CN111667193B - Coal mine gas safety evaluation method based on D-S evidence theory - Google Patents
Coal mine gas safety evaluation method based on D-S evidence theory Download PDFInfo
- Publication number
- CN111667193B CN111667193B CN202010538282.6A CN202010538282A CN111667193B CN 111667193 B CN111667193 B CN 111667193B CN 202010538282 A CN202010538282 A CN 202010538282A CN 111667193 B CN111667193 B CN 111667193B
- Authority
- CN
- China
- Prior art keywords
- evidence
- data
- evaluation
- coal mine
- expressed
- 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
- 238000011156 evaluation Methods 0.000 title claims abstract description 85
- 239000003245 coal Substances 0.000 title claims abstract description 56
- 238000005315 distribution function Methods 0.000 claims abstract description 46
- 238000000034 method Methods 0.000 claims abstract description 39
- 238000012544 monitoring process Methods 0.000 claims abstract description 37
- 238000009826 distribution Methods 0.000 claims abstract description 13
- 238000005259 measurement Methods 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims description 17
- 238000004458 analytical method Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 230000002159 abnormal effect Effects 0.000 claims description 7
- 239000000428 dust Substances 0.000 claims description 7
- 238000010220 Pearson correlation analysis Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 4
- 230000004927 fusion Effects 0.000 abstract description 27
- 230000000875 corresponding effect Effects 0.000 description 16
- 238000010219 correlation analysis Methods 0.000 description 4
- 230000002596 correlated effect Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000007499 fusion processing Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000015572 biosynthetic process Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000002715 modification method Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000009933 burial Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000739 chaotic effect Effects 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012843 least square support vector machine Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000012958 reprocessing Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000000528 statistical test Methods 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000005641 tunneling Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Educational Administration (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Agronomy & Crop Science (AREA)
- Animal Husbandry (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a coal mine gas safety evaluation method based on a D-S evidence theory, which comprises the following steps of S1, setting classification indexes and constructing an identification framework; s2, obtaining original data and calculating a plurality of evidence source data for evaluation; s3, designing a basic probability distribution function by adopting a posterior probability modeling method; s4, correcting a basic probability distribution function, reducing the conflict of evidence data for evaluation, S5, adopting a conflict distribution coefficient, improving a Dempster combination rule, and fusing the evidence data for evaluation; s6, uncertainty measurement is carried out by using Shannon entropy, and decision suggestions are provided; according to the invention, the predicted value of the monitoring data is used as evidence source data, so that the gas safety state is prejudged in advance, and the emergency level and the control capability of the gas disaster are improved; the Dempster combination rule is improved, the accuracy of fusion data is improved, and the automation and the credibility of evaluation are improved.
Description
Technical Field
The invention belongs to the technical field of safety evaluation, and particularly relates to a coal mine gas safety evaluation method based on a D-S evidence theory.
Background
The gas is taken as accompanying gas in the coal formation process, and is gushed out along with the development process, thus having great threat to the physical health of miners and the efficient and safe production of mines. Therefore, the gas safety evaluation is an important component of coal mine safety management, and the elimination and prevention of gas accidents are one of the important problems to be studied and solved in the present and future of the China coal industry. The most common method for the current safety evaluation method is to survey and analyze through a safety check list or analyze by adopting an accident tree, and also to judge the safety situation of the working face by adopting a least square vector machine, a BP neural network and other methods by students. However, the existing method has relatively simple indexes, the application of relevant data such as wind speed, temperature, dust concentration content and the like is insufficient, the reliability of evaluating the gas safety state is low, and erroneous judgment is easy to cause.
The gas safety evaluation needs to be comprehensively analyzed on the basis of data such as gas concentration, wind speed, temperature and the like, and the data information fusion technology is an information comprehensive processing technology which utilizes multi-source information cooperatively to obtain more objective and more intrinsic knowledge of things or targets. Among the many fusion models and methods, the D-S evidence theory algorithm is one of the most effective algorithms. However, the algorithm has a plurality of limitations such as focal element explosiveness, limited identification framework and independent problem between evidences, conflict evidence synthesis problem and the like.
Accordingly, there is a need to provide an improved solution to the above-mentioned deficiencies of the prior art.
Disclosure of Invention
The invention aims to provide a coal mine gas safety evaluation method based on a D-S evidence theory, which adopts the evidence theory to develop coal mine gas safety evaluation, improves the degree of automation and the reliability of the evaluation, adopts a predicted value of monitoring data as evidence source data, can realize the advanced pre-judgment of the gas safety state of a mine, and improves the emergency level and the control capability of gas disasters.
In order to achieve the above object, the present invention provides the following technical solutions:
the invention provides a coal mine gas safety evaluation method based on a D-S evidence theory, which comprises the following steps:
s1, setting classification indexes according to the safety state of gas in coal mine regulation, and constructing an identification framework;
s2, according to the safety evaluation requirement, the data of the monitoring object obtained from the coal mine monitoring system is used as the original data, the original data is processed, and a plurality of evidence source data for evaluation are obtained through calculation;
s3, designing a basic probability distribution function by adopting a posterior probability modeling method;
s4, correcting the basic probability distribution function to reduce the conflict of the evidence data for evaluation, and calculating the evidence data for evaluation by using the corrected basic probability distribution function;
s5, adopting conflict distribution coefficients, improving a Dempster combination rule, and fusing the evidence data for evaluation;
s6, selecting the state with the highest probability in the fused data as a recognition result, carrying out uncertainty measurement by using Shannon entropy, carrying out comprehensive analysis and evaluating the coal mine gas safety state, and providing decision suggestion.
According to the coal mine gas safety evaluation method based on the D-S evidence theory, as a preferred scheme, the gas safety state is divided into four states according to the coal mine safety regulations, namely, the states of no danger, light danger, moderate danger and serious danger, so that the built recognition frame is as follows:
X＝{X 1 (no danger), X 2 (mild danger), X 3 (moderate risk), X 4 (serious danger) };
according to the improved Dempster combination rule, the evidence data for evaluation are fused to the corresponding identification framework.
According to the coal mine gas safety evaluation method based on the D-S evidence theory, as a preferable scheme, the S2 body comprises:
s201, carrying out related influence factor determination on the original data by adopting Pearson correlation analysis, wherein the original data is time series data of a monitoring object obtained in a coal mine monitoring system, and the Pearson correlation analysis formula is as follows:
wherein x= [ x ] 1 ,x 2 ,…,x n ] T And y= [ y ] 1 ,y 2 ,…,y n ] T Two sets of time series data;and->Respectively corresponding average values; r is (r) xy Representing pearson correlation coefficients;
s202, acquiring a plurality of time series data of the original data from a monitoring system, and respectively performing abnormal time series data processing, missing time series data processing and time series data moving average noise reduction processing according to the quality condition of the time series data of the original data;
and S203, processing the time series data of the original data by using a prediction algorithm, and obtaining a plurality of evidence source data for evaluation according to the safety evaluation requirement.
According to the coal mine gas safety evaluation method based on the D-S evidence theory, as a preferable scheme, the S3 adopts a posterior probability modeling method to design a basic probability distribution function, and the method specifically comprises the following steps:
any evidence source data is calculated by a prediction algorithm from a plurality of evidence source data for evaluation, and an identification frame is set as X= { X 1 ,X 2 ,X 3 ,X 4 The set recognition frame contains 4 evaluation levels, and the distance between the set evidence source data and the evaluation level in the set recognition frame is:
d i (X i ，y)＝|X i -y|
wherein d i As distance value, X i In order to identify the ith value in the frame, y is evidence source data;
the correlation coefficient between the set evidence source data and the evaluation level in the set recognition frame can be expressed as:
wherein d i For distance value, c i Is a correlation coefficient;
thus, the base probability distribution function m (i) in the set recognition frame corresponding to the set evidence source data and the uncertainty m (Θ) corresponding to the set evidence source data can be expressed as:
where m (i) is the base probability distribution function, m (Θ) is the uncertainty, c i And as the correlation coefficient, y is evidence source data for evaluation obtained by calculating the time series prediction model, and x is expected output value of the time series prediction model.
According to the above method for evaluating coal mine gas safety based on D-S evidence theory, as a preferred solution, in S4, the basic probability distribution function is modified by calculating the weight to reduce the collision of the evidence data for evaluation, specifically:
for the setting event, there are N evidence source data, and the corresponding identification frame X comprises N focal elements, m k The evidence set is expressed as an evidence set formed by the basic probability distribution function values corresponding to the N evidence source data under the N focal elements, and the evidence set is expressed as follows:
m k ＝[m k (A 1 )，m k (A 2 )，…m k (A n )] T ，i＝1，2，…n
calculating m using a distance formula j And m k Distance d of (2) jk Distance formula d jk Expressed as:
m j and m k The similarity of (2) can be expressed as s jk Deriving a similarity formula s from the distance formula jk The method comprises the following steps:
s jk ＝1-d jk
the distance between evidence sets is inversely proportional to the mutual support degree, and the evidence set m k The support degree T of (2) can be expressed as:
according to the coal mine gas safety evaluation method based on the D-S evidence theory, as a preferable scheme, the evidence set m is used as a basis k The support degree T of the evidence collection is calculated, and the specific formula is as follows:
after assigning the weights, the modified base probability distribution function corresponding to the evidence data may be expressed as:
m' k (i)＝β(m k )·m k
m' k (Θ)＝β(m k )·m k +(1-β(m k ))。
according to the method for evaluating coal mine gas safety based on the D-S evidence theory, as a preferable scheme, the step S5 of fusing the evidence data for evaluation specifically comprises the following steps:
based on two independent evidences M 1 ,M 2 ，M 1 ,M 2 The focal elements of the two evidences are Bi and C respectively j (i＝1,2,3,…,n,j＝1,2,3,…,m)，M 1 ,M 2 The basic probability distribution function value after the two evidences are fused is m (A), and m (A) is expressed as:
wherein K (M 1 ,M 2 ) Called collision coefficients, representing two independent evidences M 1 ,M 2 Is a degree of conflict; when the conflict coefficient is 0, the two independent evidences are not in conflict; when the conflict coefficient is 1, the two independent evidences are completely in conflict.
According to the coal mine gas safety evaluation method based on the D-S evidence theory, as a preferable scheme, the S5 adopts a conflict distribution coefficient to improve the Dempster combination rule, and the method specifically comprises the following steps:
conflict allocation coefficient ω (A i ) Expressed as:
the improved formula of the Dempster combining rule is:
according to the above method for evaluating coal mine gas safety based on D-S evidence theory, as a preferred solution, the S6 selects the state with the largest probability in the fused data as the recognition result, uses Shannon entropy to perform uncertainty measurement, performs comprehensive analysis and evaluates the coal mine gas safety state, and proposes decision suggestion, which specifically includes:
the uncertainty measurement value is calculated by adopting the Shannon entropy principle, and the calculation process is as follows:
let n signal sources compose a signal x= { X 1 ,x 2 ,x 3 …,x n }，X＝{x 1 ,x 2 ,x 3 …,x n Probability p= { P (x) of n signal sources providing corresponding information for setting event for a set of time series data 1 ),p(x 2 ),p(x 3 ),…,p(x n ) The system structure S of the signal source can be expressed as:
the Shannon entropy of this signal source is expressed as:
according to the coal mine gas safety evaluation method based on the D-S evidence theory, preferably, the monitoring object at least comprises one of gas concentration, wind speed, temperature and dust concentration.
Compared with the closest prior art, the technical scheme provided by the invention has the following excellent effects:
according to the gas safety evaluation method, the evidence theory is adopted to carry out coal mine gas safety evaluation, so that the degree of automation and the credibility of the evaluation are improved; the predicted value of the monitoring data is used as evidence source data, so that the gas safety state of the mine can be predicted in advance, and the emergency level and the control capability of the gas disaster are improved. Aiming at the problem of evidence data conflict which is easy to exist in an evidence theory, a posterior probability distribution method is adopted to design a basic probability distribution function, a weight coefficient is utilized to correct the basic probability distribution function, and meanwhile, a conflict distribution coefficient is introduced, so that a Dempster combination rule is improved, and the accuracy of fusion data is improved.
Drawings
FIG. 1 is a schematic flow chart of a coal mine gas safety evaluation method in an embodiment of the invention;
FIG. 2 is a schematic diagram of the distribution of monitoring points of sensors in coal mine according to an embodiment of the invention;
fig. 3 is a comparison chart before and after processing data of the monitoring point a08 in the embodiment of the invention.
Detailed Description
The following description of the technical solutions in the embodiments of the present invention will be clear and complete, and it is obvious that the described embodiments are only some embodiments of the present invention, but not all embodiments. All other embodiments, which are derived by a person skilled in the art based on the embodiments of the invention, fall within the scope of protection of the invention.
The present invention will be described in detail below with reference to the drawings and embodiments, and it should be noted that the embodiments of the present invention and features of the embodiments may be combined with each other without conflict.
The invention provides a coal mine gas safety evaluation method based on a D-S evidence theory, which comprises the following steps:
s1, setting four classification indexes according to different gas safety states in coal mine regulations, and constructing an identification framework.
From the point of view of D-S evidence theory, the "gas safety state" can be regarded as a judging problem, and summary of the hypothetical results for this problem can be described as an identification framework, and the gas safety state is classified into four states of no risk, mild risk, moderate risk and serious risk according to the "coal mine safety regulations" and the parameter value range under specific conditions, so the establishment of the safety evaluation identification framework can be expressed as:
X＝{X 1 (no danger), X 2 (mild danger), X 3 (moderate risk), X 4 (serious danger) }.
S2, processing the original data according to the safety evaluation requirement, and obtaining a plurality of evidence source data for evaluation through calculation.
The change of the gas concentration of the coal mine is affected by a plurality of parameters, such as the geological structure of the coal seam, the burial depth of the coal seam, the thickness of the coal seam, the negative pressure, the wind speed, the roof pressure, the dust concentration, the gas concentration of adjacent points and the like, and in the embodiment of the invention, besides the gas concentration, time series data which can be acquired in a coal mine monitoring and controlling system such as the wind speed, the dust concentration, the temperature and the like are selected as raw data, the raw data are used as the basis of evidence source data for evaluation, and the specific steps for processing the raw data to acquire the evidence source data for evaluation are as follows:
s201, selecting evidence data variables for evaluation by adopting correlation analysis in statistical analysis, and determining correlation factors by adopting Pearson correlation analysis, wherein a calculation formula is expressed as follows:
wherein x= [ x ] 1 ,x 2 ,…,x n ] T And y= [ y ] 1 ,y 2 ,…,y n ] T For two sets of time series data,and->R is the corresponding average value xy Is the pearson correlation coefficient, which is used to represent the degree of correlation. The statistical test for the pearson correlation coefficient is to calculate the value of the t statistic, which obeys the t distribution of n-2 degrees of freedom. R can be automatically calculated by SPSS (Statistical Product and Service Solutions) software xy And a t value. By means of correlation coefficient r xy The degree of closeness of the correlation between the variables can be judged, for example, when the correlation coefficient value is less than or equal to 0.3, the correlation between the variables is not existed, and the specific correlation coefficient is shown in table 1.
TABLE 1 correlation coefficient
S202, time series data of various sensors (such as wind speed, temperature and the like) through correlation analysis are obtained from a monitoring system, and abnormal time series data processing, missing time series data processing, moving average noise reduction processing and the like are selected according to the quality condition of the time series data.
The abnormal data is mainly represented by zero values or very few high monitoring values different from normal values in time sequence, and the abnormal data is generally judged by assuming that continuous zero values occur at a certain moment, and the total number of the zero values is different from the number of moments in a selected time period, the points are called abnormal data, and the points are high monitoring value points when the number of high monitoring points exceeding the safety requirement occurring at a certain moment accounts for 5% of the total monitoring points. The missing data is supplemented by adopting an interpolation method aiming at the missing data on the time sequence, the missing data is supplemented by adopting a discarding method aiming at the abnormal data, and the interpolation method is adopted again for supplementing, so that the final aim is to avoid uncontrollable result deviation caused by data problems.
S203, reprocessing the time series data processed in the step S202 by using a prediction algorithm, wherein the prediction algorithm can be as follows: and acquiring predicted data y after a certain time, namely evidence source data y for evaluation, according to an evaluation requirement, and acquiring the evidence source data after n minutes when the evaluation is in a safe state after n minutes.
S3, designing a basic probability distribution function by adopting a posterior probability modeling method, taking a single sensor as an example in the embodiment of the invention, setting evidence source data for evaluation of a time sequence prediction model corresponding to the sensor as y, and setting an identification frame containing four evaluation grades as X= { X 1 ,X 2 ,X 3 ,X 4 The distance between the available evidence source data and the evaluation level can be expressed as:
d i (X i ，y)＝|X i -y|
wherein d i As distance value, X i In order to identify the ith value in the frame, y is evidence data;
the correlation coefficient between the set evidence source data and each evaluation level in the set recognition frame can be expressed as:
wherein d i For distance value, c i Is a correlation coefficient;
thus, the base probability distribution function m (i) in the recognition framework for each evidence and the uncertainty m (Θ) of the corresponding evidence can be expressed as:
where m (i) is the base probability distribution function, m (Θ) is the uncertainty, c i And as the correlation coefficient, y is evidence source data for evaluation obtained by calculating the time series prediction model, and x is expected output value of the time series prediction model.
S4, correcting the basic probability distribution function to reduce the conflict of the evidence data for evaluation, and calculating the evidence data for evaluation by using the corrected basic probability distribution function.
In order to make the evidence theory fusion part as accurate as possible, the evidence data needs to be modified, the influence of interference factors on the fusion result can be reduced by modifying the evidence data, the accuracy of the fusion result is improved, the traditional D-S evidence theory can lead to the failure of the fusion result when the evidence data completely conflict, the prediction result can not reflect the actual situation, in the stages of sensor data acquisition, filtering, denoising and feature extraction, the difference of various sensors is large, in the Dempster combination rule, the importance degree of the various sensors is the same by default, and different evidence data have different trust degrees on each evaluation state in the identification frame are not considered, therefore, the importance degree of each evidence data is distributed by adopting the idea of distributing weights, the reliability of the evidence data on the decision result can be increased, and the influence of the evidence data with conflict on the decision result is weakened.
For an uncertain event, there are N evidence source data, and the corresponding identification frame X contains N focal elements, m k The evidence set is represented as an evidence set composed of basic probability distribution function values corresponding to N evidence source data under N focal elements, and the evidence set is represented as:
m k ＝[m k (A 1 )，m k (A 2 )，…m k (A n )] T ，i＝1，2，…n
calculating m using a distance formula i And m j Distance of (d), distance formula d jk Expressed as:
m i and m j The similarity of (2) can be expressed as s jk Deriving a similarity formula s from the distance formula jk The concrete steps are as follows:
s jk ＝1-d jk
the smaller the distance between evidence sets, the greater the degree of mutual support, evidence set m k The support degree T of (2) can be represented by summation, then the evidence set m k The support level of (2) can be expressed as:
different weight values can be calculated by utilizing the distance similarity matrix between evidences, so that the aim of correcting the evidence data is fulfilled. In order to prevent the corrected evidence data from being too conservative, the advantages of the original evidence data are lost, and the original correct evidence data are reserved to further improve the fusion effect. Based on the idea, under the condition of retaining a group of excellent evidence data, calculating the weight beta of the evidence data according to the magnitude of the supporting degree, wherein the specific formula is as follows:
after assigning the weights, the modified base probability distribution function corresponding to the evidence data may be expressed as:
m' k (i)＝β(m k )·m k
m' k (Θ)＝β(m k )·m k +(1-β(m k ))。
s5, adopting conflict distribution coefficients, improving the original Dempster combination rule, and fusing the data.
The data fusion is to fuse the probability values calculated by the corrected basic probability distribution functions under different evidence data into the probability values of the corresponding recognition frames by using the improved combination rules, wherein the probability values calculated by the corrected basic probability distribution functions are calculated according to the probability values calculated by the original basic probability distribution functions, and the probability values calculated by the original basic probability distribution functions are calculated by the processed evidence source data.
Since the combination rule of the evidence theory is imperfect in processing the evidence data, reasonable modification of the combination rule can also improve the fusion accuracy, after the evidence data is modified, the evidence data is simply modified to prevent high conflict among evidences, and effective information of the original evidence data is likely to be lost by the modified evidence data, so that a conflict distribution coefficient is introduced on the basis of the traditional D-S evidence theory Dempster combination rule, and the accuracy of a decision stage is improved.
Conflict allocation coefficient ω (A i ) Expressed as:
the improved formula of the Dempster combining rule is:
the method comprises the steps of adopting evidence data of monitoring and monitoring data of each sensor to distribute values of basic probability of fusion results as evidence of D-S evidence theory, judging and identifying the safety state of mine gas after fusing the values of each sensor, and based on two independent evidences M 1 ,M 2 I.e. different y values, M in evidence data 1 ,M 2 The focal elements of the two evidences are B respectively i And C j (i=1, 2,3, …, n, j=1, 2,3, …, m), the basic probability distribution function value after fusion of the two is m (a), i.e. m (a) is the basic probability distribution function m (i)The probability value under the assumption of a, m (a) is expressed as:
wherein K (M 1 ,M 2 ) Is called a collision coefficient, representing two evidences M 1 ,M 2 Is a collision degree of the (c). When the conflict coefficient is 0, no conflict exists between the two evidences; when the conflict factor is 1, the conflict between the two evidences is the largest, which indicates complete conflict.
S6, selecting the state with the highest probability in the fused data as a recognition result, carrying out uncertainty measurement by using Shannon entropy, carrying out comprehensive analysis and evaluating the coal mine gas safety state, and providing decision advice, wherein after the data are fused, the gas safety index of the coal mine can be obtained, and the uncertainty measurement value is calculated according to the Shannon entropy principle, wherein the concrete calculation method comprises the following steps:
let n signal sources compose a signal x= { X 1 ,x 2 ,x 3 …,x n Probability p= { P (x) of n signal sources providing corresponding information for a set event 1 ),p(x 2 ),p(x 3 ) …, p (xn) }, the system structure S of the signal source can be expressed as:
the Shannon entropy, i.e., uncertainty measure H, of the signal source can be expressed as:
relevant work is carried out according to the uncertainty measure and the safety index:
the non-danger represents good environment of the underground working face of the coal mine.
A mild hazard means that there is a risk to the coal mine downhole face that the risk value is within an acceptable range.
The moderate danger indicates unsafe working surfaces in the coal mine, and the risk value is beyond the accepted range, so that workers are required to perform on-site inspection.
Serious danger means that the underground working face of the coal mine is very bad in environment, and workers should be evacuated.
If an uncertainty situation occurs (uncertainty measure >0.5 is defined as an uncertainty situation), the data and analysis flow are carefully examined and the evaluation work is re-conducted.
For a better understanding of the present invention, a further explanation of the invention is provided below with a set of examples.
The arrangement of the monitoring points of each gas sensor is shown in fig. 2, and in order to evaluate the gas safety state of a mining fully mechanized working face for 5 minutes, an evaluated identification frame is set as X= { X 1 (no danger), X 2 (mild danger), X 3 (moderate risk), X 4 (serious danger) }.
As shown in fig. 2, the monitoring data that can be acquired in the working surface includes: the gas concentration of the upper corner (A02), the gas concentration of the tunneling working face (A01) at 10 meters, the wind speed (A09), the dust concentration (A11), the temperature of the return air (A07) at 15 meters and the gas concentration of the return air (A08) at 15 meters.
The collected gas sensor related monitoring data are shown in table 2, and the table 2 contains gas concentration monitoring data of three monitoring points A02, A01 and A08.
Table 2 monitoring and monitoring data
The pearson correlation analysis is carried out on the monitoring points A02, A01, A09, A11, A07 and A08 by adopting SPSS (Statistical Product and Service Solutions) software, more data are selected as much as possible for processing in order to ensure the reliability of the result, the data in the embodiment of the invention are analyzed by adopting 17820 data with original sampling intervals, and the correlation analysis results of all influence factors are shown in table 3.
TABLE 3 results of correlation analysis of influence factors
According to the correlation coefficient evaluation criteria shown in table 1, it can be seen from table 3 that a01 and a08 are highly correlated, a02, a09, a07 and a08 are significantly correlated, a11 and a08 are low correlated, and the correlation gas concentration, wind speed, temperature and dust concentration selected in the embodiment of the present invention are reasonable as input parameters.
Because a large amount of noise exists in the monitoring data, the monitoring data is subjected to noise reduction treatment, the column vector y is smoothed by a moving average filter, the column vector yy with the same length as the column vector y is returned, the default window width of the moving average filter is 5, and the calculation formula of elements in yy is as follows:
yy(1)＝y(1)
yy(2)＝(y(1)+y(2)+y(3))/3
yy(3)＝(y(1)+y(2)+y(3)+y(4)+y(5))/5
yy(4)＝(y(2)+y(3)+y(4)+y(5)+y(6))/5
……
taking the data value of the monitoring point A08 as an example, the preprocessed (namely noise-reduced) data is shown in fig. 3, and the embodiment of the invention predicts the gas safety state after 5 minutes, so that the data is predicted by adopting a multi-variable self-adaptive weighted least square support vector machine method of chaotic particle swarm optimization, and the prediction result is shown in table 4.
TABLE 4 prediction results
The sensor number is used for replacing the evidence number, the posterior probability modeling method is adopted for calculating the basic probability distribution function of each sensor, and the obtained basic probability distribution function of each sensor is shown in table 5.
TABLE 5 basic probability distribution function
As can be seen from Table 5, the result A09 of the single sensor recognition is m (X 2 ) = 0.8079, a07 is m (X 4 ) = 0.2399, a11 is m (X 1 ) = 0.4939, a02 is m (X 1 ) = 0.5551, a01 is m (X 1 ) = 0.5664 and a08 is m (X 1 ) = 0.5954. Obviously, a09, a07 have a large collision with other sensors, so evidence data needs to be corrected before fusion.
And (3) correcting the evidence data by adopting the correction method described in the step (S5), and carrying out basic probability value calculation on each sensor again according to the basic probability distribution function obtained by calculation in the table (5), wherein the corrected basic probability distribution function is shown in the table (6).
TABLE 6 modified base probability distribution function
As can be seen from Table 6, A09 is defined by m (X 2 ) = 0.8079 corrected to m (X 2 ) = 0.4622, a07 is defined by m (X 4 ) = 0.2399 corrected to m (X 4 ) = 0.1997, the modified evidence indicates that the evidence source modification method proposed by the present invention is trusted, and the excellent set of evidence of a02 is preserved. Meanwhile, the basic probability distribution function shows that the decision failure can be caused by using only the sensors A09 and A07 as evaluation evidences, and the accuracy of the identification is low by using only the sensors A11, A02, A01 and A08 as the evaluation evidences, so that the reliability of the decision result is low. Therefore, the mine gas safety state evaluation is unreliable by adopting only a single sensor, and the reliability of a decision result can be increased by multi-sensor data fusion.
The comparison analysis of the conflict degree shows that the data fusion plays an important role in decision results, and the comparison data are adopted for the description of the invention to better explain the advantages:
the method for applying the conventional D-S theory is referred to herein as the D-S-0 evidence theory.
The method for modifying evidence data is referred to herein as the D-S-1 evidence theory.
The method for modifying the combining rule is referred to herein as the D-S-2 evidence theory.
The sensors A09, A07, A11, A02, A01, A08 are denoted as evidence e 1 、e 2 、e 3 、e 4 、e 5 、e 6 . The fusion process of the multiple sensors is to sequentially perform the fusion process of the two sensors, and comparison results of the fusion of the multiple sensors of the three methods are shown in the following tables 7-11.
TABLE 7 e 1 e 2 Fusion result comparison analysis
As can be seen from Table 7, the fused evidence data are all high-conflict evidence, so that the decision results of the D-S-0 evidence theory and the D-S-1 evidence theory are invalid, and the identification result of the D-S-2 evidence theory is uncertain.
Table 8 e 1 e 2 e 3 Fusion result comparison analysis
Introduction of evidence data e in Table 8 3 And then, the D-S-0 evidence theory and the D-S-1 evidence theory have error recognition results, and the D-S-2 evidence theory has accurate recognition results, so that the improved combination rule is accurate and effective, and the effective information of the corrected evidence data is reserved.
Table 9 e 1 e 2 e 3 e 4 Fusion results analysis
The D-S-0 evidence theory recognition results in the table 9 are wrong, the D-S-1 evidence theory and the D-S-2 evidence theory recognition results are accurate, and the improved evidence data modification method is proved to be accurate and effective, so that high conflict among evidences is eliminated.
Table 10 e 1 e 2 e 3 e 4 e 5 Fusion result comparison analysis
Table 11 e 1 e 2 e 3 e 4 e 5 e 6 Fusion result comparison analysis
Tables 10 and 11 show that the method for improving the evidence data modification and combination rules is reasonable, and the identification accuracy of the D-S-2 evidence theory is higher than that of the D-S-0 evidence theory and the D-S-1 evidence theory, so that the identification accuracy of the mine gas safety state is improved. Meanwhile, the combination rule meets the exchange law, so that the higher the accuracy of decision stage identification along with the increase of evidence in the fusion process can be obtained. The defect that a single sensor is difficult to accurately represent the safety state of the gas is overcome.
As can be seen from the above fusion results, the evaluation of the current working face is X 1 I.e. no dangerous situation.
Uncertainty measurements were made using Shannon entropy. The comparison data are still used here to calculate the entropy values of three different D-S evidences.
Uncertainty of the D-S-0 evidence fusion is:
-0.8198ln0.8198-0.1768ln0.1768-0.0028ln0.0028-0.0006ln0.0006＝0.4901
uncertainty of the D-S-1 evidence fusion is:
-0.9225ln0.9225-0.0720ln0.0720-0.0042ln0.0042-0.0013ln0.0013＝0.2955
uncertainty of the D-S-2 evidence fusion is:
-0.9485ln0.9485-0.0466ln0.0466-0.0038ln0.0038-0.0011ln0.0011＝0.2217
from the above, the uncertainty of the non-dangerous state is less than 0.5, and the coal mine underground working face environment can be basically determined to be good.
Meanwhile, the uncertainty degree value comparison shows that the D-S-2 evidence theory has lower uncertainty degree relative to the D-S-0 evidence theory and the D-S-2 evidence theory, so that the calculation result is more accurate and convincing, and the actual requirement can be better met.
In conclusion, the gas safety evaluation method adopts the evidence theory to develop the coal mine gas safety evaluation, so that the degree of automation and the reliability of the evaluation are improved; the predicted value of the monitoring data is used as evidence source data, so that the gas safety state of the mine can be predicted in advance, and the emergency level and the control capability of the gas disaster are improved. Aiming at the problem of evidence data conflict which is easy to exist in an evidence theory, a posterior probability distribution method is adopted to design a basic probability distribution function, a weight coefficient is utilized to correct the basic probability distribution function, and meanwhile, a conflict distribution coefficient is introduced, so that a Dempster combination rule is improved, and the accuracy of fusion data is improved.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the particular embodiments disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims.
Claims (6)
1. The coal mine gas safety evaluation method based on the D-S evidence theory is characterized by comprising the following steps of:
s1, setting classification indexes according to the safety state of gas in coal mine regulation, and constructing an identification framework;
s2, according to the safety evaluation requirement, the data of the monitoring object obtained from the coal mine monitoring system is used as the original data, the original data is processed, and a plurality of evidence source data for evaluation are obtained through calculation;
s3, designing a basic probability distribution function by adopting a posterior probability modeling method;
s4, correcting the basic probability distribution function to reduce the conflict of the evidence data for evaluation, and calculating the evidence data for evaluation by using the corrected basic probability distribution function;
s5, adopting conflict distribution coefficients, improving a Dempster combination rule, and fusing the evidence data for evaluation;
s6, selecting the state with the highest probability in the fused data as a recognition result, carrying out uncertainty measurement by using Shannon entropy, carrying out comprehensive analysis and evaluating the coal mine gas safety state, and providing decision suggestion;
in the step S4, the basic probability distribution function is modified by calculating a weight to reduce the collision of the evidence data for evaluation, specifically:
for the setting event, there are N evidence source data, and the corresponding identification frame X comprises N focal elements, m k The evidence set is expressed as an evidence set formed by the basic probability distribution function values corresponding to the N evidence source data under the N focal elements, and the evidence set is expressed as follows:
m k ＝[m k (A 1 )，m k (A 2 )，…m k (A n )] T ，i＝1，2，…n，
calculating m using a distance formula j And m k Distance d of (2) jk Distance formula d jk Expressed as:
m j and m k The similarity of (2) can be expressed as s jk Deriving a similarity formula s from the distance formula jk The method comprises the following steps:
s jk ＝1-d jk ，
the distance between evidence sets is inversely proportional to the mutual support degree, and the evidence set m k The support degree T of (2) can be expressed as:
under the condition of retaining a group of excellent evidence data, according to the evidence set m k The support degree T of the evidence collection is calculated, and the specific formula is as follows:
after assigning the weights, the modified base probability distribution function corresponding to the evidence data may be expressed as:
m' k (i)＝β(m k )·m k ，
m' k (Θ)＝β(m k )·m k +(1-β(m k ))，
wherein m' k (Θ) represents the modified uncertainty;
in the step S5, the evidence data for evaluation are fused, specifically:
based on two independent evidences M 1 ,M 2 ，M 1 ,M 2 The focal elements of the two evidences are B respectively i And C j (i＝1,2,3,…,n,j＝1,2,3,…,m)，M 1 ,M 2 The basic probability distribution function value after the two evidences are fused is m (A), and m (A) is expressed as:
wherein K (M 1 ,M 2 ) Called collision coefficients, representing two independent evidences M 1 ,M 2 Is a degree of conflict; when the conflict coefficient is 0, the two independent evidences are not in conflict; when the conflict coefficient is 1, the two independent evidences are completely in conflict;
and in the step S5, a conflict distribution coefficient is adopted to improve a Dempster combination rule, and the method specifically comprises the following steps:
conflict allocation coefficient ω (A i ) Expressed as:
the improved formula of the Dempster combining rule is:
2. the method for evaluating the safety of coal mine gas based on the D-S evidence theory according to claim 1, wherein the safety state of the gas is divided into four states according to the safety regulations of the coal mine, namely, four states of no danger, mild danger, moderate danger and serious danger, so that the constructed recognition framework is as follows:
X＝{X 1 (no danger), X 2 (mild danger), X 3 (moderate risk), X 4 (serious danger) };
according to the improved Dempster combination rule, the evidence data for evaluation are fused to the corresponding identification framework.
3. The method for evaluating coal mine gas safety based on the D-S evidence theory according to claim 1, wherein the S2 comprises:
s201, carrying out related influence factor determination on the original data by adopting Pearson correlation analysis, wherein the original data is time series data of a monitoring object obtained in a coal mine monitoring system, and the Pearson correlation analysis formula is as follows:
wherein x= [ x ] 1 ,x 2 ,…,x n ] T And y= [ y ] 1 ,y 2 ,…,y n ] T Two sets of time series data;and->Respectively corresponding average values; r is (r) xy Representing pearson correlation coefficients;
s202, acquiring a plurality of time series data of the original data from a monitoring system, and respectively performing abnormal time series data processing, missing time series data processing and time series data moving average noise reduction processing according to the quality condition of the time series data of the original data;
and S203, processing the time series data of the original data by using a prediction algorithm, and obtaining a plurality of evidence source data for evaluation according to the safety evaluation requirement.
4. The coal mine gas safety evaluation method based on the D-S evidence theory according to claim 1, wherein the step S3 of designing a basic probability distribution function by adopting a posterior probability modeling method is specifically as follows:
any evidence source data calculated by adopting a prediction algorithm is set as the evidence source datayLet the recognition frame be x= { X 1 ,X 2 ,X 3 ,X 4 The set recognition frame contains 4 evaluation levels, and the distance between the set evidence source data and the evaluation level in the set recognition frame is:
d i (X i ，y)＝|X i -y|，
wherein d i As distance value, X i In order to identify the ith value in the frame, y is evidence source data;
the correlation coefficient between the set evidence source data and the evaluation level in the set recognition frame can be expressed as:
wherein d i For distance value, c i Is a correlation coefficient;
thus, the base probability distribution function m (i) in the set recognition frame corresponding to the set evidence source data and the uncertainty m (Θ) corresponding to the set evidence source data can be expressed as:
where m (i) is the base probability distribution function, m (Θ) is the uncertainty, c i And as the correlation coefficient, y is evidence source data for evaluation obtained by calculating the time series prediction model, and x is expected output value of the time series prediction model.
5. The method for evaluating the safety of coal mine gas based on the D-S evidence theory according to claim 1, wherein the step S6 is characterized in that the state with the highest probability in the fused data is selected as the recognition result, uncertainty measurement is carried out by using Shannon entropy, comprehensive analysis is carried out, the safety state of the coal mine gas is evaluated, and decision suggestion is provided, specifically:
the uncertainty measurement value is calculated by adopting the Shannon entropy principle, and the calculation process is as follows:
let n signal sources compose a signal x= { X 1 ,x 2 ,x 3 …,x n }，X＝{x 1 ,x 2 ,x 3 …,x n Probability p= { P (x) of n signal sources providing corresponding information for setting event for a set of time series data 1 ),p(x 2 ),p(x 3 ),…,p(x n ) The system structure S of the signal source can be expressed as:
the Shannon entropy of this signal source is expressed as:
6. the D-S evidence theory-based coal mine gas safety evaluation method according to claim 1, wherein the monitoring object includes at least one of a gas concentration, a wind speed, a temperature, and a dust concentration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010538282.6A CN111667193B (en) | 2020-06-12 | 2020-06-12 | Coal mine gas safety evaluation method based on D-S evidence theory |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010538282.6A CN111667193B (en) | 2020-06-12 | 2020-06-12 | Coal mine gas safety evaluation method based on D-S evidence theory |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111667193A CN111667193A (en) | 2020-09-15 |
| CN111667193B true CN111667193B (en) | 2024-01-02 |
Family
ID=72387421
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010538282.6A Active CN111667193B (en) | 2020-06-12 | 2020-06-12 | Coal mine gas safety evaluation method based on D-S evidence theory |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111667193B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112649034B (en) * | 2020-12-28 | 2023-09-19 | 精英数智科技股份有限公司 | Method and device for identifying adjustment state of underground coal mine sensor |
| CN112800999B (en) * | 2021-02-04 | 2023-12-01 | 大连大学 | Intelligent command system target identification fusion method based on intelligent networking |
| CN113469034A (en) * | 2021-06-30 | 2021-10-01 | 鄂尔多斯应用技术学院 | Fusion recognition method for coal rock cutting state of coal mining machine |
| CN114757295B (en) * | 2022-04-28 | 2024-04-02 | 浙江科技学院 | Multi-sensor data fusion method and application based on cloud model and evidence theory |
| CN115564148B (en) * | 2022-12-01 | 2023-04-07 | 北京科技大学 | Chamber safety analysis method based on multi-source heterogeneous data fusion |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101806229A (en) * | 2010-03-15 | 2010-08-18 | 山东科技大学 | Mine roof safety evaluation method based on multi-target monitoring data fusion |
| CN110119713A (en) * | 2019-05-14 | 2019-08-13 | 云南电网有限责任公司电力科学研究院 | Mechanical Failure of HV Circuit Breaker diagnostic method based on D-S evidence theory |
| CN110533091A (en) * | 2019-08-22 | 2019-12-03 | 贵州大学 | A kind of more evident information fusion methods for improving DS evidence theory |
-
2020
- 2020-06-12 CN CN202010538282.6A patent/CN111667193B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101806229A (en) * | 2010-03-15 | 2010-08-18 | 山东科技大学 | Mine roof safety evaluation method based on multi-target monitoring data fusion |
| CN110119713A (en) * | 2019-05-14 | 2019-08-13 | 云南电网有限责任公司电力科学研究院 | Mechanical Failure of HV Circuit Breaker diagnostic method based on D-S evidence theory |
| CN110533091A (en) * | 2019-08-22 | 2019-12-03 | 贵州大学 | A kind of more evident information fusion methods for improving DS evidence theory |
Non-Patent Citations (2)
| Title |
|---|
| Host risk evaluation framework based on multi-source information;Cuixia Gao 等;《2009 ISECS International Colloquium on Computing, Communication, Control, and Management》;第249-252页 * |
| 罗小燕 等.基于加权D-S证据理论融合多域特征的岩体失稳预测方法.煤炭学报.2019,第45卷(第10期),第3446-3452页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111667193A (en) | 2020-09-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111667193B (en) | Coal mine gas safety evaluation method based on D-S evidence theory | |
| CN101806229B (en) | Mine roof safety evaluation method based on multi-target monitoring data fusion | |
| CN108399277B (en) | Bridge damage identification method based on temperature and strain correlation | |
| CN106950945B (en) | A kind of fault detection method based on dimension changeable type independent component analysis model | |
| CN110348752B (en) | Large industrial system structure safety assessment method considering environmental interference | |
| CN110703214A (en) | Weather radar state evaluation and fault monitoring method | |
| CN107272667A (en) | A kind of industrial process fault detection method based on parallel PLS | |
| JP6778132B2 (en) | Abnormality diagnosis system for equipment | |
| CN114463954B (en) | Combustible gas monitoring and early warning system and method based on Internet of things | |
| CN113705951A (en) | Method for evaluating mine ecological restoration effect | |
| CN111401784B (en) | Fire safety grade assessment method | |
| CN110598265A (en) | Bridge health data abnormity correction method and system based on wavelet analysis | |
| CN112365055A (en) | Method for predicting suitability of residual coal re-mining condition of integrated coal mine | |
| CN116092198B (en) | Mining safety helmet identification detection method, device, equipment and medium | |
| CN117523299A (en) | Image recognition method, system and storage medium based on computer network | |
| CN112561257A (en) | Bridge structure safety evaluation method and device based on big data | |
| CN117076935A (en) | Digital twin-assisted mechanical fault data lightweight generation method and system | |
| CN115600747B (en) | Tunnel state monitoring and management method and system based on Internet of things | |
| CN113375775B (en) | Weight correction method and weight correction system for shared vehicle and electronic equipment | |
| CN115526407A (en) | Power grid big data safety detection early warning method and system based on causal machine learning | |
| CN115169856A (en) | Tunnel lining state evaluation method based on multi-source information fusion | |
| Chou et al. | Economic Statistical Design of Average Control Charts for Monitoring a Process under Non-normality | |
| CN115278742B (en) | Equipment alarm-based equipment base station health assessment system and method | |
| CN117455268B (en) | Bus one-key emergency escape data analysis system and method based on Internet of things | |
| CN112949816A (en) | ANN-based method and system for identifying abnormal data of hydrostatic level |
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 |