CN117493759A - Gas methane distinguishing method and device based on principal component analysis and vector machine - Google Patents
Gas methane distinguishing method and device based on principal component analysis and vector machine Download PDFInfo
- Publication number
- CN117493759A CN117493759A CN202311341952.5A CN202311341952A CN117493759A CN 117493759 A CN117493759 A CN 117493759A CN 202311341952 A CN202311341952 A CN 202311341952A CN 117493759 A CN117493759 A CN 117493759A
- Authority
- CN
- China
- Prior art keywords
- gas
- data
- biogas
- methane
- principal component
- 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.)
- Pending
Links
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 title claims abstract description 406
- 239000013598 vector Substances 0.000 title claims abstract description 98
- 238000000513 principal component analysis Methods 0.000 title claims abstract description 89
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000005070 sampling Methods 0.000 claims abstract description 56
- 238000012544 monitoring process Methods 0.000 claims abstract description 40
- 238000007781 pre-processing Methods 0.000 claims abstract description 27
- 239000000126 substance Substances 0.000 claims abstract description 27
- 230000009467 reduction Effects 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims description 45
- 239000011159 matrix material Substances 0.000 claims description 30
- 230000006870 function Effects 0.000 claims description 24
- 238000012545 processing Methods 0.000 claims description 21
- 239000012080 ambient air Substances 0.000 claims description 20
- 238000004364 calculation method Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 12
- 238000012850 discrimination method Methods 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 5
- 238000004422 calculation algorithm Methods 0.000 abstract description 25
- 239000007789 gas Substances 0.000 description 351
- 239000002737 fuel gas Substances 0.000 description 10
- 230000008569 process Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 6
- 238000012360 testing method Methods 0.000 description 6
- 239000000203 mixture Substances 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 description 2
- 238000009412 basement excavation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000007635 classification algorithm Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- OTMSDBZUPAUEDD-UHFFFAOYSA-N Ethane Chemical compound CC OTMSDBZUPAUEDD-UHFFFAOYSA-N 0.000 description 1
- 239000003570 air Substances 0.000 description 1
- 230000036626 alertness Effects 0.000 description 1
- 229910002092 carbon dioxide Inorganic materials 0.000 description 1
- 239000001569 carbon dioxide Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000000855 fermentation Methods 0.000 description 1
- 230000004151 fermentation Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 231100000572 poisoning Toxicity 0.000 description 1
- 230000000607 poisoning effect Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000010865 sewage Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001988 toxicity Effects 0.000 description 1
- 231100000419 toxicity Toxicity 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 238000002759 z-score normalization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- 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
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- 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
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Computing Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Investigating Or Analyzing Materials By The Use Of Fluid Adsorption Or Reactions (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
Abstract
The invention provides a method and a device for discriminating gas methane based on principal component analysis and a vector machine, wherein the method comprises the following steps: collecting gas methane data to be distinguished according to a preset sampling rule at a target monitoring point, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished; preprocessing the gas methane data to be distinguished to obtain target gas methane data; performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information; and inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification and distinguishing to obtain distinguishing results. According to the invention, the acquired data are preprocessed and subjected to dimension reduction by using a principal component analysis algorithm, and then the dimension-reduced data are classified and judged by using a trained gas methane judgment model based on principal component analysis and a vector machine, so that the judgment accuracy and generalization capability of gas methane are improved.
Description
Technical Field
The invention relates to the technical field of gas biogas discrimination, in particular to a gas biogas discrimination method and device based on principal component analysis and a vector machine.
Background
With the rapid development of the economic society in China, urban fuel gas is widely applied, and the construction scale of fuel gas pipelines is continuously expanding as a main transportation mode. Urban fuel gas has the characteristics of inflammability, explosiveness and toxicity, and once a fuel gas pipeline leaks, fire, explosion and poisoning accidents are extremely easy to occur, so that national and people's lives and properties are lost. Based on this, timely monitoring of gas leakage is important in preventing gas accidents. However, with the rapid development of cities, urban building facilities are densely distributed, various underground facilities are more and more, and other municipal facilities nearby a gas pipeline are not counted, wherein the municipal facilities comprise drainage pipelines, sewage wells, septic tanks, running water wells, electric power underdrain, telecommunication pipeline wells and the like, and after long-term fermentation, biogas is often accumulated in the underdrain and inspection wells. Because the components of the fuel gas and the methane are close, the fuel gas monitor is easy to report by mistake. False alarms from gas monitors may result in gas enterprises being unable to quickly and accurately determine the location and severity of gas leaks, thereby delaying processing time. Prolonged false positives may also reduce the alertness of the pipeline monitoring personnel and even cause them to ignore real hazards. Therefore, how to rapidly and accurately distinguish the fuel gas from the biogas becomes a key problem for fuel gas leakage monitoring.
In the prior art, gas concentration change is mostly adopted to distinguish gas and methane, and the gas and methane distinguishing accuracy is low.
Disclosure of Invention
The invention provides a gas methane distinguishing method and device based on principal component analysis and a vector machine, which are used for solving the defect of low gas methane distinguishing accuracy in the prior art and realizing gas methane distinguishing with higher accuracy.
The invention provides a gas methane distinguishing method based on principal component analysis and a vector machine, which comprises the following steps:
collecting gas methane data to be distinguished according to a preset sampling rule at a target monitoring point, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished;
preprocessing the gas methane data to be distinguished to obtain target gas methane data;
performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information;
inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time.
According to the gas methane distinguishing method based on the principal component analysis and the vector machine, the preset sampling rule comprises the following steps:
sampling according to a first preset frequency under the condition that the concentration of the gas methane to be distinguished is lower than or equal to a preset concentration threshold value;
and under the condition that the concentration of the gas methane to be distinguished is higher than a preset concentration threshold value, sampling is carried out according to a second preset frequency.
According to the gas biogas discrimination method based on the principal component analysis and the vector machine, the principal component analysis is carried out on the target gas biogas data to obtain the characteristic information of a preset number of gas biogas, and the method specifically comprises the following steps:
calculating a covariance matrix of the target gas biogas data to obtain a monitoring data covariance matrix;
and obtaining the characteristic information of the preset number of gas biogas according to the covariance matrix of the monitoring data.
According to the gas biogas discrimination method based on the principal component analysis and the vector machine, provided by the invention, a gas biogas discrimination model is obtained by training a large number of gas biogas data samples based on the probability classification vector machine, and the method specifically comprises the following steps:
preprocessing the gas methane data sample to obtain a target sample;
Classifying the target sample according to sampling points, and carrying out principal component analysis on the target sample for each sampling point to obtain a covariance matrix;
integrating the covariance matrixes of all the sampling points to obtain a training data set;
dividing the training data set into positive and negative sample data sets according to a predetermined classification mode, and updating parameters of a pre-constructed probability classification vector machine by using the positive and negative sample data sets in a mode of maximizing a pre-selected loss function to obtain a gas methane discrimination model;
the predetermined classification mode comprises that the gas leakage is positive, and the gas non-leakage is negative.
According to the gas methane distinguishing method based on the principal component analysis and the vector machine, the preprocessing comprises sampling frequency alignment processing, normalization processing and noise reduction processing.
According to the gas methane distinguishing method based on the principal component analysis and the vector machine, the decision function of the gas methane distinguishing model comprises the following steps:
wherein alpha is i And beta i Is a model parameter obtained by training, K (x, x i ) Is a kernel function, beta 0 Is a variable for controlling the position of the classification face, x is an input value, x i Is a data value in the training set.
According to the gas biogas discrimination method based on the principal component analysis and the vector machine, the gas biogas characteristic information is input into a pre-trained gas biogas discrimination model for classification discrimination to obtain discrimination results, and the method specifically comprises the following steps:
calculating the probability of gas leakage and/or gas non-leakage by using a probability calculation formula according to the gas biogas characteristic information to obtain a calculation result;
and comparing and verifying the calculation result to obtain a discrimination result.
According to the gas methane distinguishing method based on the principal component analysis and the vector machine, the probability calculation formula comprises the following steps:
P(y＝1|x)+P(y＝-1|x)＝1
wherein P (y= 1|x) represents the probability of gas leakage, P (y= -1|x) represents the probability of not gas leakage, and f (x) represents the decision function of the gas biogas discrimination model.
The invention also provides a gas methane distinguishing device based on the principal component analysis and the vector machine, which comprises:
the acquisition unit is used for acquiring gas methane data to be distinguished according to a preset sampling rule at target monitoring points, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished;
The processing unit is used for preprocessing the gas methane data to be distinguished to obtain target gas methane data;
the analysis unit is used for carrying out principal component analysis on the target gas biogas data to obtain the characteristic information of a preset number of gas biogas;
the judging unit is used for inputting the characteristic information of the gas methane into a pre-trained gas methane judging model to carry out classification judgment to obtain a judging result; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time.
The invention also provides electronic equipment, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the gas methane distinguishing method based on the principal component analysis and the vector machine when executing the program.
The invention also provides a non-transitory computer readable storage medium, on which is stored a computer program which, when executed by a processor, implements a method for discriminating gas biogas based on principal component analysis and a vector machine as described in any one of the above.
The invention also provides a computer program product, which comprises a computer program, wherein the computer program is executed by a processor to realize the gas methane distinguishing method based on the principal component analysis and the vector machine.
According to the gas methane distinguishing method and device based on the principal component analysis and the vector machine, gas methane data to be distinguished are collected at target monitoring points according to the preset sampling rule, and the gas methane data to be distinguished comprise physical and chemical characteristic parameters of the gas methane to be distinguished; preprocessing the gas methane data to be distinguished to obtain target gas methane data; performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information; inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time. According to the invention, the acquired data are preprocessed by taking the data indexes such as the ambient temperature, the ambient air pressure, the alarm time, the combustible gas components and the concentration fraction into consideration, the dimension of the data set is reduced by using a principal component analysis algorithm, and the dimension-reduced data are classified and judged by using a trained gas methane judgment model based on principal component analysis and a vector machine, so that the judgment accuracy and the generalization capability of gas methane are improved.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is one of the flow diagrams of the fuel gas biogas discrimination method based on the principal component analysis and the vector machine provided by the invention;
FIG. 2 is a second flow chart of the gas methane discriminating method based on the principal component analysis and the vector machine provided by the invention;
FIG. 3 is a schematic diagram of an implementation flow of an embodiment of a gas biogas discrimination method based on principal component analysis and a vector machine provided by the invention;
FIG. 4 is a schematic structural diagram of the gas methane discriminating device based on the principal component analysis and the vector machine;
fig. 5 is a schematic structural diagram of an electronic device provided by the present invention.
Reference numerals:
410: an acquisition unit; 420: a processing unit; 430: an analysis unit; 440: and a discriminating unit.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The gas leakage and the biogas interference are accurately identified, so that gas enterprises can rapidly judge the gas pipeline leakage condition, hidden danger and dangerous situation are eliminated in time, and the life and property safety of people and the urban safe operation are ensured; meanwhile, the method is also beneficial to reducing the false detection and the false excavation of the gas pipeline, avoiding unnecessary excavation cost and damage to the road surface and avoiding inconvenience to the life of citizens.
In the prior art, gas concentration variation is generally adopted to distinguish between fuel gas and biogas, and only a specific gas component, such as methane gas, is generally considered, other gas components which can better react the characteristics of two mixed gases, such as ethane, carbon dioxide and the like, are not considered, and moreover, the accuracy is not high due to environmental factors, and meanwhile, the migration capability is also insufficient.
Based on the method, the invention provides a gas methane distinguishing method and device based on principal component analysis and a vector machine.
The method for discriminating gas biogas based on principal component analysis and a vector machine according to the present invention will be described below with reference to fig. 1 to 3, and fig. 1 is a schematic flow diagram of the method for discriminating gas biogas based on principal component analysis and a vector machine according to the present invention, as shown in fig. 1, and includes the steps of:
step 110: and collecting gas methane data to be distinguished according to a preset sampling rule at the target monitoring point, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished.
And acquiring corresponding discrimination data for each target monitoring point, wherein the discrimination data is gas methane data to be discriminated. Specifically, the gas methane data to be discriminated comprises physical and chemical characteristic parameters of the gas methane to be discriminated of the target monitoring point.
Furthermore, the gas methane data to be distinguished comprise the ambient temperature, the ambient air pressure, the alarm time, the combustible gas components and concentration fractions at the alarm time, and the like. In one embodiment, the target monitoring points are collectedAmbient temperature T, ambient air pressure P, alarm time T, and combustible gas component g at alarm time 1 ,g 2 ,…,g k And corresponding concentration c 1 ,c 2 ,…,c k . It is to be understood that the device for collecting the methane data of the gas to be discriminated can be selected according to the actual requirement, and the invention is not limited to this.
It should be noted that the gas methane data to be discriminated is collected according to the sampling rule. Each acquisition time point obtains a data point comprising physical and chemical characteristic parameters of the gas biogas to be discriminated. Further, in one embodiment, each acquisition time point obtains a combustible gas component g including an ambient temperature T, an ambient pressure P, an alarm time T and an alarm time 1 ,g 2 ,…,g k Corresponding concentration c of (2) 1 ,c 2 ,…,c k Is a data point of (c). After the data points are obtained, the data points are combined into a vector [ T, P, c ] with the dimension of i+2 according to the acquisition time 1 ,c 2 ,…,c k ]Representing the physicochemical properties and the change conditions of the methane-burning sample in the data acquisition position in the period of time. Where k is the combustible gas composition fraction.
In some embodiments, the gas methane data to be discriminated is continuously and uninterruptedly collected by adopting a collecting device installed at the target monitoring point. In other embodiments, the acquisition device mounted on the target monitoring point is used at certain time intervals t 0 And collecting the gas methane data to be distinguished.
Further, in order to be more beneficial to distinguishing the gas biogas, a preset sampling rule is constructed so as to increase the sampling frequency when the concentration of the combustible gas is greater than a certain threshold value. Based on the above embodiment, the preset sampling rule includes:
sampling according to a first preset frequency under the condition that the concentration of the gas methane to be distinguished is lower than or equal to a preset concentration threshold value;
and under the condition that the concentration of the gas methane to be distinguished is higher than a preset concentration threshold value, sampling is carried out according to a second preset frequency.
Specifically, at a certain time interval t 0 Collecting gas methane data to be distinguished of a monitoring area of a target monitoring point, and setting a flammable gas concentration threshold value c h Increasing the sampling frequency of the combustible gas to t when the sampling frequency is higher than a threshold value ‘ 0 . In one embodiment, at c h For the preset concentration threshold value, collect t 0 、t ‘ 0 For data n of interval 0 Next, each acquisition time point obtains a combustible gas component g including the ambient temperature T, the ambient air pressure P, the alarm time T and the alarm time 1 ,g 2 ,…,g k Corresponding concentration c of (2) 1 ,c 2 ,…,c k Is a data point of (c). The data points are combined into a dimension n according to the acquisition time sequence 0 X (k+3) matrixMatrix arrayThe gas methane data to be distinguished is obtained. Matrix->Representing the physicochemical and changing conditions of the methane-burning sample at the data acquisition position in the period of time.
Step 120: preprocessing the gas methane data to be distinguished to obtain target gas methane data.
The collected gas methane data to be distinguished is preprocessed, and the main purpose of the preprocessing is to format and normalize the gas methane data to be distinguished, specifically, to clean, normalize and reduce noise of the gas methane data to be distinguished, as shown in fig. 2.
In some embodiments, the preprocessing includes a sampling frequency alignment process, a normalization process, and a noise reduction process.
Specifically, in some embodiments, bilinear interpolation is performed on the acquired gas biogas data to be distinguished, different time measurement data are aligned, and then a data set consisting of all vectors is subjected to Z-score standardization to obtain target gas biogas data, so that preprocessing is completed. Further, the plurality of target gas biogas data constitutes a target data set.
Furthermore, in a specific embodiment, bilinear interpolation is utilized to convert all the gas biogas data to be discriminated into a sampling interval t ‘ 0 The data with number n of samples is sampled to align the vectors under different sampling time interval metrics. Finally, a data matrix M is obtained n×(k+3) The data set of all vector components is then Z-score normalized to yield the target data set. After the pretreatment step is completed, the gas methane data to be distinguished are converted into Gaussian distribution with the mean value of 0 and the standard deviation of 1.
Step 130: and carrying out principal component analysis on the target gas biogas data to obtain the characteristic information of a preset number of gas biogas.
After preprocessing is completed to obtain target gas biogas data, main component analysis is needed to be carried out on the target gas biogas data before the data to be distinguished is input into a gas biogas distinguishing model, so that the dimension of the high-dimensional data is reduced, the complexity and redundant information of the data are reduced, meanwhile, important characteristic information is reserved, the main component coefficients of the data are obtained, and then the dimension-reduced data are used as input of the gas biogas distinguishing model and are applied to a next recognition algorithm, as shown in fig. 2.
It will be appreciated that principal component analysis (Principal Component Analysis, PCA) is a linear dimension reduction algorithm that is capable of converting high-dimensional data to low-dimensional data while preserving the primary characteristics of the original data. The basic idea of the PCA algorithm is to find a new set of orthogonal bases such that the projection of the data onto the set of orthogonal bases has the greatest variance. This set of orthogonal bases can reflect the variability and diversity of the data to the greatest extent. The set of orthogonal bases is the principal component of the data, and is composed of the feature vectors of the original data, and the features of the original data can be effectively expressed by the set of orthogonal bases. Principal component analysis is not a classification algorithm per se, but a dimension reduction technique. Its main objective is to find the main features in the data and reduce the high-dimensional data to lower dimensions.
In some embodiments, the main component analysis is performed on the target gas biogas data to obtain a preset number of gas biogas feature information, which specifically includes:
calculating a covariance matrix of the target gas biogas data to obtain a monitoring data covariance matrix;
and obtaining the characteristic information of the preset number of gas biogas according to the covariance matrix of the monitoring data.
Specifically, the main component analysis step includes: reading target gas methane data, calculating covariance matrix of the data, and marking the covariance matrix as monitoring data covariance matrix for convenience of description, and selecting a preset number of main components from the monitoring data covariance matrix to obtain gas methane characteristic information. The gas biogas characteristic information comprises main gas component characteristics and concentration change characteristics.
In one embodiment, the covariance matrix is calculated as follows:
where n represents the number of data samples, j represents the j-th sampling time, i is i=1, 2, and 3 … l+3, which is a certain attribute (temperature, pressure, concentration, etc.) of all the acquired data. S is S i Representing covariance of the ith column data, S is a covariance matrix, its dimension is (k+3) x (k+3),represents the average of all data column i, since the target gas biogas data is normalized to a gaussian distribution in the pretreatment +. >Covariance calculation can be reduced to:
after the covariance matrix of the monitoring data is calculated, the pre-preset number of combustible gas components with the largest contribution degree of the main components in the target gas biogas data are selected as the main components, and the main gas component characteristics and the concentration change characteristics of the main components are reserved as the gas biogas characteristic information.
According to the method, the covariance matrix is calculated, the data matrix with huge sampling times of the same target monitoring point is converted into the finite dimension matrix of (k+3) x (k+3), and the change characteristics of the data are reserved. By principal component analysis, the varying correlation between the sampled data is preserved. It should be noted in particular that by principal component analysis, the linear characteristics of the target gas biogas data can be highlighted. Since the concentration variation of the gas leakage is relatively linear in the actual operation process, the characteristic can be effectively maintained by the principal component analysis.
Step 140: inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine.
And inputting the dimensionality reduction data obtained after the principal component analysis into a gas methane discrimination model to obtain a discrimination result, and completing the discrimination of the gas methane. It should be noted that the gas biogas discrimination model is based on a probability classification vector machine and is obtained by training a large number of gas biogas data samples.
It can be appreciated that the probabilistic classification vector machine (Probabilistic Classification Vector Machine, PCVM) is a bayesian-theory-based classification algorithm that combines the advantages of a Support Vector Machine (SVM) and a correlation vector machine (RVM) to provide a sparse bayesian solution. The PCVM algorithm can be used for two-class or multi-class problems, which can output the probability of each class, thereby providing more information and uncertainty metrics. The basic idea of the PCVM algorithm is to map data into a high-dimensional feature space and then find an optimal hyperplane in the feature space so that the separation between different classes of data is maximized. Unlike the SVM algorithm, the PCVM algorithm does not need to use a relaxation variable and a soft interval to process noise and nonlinear data, but uses a gaussian prior distribution to constrain the parameters of the hyperplane, thereby achieving sparsity and regularization. The PCVM algorithm may also use a kernel function to process non-linearly separable data, thereby increasing the flexibility and adaptability of the model.
Based on a pre-constructed probability classification vector machine, a large number of gas biogas data samples are utilized to obtain parameters and threshold values of the classifier through training and testing the data, so that a gas biogas discrimination model is obtained.
In some embodiments, based on a probability classification vector machine, a large number of gas biogas data samples are utilized for training to obtain a gas biogas discrimination model, which specifically comprises the following steps:
preprocessing the gas methane data sample to obtain a target sample;
classifying the target sample according to sampling points, and carrying out principal component analysis on the target sample for each sampling point to obtain a covariance matrix;
integrating the covariance matrixes of all the sampling points to obtain a training data set;
dividing the training data set into positive and negative sample data sets according to a predetermined classification mode, and updating parameters of a pre-constructed probability classification vector machine by using the positive and negative sample data sets in a mode of maximizing a pre-selected loss function to obtain a gas methane discrimination model;
the predetermined classification mode comprises that the gas leakage is positive, and the gas non-leakage is negative.
In particular, in this embodiment, the gas biogas data samples are also obtained by collection. Specifically, for gas and methane discrimination of a certain target monitoring area, firstly, gas and methane data samples are collected at each target monitoring point by utilizing collection equipment according to a preset sampling rule, and a gas and methane discrimination model is trained by utilizing the collected gas and methane data samples.
In the actual operation process, the gas methane data sample comprises physical and chemical characteristic parameters of the gas methane to be distinguished, such as ambient temperature, ambient air pressure, alarm time, combustible gas components and concentration fractions at the alarm time, and the like. In a specific embodiment, at each target detection point of the target monitoring area, the combustible gas component g including the ambient temperature T, the ambient air pressure P, the alarm time T and the alarm time is obtained at each acquisition time point 1 ,g 2 ,…,g k Corresponding concentration c of (2) 1 ,c 2 ,…,c k Is a data point of (c). After the data points are obtained, the data points are combined into a vector with the dimension of i+2 according to the acquisition time, and each vector is a gas methane data sample. In one embodiment, the data points are determined in terms of ambient temperature, ambient pressure, and combustible gas composition g 1 ,g 2 ,…,g k Corresponding concentration c of (2) 1 ,c 2 ,…,c k Form set vectors [ T, P, c ] 1 ,c 2 ,…,c k ]Representing the physicochemical properties and the change conditions of the methane-burning sample in the data acquisition position in the period of time. Where k is the combustible gas composition fraction.
Further, preprocessing is carried out on the acquired gas methane data samples, wherein the preprocessing comprises sampling frequency alignment processing, normalization processing and noise reduction processing.
Specifically, in some embodiments, bilinear interpolation is performed on the acquired gas biogas data samples, different time measurement data are aligned, and then a data set consisting of all vectors is subjected to Z-score normalization to obtain target samples, so that preprocessing is completed.
And then, carrying out principal component analysis on the target sample of each sampling point to obtain a covariance matrix S of each sampling point, and integrating the covariance matrices S obtained by the multiple sampling points to obtain a training data set.
And inputting the positive and negative samples of the training data set into a pre-constructed probability classification vector machine according to a pre-determined classification mode for training. Specifically, in the training stage, positive and negative samples of a training data set are utilized on the basis of a preselected loss function, parameters of a probability classification vector machine are solved in a mode of maximizing the loss function, and the probability classification vector machine is updated by utilizing the solved parameters, so that a gas methane discrimination model is obtained.
In some embodiments, the training data set is divided into positive and negative sample data sets according to a predetermined classification mode, the positive and negative sample data sets are utilized, the parameters of the probability classification vector machine are solved in a mode of maximizing a loss function, and the probability classification vector machine is updated by utilizing the solved parameters, so that the gas methane discrimination model is obtained.
In a specific embodiment, a log-likelihood function is selected as a loss function, and a Bayesian iterative solution method is used for solving the maximized loss function to obtain a training parameter result.
In some embodiments, the decision function of the gas biogas discrimination model is defined as:
wherein alpha is i And beta i Is a model parameter obtained by training, K (x, x i ) Is a kernel function, beta 0 Is a variable for controlling the position of the classification face, x is an input value, x i Is a data value in the training set.
The gas leakage is defined as a positive example, otherwise, the gas leakage is a negative example, and corresponding conditional probability and posterior probability can be obtained and recorded as a probability calculation formula:
P(y＝1|x)+P(y＝-1|x)＝1
wherein P (y= 1|x) represents the probability of gas leakage, P (y= -1|x) represents the probability of not gas leakage, and f (x) represents the decision function of the gas biogas discrimination model.
Further, after training is completed, the method further comprises: and selecting a part of the training data set as a test data set, extracting test data from the test data set, inputting the test data into the gas methane distinguishing model, calculating the probability that the gas methane data sample in the test data belongs to the positive example and the negative example according to the posterior probability, and comparing and verifying the probability value to judge the accuracy of the gas methane distinguishing model.
It can be appreciated that since the probability classification vector machine is a probabilistic model, the loss function and the solution process used by this model are both probabilistic models. Its output is probabilistic, e.g., 69% probability is gas leakage, etc. For the hard output of the traditional method, the gas biogas discrimination model in the gas biogas discrimination method based on the principal component analysis and the vector machine is more convenient for a decision maker.
After training, in some embodiments, the inputting the characteristic information of the gas biogas into a pre-trained gas biogas discrimination model to perform classification discrimination to obtain a discrimination result specifically includes:
calculating the probability of gas leakage and/or gas non-leakage by using a probability calculation formula according to the gas biogas characteristic information to obtain a calculation result;
and comparing and verifying the calculation result to obtain a discrimination result.
Specifically, after training is completed, the characteristic information of the gas methane is input, the probability that the gas methane data to be judged belong to the positive example and the negative example is calculated according to posterior probability in a probability calculation formula, the calculation results are compared and verified, and the judgment result is output.
Still further, in some embodiments, the computed result is validated against the actual result.
The gas methane distinguishing method based on the principal component analysis and the vector machine provided by the invention adopts a mode of combining the principal component analysis algorithm and the probability classification vector machine, has high data processing speed and high distinguishing accuracy of gas and methane. In one embodiment, the method for distinguishing the gas and the biogas based on the principal component analysis and the vector machine can distinguish the gas and the biogas by analyzing experimental data, and has higher accuracy rate which reaches more than 90 percent. The gas methane distinguishing method based on the principal component analysis and the vector machine can be flexibly adjusted according to the actual application scene, can be optimized by combining other algorithms, has strong adaptability and expansibility, and has obvious implementation effect and wide application prospect in the aspects of gas monitoring and early warning and the like.
The invention also comprises a specific embodiment for carrying out primary discrimination by adopting the provided gas methane discrimination method based on the principal component analysis and the vector machine.
The implementation requirements of this embodiment are as follows:
technical equipment: this embodiment requires the use of a computer, sensors and related software. Computers need to be equipped with sufficient processor, memory and hard disk capacity to support large-scale data processing. The sensor needs to have high accuracy and stability to ensure that the acquired data is accurate and reliable. The related software needs to support the implementation of principal component analysis algorithms and probabilistic classification vector machine algorithms, such as Python, etc.
Personnel: the present embodiment requires personnel with related technical knowledge and practical experience, including data acquisition, data processing, algorithm implementation, and the like. The advice team includes at least one data analyst and one software engineer.
Environment: this example needs to be performed in a laboratory or field environment. Laboratory environment needs to be kept clean and tidy, and interference to experimental data is avoided. The field environment needs to have gas methane sample collection conditions, so that the collected samples are representative.
And (3) data acquisition: the influence of environmental factors such as temperature, humidity, air pressure and the like need to be strictly controlled in the data acquisition process. A sufficient number of samples need to be collected to ensure the integrity and reliability of the data. At the same time, the sample needs to be marked for subsequent data processing and analysis.
The algorithm is realized: this embodiment requires the implementation of a combination of principal component analysis algorithms and probabilistic classification vector machine algorithms. When the algorithm is realized, the algorithm parameters need to be optimized so as to improve the accuracy and stability of the algorithm. Meanwhile, the algorithm results need to be evaluated and analyzed in order to further optimize the performance of the algorithm.
The steps are as follows, as shown in FIG. 3:
s1: selecting target monitoring points in a clean and tidy gas biogas collection environment, and collecting gas biogas data to be distinguished according to a preset sampling rule, wherein the gas biogas data to be distinguished comprises the concentration of combustible components of the gas biogas to be distinguished, the ambient temperature, the ambient air pressure and the alarm time;
s2: preprocessing the acquired gas methane data to be distinguished to obtain target gas methane data; performing principal component analysis on the target gas biogas data to obtain characteristic information of a preset number of gas biogas;
s3: training by utilizing gas methane data to be distinguished based on a probability classification vector machine to obtain a gas methane distinguishing model;
s4: inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification and distinguishing, and outputting distinguishing results.
According to the gas methane distinguishing method based on the principal component analysis and the vector machine, gas methane data to be distinguished are collected at target monitoring points according to a preset sampling rule, and the gas methane data to be distinguished comprise physical and chemical characteristic parameters of the gas methane to be distinguished; preprocessing the gas methane data to be distinguished to obtain target gas methane data; performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information; inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time. According to the invention, the acquired data are preprocessed by taking the data indexes such as the ambient temperature, the ambient air pressure, the alarm time, the combustible gas components and the concentration fraction into consideration, the dimension of the data set is reduced by using a principal component analysis algorithm, and the dimension-reduced data are classified and judged by using a trained gas methane judgment model based on principal component analysis and a vector machine, so that the judgment accuracy and the generalization capability of gas methane are improved. The gas biogas discrimination method based on the principal component analysis and the vector machine has the advantages of high classification accuracy, high classification speed, wide application range and the like, can be applied to the field of gas safety and emergency management, is used for identifying gas leakage and biogas interference, reduces the false alarm rate of a gas monitor, enables gas enterprises to rapidly judge the gas pipeline leakage condition, eliminates hidden danger and dangerous situation in time, and ensures people life and property safety and city safety operation.
The gas biogas distinguishing device based on the principal component analysis and the vector machine provided by the invention is described below, and the gas biogas distinguishing device based on the principal component analysis and the vector machine described below and the gas biogas distinguishing method based on the principal component analysis and the vector machine described above can be correspondingly referred to each other. Fig. 4 is a schematic structural diagram of a gas biogas discriminating apparatus based on principal component analysis and a vector machine, and as shown in fig. 4, the apparatus includes:
the acquisition unit 410 is used for acquiring gas biogas data to be distinguished according to a preset sampling rule at a target monitoring point, wherein the gas biogas data to be distinguished comprises physical and chemical characteristic parameters of the gas biogas to be distinguished;
the processing unit 420 is configured to pre-process the gas biogas data to be determined to obtain target gas biogas data;
the analysis unit 430 is configured to perform principal component analysis on the target gas biogas data to obtain a preset number of gas biogas feature information;
the distinguishing unit 440 is configured to input the characteristic information of the gas biogas to a pre-trained gas biogas distinguishing model for classification and distinguishing, so as to obtain a distinguishing result; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time.
Based on the foregoing embodiment, in the apparatus, the preset sampling rule includes:
sampling according to a first preset frequency under the condition that the concentration of the gas methane to be distinguished is lower than or equal to a preset concentration threshold value;
and under the condition that the concentration of the gas methane to be distinguished is higher than a preset concentration threshold value, sampling is carried out according to a second preset frequency.
Based on the above embodiment, in the device, the main component analysis is performed on the target gas biogas data to obtain a preset number of gas biogas feature information, which specifically includes:
calculating a covariance matrix of the target gas biogas data to obtain a monitoring data covariance matrix;
and obtaining the characteristic information of the preset number of gas biogas according to the covariance matrix of the monitoring data.
Based on the above embodiment, in the device, based on a probability classification vector machine, a large number of gas biogas data samples are utilized for training to obtain a gas biogas discrimination model, which specifically comprises:
preprocessing the gas methane data sample to obtain a target sample;
classifying the target sample according to sampling points, and carrying out principal component analysis on the target sample for each sampling point to obtain a covariance matrix;
Integrating the covariance matrixes of all the sampling points to obtain a training data set;
dividing the training data set into positive and negative sample data sets according to a predetermined classification mode, and updating parameters of a pre-constructed probability classification vector machine by using the positive and negative sample data sets in a mode of maximizing a pre-selected loss function to obtain a gas methane discrimination model;
the predetermined classification mode comprises that the gas leakage is positive, and the gas non-leakage is negative.
Based on the above embodiment, in the apparatus, the preprocessing includes sampling frequency alignment processing, normalization processing, and noise reduction processing.
Based on the above embodiment, in the device, the decision function of the gas biogas discrimination model includes:
wherein alpha is i And beta i Is a model parameter obtained by training, K (x, x i ) Is a kernel function, beta 0 Is a variable for controlling the position of the classification face, x is an input value, x i Is a data value in the training set.
Based on the above embodiment, in the apparatus, the determining unit 440 specifically includes:
calculating the probability of gas leakage and/or gas non-leakage by using a probability calculation formula according to the gas biogas characteristic information to obtain a calculation result;
And comparing and verifying the calculation result with a preset threshold value to obtain a discrimination result.
Based on the above embodiment, in the apparatus, the probability calculation formula includes:
P(y＝1|x)+P(y＝-1|x)＝1
wherein P (y= 1|x) represents the probability of gas leakage, P (y= -1|x) represents the probability of not gas leakage, and f (x) represents the decision function of the gas biogas discrimination model.
According to the gas methane distinguishing device based on the principal component analysis and the vector machine, gas methane data to be distinguished are collected at target monitoring points according to a preset sampling rule, and the gas methane data to be distinguished comprise physical and chemical characteristic parameters of the gas methane to be distinguished; preprocessing the gas methane data to be distinguished to obtain target gas methane data; performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information; inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time. According to the invention, the acquired data are preprocessed by taking the data indexes such as the ambient temperature, the ambient air pressure, the alarm time, the combustible gas components and the concentration fraction into consideration, the dimension of the data set is reduced by using a principal component analysis algorithm, and the dimension-reduced data are classified and judged by using a trained gas methane judgment model based on principal component analysis and a vector machine, so that the judgment accuracy and the generalization capability of gas methane are improved.
Fig. 5 illustrates a physical schematic diagram of an electronic device, as shown in fig. 5, which may include: processor 510, communication interface (Communications Interface) 520, memory 530, and communication bus 540, wherein processor 510, communication interface 520, memory 530 complete communication with each other through communication bus 540. Processor 510 may invoke logic instructions in memory 530 to perform a gas biogas discrimination method based on principal component analysis and a vector machine, the method comprising: collecting gas methane data to be distinguished according to a preset sampling rule at a target monitoring point, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished; preprocessing the gas methane data to be distinguished to obtain target gas methane data; performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information; inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time.
Further, the logic instructions in the memory 530 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product, where the computer program product includes a computer program, where the computer program can be stored on a non-transitory computer readable storage medium, and when the computer program is executed by a processor, the computer can execute the method for determining gas biogas based on principal component analysis and vector machine provided by the above methods, and the method includes: collecting gas methane data to be distinguished according to a preset sampling rule at a target monitoring point, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished; preprocessing the gas methane data to be distinguished to obtain target gas methane data; performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information; inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time.
In still another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the method for discriminating biogas of gas based on principal component analysis and a vector machine provided by the above methods, the method comprising: collecting gas methane data to be distinguished according to a preset sampling rule at a target monitoring point, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished; preprocessing the gas methane data to be distinguished to obtain target gas methane data; performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information; inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. The gas methane distinguishing method based on the principal component analysis and the vector machine is characterized by comprising the following steps of:
collecting gas methane data to be distinguished according to a preset sampling rule at a target monitoring point, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished;
preprocessing the gas methane data to be distinguished to obtain target gas methane data;
performing principal component analysis on the target gas biogas data to obtain a preset number of gas biogas characteristic information;
inputting the characteristic information of the gas methane into a pre-trained gas methane distinguishing model for classification distinguishing to obtain distinguishing results; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time.
2. The method for discriminating gas biogas based on principal component analysis and vector machine according to claim 1 wherein said preset sampling rule comprises:
sampling according to a first preset frequency under the condition that the concentration of the gas methane to be distinguished is lower than or equal to a preset concentration threshold value;
and under the condition that the concentration of the gas methane to be distinguished is higher than a preset concentration threshold value, sampling is carried out according to a second preset frequency.
3. The method for discriminating gas biogas based on principal component analysis and vector machine according to claim 1 wherein said performing principal component analysis on said target gas biogas data to obtain a preset number of gas biogas feature information specifically comprises:
calculating a covariance matrix of the target gas biogas data to obtain a monitoring data covariance matrix;
and obtaining the characteristic information of the preset number of gas biogas according to the covariance matrix of the monitoring data.
4. The gas biogas discrimination method based on principal component analysis and a vector machine according to claim 1, wherein the gas biogas discrimination model is trained by using a large number of gas biogas data samples based on a probability classification vector machine, and specifically comprises:
Preprocessing the gas methane data sample to obtain a target sample;
classifying the target sample according to sampling points, and carrying out principal component analysis on the target sample for each sampling point to obtain a covariance matrix;
integrating the covariance matrixes of all the sampling points to obtain a training data set;
dividing the training data set into positive and negative sample data sets according to a predetermined classification mode, and updating parameters of a pre-constructed probability classification vector machine by using the positive and negative sample data sets in a mode of maximizing a pre-selected loss function to obtain a gas methane discrimination model;
the predetermined classification mode comprises that the gas leakage is positive, and the gas non-leakage is negative.
5. The method for discriminating gas biogas based on principal component analysis and vector machine according to claim 1 or 4 wherein said preprocessing includes sampling frequency alignment processing, normalization processing and noise reduction processing.
6. The method for discriminating gas biogas based on principal component analysis and vector machine according to claim 1 or 4 wherein said decision function of said gas biogas discrimination model comprises:
wherein alpha is i And beta i Is a model parameter obtained by training, K (x, x i ) Is a kernel function, beta 0 Is a variable for controlling the position of the classification face, x is an input value, x i Is a data value in the training set.
7. The method for discriminating gas biogas based on principal component analysis and vector machine according to claim 6 wherein said inputting said gas biogas feature information into a pre-trained gas biogas discrimination model for classification discrimination, obtaining discrimination results, specifically comprising:
calculating the probability of gas leakage and/or gas non-leakage by using a probability calculation formula according to the gas biogas characteristic information to obtain a calculation result;
and comparing and verifying the calculation result to obtain a discrimination result.
8. The method for discriminating gas biogas based on principal component analysis and vector machine according to claim 7 wherein said probability calculation formula comprises:
P(y＝1|x)+P(y＝-1|x)＝1
wherein P (y= 1|x) represents the probability of gas leakage, P (y= -1|x) represents the probability of not gas leakage, and f (x) represents the decision function of the gas biogas discrimination model.
9. The utility model provides a gas marsh gas discriminating device based on principal component analysis and vector machine which characterized in that includes:
the acquisition unit is used for acquiring gas methane data to be distinguished according to a preset sampling rule at target monitoring points, wherein the gas methane data to be distinguished comprises physical and chemical characteristic parameters of the gas methane to be distinguished;
The processing unit is used for preprocessing the gas methane data to be distinguished to obtain target gas methane data;
the analysis unit is used for carrying out principal component analysis on the target gas biogas data to obtain the characteristic information of a preset number of gas biogas;
the judging unit is used for inputting the characteristic information of the gas methane into a pre-trained gas methane judging model to carry out classification judgment to obtain a judging result; the gas methane distinguishing model is obtained by training a large number of gas methane data samples based on a probability classification vector machine; the physical and chemical characteristic parameters of the gas biogas to be distinguished at least comprise the components and concentration fractions of the combustible gas at the time of the alarm, the ambient temperature, the ambient air pressure and the alarm time.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the gas biogas discrimination method based on principal component analysis and a vector machine according to any one of claims 1 to 8 when executing the program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311341952.5A CN117493759A (en) | 2023-10-17 | 2023-10-17 | Gas methane distinguishing method and device based on principal component analysis and vector machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311341952.5A CN117493759A (en) | 2023-10-17 | 2023-10-17 | Gas methane distinguishing method and device based on principal component analysis and vector machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117493759A true CN117493759A (en) | 2024-02-02 |
Family
ID=89669801
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311341952.5A Pending CN117493759A (en) | 2023-10-17 | 2023-10-17 | Gas methane distinguishing method and device based on principal component analysis and vector machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117493759A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118151614A (en) * | 2024-05-08 | 2024-06-07 | 镇江贝思特有机活性肥料有限公司 | Control method and control system for adjusting organic material fermentation process by septic tank water |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600140A (en) * | 2016-12-07 | 2017-04-26 | 河海大学常州校区 | Improved support vector machine-based gas pipeline fault prediction and early warning system and method |
-
2023
- 2023-10-17 CN CN202311341952.5A patent/CN117493759A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600140A (en) * | 2016-12-07 | 2017-04-26 | 河海大学常州校区 | Improved support vector machine-based gas pipeline fault prediction and early warning system and method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118151614A (en) * | 2024-05-08 | 2024-06-07 | 镇江贝思特有机活性肥料有限公司 | Control method and control system for adjusting organic material fermentation process by septic tank water |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110263846B (en) | Fault diagnosis method based on fault data deep mining and learning | |
| CN111882446B (en) | Abnormal account detection method based on graph convolution network | |
| CN109302410B (en) | Method and system for detecting abnormal behavior of internal user and computer storage medium | |
| CN112179691B (en) | Mechanical equipment running state abnormity detection system and method based on counterstudy strategy | |
| CN112966259B (en) | Operation and maintenance behavior security threat assessment method and equipment for power monitoring system | |
| CN117493759A (en) | Gas methane distinguishing method and device based on principal component analysis and vector machine | |
| CN103412557A (en) | Industrial fault detection and diagnostic method suitable for nonlinear process on-line monitoring | |
| Mechefske et al. | Fault detection and diagnosis in low speed rolling element bearings Part II: The use of nearest neighbour classification | |
| CN113516228B (en) | Network anomaly detection method based on deep neural network | |
| CN111738044A (en) | Campus violence assessment method based on deep learning behavior recognition | |
| CN116451142A (en) | Water quality sensor fault detection method based on machine learning algorithm | |
| CN117150283A (en) | Emergency environmental event safety early warning method based on big data analysis | |
| CN116384736A (en) | Smart city risk perception method and system | |
| CN117421616B (en) | Mine shaft detection system and method | |
| CN113704201A (en) | Log anomaly detection method and device and server | |
| CN116707918A (en) | Network security situation assessment method based on CBAM-EfficientNet anomaly detection | |
| Li et al. | Fault diagnosis and prediction of complex system based on Hidden Markov model | |
| CN116311739A (en) | Multi-sensor fire detection method based on long-short-term memory network and environment information fusion | |
| CN113657726B (en) | Personnel risk analysis method based on random forest | |
| CN106530199B (en) | Multimedia integration steganalysis method based on window type hypothesis testing | |
| Zhang et al. | An unsupervised leak detection method with aggregating prediction and reconstruction along projection pathway for natural gas gathering pipelines | |
| CN117579393B (en) | Information terminal threat monitoring method, device, equipment and storage medium | |
| CN112070761B (en) | Prawn freshness nondestructive testing method based on deep learning | |
| CN115831339B (en) | Medical system risk management and control pre-prediction method and system based on deep learning | |
| CN113904801B (en) | Network intrusion detection method and 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 |