CN118817189A

CN118817189A - Valve internal leakage detection early warning method and system

Info

Publication number: CN118817189A
Application number: CN202411305516.7A
Authority: CN
Inventors: 季刚
Original assignee: Nantong Jinyun Fluid Equipment Co ltd
Current assignee: Nantong Jinyun Fluid Equipment Co ltd
Priority date: 2024-09-19
Filing date: 2024-09-19
Publication date: 2024-10-22

Abstract

The application provides a valve internal leakage detection early warning method and a system, which relate to the technical field of valve internal leakage detection, and the method comprises the following steps: detecting sound wave signals of the valve to obtain a detected sound wave signal set; performing filtering reconstruction to obtain a reconstructed sound wave signal set; detecting the fluid characteristics of the valve, and determining a detected fluid characteristic data set; loading a valve flaw detection data set; based on a valve internal leakage joint detection algorithm, performing valve internal leakage characteristic analysis according to a reconstructed sound wave signal set, a detected fluid characteristic data set and a valve flaw detection data set, and determining a valve internal leakage generation coefficient; if the valve internal leakage generating coefficient is larger than/equal to the valve internal leakage generating threshold value, generating a valve internal leakage alarming instruction; and carrying out internal leakage tracing detection on the valve to generate an internal leakage tracing early warning signal of the valve. The application can solve the technical problem of lower accuracy of the valve internal leakage detection in the prior art, and achieves the technical effect of improving the accuracy of the valve internal leakage detection.

Description

Valve internal leakage detection early warning method and system

Technical Field

The application relates to the technical field of valve internal leakage detection, in particular to a valve internal leakage detection early warning method and system.

Background

The detection of the internal leakage of the valve refers to detecting the flowing condition of the internal fluid medium after the valve is closed so as to judge whether the valve has the leakage phenomenon caused by incomplete closing.

At present, the existing internal leakage detection of the valve is mostly carried out by only depending on single detection data or characteristic data, so that errors such as abnormality of detection equipment can influence an internal leakage detection result, the internal leakage detection is limited by a single method, and the accuracy and timeliness of the internal leakage detection of the valve are reduced. Accordingly, there is a need for a method to solve the above-mentioned problems.

In summary, in the prior art, since the internal leakage detection is mostly performed only by a single method, there may be a limitation of the single method in the internal leakage detection of the valve, and the accuracy of the internal leakage detection of the valve is further affected.

Disclosure of Invention

The application aims to provide a valve internal leakage detection early warning method and a valve internal leakage detection early warning system, which are used for solving the technical problem that in the prior art, the internal leakage detection of a valve possibly has the limitation of a single method because the internal leakage detection is mostly carried out by only depending on the single method, and the accuracy of the valve internal leakage detection is further influenced.

In view of the above problems, the application provides a valve internal leakage detection early warning method and a valve internal leakage detection early warning system.

In a first aspect, the present application provides a valve internal leakage detection early warning method, which is implemented by a valve internal leakage detection early warning system, where the method includes: detecting the sound wave signals of the valve according to the internal leakage detection end of the valve to obtain a detected sound wave signal set; performing filtering reconstruction according to the detected sound wave signal set to obtain a reconstructed sound wave signal set; detecting the fluid characteristics of the valve according to a fluid detection end in the valve, and determining a detected fluid characteristic data set; loading a valve flaw detection data set; based on a valve internal leakage joint detection algorithm, performing valve internal leakage characteristic analysis according to the reconstructed sound wave signal set, the detected fluid characteristic data set and the valve flaw detection data set, and determining a valve internal leakage generation coefficient; if the valve internal leakage generating coefficient is larger than or equal to the valve internal leakage generating threshold value, generating a valve internal leakage alarming instruction; and carrying out internal leakage tracing detection on the valve according to the valve internal leakage warning instruction to generate a valve internal leakage tracing early warning signal.

In a second aspect, the present application further provides a valve internal leakage detection early warning system, configured to perform the valve internal leakage detection early warning method according to the first aspect, where the system includes: the detection sound wave signal set acquisition module is used for detecting sound wave signals of the valve according to the internal leakage detection end of the valve to acquire a detection sound wave signal set; the reconstruction sound wave signal set obtaining module is used for carrying out filtering reconstruction according to the detection sound wave signal set to obtain a reconstruction sound wave signal set; the detection fluid characteristic data set determining module is used for detecting the fluid characteristic of the valve according to the fluid detection end in the valve and determining a detection fluid characteristic data set; the valve flaw detection data set loading module is used for loading a valve flaw detection data set; the valve internal leakage generation coefficient determining module is used for carrying out valve internal leakage characteristic analysis according to the reconstructed sound wave signal set, the detected fluid characteristic data set and the valve flaw detection data set based on a valve internal leakage joint detection algorithm to determine a valve internal leakage generation coefficient; the valve internal leakage alarm instruction generation module is used for generating a valve internal leakage alarm instruction if the valve internal leakage generation coefficient is greater than/equal to a valve internal leakage generation threshold value; and the valve internal leakage tracing early-warning signal generation module is used for carrying out internal leakage tracing detection on the valve according to the valve internal leakage warning instruction to generate a valve internal leakage tracing early-warning signal.

