CN116838947A - Fault diagnosis method for oxygen supply and saving system and oxygen supply and saving system - Google Patents
Fault diagnosis method for oxygen supply and saving system and oxygen supply and saving system Download PDFInfo
- Publication number
- CN116838947A CN116838947A CN202310801985.7A CN202310801985A CN116838947A CN 116838947 A CN116838947 A CN 116838947A CN 202310801985 A CN202310801985 A CN 202310801985A CN 116838947 A CN116838947 A CN 116838947A
- Authority
- CN
- China
- Prior art keywords
- domain image
- fault
- pressure signal
- time domain
- oxygen
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 239000001301 oxygen Substances 0.000 title claims abstract description 114
- 229910052760 oxygen Inorganic materials 0.000 title claims abstract description 114
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 title claims abstract description 112
- 238000003745 diagnosis Methods 0.000 title claims abstract description 45
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000013145 classification model Methods 0.000 claims abstract description 58
- 238000012549 training Methods 0.000 claims abstract description 17
- 238000003860 storage Methods 0.000 claims description 22
- 238000004422 calculation algorithm Methods 0.000 claims description 19
- 238000012216 screening Methods 0.000 claims description 16
- 238000009826 distribution Methods 0.000 claims description 10
- 238000004519 manufacturing process Methods 0.000 claims description 6
- 230000000737 periodic effect Effects 0.000 claims description 5
- 238000002405 diagnostic procedure Methods 0.000 claims 7
- 238000004458 analytical method Methods 0.000 abstract description 8
- 238000010586 diagram Methods 0.000 description 34
- 238000012706 support-vector machine Methods 0.000 description 17
- 238000005457 optimization Methods 0.000 description 8
- 239000007789 gas Substances 0.000 description 6
- 239000002245 particle Substances 0.000 description 6
- 238000012847 principal component analysis method Methods 0.000 description 5
- 238000000513 principal component analysis Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 230000009194 climbing Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 210000003928 nasal cavity Anatomy 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- QVGXLLKOCUKJST-NJFSPNSNSA-N oxygen-18 atom Chemical compound [18O] QVGXLLKOCUKJST-NJFSPNSNSA-N 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
- A61M16/00—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
- A61M16/00—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes
- A61M16/021—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes operated by electrical means
- A61M16/022—Control means therefor
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
- F17D—PIPE-LINE SYSTEMS; PIPE-LINES
- F17D1/00—Pipe-line systems
- F17D1/02—Pipe-line systems for gases or vapours
- F17D1/04—Pipe-line systems for gases or vapours for distribution of gas
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
- F17D—PIPE-LINE SYSTEMS; PIPE-LINES
- F17D3/00—Arrangements for supervising or controlling working operations
- F17D3/01—Arrangements for supervising or controlling working operations for controlling, signalling, or supervising the conveyance of a product
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F17—STORING OR DISTRIBUTING GASES OR LIQUIDS
- F17D—PIPE-LINE SYSTEMS; PIPE-LINES
- F17D5/00—Protection or supervision of installations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Abstract
The invention relates to a fault diagnosis method of an oxygen supply system and the oxygen supply system. And further collecting the first pressure information generated by the fault module, and converting the first pressure information into time domain and frequency domain information. The invention uses time domain and frequency domain signal analysis method to accurately select the characteristic value of fault signal and establishes SVM model as fault diagnosis tool. The feature value selected after the PCA is used for reducing the dimension is used for training the SVM model, and an SVM classification model meeting the accuracy requirement is established.
Description
Technical Field
The invention relates to the technical field of oxygen supply apparatuses, in particular to a fault diagnosis method of an oxygen supply system and the oxygen supply system.
Background
At present, the traditional oxygen supply mode is continuous flow oxygen supply, and human oxygen inhalation is pulse oxygen inhalation, so that great resource waste is caused. Particularly for oxygen inhalation people in a highland environment, a large number of heavy oxygen storage equipment cannot be carried at a time, so that the demand for light weight is required, and the time for providing oxygen inhalation is also considered, so that an oxygen supply and saving system is generated.
Currently, some technologies, such as CN114602022 a, provide an oxygen supply adjusting device, which can implement adjustment of oxygen supply according to multi-parameter control. In addition, some prior art technologies can realize pulse type oxygen supply, for example, CN113730751a discloses an oxygen supply adjusting device, which can realize automatic pulse type oxygen supply. The oxygen supply device is also very prone to failure along with the improvement of the automation degree. If the oxygen supply system fails, pulse oxygen supply can not be realized, so that great waste is caused; the oxygen supply blockage of the oxygen supply system can be caused, and the life safety of oxygen inhalation people is greatly influenced; the pressure of the air outlet of the oxygen supply device is very unstable, and the nasal cavity, the throat and the like are damaged.
At present, aiming at fault diagnosis of an oxygen supply and saving system, a learner proposes to develop a fault diagnosis machine for the oxygen supply and saving system by using an expert machine as a method so as to improve the accuracy of fault diagnosis of an oxygen supply and saving element, or a fault diagnosis method based on fuzzy mathematics, D-S evidence theory and fuzzy integral algorithm, which is applied to the oxygen supply and saving element, and improves the efficiency of fault diagnosis of the oxygen supply and saving element. And then or the provided fault tree analysis and diagnosis method can effectively solve the complex and changeable fault diagnosis of the oxygen-saving element. However, the fault diagnosis method for the oxygen supply system is limited by a theoretical mechanism and practical application, the pursuit accuracy rate ignores efficiency or the pursuit efficiency cannot compromise the accuracy rate, and the cases and the set threshold values are supplemented by human experience, so that the process is complex and has uncertainty.
