Multi-sensor information fusion gas-liquid linkage valve health state diagnosis method and assembly

Abstract

The invention discloses a gas-liquid linkage valve health state diagnosis method and assembly based on multi-sensor information fusion, and the method comprises the steps: obtaining the performance parameters of a gas-liquid linkage valve in a normal state through sensors in a daily maintenance function test process of the gas-liquid linkage valve, and building a reference curve; meanwhile, performance parameters in a fault state are obtained, an abnormal state curve is established, and abnormal state characteristic parameters are extracted through comparison between the reference state and an abnormal state difference curve; in daily maintenance, according to dynamic performance parameters obtained by a sensor, features are extracted through a difference curve between the dynamic performance parameters and a reference signal, a fuzzy clustering analysis method is used for judging the proximity between a test object and various fault states, the fault type is rapidly judged, and the fault position is accurately positioned. Therefore, the health state of the gas-liquid linkage valve can be rapidly and quantitatively diagnosed in the daily maintenance process of the valve, and powerful support is provided for the full-life-cycle energy quantification evaluation of the gas-liquid linkage valve in the future.

Description

Technical Field

[0001] The present invention relates to the technical field of natural gas pipeline transportation and control, and particularly to a method and component for diagnosing the health status of a gas-liquid actuated valve through multi-sensor information fusion. Background Art

[0002] As a natural gas transportation control device with a relatively high frequency of failures, the gas-liquid actuated valve has become an important object for maintenance by gas fields and pipeline companies. At present, the daily maintenance of gas-liquid actuated valves in natural gas pipelines mainly adopts the traditional preventive maintenance method of "regular inspection and after-the-fact maintenance". However, in daily maintenance, when a gas-liquid actuated valve fails, it takes a long time for problem analysis and item-by-item inspection to identify and judge the failure, lacking real-time performance and having the problem of delayed fault diagnosis, which is likely to cause safety accidents. At the same time, only functional tests are carried out during the conventional maintenance of gas-liquid actuated valves, and no performance data is retained, resulting in the inability to know the health status of the valve after maintenance and it is difficult to provide guidance for the special treatment of the valve in the future. Summary of the Invention

[0003] Aiming at the problems proposed in the background art, the purpose of the present invention is to provide a method and component for diagnosing the health status of a gas-liquid actuated valve through multi-sensor information fusion, which solves the problem in the prior art that the gas-liquid actuated valve cannot be identified and judged for faults in real time, thus leading to potential safety hazards.

[0004] The present invention is achieved through the following technical solutions:

[0005] In the first aspect of the present invention, a method for diagnosing the health status of a gas-liquid actuated valve through multi-sensor information fusion is provided, including the following steps:

[0006] Install multiple types of sensors based on the general reserved interface of the gas-liquid actuated valve, and perform dynamic tests on the opening and closing processes of the gas-liquid actuated valve through the multiple types of sensors to obtain performance parameters;

[0007] Perform wavelet energy calculation on the performance parameters to obtain the energy value of the gas-liquid actuated valve in the normal state and the energy value in the abnormal state;

[0008] Construct a reference curve based on the energy value in the normal state and an abnormal state curve based on the energy value in the abnormal state;

[0009] Perform difference calculation on the abnormal state curve and the reference curve to obtain an abnormal state difference curve, and use an anomaly detection algorithm to extract the abnormal state feature parameters from the abnormal state difference curve;

[0010] Conduct dynamic tests on the on - diagnosis gas - liquid actuated valve during the opening and closing processes to obtain the to - be - diagnosed dynamic signal of the gas - liquid actuated valve. Use a deep - learning anomaly detection algorithm to extract the abnormal state feature parameters from the difference curve between the to - be - diagnosed dynamic signal and the reference curve.

[0011] Perform fuzzy clustering analysis on the to - be - diagnosed abnormal state feature parameters and the abnormal state feature parameters to obtain the fault membership degree, and judge the health state of the gas - liquid actuated valve to be tested according to the fault membership degree.

[0012] In the above - mentioned technical solution, during the daily maintenance function test of the gas - liquid actuated valve, the performance parameters of the gas - liquid actuated valve in the normal state are obtained through sensors and a reference curve is established. At the same time, the performance parameters in the fault state are obtained and an abnormal state curve is established. By comparing the difference curve between the reference state and the abnormal state, the abnormal state feature parameters are extracted. In daily maintenance, according to the dynamic performance parameters obtained by sensors, features are extracted through the difference curve with the reference signal, and the fuzzy clustering analysis method is used to judge the proximity of the test object to various fault states, quickly judge the fault type and accurately locate the fault position, so as to realize the rapid quantitative diagnosis of the health state of the gas - liquid actuated valve during the daily maintenance process of the valve, providing strong support for the quantitative evaluation of the performance of the gas - liquid actuated valve throughout its life cycle in the future.

[0013] In a possible embodiment, the wavelet energy calculation of the performance parameters to obtain the energy values of the gas - liquid actuated valve in the normal state and the energy values in the abnormal state includes:

[0014] Divide the performance parameters into normal performance parameters in the normal state and abnormal performance parameters in various abnormal states;

[0015] Based on the mutation characteristics, use the wavelet packet decomposition technology to perform three - layer wavelet packet decomposition on the normal performance parameters and the abnormal performance parameters respectively to obtain signal sequences in multiple independent frequency bands;

[0016] Normalize the signal sequences in multiple independent frequency bands based on the energy change characteristics of each independent frequency band to obtain the energy values of each independent frequency band, and set the energy values of the frequency bands as state feature values;

[0017] Extract the total values of each independent frequency band, and calculate the energy values of each independent frequency band in the normal state and the energy values in each abnormal state by calculating the state feature values of each independent frequency band and their corresponding total values of the independent frequency bands.

[0018] In a possible embodiment, constructing a reference curve based on the energy values in the normal state and constructing an abnormal state curve based on the energy values in the abnormal state includes:

[0019] Plot the energy values of each independent frequency band in the normal state in the form of a curve graph to obtain a reference curve;

[0020] Plot the energy values of each independent frequency band in various abnormal states in the form of curve graphs respectively to obtain multiple abnormal state curves.

[0021] In a possible embodiment, perform a difference calculation on the abnormal state curve and the reference curve to obtain an abnormal state difference curve, including:

[0022] Calculate the energy difference between the abnormal state curve and the reference curve for each independent frequency band, and plot the energy differences for each independent frequency band in the form of a curve graph to obtain an abnormal state difference curve.

[0023] In a possible embodiment, use an anomaly detection algorithm to extract abnormal states from the abnormal state difference curve to obtain abnormal state characteristic parameters, including:

[0024] Calculate the state support degree of the energy differences for each independent frequency band, and select a set of state conditions from each independent frequency band according to the state support degree;

[0025] Calculate the state dependence degree of each abnormal state on the set of state conditions;

[0026] Construct an abnormal state characteristic graph based on the state support degree and the state dependence degree, and determine the abnormal state characteristic parameters according to the abnormal state characteristic graph.

[0027] In a possible embodiment, construct an abnormal state characteristic graph based on the state support degree and the state dependence degree, and determine the abnormal state characteristic parameters according to the abnormal state characteristic graph, including:

[0028] Use one independent frequency band as a node, use the state support degree as the node parameter, and use the absolute value of the difference between the state dependence degrees of two nodes as the connection parameter to complete the construction of the abnormal state characteristic graph;

[0029] Select the node with the largest node parameter, put its node parameter into the queue, and then select the node with the smallest connection parameter from the nodes connected to the node with the largest node parameter and put its node parameter into the queue until all nodes in the abnormal state characteristic graph are traversed to obtain a queue to be calculated;

[0030] Calculate the queue to be calculated according to the queue order to obtain the abnormal state characteristic parameters.

[0031] The second aspect of the present invention provides a health state diagnosis system for a gas-liquid linked valve with multi-sensor information fusion, including:

[0032] Dynamic test module, which is used to install sensors based on the general reserved interface of the gas-liquid actuated valve, dynamically test the opening and closing process of the gas-liquid actuated valve, and obtain the performance parameters of the gas-liquid actuated valve;

[0033] Performance calculation module, which is used to install multiple types of sensors based on the general reserved interface of the gas-liquid actuated valve, dynamically test the opening and closing process of the gas-liquid actuated valve through multiple types of sensors, and obtain performance parameters;

[0034] Energy value calculation module, which is used to perform wavelet energy calculation on the performance parameters to obtain the energy value of the gas-liquid actuated valve in the normal state and the energy value in the abnormal state;

[0035] Abnormal feature module, which is used to calculate the difference between the abnormal state curve and the reference curve to obtain the abnormal state difference curve, and use the abnormal detection algorithm to extract the abnormal state from the abnormal state difference curve to obtain the abnormal state characteristic parameters;

[0036] Module to be diagnosed, which is used to dynamically test the opening and closing process of the gas-liquid actuated valve to be diagnosed, obtain the dynamic signal to be diagnosed of the gas-liquid actuated valve to be diagnosed, and use the deep learning abnormal detection algorithm to extract the abnormal state from the difference curve between the dynamic signal to be diagnosed and the reference curve to obtain the abnormal state characteristic parameters to be diagnosed;

[0037] Fault membership module, which is used to perform fuzzy clustering analysis on the abnormal state characteristic parameters to be diagnosed and the abnormal state characteristic parameters to obtain the fault membership degree, and judge the health state of the gas-liquid actuated valve to be tested according to the fault membership degree.

[0038] In an optional embodiment, the performance calculation module includes:

[0039] Abnormal performance unit, which is used to divide the performance parameters into normal performance parameters in the normal state and abnormal performance parameters in various abnormal states;

[0040] Wavelet decomposition unit, which is used to perform three-layer wavelet packet decomposition on the normal performance parameters and abnormal performance parameters respectively based on the mutation characteristics by using the wavelet packet decomposition technology to obtain signal sequences in multiple independent frequency bands;

[0041] State feature unit, which is used to normalize the signal sequences in multiple independent frequency bands based on the energy change characteristics of each independent frequency band to obtain the energy values of each independent frequency band, and set the energy values of the frequency bands as state feature values;

[0042] An energy value calculation unit, which is used to extract the total value of each independent frequency band, calculate the state characteristic value of each independent frequency band with its corresponding total value of the independent frequency band, and obtain the energy value of each independent frequency band in the normal state and the energy value in each abnormal state.

[0043] The third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements a health state diagnosis method for a gas-liquid linked valve with multi-sensor information fusion.

[0044] The fourth aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, it implements a health state diagnosis method for a gas-liquid linked valve with multi-sensor information fusion.

[0045] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0046] The rapid health state diagnosis method for the gas-liquid linked valve proposed by the present invention can quantitatively evaluate and analyze the performance state, fault type and location of the valve based on the valve action signals collected during the daily maintenance of the valve, and can realize the health monitoring of the entire life cycle of the gas-liquid linked valve. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following will briefly introduce the drawings required for the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings. In the drawings:

[0048] Figure 1 It is a schematic structural diagram of the gas-liquid linked valve structure principle and sensor installation position provided in Embodiment 1 of the present invention;

[0049] Figure 2 It is a schematic diagram of the dynamic test acquisition signal parameters and calculation verification parameters provided in Embodiment 1 of the present invention;

[0050] Figure 3 It is a schematic flow diagram of the rapid health state diagnosis method for the gas-liquid linked valve provided in Embodiment 1 of the present invention;

[0051] Figure 4 It is a schematic diagram of the power curves of the solenoid valve in different states provided in Embodiment 1 of the present invention;

[0052] Figure 5 It is a schematic diagram of the energy ratio of each frequency band decomposed by wavelet packet provided in Embodiment 1 of the present invention;

[0053] Figure 6 Schematic diagram of the solenoid valve fault clustering effect provided by Embodiment 1 of the present invention;

[0054] Figure 7 Schematic diagram of the structure of an electronic device provided by Embodiment 3 of the present invention. Specific embodiments

[0055] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with embodiments and the accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0056] Embodiment 1

[0057] Figure 3 Schematic diagram of the flow of the method for rapid diagnosis of the health status of a gas-liquid actuated valve provided by Embodiment 1 of the present invention, as Figure 3 shown. The method for diagnosing the health status of a gas-liquid actuated valve by multi-sensor information fusion includes the following steps:

[0058] Step S1: Install multiple types of sensors based on the general reserved interface of the gas-liquid actuated valve, and perform dynamic tests on the opening and closing processes of the gas-liquid actuated valve through the multiple types of sensors to obtain performance parameters.

[0059] In a possible embodiment, installing sensors based on the general reserved interface of the gas-liquid actuated valve includes:

[0060] Install a gas storage tank pressure sensor on the reserved interface of the gas-liquid actuated valve, install a liquid level gauge and a pressure sensor on the top of the gas-liquid tank, install an oil pressure sensor and an oil cylinder oil pressure sensor at the bottom of the gas-liquid tank, install a valve position sensor at the valve stem, and connect a current and voltage sensor in series to the DC coil of the solenoid valve.

[0061] It should be noted that since there may be safety hazards in modifying the gas-liquid actuated valve, and at the same time, there may be structural changes between the new and old gas-liquid actuated valves. In order to make the structures of the new and old gas-liquid actuated valves safe and applicable, in the present invention, various sensors are installed on the existing interfaces of the gas-liquid actuated valve without modifying the structure of the gas-liquid actuated valve.

[0062] Specifically, Figure 1 Schematic diagram of the structural principle of the gas-liquid actuated valve and the installation positions of sensors provided by Embodiment 1 of the present invention, as Figure 1As shown in the figure, the air storage tank pressure sensor P1 is installed on the reserved interface 5.2 of the gas-liquid linkage valve; the liquid level gauge L1 and the pressure sensor P2 are installed on the top of the gas-liquid tank 4-C, and the liquid level gauge L2 and the pressure sensor P3 are installed on the top of the gas-liquid tank 4-O; the oil pressure sensor P4 is installed at the bottom of the gas-liquid tank 4-C, and the oil pressure sensor P5 is installed at the bottom of the gas-liquid tank 4-O; the oil cylinder oil pressure sensor P6 is installed at the bottom of the gas-liquid tank 4-C, and the oil cylinder oil pressure sensor P7 is installed at the bottom of the gas-liquid tank 4-O; the valve position sensor S1 is installed at the valve stem; the DC coil of the solenoid valve is connected in series with the current-voltage sensor, where the current-voltage sensor includes C1 / V1, C2 / V2, C3 / V3, C4 / V4 in the figure.

[0063] In a possible embodiment, the performance parameters include: the air storage tank air pressure signal, the gas-liquid tank top air pressure signal, the gas-liquid tank bottom oil pressure signal, the oil pressure signals on both sides of the oil cylinder piston, the gas-liquid tank liquid level signal, the valve opening signal, the valve opening and closing time, the solenoid valve current signal, the solenoid valve voltage signal, and the solenoid valve action time.

[0064] Step S2: Perform wavelet energy calculation on the performance parameters to obtain the energy value of the gas-liquid linkage valve in the normal state and the energy value in the abnormal state.

[0065] Before performing the wavelet energy calculation, it is necessary to calculate the calibration parameters of the performance parameters, where the calibration parameters include: the valve opening and closing stroke torque, the solenoid valve inductance coefficient, and the solenoid valve power. As Figure 2 shown, the opening and closing stroke torque can be calculated through the air storage tank air pressure signal, the gas-liquid tank top air pressure signal, the gas-liquid tank bottom oil pressure signal, the oil pressure signals on both sides of the oil cylinder piston, the gas-liquid tank liquid level signal, the valve opening signal, and the valve opening and closing time; the solenoid valve inductance coefficient and the solenoid valve power can be calculated through the solenoid valve current signal, the solenoid valve voltage signal, and the solenoid valve action time.

[0066] Among them, the valve opening and closing stroke torque can be obtained by plotting the torque at each time point in the form of a curve graph.

[0067] The calculation process of the solenoid valve inductance coefficient includes:

[0068]

[0069] In the above formula, L(t) is the inductance coefficient, U(t) is the voltage passing through the solenoid valve, I(t) is the current passing through the solenoid valve, R is the solenoid valve resistance, and t is the energization time.

[0070] Among them, U(t) and I(t) can be obtained through SIPLUG measurement, dI / dt can be calculated from I(t), and R is known, so the differential equation about L(t) can be constructed through the above formula. The differential equation can be solved by ADAM software to obtain the inductance coefficient L(t).

[0071] The calculation process of solenoid valve power includes:

[0072] P=UI

[0073] In the above formula, P is the solenoid valve power.

[0074] Among them, the power of the 24V solenoid valve is about 12W and the current is about 500mA.

[0075] The current and voltage under the three states of normal state, solenoid valve core stuck and spring deterioration are collected, and the calibration parameter power P is obtained based on the formula P = UI. The power P is transformed into the curve of the action cycle of the solenoid valve under each state. Figure 4 as shown.

[0076] At this time, the performance curve of each calibration parameter of the performance parameter can be obtained. At this time, the performance curve is decomposed into three layers of wavelet packets using the db wavelet basis function.

[0077] In an optional embodiment, wavelet energy calculation is performed on the performance parameter to obtain the energy value of the gas-liquid linkage valve in a normal state and the energy value in an abnormal state, including:

[0078] The performance parameters are divided into normal performance parameters under normal conditions and abnormal performance parameters under various abnormal conditions;

[0079] Based on the mutation characteristics, the normal performance parameters and the abnormal performance parameters are respectively decomposed into three layers of wavelet packets using wavelet packet decomposition technology to obtain signal sequences of multiple independent frequency bands;

[0080] Normalizing the signal sequences of the multiple independent frequency bands based on the energy change characteristics of each independent frequency band to obtain a frequency band energy value of each independent frequency band, and setting the frequency band energy value as a state characteristic value;

[0081] The independent frequency band total value of each independent frequency band is extracted, and the state characteristic value of each independent frequency band is calculated with the corresponding independent frequency band total value to obtain the energy value of each independent frequency band in a normal state and the energy value of each independent frequency band in each abnormal state.

[0082] It should be noted that since there are inflection points in the performance parameters in the abnormal state, the inflection points are used as the mutation characteristics during wavelet packet decomposition. Based on this mutation characteristic, the performance parameters are decomposed by three-layer wavelet packet decomposition, and 8 frequency bands can be obtained. After normalizing the frequency band energy values, the frequency band energy values can be obtained, and the total energy value of each frequency band is extracted. Calculate the ratio of the energy of each frequency band to the total energy of the frequency bands, and use this energy ratio as the energy value. This energy value can be used as the basis for subsequent abnormal feature extraction.

[0083] The abnormal state characteristic parameters include air pipeline blockage, air pipeline leakage, air control valve group jamming, internal leakage of the air control valve group, external leakage of the air control valve group, oil pipeline blockage, oil pipeline external leakage, failure of the oil cylinder piston seal, damage to the transmission device, excessive torque for valve opening and closing, deterioration of the solenoid valve spring, and solenoid valve power supply failure. The present invention uses step S2 to calculate the above abnormal state characteristic parameters. Taking the abnormal state of the solenoid valve as an example, step S2 is used to calculate the solenoid valve data, and a schematic diagram of the energy ratio of each frequency band obtained by wavelet packet decomposition as shown in Figure 5 can be obtained.

[0084] Step S3: Construct a reference curve based on the energy values in the normal state, and construct an abnormal state curve based on the energy values in the abnormal state.

[0085] In an optional embodiment, constructing a reference curve based on the energy values in the normal state and constructing an abnormal state curve based on the energy values in the abnormal state includes:

[0086] Plot the energy values of each independent frequency band in the normal state in the form of a curve graph to obtain a reference curve;

[0087] Plot the energy values of each independent frequency band in various abnormal states in the form of a curve graph respectively to obtain multiple abnormal state curves.

[0088] Step S4: Calculate the difference between the abnormal state curve and the reference curve to obtain an abnormal state difference curve, and use an anomaly detection algorithm to extract the abnormal state from the abnormal state difference curve to obtain abnormal state characteristic parameters.

[0089] In an optional embodiment, calculating the difference between the abnormal state curve and the reference curve to obtain an abnormal state difference curve includes:

[0090] Calculate the energy difference between the abnormal state curve and the reference curve for each independent frequency band, and plot the energy differences for each independent frequency band in the form of a curve graph to obtain an abnormal state difference curve.

[0091] In an optional embodiment, using an anomaly detection algorithm to extract the abnormal state from the abnormal state difference curve to obtain abnormal state characteristic parameters includes:

[0092] Calculate the state support degree of the energy difference under each independent frequency band, and select a set of state conditions from each independent frequency band according to the state support degree;

[0093] Calculate the state dependence degree of each abnormal state on the set of state conditions;

[0094] Construct an abnormal state feature map based on the state support degree and the state dependence degree, and determine abnormal state feature parameters according to the abnormal state feature map.

[0095] It should be noted that the state support degree refers to the importance of this independent frequency band for each abnormal state. Select the first n important frequency bands as state conditions, which are used as the nodes for constructing the abnormal state feature map subsequently. The state dependence degree refers to the proportion of this state condition in the set of state conditions under this abnormal state, and this proportion is used as the state dependence degree.

[0096] In an optional embodiment, constructing an abnormal state feature map based on the state support degree and the state dependence degree, and determining abnormal state feature parameters according to the abnormal state feature map includes:

[0097] Take an independent frequency band as a node, take the state support degree as the node parameter, and take the absolute value of the difference between the state dependence degrees of two nodes as the connection parameter to complete the construction of the abnormal state feature map;

[0098] Select the node with the largest node parameter, put its node parameter into the queue, and then select the node with the smallest connection parameter from the nodes connected to the node with the largest node parameter, put its node parameter into the queue, until all nodes in the abnormal state feature map are traversed to obtain a queue to be calculated;

[0099] Calculate according to the queue order for the queue to be calculated to obtain abnormal state feature parameters.

[0100] Step S5: Conduct dynamic tests on the gas-liquid actuated valve to be diagnosed during the opening and closing process, obtain the dynamic signal to be diagnosed of the gas-liquid actuated valve to be diagnosed, and use the deep learning anomaly detection algorithm to extract the abnormal state from the difference curve between the dynamic signal to be diagnosed and the reference curve to obtain the abnormal state feature parameters to be diagnosed.

[0101] Step S6: Conduct fuzzy clustering analysis on the abnormal state feature parameters to be diagnosed and the abnormal state feature parameters to obtain the fault membership degree, and judge the health state of the gas-liquid actuated valve to be tested according to the fault membership degree.

[0102] Among them, for the dynamic test of the gas-liquid actuated valve to be tested, extract feature parameters from the difference curve between the dynamic signal of the test object and the reference state signal, conduct fuzzy clustering analysis on the obtained feature parameters of the test object and the abnormal feature parameters, and the fault clustering effect is asFigure 6 As shown, the fault membership degree is discriminated, the possible faults are classified and isolated, and the health status of the measured object is quickly judged.

[0103] Embodiment 2

[0104] Based on Embodiment 1, Embodiment 2 of the present invention provides a health status diagnosis system for a gas-liquid linked valve with multi-sensor information fusion, including:

[0105] A dynamic test module, which is used to install sensors based on the general reserved interface of the gas-liquid linked valve, perform dynamic tests on the gas-liquid linked valve during the opening and closing processes, and obtain the performance parameters of the gas-liquid linked valve;

[0106] A performance calculation module, which is used to install multiple types of sensors based on the general reserved interface of the gas-liquid linked valve, perform dynamic tests on the gas-liquid linked valve during the opening and closing processes through multiple types of sensors, and obtain performance parameters;

[0107] An energy value calculation module, which is used to perform wavelet energy calculation on the performance parameters to obtain the energy value of the gas-liquid linked valve in the normal state and the energy value in the abnormal state;

[0108] An abnormal feature module, which is used to calculate the difference between the abnormal state curve and the reference curve to obtain an abnormal state difference curve, and use an anomaly detection algorithm to extract the abnormal state from the abnormal state difference curve to obtain abnormal state characteristic parameters;

[0109] A module to be diagnosed, which is used to perform dynamic tests on the gas-liquid linked valve to be diagnosed during the opening and closing processes to obtain the dynamic signal to be diagnosed of the gas-liquid linked valve to be diagnosed, and use a deep learning anomaly detection algorithm to extract the abnormal state from the difference curve between the dynamic signal to be diagnosed and the reference curve to obtain the abnormal state characteristic parameters to be diagnosed;

[0110] A fault membership module, which is used to perform fuzzy clustering analysis on the abnormal state characteristic parameters to be diagnosed and the abnormal state characteristic parameters to obtain the fault membership degree, and judge the health status of the gas-liquid linked valve to be measured according to the fault membership degree.

[0111] In an optional embodiment, the performance calculation module includes:

[0112] An abnormal performance unit, which is used to divide the performance parameters into normal performance parameters in the normal state and abnormal performance parameters in various abnormal states;

[0113] A wavelet decomposition unit, wherein the wavelet decomposition unit is used to perform three-layer wavelet packet decomposition on the normal performance parameters and the abnormal performance parameters respectively based on the mutation characteristics by using the wavelet packet decomposition technology to obtain signal sequences of multiple independent frequency bands;

[0114] A state characteristic unit, the state characteristic unit is used to normalize the signal sequences of multiple independent frequency bands based on the energy change characteristics of each independent frequency band, obtain the frequency band energy value of each independent frequency band, and set the frequency band energy value as the state characteristic value;

[0115] The energy value calculation unit is used to extract the independent frequency band total value of each independent frequency band, calculate the state characteristic value of each independent frequency band and its corresponding independent frequency band total value, and obtain the energy value of each independent frequency band in a normal state and the energy value in each abnormal state.

[0116] Example 3

[0117] Figure 7 A schematic diagram of the structure of an electronic device provided in Embodiment 2 of the present invention, such as Figure 7 As shown, the electronic device includes a processor 21, a memory 22, an input device 23 and an output device 24; the number of processors 21 in the computer device can be one or more. Figure 7 A processor 21 is taken as an example; the processor 21, the memory 22, the input device 23 and the output device 24 in the electronic device can be connected by a bus or other means. Figure 7 The example of connecting through bus is taken in the following.

[0118] The memory 22 is a computer-readable storage medium that can be used to store software programs, computer executable programs, and modules. The processor 21 executes various functional applications and data processing of the electronic device by running the software programs, instructions, and modules stored in the memory 22, that is, the gas-liquid linkage valve health status diagnosis method based on multi-sensor information fusion of Example 1 is implemented.

[0119] The memory 22 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 22 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some instances, the memory 22 may further include a memory remotely arranged relative to the processor 21, and these remote memories may be connected to the electronic device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0120] The input device 23 can be used to receive the ID and password input by the user, etc. The output device 24 is used to output the network configuration page.

[0121] Example 4

[0122] Embodiment 4 of the present invention further provides a computer-readable storage medium, wherein the computer-executable instructions, when executed by a computer processor, are used to implement the method for diagnosing the health status of a gas-liquid linkage valve by multi-sensor information fusion as provided in Embodiment 1.

[0123] An embodiment of the present invention provides a storage medium containing computer executable instructions, and its computer executable instructions are not limited to the method operations provided in Example 1, and can also execute related operations in the gas-liquid linkage valve health status diagnosis method based on multi-sensor information fusion provided in any embodiment of the present invention.

[0124] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for diagnosing the health status of a gas-liquid actuated valve based on multi-sensor information fusion, characterized in that, It includes the following steps: Install multiple types of sensors based on the general reserved interface of the gas-liquid actuated valve, and dynamically test the opening and closing process of the gas-liquid actuated valve through the multiple types of sensors to obtain performance parameters; Perform wavelet energy calculation on the performance parameters to obtain the energy value of the gas-liquid actuated valve in the normal state and the energy value in the abnormal state; Construct a reference curve based on the energy value in the normal state and construct an abnormal state curve based on the energy value in the abnormal state; Perform difference calculation on the abnormal state curve and the reference curve to obtain an abnormal state difference curve, and use an anomaly detection algorithm to extract the abnormal state from the abnormal state difference curve to obtain abnormal state characteristic parameters; Perform dynamic testing on the gas-liquid actuated valve to be diagnosed during the opening and closing process to obtain the dynamic signal to be diagnosed of the gas-liquid actuated valve to be diagnosed, and use a deep learning anomaly detection algorithm to extract the abnormal state from the difference curve between the dynamic signal to be diagnosed and the reference curve to obtain the abnormal state characteristic parameters to be diagnosed; Perform fuzzy clustering analysis on the abnormal state characteristic parameters to be diagnosed and the abnormal state characteristic parameters to obtain a fault membership degree, and judge the health state of the gas-liquid actuated valve to be tested according to the fault membership degree.

2. The method for diagnosing the health state of the gas-liquid actuated valve with multi-sensor information fusion according to claim 1, characterized in that, Performing wavelet energy calculation on the performance parameters to obtain the energy value of the gas-liquid actuated valve in the normal state and the energy value in the abnormal state includes: Divide the performance parameters into normal performance parameters in the normal state and abnormal performance parameters in various abnormal states; Based on the mutation characteristics, use wavelet packet decomposition technology to perform three-layer wavelet packet decomposition on the normal performance parameters and abnormal performance parameters respectively to obtain signal sequences in multiple independent frequency bands; Normalize the signal sequences in multiple independent frequency bands based on the energy change characteristics of each independent frequency band to obtain the energy values of each independent frequency band, and set the energy values of the frequency bands as state characteristic values; Extract the total value of each independent frequency band, and calculate the energy value of each independent frequency band in the normal state and the energy value in each abnormal state by calculating the state characteristic value of each independent frequency band and its corresponding total value of the independent frequency band.

3. The method for diagnosing the health state of the gas-liquid actuated valve with multi-sensor information fusion according to claim 2, wherein Constructing a reference curve based on the energy value in the normal state and constructing an abnormal state curve based on the energy value in the abnormal state includes: Plot the energy values of each independent frequency band in the normal state in the form of a curve graph to obtain a reference curve; Plot the energy values of each independent frequency band in various abnormal states in the form of curve graphs respectively to obtain multiple abnormal state curves.

4. The method for diagnosing the health state of the gas-liquid actuated valve with multi-sensor information fusion according to claim 3, wherein, Performing difference calculation on the abnormal state curve and the reference curve to obtain an abnormal state difference curve includes: Calculate the energy difference between the abnormal state curve and the reference curve in each independent frequency band, and plot the energy differences in each independent frequency band in the form of a curve graph to obtain an abnormal state difference curve.

5. The method for diagnosing the health state of the gas-liquid actuated valve with multi-sensor information fusion according to claim 4, wherein, Using an anomaly detection algorithm to extract the abnormal state from the abnormal state difference curve to obtain abnormal state characteristic parameters includes: Calculate the state support degree of the energy difference in each independent frequency band, and select a state condition set from each independent frequency band according to the state support degree; Calculate the state dependence degree of each abnormal state on the state condition set; Construct an abnormal state feature map based on the state support degree and the state dependence degree, and determine abnormal state feature parameters according to the abnormal state feature map.

6. The method for diagnosing the health state of a gas-liquid actuated valve with multi-sensor information fusion according to claim 5, wherein, Constructing an abnormal state feature map based on the state support degree and the state dependence degree, and determining abnormal state feature parameters according to the abnormal state feature map includes: Taking an independent frequency band as a node, using the state support degree as the node parameter, and using the absolute value of the difference between the state dependence degrees of two nodes as the connection parameter to complete the construction of the abnormal state feature map; Select the node with the largest node parameter, put its node parameter into the queue, and then select the node with the smallest connection parameter from the nodes connected to the node with the largest node parameter and put its node parameter into the queue until all nodes in the abnormal state feature map are traversed to obtain a queue to be calculated; Calculate the queue to be calculated according to the queue order to obtain abnormal state feature parameters.

7. A health state diagnosis system for a gas-liquid actuated valve with multi-sensor information fusion, characterized in that, Including: A dynamic test module, which is used to install sensors based on the general reserved interface of the gas-liquid actuated valve, conduct dynamic tests on the opening and closing processes of the gas-liquid actuated valve, and obtain the performance parameters of the gas-liquid actuated valve; A performance calculation module, which is used to install multiple types of sensors based on the general reserved interface of the gas-liquid actuated valve, conduct dynamic tests on the opening and closing processes of the gas-liquid actuated valve through multiple types of sensors, and obtain performance parameters; An energy value calculation module, which is used to perform wavelet energy calculation on the performance parameters to obtain the energy value of the gas-liquid actuated valve in the normal state and the energy value in the abnormal state; An abnormal feature module, which is used to calculate the difference between the abnormal state curve and the reference curve to obtain an abnormal state difference curve, and use an anomaly detection algorithm to extract the abnormal state from the abnormal state difference curve to obtain abnormal state feature parameters; A module to be diagnosed, which is used to conduct dynamic tests on the gas-liquid actuated valve to be diagnosed during the opening and closing processes, obtain the dynamic signal to be diagnosed of the gas-liquid actuated valve to be diagnosed, and use a deep learning anomaly detection algorithm to extract the abnormal state from the difference curve between the dynamic signal to be diagnosed and the reference curve to obtain the abnormal state feature parameters to be diagnosed; A fault membership module, which is used to perform fuzzy clustering analysis on the abnormal state feature parameters to be diagnosed and the abnormal state feature parameters to obtain a fault membership degree, and judge the health state of the gas-liquid actuated valve to be tested according to the fault membership degree.

8. The multi-sensor information fusion-based health status diagnosis system for gas-liquid actuated valves according to claim 7, characterized in that, The performance calculation module includes: An abnormal performance unit, which is used to divide the performance parameters into normal performance parameters in the normal state and abnormal performance parameters in various abnormal states; A wavelet decomposition unit, which is used to perform three-layer wavelet packet decomposition on the normal performance parameters and abnormal performance parameters respectively based on the mutation characteristics using the wavelet packet decomposition technology to obtain signal sequences in multiple independent frequency bands; A state feature unit, which is used to normalize the signal sequences in multiple independent frequency bands based on the energy change characteristics of each independent frequency band to obtain the energy value of each independent frequency band, and set the energy value of the frequency band as the state feature value; The energy value calculation unit is used to extract the independent frequency band total value of each independent frequency band, calculate the state characteristic value of each independent frequency band and its corresponding independent frequency band total value, and obtain the energy value of each independent frequency band in a normal state and the energy value in each abnormal state.

9. An electronic device, characterized in that, The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for diagnosing the health status of a gas-liquid linkage valve by multi-sensor information fusion as described in claims 1 to 6 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, the method for diagnosing the health status of a gas-liquid linkage valve by multi-sensor information fusion as described in claims 1 to 6 is implemented.

Images