One or more technical schemes provided by the application have at least the following technical effects or advantages:

The method comprises the steps that a detection sound wave signal set is obtained by detecting sound wave signals of a valve according to a valve inner leakage detection end; performing filtering reconstruction according to the detected sound wave signal set to obtain a reconstructed sound wave signal set; detecting the fluid characteristics of the valve according to a fluid detection end in the valve, and determining a detected fluid characteristic data set; loading a valve flaw detection data set; based on a valve internal leakage joint detection algorithm, performing valve internal leakage characteristic analysis according to the reconstructed sound wave signal set, the detected fluid characteristic data set and the valve flaw detection data set, and determining a valve internal leakage generation coefficient; if the valve internal leakage generating coefficient is larger than or equal to the valve internal leakage generating threshold value, generating a valve internal leakage alarming instruction; and carrying out internal leakage tracing detection on the valve according to the valve internal leakage warning instruction to generate a valve internal leakage tracing early warning signal, namely, mutually supplementing and verifying through a plurality of valve internal leakage detection methods, so that the technical aim of improving the accuracy and reliability of the valve internal leakage detection is finally realized, and the technical effect of comprehensively knowing the working state and the performance change of the valve is achieved.

The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent. It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the application or to delineate the scope of the application. Other features of the present application will become apparent from the description that follows.

Drawings

In order to more clearly illustrate the application or the technical solutions of the prior art, the following brief description will be given of the drawings used in the description of the embodiments or the prior art, it being obvious that the drawings in the description below are only exemplary and that other drawings can be obtained from the drawings provided without the inventive effort for a person skilled in the art.

FIG. 1 is a schematic flow chart of a valve internal leakage detection and early warning method of the application;

fig. 2 is a schematic structural diagram of a valve internal leakage detection early warning system according to the present application.

Reference numerals illustrate:

the system comprises a sound wave detection signal set obtaining module 11, a sound wave reconstruction signal set obtaining module 12, a fluid detection characteristic data set determining module 13, a valve flaw detection data set loading module 14, a valve internal leakage generation coefficient determining module 15, a valve internal leakage alarm instruction generating module 16 and a valve internal leakage tracing early-warning signal generating module 17.

Detailed Description

The application provides a valve internal leakage detection early warning method and a valve internal leakage detection early warning system, which solve the technical problem that the accuracy of the valve internal leakage detection is further affected due to the fact that the valve internal leakage detection is limited by a single method in the prior art. The technical aim of improving the accuracy and reliability of the internal leakage detection of the valve is fulfilled, and the technical effect of comprehensively knowing the working state and the performance change of the valve is achieved.

In the following, the technical solutions of the present application will be clearly and completely described with reference to the accompanying drawings, and it should be understood that the described embodiments are only some embodiments of the present application, but not all embodiments of the present application, and that the present application is not limited by the exemplary embodiments described herein. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application. It should be further noted that, for convenience of description, only some, but not all of the drawings related to the present application are shown.

Example 1

Referring to fig. 1, the application provides a valve internal leakage detection early warning method, wherein the method is applied to a valve internal leakage detection early warning system, and the method specifically comprises the following steps:

step one: detecting the sound wave signals of the valve according to the internal leakage detection end of the valve to obtain a detected sound wave signal set;

Specifically, the valve internal leakage detector is a device for detecting leakage inside the valve. The leak detector in the valve generates turbulence when gas leaks, so that sound waves with specific frequencies are generated. And installing a valve inner leakage detector at the valve inner leakage detection end, and detecting the sound wave signals of the valve to obtain a detection sound wave signal set.

Step two: performing filtering reconstruction according to the detected sound wave signal set to obtain a reconstructed sound wave signal set;

Specifically, the complexity of the sound source is identified by the sound source detection signal set, the noise reduction processing is carried out on the high-complexity signals according to the complexity classification, and the reconstructed sound wave signal set is obtained by integrating all the signals, so that the quality and the signal-to-noise ratio of the sound wave signals are improved, and a better data base is provided for subsequent signal analysis or application.

Step three: detecting the fluid characteristics of the valve according to a fluid detection end in the valve, and determining a detected fluid characteristic data set;

Specifically, a sensor and other devices are arranged at the fluid detection end in the valve, and the fluid characteristic detection is carried out on the fluid detection end in the valve, such as the flow rate, the pressure and the like of the fluid, so that a detected fluid characteristic data set is determined.