In view of the foregoing, there is a need for a fault diagnosis method for an oxygen supply and an oxygen supply system, so as to accurately, quickly, efficiently and comprehensively diagnose the fault of the oxygen supply and the oxygen supply system.
Disclosure of Invention
The present invention is directed to alleviating or solving at least one aspect or point of the problems discussed above.
The invention provides a fault diagnosis method of an oxygen supply and saving system, which comprises the following steps: comprises a regulation and control module, an oxygen production module, a total storage module and an oxygen saving module; the method comprises the following steps:
s1: collecting a first pressure signal, wherein the first pressure signal reflects a working pressure signal of an oxygen-saving system;
s2: converting the first pressure signal value into a first time domain image, and performing fast Fourier transform on the first time domain image to output a first frequency domain image;
s3: obtaining information of a fault module according to the first time domain image and the first frequency domain image;
s4: acquiring fault module information and corresponding first time domain image information and first frequency domain image information of a first pressure signal;
s5: extracting a first quantity of first time domain image information and characteristic values of the first frequency domain image information;
s6: screening the first quantity of characteristic values through PCA to obtain a second quantity of characteristic values, wherein the second quantity is smaller than the first quantity;
s7: training the SVM classification model by adopting a second number of characteristic values;
s8: a final classification model is determined.
According to the fault diagnosis method of the oxygen supply system, the first pressure signal is collected and analyzed, so that the information of a fault module is determined.
And further collecting first pressure information when generating a fault module, converting the first pressure information into time domain and frequency domain information, extracting a first number of characteristic values, screening a second number of characteristic values through PCA to represent the characteristics when generating the fault, and further training through SVM classification models, so that the fault classification models can be accurately represented by the minimum number of characteristics (the second number), and the accuracy is sufficient. Therefore, fault diagnosis is realized in a mode of small calculated amount and extremely high accuracy.
The invention also provides a fault diagnosis method of the oxygen supply and saving system, which comprises the following steps: comprises a regulation and control module, an oxygen production module, a total storage module and an oxygen saving module; the method comprises the following steps:
s1: collecting a first pressure signal, wherein the first pressure signal reflects a working pressure signal of an oxygen-saving system;
s2: converting the first pressure signal value into a first time domain image, and performing fast Fourier transform on the first time domain image to output a first frequency domain image;
s3: obtaining information of a fault module according to the first time domain image and the first frequency domain image;
s31: collecting a second pressure signal, wherein the second pressure signal reflects a working pressure signal of an internal device of the fault module;
s32: converting the second pressure signal value into a second time domain image, and performing fast Fourier transform on the second pressure signal to output a second frequency domain image;
s33: and obtaining information of the internal device with the fault module according to the second time domain image and the second frequency domain image.
S41: acquiring second time domain image information and second frequency domain image information of a corresponding second pressure signal when faults occur;
s51: extracting a third quantity of second time domain image information and characteristic values of the second frequency domain image information;
s61: screening to obtain a fourth number of characteristic values from the third number of characteristic values through PCA screening, wherein the fourth number is smaller than the third number;
s71: training the SVM classification model by adopting a fourth number of second pressure signal characteristic values;
s8: a final classification model is determined.
According to the method, based on the first diagnosis method, the fault can be further and accurately positioned to the components in each module through measurement of the second pressure signal, feature extraction, PCA screening and SVM classification training, and the accuracy is greatly improved.
In order to achieve better effects, the invention also provides the following preferable technical scheme:
preferably, determining the final classification model in step S8 includes: judging whether the classification accuracy is more than 90%, if so, determining a final classification model; if not, optimizing the SVM classification model through a PSO algorithm until the classification accuracy reaches 90%, and determining a final classification model.
Preferably, in step S8, if the classification accuracy rate is not more than 90%, optimizing the penalty factor C and the kernel function g in the SVM classification model by using the PSO algorithm, and finally obtaining the SVM classification model applicable to fault diagnosis.
Preferably, step S8 further comprises the step of: and inputting the pressure signal to be detected into the final classification model, and diagnosing fault information.
Preferably, step S3 includes: observing the overall fluctuation condition of the first time domain image signal curve and the time domain characteristics of the highest point, the lowest point and the periodic signal thereof; and analyzing and positioning information of the fault module by matching with the energy distribution condition of the first frequency domain image signal and harmonic distribution, energy of each frequency and frequency amplitude.
Preferably, step S8 further comprises the step of: and setting respective fault labels for the faults, and displaying fault information by the classification model in a fault label displaying mode.
Preferably, the first number of characteristic values comprises:
the time domain selected characteristic values are maximum value, minimum value, peak value, variance, kurtosis value, skewness value, waveform factor, pulse factor and margin factor;
the eigenvalues selected in the frequency domain are the center-of-gravity frequency, root mean square frequency, and standard deviation of frequency.
