtain k detection indicators JB of the valve, JB = [JB1, JB2, JB3,..., JB k , where the z-th detection indicator is JB z ; Set that the actual detected value of the detection indicator JB z obtained at the first detection time IT1 is Zs. According to the set judgment threshold Zc of the detection indicator JB z , the index deviation θ z of the detection indicator JB z in this detection can be obtained, θ = Zs - Zc; for example, in the embodiment, an important detection and evaluation indicator corresponding to the valve fault type "jamming" is the deviation between the actual opening and the commanded opening of the valve. Once the deviation between the valve opening and the commanded opening is large, it indicates that the risk of "jamming" is also large. In addition, the deviations of detection indicators such as valve flow rate and pressure difference can also be used as the judgment basis for this fault type.

[0020] At the first detection time IT1, according to the obtained index deviation θ z of the detection indicator JB z , the first prediction score i of the valve FM j for the occurrence of the fault type IQ can be obtained; According to the set risk threshold IW0 of the fault type IQ j ; When IW1 > IW0, it indicates that the valve FM i has a risk of occurring the fault type IQ jThe risk, according to the design concept of this application, is to anticipate faults in advance, that is, prevention should be carried out in advance before the faults occur. Therefore, when initially setting, the value of the risk threshold IW0 should be set relatively small. However, once the value of IW0 is set small, misjudgment may occur frequently. To solve the problem of misjudgment, it is necessary to verify the anticipation again, and then transfer to the associated verification step: Set the detection index JB z The maximum deviation θ max ; When the detection index JB obtained at the IT1 detection time z The index deviation θ z > θ max , then take JB z As the risk associated index of the valve FM i And its associated valve; Set a total of r risk associated indexes IJB = [IJB1, IJB2, IJB3,..., IJB r are obtained at the IT1 detection time, r ≤ k, where the p-th risk associated index is IJB p , IJB p The index deviation of is θ p ; Obtain the risk associated distribution λ = r / k; Obtain the associated risk threshold ; Among the remaining n - 1 valves, obtain the associated valve of the valve FM i , and the associated valve refers to the valve that will be affected by the associated failure and have a certain risk when the valve FM i Has a failure type IQ j .

[0021] Specifically, the valve FM i And its associated valve are usually located in the same pipeline system, with relatively close physical positions, and control parameters such as flow rate, direction, or pressure on the same operation platform. The more similar the valve and its associated valve are in terms of type, function, medium, etc., and the more operationally dependent they are on each other, the higher their correlation; For example, in an embodiment of a hydraulic application system, according to historical data, it can be found that among the data of two valves with a spacing L ≤ S Δt (where S is the pressure wave velocity, about 1480 m / s for water medium; Δt is the system response time threshold, usually taken as < 50 ms) and a flow interaction coefficient K > 0.3, it shows that a change in the opening of one valve can cause a change in the flow rate of the other valve by more than 30%. Therefore, the valves that meet the above conditions are used as associated valves, and the situation of flow rate change is used as a quantitative assignment of their correlation.

[0022] In other embodiments, there are similar situations. For example, when a valve and its associated valve handle the same medium (such as slurry containing solid particles), the probability of synchronous seat erosion increases by 3 - 5 times; for valves coexisting in a high-vibration area (such as the compressor outlet), mechanical fatigue failures show a linear correlation (the risk is significant when the vibration acceleration > 5g). Another example is in a steam generator system, where multiple steam valves and condensate valves usually need to be opened or closed in a certain order, and their operations are complementary or dependent on each other. For instance, after one valve is opened, another valve will close to prevent reverse water flow. At this time, if this valve fails, it will directly affect the operation of the other valve, and these two valves are regarded as associated valves with a relatively high degree of association.

[0023] Therefore, the above-mentioned associated data can be quantified as the degree of association of associated valves.

[0024] In addition, the valve association characteristics and the degree of association can also be used as input and output data samples (x, y) to train a neural network model.

[0025] Among them, the association characteristics include valve position, material, function, medium, flow rate, pressure, vibration frequency, and operation sequence, and the output sample is the degree of association of the valve.

[0026] The specific algorithm is as follows: divide a hyperplane in the sample space , denoted as (w, b), and the functional margin of the sample point to the divided hyperplane is , and the geometric margin is:

[0027] If , it can be seen that the functional margin and the geometric margin are equal. If the hyperplane parameters w and b change proportionally, the functional margin also changes proportionally, while the geometric margin remains unchanged.

[0028] The mathematical formula of the support vector machine is:

[0029] The value of the functional margin does not affect the solution of the optimization problem. The functional margin has no influence on the inequality of the above optimization problem and also has no influence on the objective function. Therefore, take , and at the same time, maximizing is equivalent to minimizing , and the optimization problem of the support vector machine for linearly separable data can be obtained;

[0030] Solve the dual problem by Lagrangian duality; establish the Lagrangian function and introduce the Lagrange multiplier , define the Lagrangian function:

[0031] According to the duality of the original problem, the duality of the original problem is a max-min problem, that is

[0032] When finding the minimum, Taking the derivatives with respect to \(w\) and \(b\) and setting them to zero respectively, we can get

[0033] Substituting it into the dual problem, we can get

[0034] Solving for After that, then \(w\) and \(b\) can also be obtained accordingly

[0035] The KKT conditions for the above inequality constraint

[0036] For any training sample There is always Or ; The dual problem is:

[0037] Introduce the kernel function

[0038] Let the original space be , and the new space is , define the transformation (mapping) from the original space to the new space:

[0039] Let be a symmetric function, then The necessary and sufficient condition for to be a positive definite kernel function is that for any The corresponding Gram matrix (the matrix composed of vector inner products): is a positive semi-definite matrix: all eigenvalues are not less than 0 The objective function of SVM is:

[0040] Here \(C\) represents the degree of penalty for classification errors in the case of linear inseparability; in principle, \(C\) can be chosen as any number greater than 0 according to needs. The larger \(C\) is, the greater the total error in the entire optimization process The higher the degree of attention, the higher the requirement for reducing errors, and the interval may even be reduced at all costs.

[0041] According to the above model, c associated valves IG = [IG1, IG2, IG3, …, IG c with fault propagation relevance for valve FM are obtained, where the u-th associated valve is IG u , and the correlation coefficient between FM i and IG u is ω u ; At the first associated time GT1 (GT1 = IT1 + △T1), the actual detected value of the risk association index IJB u of the u-th associated valve IG p is obtained. Then, according to the judgment threshold of IJB p , the index deviation θ u of the risk association index IJB p of the associated valve IG u can be obtained; Thus, the associated index deviation p of IJB is obtained; The associated prediction scores of r risk association indexes IJB are obtained ; When GW1 ≥ GW0, it means that after passing the associated verification, it is confirmed that there is a risk. At this time, a system warning is issued for the fault type IQ j ; When GW1 < GW0, it means that after passing the associated verification, it is suspected that the risk may be a false alarm. Therefore, the valve needs to be detected again, and the second prediction step is entered: At the second detection time IT2 (IT2 = GT1 + △T2), k detection indexes JB of valve FM i are obtained. According to the index deviation θ z of the detection index JB z in this detection, the second prediction score i of valve FM j for the fault type IQ can be obtained; When IW2 ≥ IW1, it is confirmed that there is a risk. At this time, a system warning is issued for the fault type IQ j ; When IW2 ≤ IW0, it is confirmed that the risk is lifted, and this step ends, and the process transfers to the first prediction step; When IW0 < IW2 < IW1, the risk situation cannot be confirmed, and the process transfers to the associated detection step.

[0042] Furthermore, the present invention also provides a computer-readable storage medium, which includes a stored program. When the program runs, it executes the method described in the above method embodiment.

[0043] Furthermore, the present invention also provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to execute the method described in the above method embodiment through the computer program.

[0044] Furthermore, it should be understood that since the setting of each module is only to illustrate the functional units of the device of the present invention, the physical devices corresponding to these modules can be the processor itself, or a part of the software in the processor, a part of the hardware, or a part of the combination of software and hardware. Therefore, the number of each module in the figure is only illustrative.

[0045] Those skilled in the art can understand that all or part of the processes of the methods in the above embodiments of the present invention can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form. The computer-readable storage medium can include: any entity or device, medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory, random access memory, electrical carrier signal, telecommunication signal, and software distribution medium that can carry the computer program code.

[0046] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is implemented by technologies such as a network, NFC (Near Field Communication), or others. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, trackball, or touchpad set on the shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0047] Those skilled in the art can understand that each module in the device can be adaptively split or combined, and such splitting or combining of specific modules will not cause the technical solution to deviate from the principle of the present invention. Therefore, the technical solutions after splitting or combining will all fall within the protection scope of the present invention.

Claims

1. A valve detection system, characterized in that, It includes a valve data module, an associated valve module, a first prediction module, an associated verification module, and a second prediction module; The valve data module includes the associated features and failure types of the valves in the system; The associated valve module can obtain the associated valves that have a failure propagation correlation with the valve to be inspected according to the associated features of the valve; The first prediction module can detect the valve, and judge whether the valve has the risk of failure through the deviation between the actual detection data and the theoretical data of the valve; when the judgment is yes, the associated verification module can detect the associated valves of the valve to be detected, and verify the judgment result of the first prediction module through the obtained detection data of the associated valves; when the verification result shows that the first prediction module makes a misjudgment, the second prediction module re-judges the failure risk of the valve.

2. The valve detection system according to claim 1, wherein The associated features include valve position, material, function, medium, flow rate, pressure, vibration frequency, and operation sequence.

3. The valve detection system according to claim 1, wherein, The failure types include jamming, blockage, adhesion, leakage, and valve body fracture.

4. The valve detection system according to claim 1, characterized in that, When the first prediction module and the second prediction module detect, they judge the risk of failure by obtaining the deviation of several detection indexes corresponding to the failure type.

5. The valve detection system according to claim 1, characterized in that, The associated verification module verifies the judgment result of the first prediction module by obtaining the deviation of the risk-associated indexes with risks among the detection indexes.

6. A valve detection method using the valve detection system according to any one of claims 1-5, characterized in that, It includes the following steps: S1, the system provides the detection of n valves FM, FM = [FM1, FM2, FM3, …, FM n , where the i-th valve is FM i ; Set that the valve FM i can have m types of failures IQ, IQ = [IQ1, IQ2, IQ3, …, IQ m , where the j-th type of failure is IQ j ; For valve FM i Perform the first prediction step: Set the judgment of the fault type IQ j requires obtaining k detection indicators JB of the valve, JB = [JB1, JB2, JB3, …, JB k , where the z-th detection indicator is JB z ; When the actual detected value Zs of the detection index JB is obtained at the first detection time IT1, according to the judgment threshold Zc of the set detection index JB z , the index deviation θ of the detection index JB in this detection can be obtained z =Zs - Zc; z z ​​ At the first detection time IT1, according to the obtained detection index JB z of the index deviation θ z , the first prediction score of the failure type IQ i of the valve FM j can be obtained ; According to the set fault type IQ j risk threshold IW0; When IW1 > IW0, transfer to the associated verification step S2: S2, obtain the valve FM according to the index deviation obtained from the IT1 detection time i r risk-associated indicators IJB = [IJB1, IJB2, IJB3, …, IJB r of its associated valve, r ≤ k, where the p-th risk-associated indicator is IJB p , IJB p has an index deviation of θ p ; Obtain the associated risk threshold GW0; Obtain valve FM i c associated valves IG = [IG1, IG2, IG3, …, IG c with fault propagation, where the u-th associated valve is IG u , FM i and the correlation coefficient between FM u and IG u is ω At the first associated time GT1, where GT1 = IT1 + △T1, obtain the risk associated index IJB of the u-th associated valve IG u of the actual detected value, and then according to the judgment threshold of IJB p to obtain the risk associated index IJB of the associated valve IG p and the index deviation θ u of the risk associated index IJB of the associated valve IG p ; u ; Thus, the IJB is obtained. p The associated index deviation ; Obtain the correlation prediction scores of r risk-related indicators IJB ; When GW1 ≥ GW0, issue a system warning for fault type IQ j and conduct a system warning; When GW1 < GW0, enter the second prediction step S3: S3. At the second detection time IT2, where IT2 = GT1 + △T2, obtain k detection indicators JB of the valve FM i and, based on the obtained detection indicators JB z and the index deviation θ z ' in this detection, be able to obtain the failure type IQ i of the valve FM j and the second prediction score ; When IW2 ≥ IW1, it is confirmed that there is a risk of occurrence. At this time, for the fault type IQ j a system warning is issued; When IW2 ≤ IW0, confirm that the risk is lifted, end this step, and transfer to the first prediction step S1; When IW0 < IW2 < IW1, the risk situation cannot be confirmed, and transfer to the associated detection step S2.

7. A valve detection method as described in claim 6, characterized in that, In the associated detection step S2, the method for obtaining the risk-associated indexes is: Set the detection index JB z The maximum deviation θ max ; When the detection index JB obtained at the IT1 detection time z has an index deviation θ z > θ max then JB z is used as the risk correlation index of the valve FM i and its associated valves.

8. A valve detection method as described in claim 6, characterized in that, In the associated detection step S2, the method for obtaining the associated risk threshold is: Obtain the risk-associated distribution λ = r / k; Obtain the associated risk threshold .

9. A valve detection method as described in claim 6, characterized in that, In the association detection step S2, the associated valve FM is obtained i The method for obtaining the associated valve is as follows: By establishing a neural network model, using the association features and the degree of association as input and output data samples, the neural network model is trained.

Images