Step four: loading a valve flaw detection data set;

In particular, valve inspection is used to detect internal or external defects of a valve, which may include cracks, corrosion, pinholes, inclusions, and the like. Through the valve flaw detection, the potential safety hazard of the valve can be found in time, and the normal operation and the safety of the valve are ensured. For example, a valve flaw detection data set is obtained by flaw detection of a valve by a valve flaw detection device.

Step five: based on a valve internal leakage joint detection algorithm, performing valve internal leakage characteristic analysis according to the reconstructed sound wave signal set, the detected fluid characteristic data set and the valve flaw detection data set, and determining a valve internal leakage generation coefficient;

Specifically, the reconstructed acoustic wave signal set, the detected fluid characteristic data set and the valve flaw detection data set are integrated to form a comprehensive data set for subsequent valve internal leakage characteristic analysis. Key features associated with valve leaks are extracted from the integrated dataset. The integrated data set may include the frequency, amplitude and waveform characteristics of the acoustic signal, flow changes in the fluid characteristics, pressure fluctuations, etc., as well as abnormal signals or damage conditions detected in the valve detection. And analyzing the extracted characteristics by utilizing a valve internal leakage joint detection algorithm, and identifying a mode or a rule related to the valve internal leakage. And outputting a valve internal leakage generation coefficient based on the result of the feature analysis.

Step six: if the valve internal leakage generating coefficient is larger than or equal to the valve internal leakage generating threshold value, generating a valve internal leakage alarming instruction;

Specifically, the calculated valve internal leakage generation coefficient is compared with a valve internal leakage generation threshold value. The valve internal leakage generation threshold is set according to actual operation experience, safety standard or related regulations and is used for judging whether the valve has internal leakage risk or not. If the valve internal leakage generation coefficient is larger than or equal to the valve internal leakage generation threshold value, indicating that the valve has an internal leakage risk, and generating a valve internal leakage alarm instruction.

Step seven: and carrying out internal leakage tracing detection on the valve according to the valve internal leakage warning instruction to generate a valve internal leakage tracing early warning signal.

Specifically, the valve is subjected to internal leakage tracing detection according to an internal leakage warning instruction of the valve. The tracing detection comprises pressurizing trace gas inside the detected valve, and detecting leaked trace gas outside, so that the leakage position and the leakage amount can be detected. Based on the result of the tracer detection, a valve internal leakage tracer early warning signal is generated, which contains specific information about the internal leakage, such as position, degree or potential risk, for taking further measures, such as emergency maintenance, valve replacement or operating parameter adjustment, etc.

The valve internal leakage detection early warning method is applied to a valve internal leakage detection early warning system, can achieve the technical aim of improving the accuracy and reliability of valve internal leakage detection, and achieves the technical effect of comprehensively knowing the working state and performance change of the valve.

Further, the application also comprises the following steps:

Performing sound source complexity identification according to each detected sound wave signal in the detected sound wave signal set, and determining the complexity of a plurality of signal sound sources;

judging whether the complexity of the plurality of signal sound sources is smaller than a sound source complexity threshold value or not, and obtaining a plurality of sound source complexity judgment results;

Classifying the sound wave detection signal sets according to the sound source complexity judgment results to obtain a first sound wave detection signal set and a second sound wave detection signal set;

performing adaptive filtering and noise reduction according to the second detection sound wave signal group to obtain a second noise reduction detection sound wave signal group;

and integrating the first detection sound wave signal group and the second noise reduction detection sound wave signal group according to the sound wave signal generation time sequence characteristics to obtain the reconstruction sound wave signal set.

Specifically, the identification of the complexity of the sound source is performed based on each of the detected sound signals within the set of detected sound signals. The sound source complexity may refer to detecting the number and degree of variation of different frequency, amplitude or phase components contained in the acoustic signal, thereby determining the respective sound source complexity of the plurality of signals.

Then, it is determined whether the plurality of signal sound source complexity is less than a sound source complexity threshold. For example, the sound source complexity threshold is set based on experience or a specific application scenario to distinguish whether the signal is so complex that further noise reduction processing is required. And obtaining complexity judgment results of a plurality of sound sources through judgment. The plurality of sound source complexity determination results include a plurality of signal sound source complexity results that are less than a sound source complexity threshold and a plurality of signal sound source complexity results that are greater than or equal to the sound source complexity threshold.

And classifying the sound source detection signal sets according to the sound source complexity judgment results, classifying the results with the signal sound source complexity smaller than the sound source complexity threshold as a first sound source detection signal set, and classifying the results with the signal sound source complexity larger than or equal to the sound source complexity threshold as a second sound source detection signal set.