Preferably, step S6 includes: the first number of eigenvalues is screened and fused by PCA seeking to represent more than 95% of the eigenvalues with the smallest number of eigenvalues.
In addition, the invention also provides an oxygen supply and saving system, and the oxygen supply and saving system adopts the fault diagnosis method.
The fault diagnosis method of the oxygen supply and saving system can quickly and accurately detect the fault of the oxygen supply and saving system, prevent the occurrence of fault risks, is particularly suitable for medical oxygen supply, and can accurately and quickly respond and avoid the occurrence of medical fault risks by combining a time domain and frequency domain signal analysis method, a principal component analysis method, a particle swarm optimization algorithm and a support vector machine classification model.
Drawings
FIG. 1 is a schematic diagram of an organization of an oxygen supply system according to an exemplary embodiment of the present invention.
FIG. 2 is a schematic diagram of a fault diagnosis flow of an oxygen supply system according to an exemplary embodiment of the present invention.
FIG. 3 is a schematic diagram of a time domain diagram of an oxygen supply system according to an exemplary embodiment of the present invention when operating normally.
FIG. 4 is a schematic diagram of a time domain diagram of an oxygen supply system with an oxygen saving module operating abnormally, in accordance with an exemplary embodiment of the present invention.
Fig. 5 is a schematic diagram of a normal operation time-frequency domain of an oxygen supply system according to an exemplary embodiment of the present invention.
Fig. 6 is a schematic diagram of an abnormal operation time-frequency domain of an oxygen saving module in an oxygen supplying system according to an exemplary embodiment of the present invention.
Fig. 7 is a principal component analysis eigenvalue generation diagram of an exemplary embodiment of the present invention.
FIG. 8 is a schematic diagram of a fitness curve of a particle swarm optimization algorithm according to an exemplary embodiment of the invention.
FIG. 9 is a diagram of training accuracy of a support vector machine classification model according to an exemplary embodiment of the invention.
FIG. 10 is a diagram of support vector machine classification model prediction accuracy according to an exemplary embodiment of the present invention.
FIG. 11 is a diagram illustrating actual prediction results of a support vector machine classification model according to an exemplary embodiment of the present invention.
Wherein: the device comprises a 1-regulation and control module, a 2-oxygen generation module, a 3-total storage module, a 4-oxygen-saving module, a 21-air control unit, a 22-oxygen generation unit, a 23-exchange unit, a 24-exhaust unit, a 31-left branch gas storage unit, a 32-right branch gas storage unit, a 41-electromagnetic linkage structure, a 42-oxygen-saving control structure and a 43-air path structure.
Detailed Description
The following description of embodiments of the present invention with reference to the accompanying drawings is intended to illustrate the general inventive concept and should not be taken as limiting the invention. In the present invention, the same reference numerals denote the same or similar components.
The features described herein may be embodied in different forms and should not be construed as limited to the examples described herein. Rather, the examples described herein have been provided to illustrate only some of the many possible ways to implement the methods, devices, and/or systems described herein that will be apparent after an understanding of the present disclosure.
Although terms such as "first," "second," and "third" may be used herein to describe various elements, components, regions, layers or sections, these elements, components, regions, layers or sections should not be limited by these terms. Rather, these terms are only used to distinguish one element, component, region, layer or section from another element, component, region, layer or section.
In the description, when an element (such as a layer, region or substrate) is referred to as being "on" another element, "connected to" or "coupled to" the other element, it can be directly "on" the other element, be directly "connected to" or be "coupled to" the other element, or one or more other elements intervening elements may be present. In contrast, when an element is referred to as being "directly on" or "directly connected to" or "directly coupled to" another element, there may be no other element intervening elements present.
The terminology used herein is for the purpose of describing various examples only and is not intended to be limiting of the disclosure. Singular forms also are intended to include plural forms unless the context clearly indicates otherwise. The terms "comprises," "comprising," and "having" specify the presence of stated features, amounts, operations, components, elements, and/or combinations thereof, but do not preclude the presence or addition of one or more other features, amounts, operations, components, elements, and/or combinations thereof.
In order to enable one skilled in the art to utilize the teachings of the present invention, the following exemplary embodiments are presented in terms of particular application scenarios, particular system, device and component parameters and particular manner of connection. However, it will be apparent to those having ordinary skill in the art that these embodiments are merely examples, and that the general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the invention.
According to an exemplary embodiment of the present invention: as shown in FIG. 1, the oxygen supply and saving system comprises a regulation and control module 1, an oxygen generation module 2, a total storage module 3 and an oxygen saving module 4. Optionally, the air outlet of the oxygen generating module 2 is connected with the total storage module 3, and after oxygen is generated by the oxygen generating module 2, the oxygen is stored in two units of the total storage module 3 through pipelines, namely a left branch air storage unit 31 and a right branch air storage unit 32. The air outlet of the total storage module 3 is connected with the air inlet of the intelligent oxygen-saving module 4. When oxygen is needed, the total storage module 3 provides oxygen to the outside of the system through the oxygen-saving module 4, wherein the oxygen-saving module 4 is an intelligent oxygen-saving module.
As shown in fig. 1, a regulation module 1 is connected with an oxygen generation module 2 and a total storage module 3 through circuit control, and the regulation module 1 integrally coordinates oxygen generation and storage of the system. The oxygen generating module 2 comprises a pneumatic control unit 21, an oxygen generating unit 22, an exchange unit 23 and an exhaust unit 24, and the four units are connected through pipelines. As shown in fig. 1, the total gas storage module 3 includes a left branch gas storage unit 31 and a right branch gas storage unit 32, and the two gas storage units are respectively and independently connected with the oxygen generating module 2 and the intelligent oxygen saving module 4. The intelligent oxygen-saving module 4 comprises an electromagnetic linkage structure 41, an oxygen-saving control structure 42 and an air channel structure 43, wherein the electromagnetic linkage structure 41 and the oxygen-saving control structure 42 are connected through the air channel structure 43.
According to an exemplary embodiment of the present invention: as shown in fig. 2, the fault diagnosis method for an oxygen saving system provided by the invention comprises the following steps:
s1: collecting a first pressure signal, wherein the first pressure signal reflects a pressure signal when the oxygen-saving system works;
s2: converting the first pressure signal value into a first time domain image, and performing fast Fourier transform on the first time domain image to output a first frequency domain image;
s3: obtaining information of a fault module according to the first time domain image and the first frequency domain image;
s4: acquiring fault module information and corresponding first time domain image information and first frequency domain image information of a first pressure signal;
s5: extracting a first quantity of first time domain image information and characteristic values of the first frequency domain image information; the first number of feature values may be specifically selected by those skilled in the art according to practical situations, and features that can characterize the first time domain image information and the first frequency domain image information are common in the art.
S6: screening the first quantity of characteristic values through PCA to obtain a second quantity of characteristic values, wherein the second quantity is smaller than the first quantity;
s7: training the SVM classification model by adopting a second number of characteristic values;
s8: a final classification model is determined.
According to the fault diagnosis method of the oxygen supply system, the first pressure signal is collected and analyzed, so that the information of a fault module is determined.
Optionally, a pressure sensor is used for monitoring a working pressure output port of the oxygen supply system, a first time domain diagram is generated after data is obtained, and the monitored data is subjected to time-frequency conversion through a fast Fourier transform algorithm to generate the first frequency domain diagram. And performing feature analysis on the obtained time domain diagram and the obtained frequency domain diagram to determine a fault module of the oxygen supply system. The time domain diagram and the frequency domain diagram when faults occur can be compared with the time domain diagram and the frequency domain diagram when normal operation occurs, and the fault module is judged.
The invention converts the first pressure information generated by the fault module into time domain and frequency domain information, extracts the first number of characteristic values, screens the second number of characteristic values by PCA to represent the characteristics generated by the fault module, and further realizes the accurate representation of the fault classification model by using the minimum number of characteristics (the second number) in a training mode of the SVM classification model, and has enough accuracy. Therefore, fault diagnosis is realized in a mode of small calculated amount and extremely high accuracy. The first pressure information when the fault module is collected can be collected through the testing machine, for example, different modules of the testing machine respectively generate faults in a manual intervention mode, and the information of the different fault modules and the corresponding first pressure information are collected. The system can also be collected during actual use, and during actual use, the fault information is determined by analyzing the first time domain diagram and the first frequency domain diagram and comparing the first time domain diagram and the first frequency domain diagram with the image information in a normal state, and fault module information and corresponding first pressure information are collected. And preparing for subsequent establishment of the classification model through information acquisition.
According to an exemplary embodiment of the present invention: if finer fault information is required to be obtained, such as analysis of the first pressure information, only the fault module can be judged to be an oxygen generating module, a total storage module, an intelligent oxygen saving module and a regulating module. If it judges that it is an oxygen-saving module, it cannot judge which unit or structure in the oxygen-saving module is faulty, if it cannot determine that it is an electromagnetic linkage structure fault, or an oxygen-saving control structure and an air circuit structure fault.
According to the above situation, the present invention further provides a fault diagnosis method for an oxygen supply and saving system, where the oxygen supply and saving system: the method comprises the following steps:
s1: collecting a first pressure signal, wherein the first pressure signal reflects a working pressure signal of an oxygen-saving system;
s2: converting the first pressure signal value into a first time domain image, and performing fast Fourier transform on the first time domain image to output a first frequency domain image;
s3: obtaining information of a fault module according to the first time domain image and the first frequency domain image;
s31: collecting a second pressure signal, wherein the second pressure signal reflects a working pressure signal of an internal device of the fault module;
s32: converting the second pressure signal value into a second time domain image, and performing fast Fourier transform on the second pressure signal to output a second frequency domain image;
s33: and obtaining information of the internal device with the fault of the fault module according to the second time domain image and the second frequency domain image.
S41: acquiring second time domain image information and second frequency domain image information of a corresponding second pressure signal when faults occur;
s51: extracting a third quantity of second time domain image information and characteristic values of the second frequency domain image information;
s61: screening to obtain a fourth number of characteristic values from the third number of characteristic values through PCA screening, wherein the fourth number is smaller than the third number;
s71: training the SVM classification model by adopting a fourth number of second pressure signal characteristic values;
s8: a final classification model is determined.
According to the method, based on the first diagnosis method, the fault can be further and accurately positioned to the components in each module through measurement of the second pressure signal, feature extraction, PCA screening and SVM classification training, and the accuracy is greatly improved.
The invention can monitor and collect the pressure data again aiming at the fault module of the oxygen-saving system, extract the time domain and frequency domain characteristic values of the collected data, and apply the principal component analysis method, namely PCA screening, to reduce the dimension of the characteristic values, eliminate the redundancy of the characteristic values, improve the efficiency and ensure the accuracy. And replacing or characterizing information of the characteristic value which is more than or equal to 95% by using the minimum characteristic value quantity. And training a support vector machine classification model, namely an SVM classification model, by using a feature value obtained by a principal component analysis method, and optionally, if the classification accuracy rate is not more than 90%, optimizing a penalty factor C and a kernel function g in the support vector machine classification model by using a particle swarm optimization algorithm, namely a PSO algorithm, so as to finally obtain the support vector machine fault classification model applicable to fault diagnosis.
According to an exemplary embodiment of the present invention: the determining of the final classification model in step S8 comprises: judging whether the classification accuracy is more than 90%, if so, determining a final classification model; if not, optimizing the SVM classification model through a PSO algorithm until the classification accuracy reaches 90%, and determining a final classification model. By controlling the accuracy of the classification model to be more than 90%, fault information can be accurately and rapidly given. Particularly suitable for oxygen supply systems requiring rapid response as in the present invention. Further comprising step S9: and inputting the pressure signal to be detected into the final classification model, and diagnosing fault information.
According to an exemplary embodiment of the present invention: the step S3 includes: and observing the overall fluctuation condition of the first time domain image signal curve, the highest point, the lowest point and the time domain characteristics of the periodic signal, and analyzing and positioning the information of the fault module by matching with the energy distribution condition of the first frequency domain image signal, the harmonic distribution, the energy of each frequency and the frequency amplitude.
According to an exemplary embodiment of the present invention: the step S3 includes: and observing the integral fluctuation condition of the first time domain image signal curve and the time domain obvious characteristics such as the highest point, the lowest point and the periodic signal thereof, and analyzing and positioning the information of the fault module by matching with the energy distribution condition of the first frequency domain image signal and the harmonic distribution, energy of each frequency, frequency amplitude and the like.
According to an exemplary embodiment of the present invention: the first number of feature values includes: the time domain selected characteristic values are maximum value, minimum value, peak value, variance, kurtosis value, skewness value, waveform factor, pulse factor and margin factor; the frequency domain selects characteristic values such as a gravity center frequency, a root mean square frequency, a frequency standard deviation and the like.
According to an exemplary embodiment of the present invention: the step S6 includes: the screening fusion of a large number of eigenvalues by PCA seeks to represent more than 95% of the eigenvalues with the smallest number of eigenvalues. I.e. information characterizing at least 95% of the feature values before screening, among the selected feature values.
According to an exemplary embodiment of the present invention: in the step S8, if the accuracy rate of classification is not more than 90%, optimizing the penalty factor C and the kernel function g in the SVM classification model through a PSO algorithm, and finally obtaining the support vector machine fault classification model applicable to fault diagnosis.
The operation of the invention is briefly described below with reference to fig. 1-11: the method is practically applied, and the experimental object is an oxygen supply and saving system which is currently applied by the unit.
The first step: and collecting the whole pressure signal of the oxygen supply system and storing the whole pressure signal in a PC terminal. Wherein the global pressure signal may be the first pressure signal.
To ensure that the measured data is the normal working pressure of the oxygen supply and saving system, a three-way joint is arranged at the air outlet of the total working pressure, two ends of the three-way joint are normally connected with the air outlet air circuit of the total working pressure of the oxygen supply and saving system, and the other end of the three-way joint is connected with the pressure sensor. In the example, a Liff T2000.2526 pressure sensor is adopted, and the measuring range is 0-0.4 MPa. The other end of the pressure sensor is connected with a PC terminal, and the PC terminal stores pressure data acquired by the pressure sensor.
And a second step of: the original integral pressure signal is input into a time domain generation algorithm to obtain a time domain diagram of the total working pressure of the oxygen-saving system, and then the time domain signal is converted into a frequency domain signal through fast Fourier transform, and the frequency domain diagram is generated.
And observing the overall fluctuation condition of the signal curve of the time domain graph and the time domain obvious characteristics such as the highest point, the lowest point and the periodic signal thereof, and analyzing and positioning the overall obvious fault characteristics or fault modules of the oxygen supply system by matching with the energy distribution condition of the signal of the frequency domain graph and the harmonic distribution, the energy of each frequency, the frequency amplitude and the like.
As shown in fig. 3 and 5, the time-domain and frequency-domain diagrams of the overall pressure signal for normal operation of the oxygen-saving system are shown.
As shown in fig. 4 and 6, the time-domain diagram and the frequency-domain diagram of the entire pressure signal when some kind of failure occurs. The time domain image shows that the pressure fluctuation curve has obvious climbing characteristics, and the frequency domain image shows that first harmonic wave is generated in two adjacent wave peaks.
According to analysis, when the oxygen supply and saving system fails in the oxygen saving module 4, the air path on-off structure is not timely reciprocated, which is also the reason that the time domain diagram forms climbing and the first harmonic wave is generated in the frequency domain diagram, so that the failure module of the oxygen supply and saving system can be judged to be the oxygen saving module 4.
And a third step of: and (3) collecting the pressure signal of the fault module positioned in the second step, and extracting the characteristic values of the time domain and the frequency domain of the collected pressure signal. If the fault module is positioned as the oxygen-saving module 4, the second pressure signal can collect the pressure of the connecting pipeline between the oxygen-saving module 4 and the breathing port of the human body.
The principle of the classification algorithm diagnosis is that the requirement of high accuracy is met by training sample data, and then fault diagnosis and classification are carried out on actual data. The selection of sample data is crucial, and sample data with obvious characteristics is beneficial to improving the classification accuracy. Considering the total working pressure signal of the oxygen supply system acquired in the first step, a great amount of interference exists, and the fault classification accuracy is limited and is only suitable for diagnosing a fault module.
Aiming at the fine fault characteristics of the oxygen supply and saving system, the fault pressure signals of the positioned fault modules need to be acquired again, so that the effect of comprehensive fault diagnosis of the oxygen supply and saving system is achieved. Wherein the fault pressure signal may be a second pressure signal.
And selecting the characteristic values of the time domain and the frequency domain of the collected fault pressure signal, wherein the characteristic values selected by the time domain are the characteristic values of the maximum value, the minimum value, the peak value, the variance, the kurtosis value, the skewness value, the waveform factor, the pulse factor, the margin factor and the like. The frequency domain selects characteristic values such as a gravity center frequency, a root mean square frequency, a frequency standard deviation and the like.
Fourth step: a large number of secondary and interfering features are generated during feature extraction. If the method is not used for eliminating, the calculated amount is greatly increased, and the diagnosis result is possibly interfered, so that the method adopts a principal component analysis algorithm to screen and fuse a large number of characteristic values, and seeks to adopt the least characteristic value quantity to represent more than 95% of characteristic value information.
As shown in fig. 7, the principal component analysis method performs feature screening and fusion on 9 time domain features and 3 frequency domain features, and finally displays that 4 feature values generated by the principal component analysis method can represent more than 95% of the features of the 12 feature values, so that the training speed of the model is greatly increased, and the classification of the model is finer.
Fifth step: a small amount of characteristic values obtained by applying a principal component analysis optimization algorithm are put into a support vector machine model, the support vector machine classification model is trained to lead the final classification accuracy to be more than 90 percent, if the accuracy is not more than 90 percent, a particle swarm optimization algorithm is selected to optimize an important parameter penalty factor C and a kernel function g of the support vector machine classification model,
as shown in fig. 8, the final optimization fitness can be seen to be 98% in the particle swarm optimization algorithm.
As shown in fig. 9 and 10, the accuracy of fault diagnosis of 244 data points by the support vector machine training model reaches 100%, the accuracy of fault diagnosis of 76 data points by the prediction model reaches 94.7%, the requirement of fault diagnosis classification of the oxygen supply and saving system is met, and the fault of the oxygen supply and saving system can be accurately, efficiently and comprehensively diagnosed in practical application.
Sixth step: and classifying actual data features after obtaining a support vector machine classification model meeting the accuracy requirement, setting respective fault labels for each fault unit or structure in the fault module, and considering that the probability of occurrence of the fault '2' is maximum when the result of the support vector machine classification model gives the result shown in fig. 11. For example, the air path structure 43 in the oxygen saving module 4 corresponding to the label "1" fails, the electromagnetic linkage structure 41 in the oxygen saving module 4 corresponding to the label "2" fails, the oxygen saving control structure 42 in the oxygen saving module 4 corresponding to the label "3" fails, and the oxygen saving module 4 corresponding to the label "4" operates normally. As shown in fig. 11, in the oxygen supply system, the electromagnetic linkage 41 may be considered to be defective.
According to one of the fault diagnosis methods disclosed by the invention, if the production of an enterprise is urgent, a time domain and frequency domain graph analysis mode can be directly adopted, so that a fault module of the oxygen-saving system is positioned, the fault module is directly replaced, the repair efficiency is improved, and the normal production of the enterprise is ensured.
The second fault diagnosis method of the invention is to monitor the total working pressure of the oxygen-saving system by using a pressure sensor and collect data, to generate time domain and frequency domain diagrams after obtaining the pressure data, to analyze the time domain and frequency domain diagrams to determine a fault module, to collect the pressure value again by the fault module, to extract a small amount of characteristic values by using a principal component analysis algorithm to eliminate the redundancy of the characteristic values, to train a support vector machine classification model by using the extracted characteristic values, to optimize the important parameters of the support vector machine classification model by using a particle swarm optimization algorithm if the required accuracy is not achieved, and to diagnose the actual fault by using the model after the support vector machine classification model meets the accuracy requirement.
The fault diagnosis method of the oxygen supply system and the oxygen supply system accurately select the characteristic value of the fault signal by using the time domain and frequency domain signal analysis method, and establish an SVM model as a fault diagnosis tool. The feature value selected after the PCA is used for reducing the dimension is used for training the SVM model, and an SVM classification model meeting the accuracy requirement is established.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes may be made and equivalents may be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (10)
1. A fault diagnosis method for an oxygen-saving system is characterized in that:
the oxygen supply and saving system comprises: comprises a regulation and control module (1), an oxygen production module (2), a total storage module (3) and an oxygen saving module (4); the method comprises the following steps:
s1: collecting a first pressure signal, wherein the first pressure signal reflects a working pressure signal of an oxygen-saving system;
s2: converting the first pressure signal value into a first time domain image, and performing fast Fourier transform on the first time domain image to output a first frequency domain image;
s3: obtaining information of a fault module according to the first time domain image and the first frequency domain image;
s4: acquiring fault module information and corresponding first time domain image information and first frequency domain image information of a first pressure signal;
s5: extracting a first quantity of first time domain image information and characteristic values of the first frequency domain image information;
s6: screening the first quantity of characteristic values through PCA to obtain a second quantity of characteristic values, wherein the second quantity is smaller than the first quantity;
s7: training the SVM classification model by adopting a second number of characteristic values;
s8: a final classification model is determined.
2. The diagnostic method of claim 1, wherein: the determining of the final classification model in step S8 comprises: judging whether the classification accuracy is more than 90%, if so, determining a final classification model; if not, optimizing the SVM classification model through a PSO algorithm until the classification accuracy reaches 90%, and determining a final classification model.
3. The diagnostic method of claim 2, wherein: in the step S8, if the classification accuracy rate is not more than 90%, optimizing the penalty factor C and the kernel function g in the SVM classification model through a PSO algorithm, and finally obtaining the SVM classification model applicable to fault diagnosis.
4. The diagnostic method of claim 1, wherein: the step S8 is followed by the steps of: and inputting the pressure signal to be detected into the final classification model, and diagnosing fault information.
5. The diagnostic method of claim 1, wherein: the step S3 includes: observing the overall fluctuation condition of the first time domain image signal curve and the time domain characteristics of the highest point, the lowest point and the periodic signal thereof; and analyzing and positioning information of the fault module by matching with the energy distribution condition of the first frequency domain image signal and harmonic distribution, energy of each frequency and frequency amplitude.
6. The diagnostic method of claim 1, wherein: the steps between the step S7 and the step S8 also comprise the steps of: S7A: and setting respective fault labels for the faults, and displaying fault information by the classification model in a fault label displaying mode.
7. The diagnostic method of claim 1, wherein: the first number of feature values includes:
the time domain selected characteristic values are maximum value, minimum value, peak value, variance, kurtosis value, skewness value, waveform factor, pulse factor and margin factor;
the eigenvalues selected in the frequency domain are the center-of-gravity frequency, root mean square frequency, and standard deviation of frequency.
8. The diagnostic method of claim 1, wherein: the step S6 includes: the first number of eigenvalues is screened and fused by PCA seeking to represent more than 95% of the eigenvalues with the smallest number of eigenvalues.
9. A fault diagnosis method for an oxygen-saving system is characterized in that:
the oxygen supply and saving system comprises: comprises a regulation and control module (1), an oxygen production module (2), a total storage module (3) and an oxygen saving module (4); the method comprises the following steps:
s1: collecting a first pressure signal, wherein the first pressure signal reflects a working pressure signal of an oxygen-saving system;
s2: converting the first pressure signal value into a first time domain image, and performing fast Fourier transform on the first time domain image to output a first frequency domain image;
s3: obtaining information of a fault module according to the first time domain image and the first frequency domain image;
s31: collecting a second pressure signal, wherein the second pressure signal reflects a working pressure signal of an internal device of the fault module;
s32: converting the second pressure signal value into a second time domain image, and performing fast Fourier transform on the second pressure signal to output a second frequency domain image;
s33: obtaining information of an internal device with a fault module according to the second time domain image and the second frequency domain image;
s41: acquiring second time domain image information and second frequency domain image information of a corresponding second pressure signal when faults occur;
s51: extracting a third quantity of second time domain image information and characteristic values of the second frequency domain image information;
s61: screening the characteristic values of the fourth quantity from the characteristic values of the third quantity through PCA screening, wherein the fourth quantity is smaller than the third quantity;
s71: training the SVM classification model by adopting a fourth number of second pressure signal characteristic values;
s8: a final classification model is determined.
10. An oxygen supply system, characterized in that: the oxygen supply system employs the failure diagnosis method according to any one of claims 1 to 9.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310801985.7A CN116838947B (en) | 2023-06-30 | 2023-06-30 | Fault diagnosis method for oxygen supply and saving system and oxygen supply and saving system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310801985.7A CN116838947B (en) | 2023-06-30 | 2023-06-30 | Fault diagnosis method for oxygen supply and saving system and oxygen supply and saving system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116838947A true CN116838947A (en) | 2023-10-03 |
CN116838947B CN116838947B (en) | 2024-02-13 |
Family
ID=88162863
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310801985.7A Active CN116838947B (en) | 2023-06-30 | 2023-06-30 | Fault diagnosis method for oxygen supply and saving system and oxygen supply and saving system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116838947B (en) |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101968644A (en) * | 2010-08-24 | 2011-02-09 | 苏州艾隆科技有限公司 | Unattended remote medical oxygenerator monitoring system |
CN107037800A (en) * | 2017-06-15 | 2017-08-11 | 中国人民解放军军事医学科学院卫生装备研究所 | Based on client oxygen generating plant remote failure diagnosis system and method for diagnosing faults |
CN107091737A (en) * | 2017-06-06 | 2017-08-25 | 太原理工大学 | A kind of study of typical faults of rotor systems combining diagnostic method based on current signal |
JP2018205292A (en) * | 2017-06-05 | 2018-12-27 | 瀏陽 宋 | State identification method by characteristic analysis of histogram in time region and frequency region |
CN113730751A (en) * | 2021-08-26 | 2021-12-03 | 广东汉泓医疗科技有限公司 | Oxygen supply adjusting device, oxygen generator and oxygen supply adjusting method |
CN114602022A (en) * | 2022-03-17 | 2022-06-10 | 中国人民解放军陆军边海防学院乌鲁木齐校区 | Oxygen supply adjusting device |
CN115221930A (en) * | 2022-09-20 | 2022-10-21 | 苏州鸿哲智能科技有限公司 | Fault diagnosis method for rolling bearing |
CN115481657A (en) * | 2022-08-25 | 2022-12-16 | 西安热工研究院有限公司 | Wind generating set communication slip ring fault diagnosis method based on electric signals |
CN115957410A (en) * | 2022-12-05 | 2023-04-14 | 新加法成都智能科技有限责任公司 | Pulse oxygen supply system and pulse oxygen supply control method |
-
2023
- 2023-06-30 CN CN202310801985.7A patent/CN116838947B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101968644A (en) * | 2010-08-24 | 2011-02-09 | 苏州艾隆科技有限公司 | Unattended remote medical oxygenerator monitoring system |
JP2018205292A (en) * | 2017-06-05 | 2018-12-27 | 瀏陽 宋 | State identification method by characteristic analysis of histogram in time region and frequency region |
CN107091737A (en) * | 2017-06-06 | 2017-08-25 | 太原理工大学 | A kind of study of typical faults of rotor systems combining diagnostic method based on current signal |
CN107037800A (en) * | 2017-06-15 | 2017-08-11 | 中国人民解放军军事医学科学院卫生装备研究所 | Based on client oxygen generating plant remote failure diagnosis system and method for diagnosing faults |
CN113730751A (en) * | 2021-08-26 | 2021-12-03 | 广东汉泓医疗科技有限公司 | Oxygen supply adjusting device, oxygen generator and oxygen supply adjusting method |
CN114602022A (en) * | 2022-03-17 | 2022-06-10 | 中国人民解放军陆军边海防学院乌鲁木齐校区 | Oxygen supply adjusting device |
CN115481657A (en) * | 2022-08-25 | 2022-12-16 | 西安热工研究院有限公司 | Wind generating set communication slip ring fault diagnosis method based on electric signals |
CN115221930A (en) * | 2022-09-20 | 2022-10-21 | 苏州鸿哲智能科技有限公司 | Fault diagnosis method for rolling bearing |
CN115957410A (en) * | 2022-12-05 | 2023-04-14 | 新加法成都智能科技有限责任公司 | Pulse oxygen supply system and pulse oxygen supply control method |
Also Published As
Publication number | Publication date |
---|---|
CN116838947B (en) | 2024-02-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP3221579B1 (en) | Wind turbine condition monitoring method and system | |
CN104390657B (en) | A kind of Generator Unit Operating Parameters measurement sensor fault diagnosis method and system | |
CN104573850A (en) | Method for evaluating state of thermal power plant equipment | |
WO2016086360A1 (en) | Wind farm condition monitoring method and system | |
KR102285987B1 (en) | Method, system and computer program for detecting error of facilities in building | |
CN110163075A (en) | A kind of multi-information fusion method for diagnosing faults based on Weight Training | |
CN116320832A (en) | Monitoring equipment fault monitoring method and device | |
CN116308304B (en) | New energy intelligent operation and maintenance method and system based on meta learning concept drift detection | |
CN108287327A (en) | Metering automation terminal fault diagnostic method based on Bayes's classification | |
CN107015486A (en) | A kind of air-conditioner water system regulating valve intelligent fault diagnosis method | |
CN113177646A (en) | Power distribution equipment online monitoring method and system based on self-adaptive edge proxy | |
CN109084971B (en) | A kind of pneumatic control valve method for diagnosing faults based on particle group optimizing | |
CN116838947B (en) | Fault diagnosis method for oxygen supply and saving system and oxygen supply and saving system | |
CN112836396B (en) | Building real-time energy consumption abnormity diagnosis system | |
CN105279553B (en) | A kind of height adds to water system fault degree recognition methods | |
JP7330754B2 (en) | Abnormality diagnosis device and method | |
CN112489841A (en) | Water level fault-tolerant control method for steam generator of nuclear power unit | |
US11339763B2 (en) | Method for windmill farm monitoring | |
KR102411915B1 (en) | System and method for froviding real time monitering and ai diagnosing abnormality sign for facilities and equipments | |
CN114299338A (en) | Fault prediction and health management system for system management and related method | |
Duan et al. | Diagnosis strategy for micro-computer controlled straight electro-pneumatic braking system using fuzzy set and dynamic fault tree | |
CN111986469A (en) | Intelligent diagnosis method for field terminal fault | |
Saucedo-Dorantes et al. | Novelty Detection Methodology Based on Self-Organizing Maps for Power Quality Monitoring | |
CN117632664B (en) | Machine room equipment monitoring method and system based on automatic comparison | |
US20220349771A1 (en) | Method of detecting a leak in a hydraulic pitch system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |