CN109325470B - Intelligent underground working face operation type identification method based on gas concentration parameters - Google Patents
Intelligent underground working face operation type identification method based on gas concentration parameters Download PDFInfo
- Publication number
- CN109325470B CN109325470B CN201811243460.1A CN201811243460A CN109325470B CN 109325470 B CN109325470 B CN 109325470B CN 201811243460 A CN201811243460 A CN 201811243460A CN 109325470 B CN109325470 B CN 109325470B
- Authority
- CN
- China
- Prior art keywords
- gas concentration
- signal sequence
- operation type
- local
- working face
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Sampling And Sample Adjustment (AREA)
Abstract
The invention discloses an underground working face operation type intelligent identification method based on gas concentration parameters. The method comprises the steps of firstly obtaining working face gas concentration data from a mine safety monitoring system, storing the working face gas concentration data into a gas parameter database, carrying out adaptive noise reduction filtering processing on the gas concentration data in the database, then establishing a historical gas concentration data sample set on the basis of operation type division and sensitive characteristic parameter extraction, constructing an adaptive modular rough neural network prediction model, and identifying the corresponding operation type by utilizing the constructed prediction model. The method has obvious noise reduction effect, can effectively remove noise, simultaneously can keep and obtain effective information in a gas concentration time sequence, can realize the analysis and identification of the working type of a working face, and can better realize the deep excavation and secondary utilization of gas concentration data.
Description
Technical Field
The invention relates to the field of monitoring and control of mine gas disasters, in particular to an intelligent analysis and identification method for operation types of a mine excavation working face based on gas concentration data characteristics.
Background
China is the first major coal country in the world, and according to the outline of the middle-term development of energy planning (2004) -2020 in China, it is clearly pointed out that China will persist in using coal as the main energy in the future period. Currently, coal mining in China has entered deep mining periods, and mining depths are increasing year by year. In the deep mining period, the pressure of mine gas is increased, the ground temperature is increased, the ground stress is increased, and the mine disasters are increasingly serious. Coal mine accident disasters seriously impair the safe coal production situation in China, and hinder the healthy, sustained and stable development of the coal industry. The coal mine accident disaster prevention and control safety management strategy is perfected, the coal mine accident disaster prevention and control consciousness idea is deepened, the coal mine accident disaster supervision and control technology is promoted, and the coal mine accident disaster prevention and control safety management strategy has important significance for ensuring coal mine safety production and promoting comprehensive and healthy development of national economy.
The major disasters such as gas accidents, outburst, rock burst, coal dust accidents, flood disasters, fire disasters, roof accidents and the like in coal mine production mostly occur on a mining working face, are closely related to operations such as mining, drilling and the like, and have more accidents during the tunneling operation. Mine safety monitoring is an important means for mine accident disaster prevention and control, effective information of an underground working face is timely acquired and effectively utilized, and the working quality of accident prevention and disaster control is favorably improved. Currently, the main focus of the domestic and foreign research on the working environment monitoring technology of the working face is on monitoring the working environment parameters such as gas concentration, wind speed and temperature. The working face is tunneled in the high gas coal roadway, the gas condition of the working face is an environmental parameter closely related to the operation condition, and the monitoring system is operated for a long time, so that a large amount of gas monitoring data are accumulated on a coal mine site. Researchers at home and abroad pay attention to the utilization of gas monitoring data, and carry out a great deal of research by advanced technical means such as data mining, artificial intelligence and the like, for example, the gas concentration is predicted by using a gas parameter sequence, and outburst danger is judged by using gas emission quantity, so that the research has great promotion effect on mine safety monitoring work.
However, the previous people have less direct attention to the operation condition in front of the working face, the operation and construction condition in the well is mainly derived from the work scheduling plan or artificial statistical record, errors and carelessness exist inevitably, and the timeliness is greatly limited. At present, as a normalized means for accident disaster prevention and control, mine safety monitoring is developed in the direction of diversification, synthesis and intellectualization of functions such as disaster diagnosis, operation scheduling and production management by pure safety monitoring at the present stage, and has the characteristics of non-contact, dynamism, multi-parameter synthesis and module integration. The influence of mining operation is often the direct reason of many accidents, and the operation process of the working face is actively, accurately and rapidly identified through related information in the pit and is brought into a system database, so that the method has great practical significance for the work such as accident mechanism analysis, disaster diagnosis, emergency treatment, accident investigation and the like, and provides basic information support for the development of the coal mine safety monitoring and early warning technology.
Disclosure of Invention
Aiming at the defects of the application of the prior art, the invention provides an underground working face operation type intelligent identification method based on mine gas concentration parameters. The method is characterized in that concentration data of gas monitoring is specially preprocessed, noise signals of an original time sequence are effectively reduced, effective gas concentration sequence characteristics are reserved and highlighted, and on the basis of a secondary time, the operation process type of the underground working face is intelligently identified by constructing a gas parameter sensitive parameter system and an operation type sample training library and adopting a multilayer feedforward neural network trained by an error reverse propagation algorithm.
The technical scheme of the invention specifically comprises the following main steps:
step 1, acquiring an original signal sequence x (t) of gas concentration through an underground gas concentration sensor, and storing the original signal sequence x (t) as historical data into a gas concentration database;
step 3, performing special self-adaptive filtering noise reduction processing based on local mean decomposition on the gas concentration data in the gas concentration database to obtain a noise-reduced gas concentration sequence;
step 5, constructing an operation type neural network intelligent recognition model of the underground working face, and inputting the operation type neural network intelligent recognition model into the model for training according to the established gas concentration database and the sensitive characteristic parameters;
and 6, intelligently identifying the operation type of the newly acquired gas concentration data to be detected by using the trained model, and obtaining an operation type result corresponding to the gas concentration data to be detected.
The method creatively utilizes the gas concentration parameter to judge the operation type of the underground working face, plays a role of data, wherein the self-adaptive filtering noise reduction processing makes special noise reduction processing so that the gas concentration parameter which seems to be irrelevant to the operation type can represent the operation type of the underground working face.
In the step 2, the operation types of the working surface are divided into blasting, coal cutting, coal dropping, shift switching, drilling operation, temporary support, anchor rod installation, overhaul and other unrelated operations. Other operation forms or regular combination of several operation types can be added, which are mainly determined according to the practical requirements of safety monitoring and early warning and the feasibility of operation type identification.
In the step 4, the sensitive characteristic parameters include a mean value, a mean square error, a variance, a range, a difference characteristic value and the like of the gas concentration data. In particular, a specific product component may also be selected as the sensitive characteristic parameter. In specific implementation, or improved parameters or other secondary mining parameters based on such parameters; other sensitive characteristic parameters including product component sequences and the like generated during data noise reduction preprocessing may also be used.
In the step 3, the adaptive filtering and denoising specifically comprises:
3.1, obtaining all extreme points n of the original signal sequence x (t)iI is 1,2,3 …, I is the number of extreme points, and then according to the extreme points, the average value of all adjacent local extreme points is obtained as the local average value mi＝(ni+ni+1) (ii)/2 and the local envelope estimate as a local envelope value ai＝(ni-ni+1)/2；
3.2, average value m for all local areasiAnd all local envelope values aiPerforming multiple smoothing treatments by adopting a sliding average method to respectively obtain local mean value functions m'1(t)、m′2(t), … and local envelope function a'1(t)、a′2(t), …, subscripts 1 and 2 represent the function results obtained by the first and second smoothing processes;
3.3, each time of smoothing in step 3.2 is followed by processing iteration until a pure frequency-modulated signal s is obtainedk(t)：
After the first smoothing in step 3.2, separating the local mean function m 'obtained by the first smoothing from the original signal sequence x (t)'1(t) leaving the first intermediate separation signal sequence h1(t)＝x(t)-m′1(t), and then obtaining the local envelope function a 'according to the first smoothing processing'1(t) demodulating to obtain a first intermediate demodulated signal sequence s1(t)＝h1(t)/a′1(t)；
Then the first intermediate demodulated signal sequence s1(t) repeating the steps as the original signal sequence x (t) in combination with the local mean function m 'obtained by the second smoothing'1(t) and local envelope function a'1(t) calculating to obtain a second intermediate separation signal sequence and a second intermediate demodulation signal sequence;
the iterative processing is carried out until the K-th intermediate demodulation signal sequence satisfies s is more than or equal to-1K(t) is less than or equal to 1 and local area envelope function a 'of the next time'K+1If (t) is 1 and K is the number of iterations, the K-th intermediate demodulation signal sequence s is obtainedK(t) stopping iterative processing as a pure frequency modulation signal; the specific iterative process is represented as:
in the formula, K is the total number of times of iteration;
3.4, mixing the pure frequency-modulated signal sK(t) results of an iterative processHas a local envelope function a'k(t) multiplying to obtain an envelope signal A of a first product component (PF)1(t)＝a′1(t)a′2(t)…a′K(t), K is 1,2,3 …, K, and the envelope signal a of the first product component is then used1(t) and its corresponding pure FM signal sK(t) multiplying to obtain a first product component PF of the original signal sequence1(t)＝A1(t)sk(t)；
3.5, PF the first product component1(t) separating from the original signal sequence x (t) and retaining the first residual signal u1(t) and with the first residual signal u1(t) repeating the above steps 3.1-3.4 for the original signal sequence x (t) until all product components are separated by J cycles and the J-th residual signal u is satisfiedJ(t) is a monotonic function whereby the decomposed expression of the following original signal sequence x (t) is constructed:
wherein J represents the number of the iterative processing in steps 3.1 to 3.4, J is 1,2,3 …, J represents the total number of iterative processing in steps 3.1 to 3.4;
the original gas concentration signal is transformed into a sum of a set of product components and a residual component.
And 3.6, carrying out correlation analysis on each obtained product component and the original signal sequence x (t), taking the energy values of all the product components to form a global minimum value of the sequence as a frequency modulation boundary point of high and low frequency components, taking all high-frequency components of the multiplication integral quantity before the frequency modulation boundary point as noise product components, and carrying out reconstruction processing on the signal after removing the noise product components to obtain a noise-reduced gas concentration sequence dx (t).
The reconstruction processing of the signal means that the product components are added.
The operation type neural network intelligent recognition model adopts a rough neural network (RMNN) combining a rough set theory and an artificial neural network to realize the intelligent recognition of the working type of the underground mining working surface, the rough neural elements of the neural network input layer are utilized to respectively calculate the sensitive characteristic parameters of gas concentration data in each operation type, the fuzzy neural elements are adopted to replace the common neural elements, the sensitive characteristic parameters of partial historical gas concentration data are adopted as training samples, and the intelligent recognition of the operation type corresponding to the recent gas data is completed.
Aiming at the problem that the prediction accuracy of a prediction model is influenced by a large amount of noise in current mine gas concentration data, the method decomposes historical gas concentration sequence data into a plurality of pure frequency modulation signal components through local mean decomposition, and then self-adaptive threshold noise reduction filtering is carried out on decomposed frequency modulation signal functions through a low-pass filter, so that effective information in a gas concentration time sequence can be retained to the maximum extent while noise is effectively removed, and the noise removal effect is ideal.
The invention has the beneficial effects that:
aiming at the defects of time delay, inaccuracy and the like of manual recording operation procedures and types, the invention constructs a self-adaptive modular rough neural network prediction model which can be used for realizing the analysis and identification of the operation types of the working face based on the statistical parameters of various gas concentration parameters or the comprehensive parameters formed by the statistical parameters according to the gas concentration sequence, and can better realize the deep excavation and secondary utilization of the gas concentration data.
The method has obvious noise reduction effect, and can retain and obtain effective information in the gas concentration time sequence while effectively removing noise.
The method can realize accurate judgment of the actual underground working type and meet the development requirements of the current mine safety monitoring system.
Drawings
FIG. 1 is a flow chart of a method according to an embodiment of the present invention;
FIG. 2 is a time-series diagram of original gas concentration data according to an embodiment of the present invention;
FIG. 3 is a time-series diagram of noise-reduced gas concentration data according to an embodiment of the present invention;
fig. 4 is a diagram of operation type classification of a coarse neural network model based on comprehensive parameters according to an embodiment of the present invention.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
The embodiment of the invention is as follows:
the method comprises the steps of acquiring gas concentration data of a coal mine excavation working face through a gas concentration sensor, using the data as an embodiment research object, manually tracking and recording operation types in a period section, using the data as a sample library, and specifically implementing the flow as shown in figure 1.
Step 1, a mine safety monitoring system is used for automatically and continuously acquiring dynamic time series data of gas emission of a tunneling face, in the embodiment, the mine adopts gas data acquired by a KJ90 (substation-F16) safety monitoring system, a gas concentration sensor is a KG9701A methane sensor, and the acquisition frequency of the gas concentration data is one data per 5 minutes. Recording an original gas concentration sequence x (t) obtained by an underground gas concentration sensor, storing the original gas concentration sequence as historical data into a database, and randomly intercepting a gas concentration signal in a certain time period as shown in figure 2;
and 2, reasonably dividing the operation types of the working face according to the actual conditions of mine production. In the embodiment, the mine operation mode adopts four-six system, the whole operation period is blasting, coal cutting, coal dropping, shift switching, drilling operation, temporary support, anchor rod installation, maintenance, other irrelevant operations and other operation modes, and for convenience of explanation, the operation types are divided into three modes of 'no operation', 'drilling operation' and 'tunneling operation' according to the intensity of mining operation.
Step 3, according to the manual record of the operation type, respectively screening n groups of gas concentration sequence samples in corresponding time intervals, and carrying out self-adaptive filtering noise reduction processing on the data based on local mean decomposition, wherein the specific implementation process is as follows;
3.1 obtaining the entirety of the original signal sequence x (t)Extreme point niI is 1,2,3 …, I is the number of extreme points, and then according to the extreme points, the average value of all adjacent local extreme points is obtained as the local average value mi＝(ni+ni+1) (ii)/2 and the local envelope estimate as a local envelope value ai＝(ni-ni+1)/2；
3.2, average value m for all local areasiAnd all local envelope values aiPerforming multiple smoothing treatments by adopting a sliding average method to respectively obtain local mean value functions m'1(t)、m′2(t), … and local envelope function a'1(t)、a′2(t)、…；
3.3, each time of smoothing in step 3.2 is followed by processing iteration until a pure frequency-modulated signal s is obtainedk(t)：
After the first smoothing in step 3.2, separating the local mean function m 'obtained by the first smoothing from the original signal sequence x (t)'1(t) leaving the first intermediate separation signal sequence h1(t)＝x(t)-m′1(t), and then obtaining the local envelope function a 'according to the first smoothing processing'1(t) demodulating to obtain a first intermediate demodulated signal sequence s1(t)＝h1(t)/a′1(t)；
Then the first intermediate demodulated signal sequence s1(t) repeating the steps as the original signal sequence x (t) in combination with the local mean function m 'obtained by the second smoothing'1(t) and local envelope function a'1(t) calculating to obtain a second intermediate separation signal sequence and a second intermediate demodulation signal sequence;
the iterative processing is carried out until the K-th intermediate demodulation signal sequence satisfies s is more than or equal to-1K(t) is less than or equal to 1 and local area envelope function a 'of the next time'K+1If (t) is 1 and K is the number of iterations, the K-th intermediate demodulation signal sequence s is obtainedK(t) stopping iterative processing as a pure frequency modulation signal; the specific iterative process is represented as:
in the formula, K is the total number of times of iteration;
3.4, mixing the pure frequency-modulated signal sK(t) all local envelope functions a 'generated in an iterative process'k(t) multiplying to obtain an envelope signal A of the first product component1(t)＝a′1(t)a′2(t)…a′K(t), K is 1,2,3 …, K, and the envelope signal a of the first product component is then used1(t) and its corresponding pure FM signal sK(t) multiplying to obtain a first product component PF of the original signal sequence1(t)＝A1(t)sk(t)；
3.5, PF the first product component1(t) separating from the original signal sequence x (t) and retaining the first residual signal u1(t) and with the first residual signal u1(t) repeating the above steps 3.1-3.4 for the original signal sequence x (t) until all product components are separated by J cycles and the J-th residual signal u is satisfiedJ(t) is a monotonic function whereby the original gas concentration signal is transformed into the sum of a set of product components and a residual component.
3.6, performing correlation analysis on each obtained product component and the original signal sequence x (t), taking the global minimum value of the product component as a frequency modulation boundary point of high and low frequency components, taking all high frequency components of the multiplication integral quantity before the frequency modulation boundary point as noise product components, reconstructing the signal after removing the noise product components, and adding each product component to obtain a noise-reduced gas concentration sequence dx (t), as shown in fig. 3.
And 4, taking the gas concentration data with the time step length T after noise reduction, and constructing an operation type division sensitive parameter system. The sensitive parameters selected in this embodiment mainly include statistical parameters such as a gas concentration mean, a mean square error, a variance, an extreme difference, a difference characteristic value, and the like, which respectively represent gas emission characteristics under different operation types, and each statistical parameter is shown in table 1.
TABLE 1 gas concentration sequence sensitivity characteristic parameter set
Wherein, T represents the time step for calculating the gas data of each sensitive parameter, Xmax represents the maximum value of the gas concentration in the period of time and Xmin represents the minimum value of the gas concentration in the period of time.
And 5, constructing an operation type neural network intelligent recognition model of the underground working face, and inputting the model into the model for training according to the established gas concentration database and the sensitive characteristic parameters.
In the embodiment, three operation types, namely three conditions of tunneling operation, drilling operation and no operation, are selected and respectively marked as w1、w2And w3And respectively calculating the statistical parameter characteristics of the samples according to different operation types, wherein the summary is shown in table 2. And extracting gas concentration mean value, mean square deviation, variance, range difference, difference characteristic value and the like as classified characteristic quantities by adopting a rough neural network (RMNN) combining a rough set theory and an artificial neural network, and classifying the operation types according to the characteristics of the characteristic quantities. In the embodiment, sample training is carried out according to different parameter parameters, a weight i is given according to the precision and contribution of each parameter, and then a comprehensive parameter Y-d is determined1i1+d2i2+…+d8i8. Wherein i 1-i 8 respectively represent the contribution weights of the 8 sensitive indexes.
And (3) using the comprehensive Y value and the neural network model constructed in the foregoing, and performing sample training again until accurate identification of the operation type can be well realized, wherein the identification result according to the comprehensive Y value in the embodiment is shown in FIG. 4.
TABLE 2 summary of statistical parameter characteristic values for different job type samples
And 6, intelligently identifying the operation type of the newly acquired gas concentration data to be detected by using the trained model, and improving the operation type in a self-adaptive manner.
The algorithm scheme provided by the invention can be seen that the intelligent identification method for the operation process based on the gas concentration characteristics mainly takes the gas concentration parameters as a research object, aims to realize the intelligent identification of the operation type of the working face, carries out special noise reduction treatment, adopts a coarse neural network model to realize the intelligent identification of the working type of the underground mining working face, is beneficial to quickly and effectively acquiring the underground operation condition, can widen the function diversification and intellectualization of the mine safety monitoring system in the aspects of disaster diagnosis, operation scheduling, production management and the like, and has reference value for the identification result on the work such as accident mechanism analysis, emergency treatment, accident investigation and the like.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present 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 solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions and scope of the present invention as defined in the appended claims.
Claims (5)
1. An underground working face operation type intelligent identification method based on gas concentration parameters is characterized by mainly comprising the following main steps:
step 1, acquiring an original signal sequence x (t) of gas concentration through an underground gas concentration sensor, and storing the original signal sequence x (t) into a gas concentration database;
step 2, dividing each original signal sequence x (t) according to the operation type of a working surface;
step 3, performing self-adaptive filtering noise reduction processing on the gas concentration data in the gas concentration database to obtain a noise-reduced gas concentration sequence;
step 4, taking the gas concentration sequence after noise reduction as a chaotic time sequence, and respectively extracting sensitive characteristic parameters from data corresponding to various operation types according to the division of the operation types;
step 5, constructing an operation type neural network intelligent recognition model of the underground working face, and inputting the operation type neural network intelligent recognition model into the model for training according to the established gas concentration database and the sensitive characteristic parameters;
and 6, intelligently identifying the operation type of the newly acquired gas concentration data to be detected by using the trained model.
2. The method for intelligently identifying the operation type of the underground working face based on the gas concentration parameter as claimed in claim 1, wherein: in the step 2, the operation types of the working surface are divided into blasting, coal cutting, coal dropping, shift switching, drilling operation, temporary support, anchor rod installation, maintenance and other unrelated operations.
3. The method for intelligently identifying the operation type of the underground working face based on the gas concentration parameter as claimed in claim 1, wherein: in the step 4, the sensitive characteristic parameters include a mean value, a mean square error, a variance, a range and a difference characteristic value of the gas concentration data.
4. The method for intelligently identifying the operation type of the underground working face based on the gas concentration parameter as claimed in claim 1, wherein: in the step 3, the adaptive filtering and denoising specifically comprises:
3.1, obtaining all extreme points n of the original signal sequence x (t)iI is 1,2,3 …, I is the number of extreme points, and then according to the extreme points, the average value of all adjacent local extreme points is obtained as the local average value mi＝(ni+ni+1) (ii)/2 and the local envelope estimate as a local envelope value ai＝(ni-ni+1)/2；
3.2, average value m for all local areasiAnd all local area envelopesValue aiPerforming multiple smoothing treatments by adopting a sliding average method to respectively obtain local mean value functions m'1(t)、m′2(t), … and local envelope function a'1(t)、a′2(t)、…；
3.3, each time of smoothing in step 3.2 is followed by processing iteration until a pure frequency-modulated signal s is obtainedk(t)：
After the first smoothing in step 3.2, separating the local mean function m 'obtained by the first smoothing from the original signal sequence x (t)'1(t) leaving the first intermediate separation signal sequence h1(t)＝x(t)-m′1(t), and then obtaining the local envelope function a 'according to the first smoothing processing'1(t) demodulating to obtain a first intermediate demodulated signal sequence s1(t)＝h1(t)/a′1(t)；
Then the first intermediate demodulated signal sequence s1(t) repeating the steps as the original signal sequence x (t) in combination with the local mean function m 'obtained by the second smoothing'1(t) and local envelope function a'1(t) calculating to obtain a second intermediate separation signal sequence and a second intermediate demodulation signal sequence;
the iterative processing is carried out until the K-th intermediate demodulation signal sequence satisfies s is more than or equal to-1K(t) is less than or equal to 1 and local area envelope function a 'of the next time'K+1If (t) is 1 and K is the number of iterations, the K-th intermediate demodulation signal sequence s is obtainedK(t) stopping iterative processing as a pure frequency modulation signal; the specific iterative process is represented as:
in the formula, K is the total number of times of iteration;
3.4, mixing the pure frequency-modulated signal sK(t) all local envelope functions a 'generated in an iterative process'k(t) multiplying to obtain an envelope signal A of the first product component1(t)＝a′1(t)a′2(t)…a′K(t), K is 1,2,3 …, K, thenThen envelope signal A of the first product component is converted into1(t) and its corresponding pure FM signal sK(t) multiplying to obtain a first product component PF of the original signal sequence1(t)＝A1(t)sk(t)；
3.5, PF the first product component1(t) separating from the original signal sequence x (t) and retaining the first residual signal u1(t) and with the first residual signal u1(t) repeating the above steps 3.1-3.4 for the original signal sequence x (t) until all product components are separated by J cycles and the J-th residual signal u is satisfiedJ(t) is a monotonic function whereby the decomposed expression of the following original signal sequence x (t) is constructed:
wherein J represents the number of the iterative processing in steps 3.1 to 3.4, J is 1,2,3 …, J represents the total number of iterative processing in steps 3.1 to 3.4;
and 3.6, carrying out correlation analysis on each obtained product component and the original signal sequence x (t), taking the energy values of all product components to form a global minimum value of the sequence as a frequency modulation boundary point of high and low frequency components, taking all high-frequency components of the multiplication integral quantity before the frequency modulation boundary point as noise product components, and carrying out reconstruction processing on the signal after removing the noise product components to obtain a noise-reduced gas concentration sequence dx (t), wherein the energy values of the product components are obtained by calculation according to the sum of squares of all sequence data of the components.
5. The method for intelligently identifying the operation type of the underground working face based on the gas concentration parameter as claimed in claim 1, wherein:
the operation type neural network intelligent identification model adopts a rough set theory and an artificial neural network (RMNN) to calculate gas concentration data sensitive characteristic parameters of each operation type by utilizing a rough neural element of a neural network input layer, and adopts a fuzzy neural element to replace a common neural element.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811243460.1A CN109325470B (en) | 2018-10-24 | 2018-10-24 | Intelligent underground working face operation type identification method based on gas concentration parameters |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811243460.1A CN109325470B (en) | 2018-10-24 | 2018-10-24 | Intelligent underground working face operation type identification method based on gas concentration parameters |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109325470A CN109325470A (en) | 2019-02-12 |
| CN109325470B true CN109325470B (en) | 2021-09-03 |
Family
ID=65263225
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811243460.1A Active CN109325470B (en) | 2018-10-24 | 2018-10-24 | Intelligent underground working face operation type identification method based on gas concentration parameters |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109325470B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112308306A (en) * | 2020-10-27 | 2021-02-02 | 贵州工程应用技术学院 | Multi-mode input coal and gas outburst risk prediction method |
| CN114046179B (en) * | 2021-09-15 | 2023-09-22 | 山东省计算中心（国家超级计算济南中心） | Method for intelligently identifying and predicting underground safety accidents based on CO monitoring data |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6289328B2 (en) * | 1998-04-17 | 2001-09-11 | The United States Of America As Represented By The Secretary Of The Navy | Chemical sensor pattern recognition system and method using a self-training neural network classifier with automated outlier detection |
| CN104537199A (en) * | 2014-10-20 | 2015-04-22 | 天地（常州）自动化股份有限公司 | Method for recognizing production processes of mine working surface by gas density time sequence |
| CN106296435A (en) * | 2016-08-18 | 2017-01-04 | 西安科技大学 | A kind of mine gas monitoring disorder data recognition method |
| EP3126787A2 (en) * | 2014-04-01 | 2017-02-08 | Saudi Arabian Oil Company | Tomographic imaging of multiphase flows |
| CN106503798A (en) * | 2015-09-08 | 2017-03-15 | 中国科学院沈阳自动化研究所 | Based on rough set and the method for diagnosing faults of the pump of BP neural network |
| US9652684B2 (en) * | 2012-11-08 | 2017-05-16 | Total Sa | Image processing for classification and segmentation of rock samples |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8374974B2 (en) * | 2003-01-06 | 2013-02-12 | Halliburton Energy Services, Inc. | Neural network training data selection using memory reduced cluster analysis for field model development |
-
2018
- 2018-10-24 CN CN201811243460.1A patent/CN109325470B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6289328B2 (en) * | 1998-04-17 | 2001-09-11 | The United States Of America As Represented By The Secretary Of The Navy | Chemical sensor pattern recognition system and method using a self-training neural network classifier with automated outlier detection |
| US9652684B2 (en) * | 2012-11-08 | 2017-05-16 | Total Sa | Image processing for classification and segmentation of rock samples |
| EP3126787A2 (en) * | 2014-04-01 | 2017-02-08 | Saudi Arabian Oil Company | Tomographic imaging of multiphase flows |
| CN104537199A (en) * | 2014-10-20 | 2015-04-22 | 天地（常州）自动化股份有限公司 | Method for recognizing production processes of mine working surface by gas density time sequence |
| CN106503798A (en) * | 2015-09-08 | 2017-03-15 | 中国科学院沈阳自动化研究所 | Based on rough set and the method for diagnosing faults of the pump of BP neural network |
| CN106296435A (en) * | 2016-08-18 | 2017-01-04 | 西安科技大学 | A kind of mine gas monitoring disorder data recognition method |
Non-Patent Citations (3)
| Title |
|---|
| 基于多因素模式识别的煤与瓦斯突出预测研究;毕慧杰等;《中国安全生产科学技术》;20170630;98-103 * |
| 基于瓦斯涌出特征的机掘面作业类型判识技术;王川等;《煤矿安全》;20131231;71-74 * |
| 瓦斯浓度的分形分析与混沌预测模型研究;李刚;《中国博士学位论文全文数据库工程科技Ⅰ辑》;20110315;B021-11 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109325470A (en) | 2019-02-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109325470B (en) | Intelligent underground working face operation type identification method based on gas concentration parameters | |
| CN112986538A (en) | Large-area soil heavy metal detection and space-time distribution characteristic analysis method and system | |
| CN104778245A (en) | Similar trajectory mining method and device on basis of massive license plate identification data | |
| CN113886917B (en) | CNN-LSTM model-based railway line region ground settlement prediction early warning method | |
| CN111832506B (en) | Remote sensing discrimination method for rebuilding vegetation based on long time sequence vegetation index | |
| US11060899B2 (en) | Method for determining a maximum allowable volume of water that can be removed over time from an underground water source | |
| CN103646167A (en) | Satellite abnormal condition detection system based on telemeasuring data | |
| CN112990739B (en) | Phosphorite area water environment risk assessment method based on multi-medium pollution characteristic identification | |
| CN104376384A (en) | Typhoon day maximum daily load prediction system based on power big data analysis | |
| CN113095694B (en) | Rainfall sand transportation model construction method suitable for multiple landform type areas | |
| CN103942325A (en) | Method for association rule mining of ocean-land climate events with combination of climate subdivision thought | |
| CN114429189A (en) | High-universality landslide displacement rate early warning method | |
| CN113720639A (en) | Method for investigation, supplement and sampling of polluted site | |
| CN116658246A (en) | Coal mine water damage monitoring and early warning system and method | |
| CN111199298A (en) | Flood forecasting method and system based on neural network | |
| CN114611626A (en) | Water environment information processing method, device, equipment and storage medium | |
| CN117314218B (en) | Coal mine water damage real-time early warning method based on big data analysis | |
| CN110956155B (en) | Fuzzy clustering identification method for fully mechanized mining face operation procedure based on CO data | |
| Hosseini Duki et al. | Evaluation of base flow separation methods for determining water extraction (case study: Gorganroud River Basin) | |
| CN114782211B (en) | Sea mountain distribution range information acquisition method and system | |
| CN110348510B (en) | Data preprocessing method based on staged characteristics of deepwater oil and gas drilling process | |
| CN113175949B (en) | Method and system for inverting water release coefficient by combining surface deformation and water level information | |
| CN110991363B (en) | Method for extracting CO emission characteristics of coal mine safety monitoring system in different coal mining processes | |
| CN116165708A (en) | Accurate positioning method for uranium-bearing ore-forming sand body with beneficial coverage area | |
| Wang et al. | Neural networks approaches for modelling river suspended sediment concentration due to tropical storms |
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 |