And then, carrying out self-adaptive filtering noise reduction processing on the second detection sound wave signal group, namely the signal group with higher sound source complexity, and dynamically adjusting filter parameters according to signal characteristics to remove noise so as to obtain a second noise reduction detection sound wave signal group, namely a signal subjected to noise reduction processing.

The timing characteristics of the generation of the acoustic wave signals then refer to the relative time relationship or time sequence between the signals. And integrating the first detection sound wave signal group and the second noise reduction detection sound wave signal group together according to the generation time sequence characteristics of the sound wave signals, namely integrating the non-noise reduction signals and the noise reduction signals to obtain a reconstructed sound wave signal set.

By identifying the complexity of the sound source, carrying out noise reduction processing on the high-complexity signals according to complexity classification, integrating all the signals to obtain a reconstructed sound wave signal set, improving the quality and the signal-to-noise ratio of the sound wave signals, and providing a better data base for subsequent signal analysis or application.

Further, the application also comprises the following steps:

Performing interference sound wave signal characteristic identification according to the second detection sound wave signal group to obtain an interference sound wave signal characteristic data set;

clustering the second detection sound wave signal group according to the interference sound wave signal characteristic data set to obtain a plurality of sound wave signal areas;

activating an adaptive filter, wherein the adaptive filter filters and reduces noise of the plurality of sound wave signal areas based on the interference sound wave signal characteristic data set to obtain a noise reduction sound wave signal set;

carrying out interference signal influence identification according to the noise reduction sound wave signal set to obtain a signal interference influence coefficient;

When the signal interference influence coefficient is smaller than a signal interference influence threshold, adding the noise reduction sound wave signal set to the second noise reduction detection sound wave signal set;

and when the signal interference influence coefficient is larger than/equal to the signal interference influence threshold, performing depth filtering on the noise reduction sound wave signal set according to the adaptive filter to generate the second noise reduction detection sound wave signal set.

Specifically, the second sound wave detection signal group is subjected to feature recognition of the sound wave interference signals, and features of interference components possibly existing in the signals, such as frequency, amplitude, phase and the like, are extracted, so that an interference sound wave signal feature data set is formed.

And then, clustering the second detection sound wave signal group according to the interference sound wave signal characteristic data set, grouping similar second detection sound wave signals together, and further obtaining a plurality of sound wave signal areas, which is beneficial to classifying signals with the same or similar interference characteristics and provides convenience for subsequent processing.

Then, an adaptive filter is activated, filtering and noise reduction processing is carried out on the plurality of sound wave signal areas based on the interference sound wave signal characteristic data set, and a noise reduction sound wave signal set is obtained. The filtering parameters can be automatically adjusted according to the characteristics of the signals through the self-adaptive filter, so that the optimal noise reduction effect is achieved.

And then, after filtering and noise reduction, carrying out interference signal influence identification on the noise reduction sound wave signal set, and evaluating the influence degree of interference components in the noise reduction processed signal to obtain a signal interference influence coefficient. The degree of influence is quantified, for example, by calculating a signal-to-interference influence coefficient.

And then, comparing the signal interference influence coefficient with a signal interference influence threshold value to judge whether the noise reduction effect meets the requirement. If the signal interference influence coefficient is smaller than the signal interference influence threshold value, the noise reduction effect is good, and the noise reduction sound wave signal set can be directly added into the second noise reduction detection sound wave signal set. If the signal interference influence coefficient is greater than or equal to the signal interference influence threshold, the interference still exists or the noise reduction effect is not ideal, and at the moment, the noise reduction sound wave signal set needs to be subjected to deep filtering processing according to the adaptive filter so as to obtain better noise reduction effect, and a final second noise reduction detection sound wave signal set is generated.

The signals in the second detection sound wave signal group are subjected to finer interference identification and processing, so that cleaner and more accurate sound wave signals are obtained and used for subsequent signal analysis or feature extraction.

Further, the application also comprises the following steps:

Predicting the valve internal leakage probability according to the valve flaw detection data set to obtain a first valve internal leakage probability coefficient;

Predicting the valve internal leakage probability according to the reconstructed sound wave signal set to obtain a second valve internal leakage probability coefficient;

Performing valve internal leakage probability prediction according to the detected fluid characteristic data set to obtain a third valve internal leakage probability coefficient;

Based on the expected valve internal leakage probability coefficient, respectively carrying out difference calculation on the first valve internal leakage probability coefficient, the second valve internal leakage probability coefficient and the third valve internal leakage probability coefficient to obtain a first expected difference internal leakage probability coefficient, a second expected difference internal leakage probability coefficient and a third expected difference internal leakage probability coefficient;

Performing duty ratio calculation according to the first expected difference inner leakage probability coefficient, the second expected difference inner leakage probability coefficient and the third expected difference inner leakage probability coefficient to generate a first inner leakage probability gain coefficient, a second inner leakage probability gain coefficient and a third inner leakage probability gain coefficient;

And performing gain optimization on the first valve internal leakage probability coefficient, the second valve internal leakage probability coefficient and the third valve internal leakage probability coefficient according to the first internal leakage probability gain coefficient, the second internal leakage probability gain coefficient and the third internal leakage probability gain coefficient to generate the valve internal leakage generation coefficient.

Specifically, a flaw detection feature internal leakage probability identification model is constructed, a valve flaw detection data set is processed, the probability of valve internal leakage is predicted, and a first valve internal leakage probability coefficient is obtained.

And then, constructing an acoustic wave characteristic internal leakage probability identification model, processing the reconstructed acoustic wave signal set, predicting the probability of internal leakage of the valve, and obtaining a second valve internal leakage probability coefficient.

And then, constructing a fluid characteristic internal leakage probability identification model, processing a detected fluid characteristic data set, predicting the probability of internal leakage of the valve, and obtaining a third valve internal leakage probability coefficient.

Next, it is expected that the valve leak probability coefficient is infinitely close to 0. And respectively calculating the difference between the first valve internal leakage probability coefficient, the second valve internal leakage probability coefficient and the third valve internal leakage probability coefficient and the expected value based on the expected valve internal leakage probability coefficient. The difference values are referred to as a first desired intra-difference leakage probability coefficient, a second desired intra-difference leakage probability coefficient, and a third desired intra-difference leakage probability coefficient, respectively.

And then, according to the magnitudes of the first expected difference internal leakage probability coefficient, the second expected difference internal leakage probability coefficient and the third expected difference internal leakage probability coefficient, calculating the duty ratio of each in the total difference, and reflecting the deviation degree of different prediction methods relative to the expected valve internal leakage probability coefficient so as to determine the gain weight of each internal leakage probability prediction. The calculated duty cycle is a first inner leakage probability gain coefficient, a second inner leakage probability gain coefficient, and a third inner leakage probability gain coefficient.

And then, weighting and calculating the original first valve internal leakage probability coefficient, the second valve internal leakage probability coefficient and the third valve internal leakage probability coefficient by using the calculated first internal leakage probability gain coefficient, second internal leakage probability gain coefficient and third internal leakage probability gain coefficient, and fusing the results of different prediction methods to obtain a more accurate and reliable valve internal leakage generation coefficient as the valve internal leakage generation coefficient.

The optimal valve internal leakage generation coefficient is finally obtained through the difference between the results and expected values of various prediction methods, the internal leakage condition of the valve can be reflected more comprehensively, and powerful support is provided for safe operation and maintenance of the valve.

Further, the application also comprises the following steps:

loading a sample valve flaw detection data record and a sample flaw detection characteristic internal leakage probability identification coefficient record;

constructing a flaw detection characteristic internal leakage probability identification model meeting convergence conditions according to the sample valve flaw detection data record and the sample flaw detection characteristic internal leakage probability identification coefficient record;

Inputting the valve flaw detection data set into the flaw detection feature internal leakage probability identification model to obtain flaw detection feature internal leakage probability identification coefficients;

Acquiring state data of flaw detection equipment corresponding to the valve flaw detection data set;

Performing abnormality detection according to the state data of the flaw detection equipment to obtain state abnormality characteristic data of the flaw detection equipment;

Carrying out valve internal leakage probability interference identification according to the state abnormal characteristic data of the flaw detection equipment to obtain an abnormal internal leakage probability interference coefficient of the flaw detection equipment;

Correcting the flaw detection characteristic internal leakage probability identification coefficient according to the flaw detection equipment abnormal internal leakage probability interference coefficient, and generating the first valve internal leakage probability coefficient.

Specifically, a sample valve flaw detection data record and a sample flaw detection characteristic endoleak probability identification coefficient record are read from a storage medium and used for constructing a flaw detection characteristic endoleak probability identification model. For example, the sample valve inspection data record is a historical time valve inspection data record, and the sample flaw detection feature leak probability identification coefficient record is a historical time flaw detection feature leak probability identification coefficient record. The sample valve inspection test data record may comprise a signal or image or measurement of valve inspection, and the sample inspection feature endoleak probability identification coefficient record may comprise an endoleak probability identification coefficient corresponding to the sample valve inspection test data record.

And then, training a flaw detection characteristic endoleak probability identification model by machine learning by using the sample valve flaw detection data record and the sample flaw detection characteristic endoleak probability identification coefficient record, and predicting the endoleak probability of the valve according to the flaw detection characteristic. In the training process, parameters of the flaw detection feature internal leakage probability recognition model are continuously adjusted to meet convergence conditions, namely when the prediction result of the flaw detection feature internal leakage probability recognition model is gradually stable and the error reaches an acceptable range, training of the flaw detection feature internal leakage probability recognition model is completed. The convergence condition may be that the prediction result of the flaw detection feature internal leakage probability identification model is not changed significantly any more.

And after the training of the flaw detection feature internal leakage probability recognition model is finished, inputting a valve flaw detection data set, namely valve flaw detection data containing a series of unknown internal leakage probabilities, into the flaw detection feature internal leakage probability recognition model, predicting each valve flaw detection data through the flaw detection feature internal leakage probability recognition model, and outputting a corresponding flaw detection feature internal leakage probability recognition coefficient.

Next, flaw detection device state data corresponding to the valve flaw detection data set is acquired. The flaw detection device status data may include operating parameters, temperature, pressure, voltage, etc. of the device to reflect the status and performance of the flaw detection device during operation.

Then, through analysis of the state data of the flaw detection equipment, the possible abnormal state of the equipment is identified, and the state abnormal characteristic data of the flaw detection equipment is obtained. For example by comparison with normal state data.

Then, the fault detection data may be error-caused due to the fault detection equipment state abnormal characteristic data, thereby influencing the prediction of the internal leakage probability. And then, according to the detected abnormal characteristic data of the state of the flaw detection equipment, evaluating the interference degree of the abnormal data on the valve internal leakage probability identification result to obtain an abnormal internal leakage probability interference coefficient of the flaw detection equipment.

And then, based on the influence of equipment abnormality on internal leakage probability identification, correcting an original flaw detection characteristic internal leakage probability identification coefficient by using a flaw detection equipment abnormality internal leakage probability interference coefficient, so as to eliminate or reduce the influence of equipment state abnormality on a prediction result, and generating a first valve internal leakage probability coefficient.

By combining flaw detection data, equipment state data and machine learning technology, the accuracy and reliability of the identification of the internal leakage probability of the valve are improved, and the method has important significance for safety monitoring and fault prevention of the valve.

Further, the application also comprises the following steps:

building an acoustic wave characteristic internal leakage probability identification model;

Inputting the reconstructed sound wave signal set into the sound wave feature endoleak probability identification model to obtain a sound wave feature endoleak probability identification coefficient;

acquiring detection end state data corresponding to the reconstructed sound wave signal set;

and correcting the sound wave characteristic internal leakage probability identification coefficient according to the state data of the detection end to obtain the second valve internal leakage probability coefficient.

Specifically, a plurality of valve acoustic detection data records are collected, including acoustic characteristics of different valve states, namely internal leakage and normal, for training and verification of an acoustic characteristic internal leakage probability recognition model. And acquiring a corresponding sound wave characteristic endoleak probability identification coefficient record. A deep learning network may be used to construct an acoustic wave feature endoleak probability recognition model. And training an acoustic feature internal leakage probability recognition model by using a plurality of valve acoustic detection data records and acoustic feature internal leakage probability recognition coefficient records, and adjusting parameters of the acoustic feature internal leakage probability recognition model through an optimization algorithm to enable the acoustic feature internal leakage probability recognition model to accurately recognize acoustic features of internal leakage of the valve. And evaluating the performance of the acoustic wave characteristic endoleak probability recognition model by a cross verification method and the like, so as to ensure that the acoustic wave characteristic endoleak probability recognition model has good generalization capability.

After the training of the acoustic wave feature internal leakage probability recognition model is finished, a reconstructed acoustic wave signal set, namely valve acoustic wave detection data containing a plurality of unknown internal leakage probabilities, is input into the acoustic wave feature internal leakage probability recognition model, each reconstructed acoustic wave signal is predicted through the acoustic wave feature internal leakage probability recognition model, and corresponding acoustic wave feature internal leakage probability recognition coefficients are output.

Then, acquiring sound wave detection equipment state data corresponding to the reconstructed sound wave signal set, namely detecting end state data corresponding to the reconstructed sound wave signal set. The acoustic wave inspection apparatus status data may include operating parameters of the apparatus, temperature, pressure, voltage, etc. to reflect the status and performance of the inspection apparatus during operation.

And then, identifying the possible abnormal state of the equipment by analyzing the detection end state data corresponding to the reconstructed sound wave signal set, and obtaining the detection end state abnormal characteristic data corresponding to the reconstructed sound wave signal set. For example by comparison with normal state data. The error of the acoustic wave data may be caused by reconstructing abnormal characteristic data of the state of the detection end corresponding to the acoustic wave signal set, so that the prediction of the internal leakage probability is affected. And then, according to the detected abnormal characteristic data of the detection end state corresponding to the reconstructed sound wave signal set, evaluating the interference degree of the abnormal data on the valve internal leakage probability identification result to obtain the abnormal internal leakage probability interference coefficient of the sound wave equipment. Based on the influence of equipment abnormality on internal leakage probability identification, the original acoustic wave characteristic internal leakage probability identification coefficient is corrected by using the acoustic wave equipment abnormality internal leakage probability interference coefficient, so that the influence of equipment state abnormality on a prediction result is eliminated or reduced, and a second valve internal leakage probability coefficient is generated.

The accuracy of identification is improved by setting up a valve internal leakage probability identification model based on sound wave characteristics and correcting state data of a detection end, so that the real-time monitoring and fault early warning of the valve state are realized, and the safety and reliability of valve operation are improved.

Further, the application also comprises the following steps:

Constructing a fluid characteristic internal leakage probability identification model;

Inputting the detected fluid characteristic data set into the fluid characteristic endoleak probability identification model to obtain a fluid characteristic endoleak probability identification coefficient;

Acquiring detection end state data corresponding to the detection fluid characteristic data set;

And correcting the fluid characteristic internal leakage probability identification coefficient according to the detection end state data to obtain the third valve internal leakage probability coefficient.

Specifically, a plurality of valve fluid detection data records are collected. And acquiring a corresponding fluid characteristic internal leakage probability identification coefficient record. A deep learning network may be used to construct a fluid feature leak probability recognition model. And training a fluid characteristic internal leakage probability identification model by using a plurality of valve fluid detection data records and fluid characteristic internal leakage probability identification coefficient records, and adjusting parameters of the fluid characteristic internal leakage probability identification model through an optimization algorithm, so that the fluid characteristic internal leakage probability identification model can accurately identify the fluid characteristic of internal leakage of the valve. And the performance of the fluid characteristic internal leakage probability identification model is evaluated through methods such as cross verification and the like, so that the fluid characteristic internal leakage probability identification model is ensured to have good generalization capability.

And after the fluid characteristic endoleak probability recognition model is trained, inputting a detected fluid characteristic data set, namely valve fluid detection data containing a plurality of unknown endoleak probabilities, into the fluid characteristic endoleak probability recognition model, predicting each detected fluid characteristic data through the fluid characteristic endoleak probability recognition model, and outputting a corresponding fluid characteristic endoleak probability recognition coefficient.

Then, the detection end state data corresponding to the detection fluid characteristic data set, namely the detection end state data corresponding to the detection fluid characteristic data set, is acquired. The fluid detection apparatus status data may include operating parameters of the apparatus, temperature, pressure, voltage, etc. for reflecting the status and performance of the fluid detection apparatus during operation.

Then, through analyzing the detection end state data corresponding to the detection fluid characteristic data set, the possible abnormal state of the equipment is identified, and the detection end state abnormal characteristic data corresponding to the detection fluid characteristic data set is obtained. For example by comparison with normal state data. The abnormal characteristic data of the state of the detection end corresponding to the detected fluid characteristic data set may cause errors of the fluid data, so that the prediction of the internal leakage probability is affected. And then, according to the detected abnormal characteristic data of the detection end state corresponding to the detected fluid characteristic data set, evaluating the interference degree of the abnormal data on the valve internal leakage probability identification result to obtain an abnormal internal leakage probability interference coefficient of the fluid detection equipment. Based on the influence of equipment abnormality on internal leakage probability identification, correcting an original fluid characteristic internal leakage probability identification coefficient by using an abnormal internal leakage probability interference coefficient of fluid detection equipment, so as to eliminate or reduce the influence of equipment state abnormality on a prediction result and generate a third valve internal leakage probability coefficient.

The accuracy of identification is improved by constructing a valve internal leakage probability identification model based on fluid characteristics and correcting state data of a detection end, so that the real-time monitoring and fault early warning of the valve state are realized, and the safety and reliability of valve operation are improved.

In summary, the valve internal leakage detection early warning method provided by the application has the following technical effects:

Example two

Based on the same inventive concept as the valve internal leakage detection and early warning method in the foregoing embodiment, the application also provides a valve internal leakage detection and early warning system, referring to fig. 2, the system includes:

The detection sound wave signal set obtaining module 11 is used for detecting sound wave signals of the valve according to the internal leakage detection end of the valve to obtain a detection sound wave signal set;

the reconstruction sound wave signal set obtaining module 12 is used for carrying out filtering reconstruction according to the detection sound wave signal set to obtain a reconstruction sound wave signal set;

the detection fluid characteristic data set determining module 13 is used for detecting the fluid characteristic of the valve according to the fluid detection end in the valve, and determining a detection fluid characteristic data set;

the valve flaw detection data set loading module 14, wherein the valve flaw detection data set loading module 14 is used for loading a valve flaw detection data set;

The valve internal leakage generation coefficient determining module 15 is used for performing valve internal leakage characteristic analysis according to the reconstructed sound wave signal set, the detected fluid characteristic data set and the valve flaw detection data set based on a valve internal leakage joint detection algorithm to determine a valve internal leakage generation coefficient;

The valve internal leakage alarm instruction generating module 16, wherein the valve internal leakage alarm instruction generating module 16 is configured to generate a valve internal leakage alarm instruction if the valve internal leakage generating coefficient is greater than or equal to a valve internal leakage generating threshold;

And the valve internal leakage tracing early-warning signal generating module 17 is used for carrying out internal leakage tracing detection on the valve according to the valve internal leakage warning instruction to generate a valve internal leakage tracing early-warning signal.

Further, the reconstructed acoustic signal set obtaining module 12 in the system is further configured to:

Further, the valve internal leakage generation coefficient determining module 15 in the system is further configured to:

In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and a valve internal leakage detection early warning method and a specific example in the first embodiment are also applicable to a valve internal leakage detection early warning system of the present embodiment, through the foregoing detailed description of a valve internal leakage detection early warning method, those skilled in the art can clearly know a valve internal leakage detection early warning system in this embodiment, so for brevity of description, the detailed description will not be repeated here. For the system disclosed in the embodiment, since the system corresponds to the method disclosed in the embodiment, the description is simpler, and the relevant points refer to the description of the method section.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the present application and the equivalent techniques thereof, the present application is also intended to include such modifications and variations.

Claims

1. The method for detecting and early warning the internal leakage of the valve is characterized by comprising the following steps:

detecting the sound wave signals of the valve according to the internal leakage detection end of the valve to obtain a detected sound wave signal set;

performing filtering reconstruction according to the detected sound wave signal set to obtain a reconstructed sound wave signal set;

Detecting the fluid characteristics of the valve according to a fluid detection end in the valve, and determining a detected fluid characteristic data set;

loading a valve flaw detection data set;

based on a valve internal leakage joint detection algorithm, performing valve internal leakage characteristic analysis according to the reconstructed sound wave signal set, the detected fluid characteristic data set and the valve flaw detection data set, and determining a valve internal leakage generation coefficient;

if the valve internal leakage generating coefficient is larger than or equal to the valve internal leakage generating threshold value, generating a valve internal leakage alarming instruction;

performing internal leakage tracing detection on the valve according to the valve internal leakage warning instruction to generate a valve internal leakage tracing early warning signal;

Based on a valve internal leakage joint detection algorithm, performing valve internal leakage feature analysis according to the reconstructed acoustic wave signal set, the detected fluid feature data set and the valve flaw detection data set, determining a valve internal leakage generation coefficient, including:

2. The method of claim 1, wherein performing filtering reconstruction from the set of detected acoustic signals to obtain a set of reconstructed acoustic signals comprises:

3. The method of claim 2, wherein adaptively filtering and denoising the second set of detected acoustic signals to obtain a second set of denoised detected acoustic signals, comprising:

4. The method of claim 1, wherein performing valve endoleak probability prediction from the valve inspection data set to obtain a first valve endoleak probability coefficient comprises:

5. The method of claim 1, wherein performing valve leak probability prediction from the reconstructed acoustic signal set to obtain a second valve leak probability coefficient comprises:

6. The method of claim 1, wherein performing valve endoleak probability prediction from the detected fluid signature dataset to obtain a third valve endoleak probability coefficient comprises:

7. A valve leak detection warning system for performing the steps of the method of any one of claims 1 to 6, the system comprising:

The detection sound wave signal set acquisition module is used for detecting sound wave signals of the valve according to the internal leakage detection end of the valve to acquire a detection sound wave signal set;

the reconstruction sound wave signal set obtaining module is used for carrying out filtering reconstruction according to the detection sound wave signal set to obtain a reconstruction sound wave signal set;

The detection fluid characteristic data set determining module is used for detecting the fluid characteristic of the valve according to the fluid detection end in the valve and determining a detection fluid characteristic data set;

The valve flaw detection data set loading module is used for loading a valve flaw detection data set;

the valve internal leakage generation coefficient determining module is used for carrying out valve internal leakage characteristic analysis according to the reconstructed sound wave signal set, the detected fluid characteristic data set and the valve flaw detection data set based on a valve internal leakage joint detection algorithm to determine a valve internal leakage generation coefficient;

The valve internal leakage alarm instruction generation module is used for generating a valve internal leakage alarm instruction if the valve internal leakage generation coefficient is greater than/equal to a valve internal leakage generation threshold value;

the valve internal leakage tracing early-warning signal generation module is used for carrying out internal leakage tracing detection on the valve according to the valve internal leakage warning instruction to generate a valve internal leakage tracing early-warning signal;

The valve internal leakage generation coefficient determining module is further used for: