CN103760597B - A kind of Automatic mine fault identification method - Google Patents

Abstract

The Automatic mine fault identification method that the present invention provides, according to mine microseism positioning result data, all micro-seismic event are carried out the cluster analysis of locus, under mine fault face is planar assumption, by to each bunch method of Eigenvalues analysis, drawing the orientation of each fault plane, the geometric parameter such as length and width, provide the tomography being likely to occur, and draw position of fault figure.The advantage of Automatic mine fault identification method of the present invention is: compared to the manual interpretation method behind general microseism location, this method avoid the subjective impact in manual interpretation, the advantage having quantification, automatization, Semi-intelligent Modular.

Description

A kind of Automatic mine fault identification method

Technical field

The present invention relates to the method that data process, be specifically related to a kind of Automatic mine fault identification method.

Background technology

The explanation of microseism positioning result typically by manually realizing, but along with mine microseism location ability more and more higher, Earthquakes location number gets more and more, and the artificial factor by experience is more and more, and simple is difficult to specifically by manual identification Mark off a reliable result, be the most just increasingly difficult to identify tomography, thus provide rationalization to build to Mine Safety in Production View.

Summary of the invention

The present invention is the defect for overcoming prior art, it is provided that a kind of Automatic mine fault identification method, can be according to ore deposit Mountain microseism positioning result data, under mine fault face is planar assumption, quantification, automatization, Semi-intelligent Modular ground Draw the geometric parameters such as the orientation of each fault plane, length and width, provide the tomography being likely to occur, and draw position of fault figure.

A kind of Automatic mine fault identification method that the present invention provides, arranges identification computer and arranges in identifying computer The micro-seismic event cloud of mine microseism positioning result, the method comprises the steps:

1) reading micro-seismic event cloud, set and wherein have N0 tomography, each tomography has random position and orientation, And the initial value of this parameter is 1；

2) to each seismic events, find out closest with it tomography, constitute initial clustering, then in class Carry out cluster analysis under the constraints of the seismic events distance sum minimum belonging to center and apoplexy due to endogenous wind, i.e. carry out overall situation distance The optimization that sum is minimum；

3) obtain the covariance matrix of the seismic events of each apoplexy due to endogenous wind, and this covariance matrix is carried out eigenvalue and feature The calculating of vector, on the premise of seismic events is uniformly distributed at random, provides tomography corresponding to such by eigenvalue Locus, the length that represent tomography with the eigenvalue of covariance matrix respectively is generous, and Length x Width is its character pair value 'sTimes, characteristic vector represents the spatial orientation of tomography；

4) judge whether the maximum of the eigenvalue of minimum is less than the eigenvalue set, if the maximum of the eigenvalue of minimum Value is all little than the eigenvalue set, then calculates and terminates, draw the locus of each tomography, performs the 6th) step；As Fruit still has a minimal eigenvalue more than the eigenvalue set after cluster computing, then this bunch is divided into two sections, always breaks The number of plies increases by one, performs next step；

5) the 2nd is returned to) step calculates again, until calculating terminates；

6) last, delete comprise hypocentral location less than 4 bunch, according to the spatial orientation of fault plane, i.e. each bunch Eigenvalue and characteristic vector value draw position of fault figure.

One Automatic mine fault identification method of the present invention, in described step 2) in, the optimization that overall situation distance sum is minimum Use following computing formula:

$J=\frac{1}{N_d}\underset{i_c=1}{\overset{N_c}{\mathrm{\Σ}}}\underset{j=1}{\overset{N_{\mathrm{ic}}}{\mathrm{\Σ}}}{D}_j^2$

Wherein, Nc is the number of earthquake bunch, and Nic is the number of earthquake in each earthquake bunch, and Dj is that this bunch of center is arrived in earthquake Distance, Nd is the number of whole event.

One Automatic mine fault identification method of the present invention, in described step 3), including for each class, calculates The center of apoplexy due to endogenous wind event and covariance matrix, its covariance matrix is:

Wherein σ²For it in the variance of respective coordinates value, (a b) is the covariance of respective coordinates value to cov.

The advantage of a kind of Automatic mine fault identification method of the present invention is: according to mine microseism positioning result data, All micro-seismic event are carried out the cluster analysis of locus, under mine fault face is planar assumption, by often The geometric parameters such as one bunch, by the method for Eigenvalues analysis, draws the orientation of each fault plane, length and width, are given and may go out Existing tomography, and draw position of fault figure.Compared to the manual interpretation method behind general microseism location, the method is kept away Exempt from the subjective impact in manual interpretation, the advantage having quantification, automatization, Semi-intelligent Modular.

Accompanying drawing explanation

Fig. 1 is the flow chart of a kind of Automatic mine fault identification method of the present invention；

Fig. 2 is the microseism location figure before not using the present invention；

Fig. 3 is the tomography identification figure automatically after using the present invention.

Detailed description of the invention

Below in conjunction with embodiment, technical scheme is described further.

With reference to Fig. 1, a kind of Automatic mine fault identification method that the present invention provides, identification computer is set and is identifying meter Arranging the micro-seismic event cloud of mine microseism positioning result in calculation machine, the method comprises the steps:

1) reading micro-seismic event cloud, set and wherein have N0 tomography, each tomography has random position and orientation, And the initial value of this parameter is 1.

2) to each seismic events, find out closest with it tomography, constitute initial clustering, then in class Carry out cluster analysis under the constraints of the seismic events distance sum minimum belonging to center and apoplexy due to endogenous wind, i.e. carry out overall situation distance The optimization that sum is minimum.The optimization following computing formula of employing that overall situation distance sum is minimum:

$J=\frac{1}{N_d}\underset{i_c=1}{\overset{N_c}{\mathrm{\Σ}}}\underset{j=1}{\overset{N_{\mathrm{ic}}}{\mathrm{\Σ}}}{D}_j^2$

Wherein, Nc is the number of earthquake bunch, and Nic is the number of earthquake in each earthquake bunch, and Dj is that this bunch of center is arrived in earthquake Distance, Nd is the number of whole event.

3) obtain the covariance matrix of the seismic events of each apoplexy due to endogenous wind, and this covariance matrix is carried out eigenvalue and feature The calculating of vector, including for each class, calculates center and covariance matrix, its covariance matrix of apoplexy due to endogenous wind event For:

$C=\left(\begin{array}{ccc}{\mathrm{\σ}}_x^2& \mathrm{cov}(x,y)& \mathrm{cov}(x,z)\\ \mathrm{cpv}(y,x)& {\mathrm{\σ}}_y^2& \mathrm{cov}(y,z)\\ \mathrm{cov}(z,x)& \mathrm{cov}(z,y)& {\mathrm{\σ}}_z^2\end{array}\right)$

Wherein σ²For it in the variance of respective coordinates value, (a b) is the covariance of respective coordinates value to cov.

On the premise of seismic events is uniformly distributed at random, provided the space bit of tomography corresponding to such by eigenvalue Putting, the length that represent tomography with the eigenvalue of covariance matrix respectively is generous, and Length x Width is its character pair value Times, characteristic vector represents the spatial orientation of tomography.

4) judge whether the maximum of the eigenvalue of minimum is less than the eigenvalue set, if the maximum of the eigenvalue of minimum Value is all little than the eigenvalue set, then calculates and terminates, draw the locus of each tomography, performs the 6th) step；As Fruit still has a minimal eigenvalue more than the eigenvalue set after cluster computing, then this bunch is divided into two sections, always breaks The number of plies increases by one, makes hypocentral location to be preferably attached on tomography, performs next step；

5) the 2nd is returned to) step calculates again, until calculating terminates.

6) last, delete comprise hypocentral location less than 4 bunch, with reference to Fig. 2 and Fig. 3, according to the space of fault plane The eigenvalue of orientation, i.e. each bunch and the value of characteristic vector draw position of fault figure.

Embodiment described above is only to be described the preferred embodiment of the present invention, not to the scope of the present invention It is defined, on the premise of without departing from the present invention relates to spirit, this area ordinary skill technical staff skill to the present invention Various deformation that art scheme is made and improvement, all should fall in the protection domain that claims of the present invention determines.

Claims (3)

1. an Automatic mine fault identification method, arranges identification computer and arranges mine in identifying computer micro-ly The micro-seismic event cloud of shake positioning result, it is characterised in that the method comprises the steps:

1) read micro-seismic event cloud, set and wherein there is N₀Individual tomography, each tomography has random position and orientation, And N₀Initial value be 1；

2) to each seismic events, find out closest with it tomography, constitute initial clustering, then in class Carry out cluster analysis under the constraints of the seismic events distance sum minimum belonging to center and apoplexy due to endogenous wind, i.e. carry out overall situation distance The optimization that sum is minimum；

3) obtain the covariance matrix of the seismic events of each apoplexy due to endogenous wind, and this covariance matrix is carried out eigenvalue and feature The calculating of vector, on the premise of seismic events is uniformly distributed at random, provides tomography corresponding to such by eigenvalue Locus, the length that represent tomography with the eigenvalue of covariance matrix respectively is generous, and Length x Width is its character pair value 'sTimes, characteristic vector represents the spatial orientation of fault plane；

4) judge whether the maximum of the eigenvalue of minimum is less than the eigenvalue set, if the maximum of the eigenvalue of minimum Value is all little than the eigenvalue set, then calculates and terminates, draw the locus of each tomography, performs the 6th) step；As Fruit still has a minimal eigenvalue more than the eigenvalue set after cluster computing, then bunch will be divided into two sections, total tomography Number increases by one, performs next step；

5) the 2nd is returned to) step calculates again, until calculating terminates；

6) last, delete comprise hypocentral location less than 4 bunch, generous according to the length of tomography and fault plane space takes To, i.e. the eigenvalue of each bunch and the value of characteristic vector draws position of fault figure.

Automatic mine fault identification method the most according to claim 1, it is characterised in that in described step 2) in, The optimization following computing formula of employing that overall situation distance sum is minimum:

$J=\frac{1}{N_d}\underset{i_c=1}{\overset{N_c}{\mathrm{\Σ}}}\underset{j=1}{\overset{N_{ic}}{\mathrm{\Σ}}}{D}_j^2$

Wherein, J is target function value, is the meansigma methods of overall situation square distance sum, and Nc is the number of earthquake bunch, and Nic is every The number of earthquake in individual earthquake bunch, Dj is the distance that this bunch of center is arrived in earthquake, and Nd is the number of whole event.

Automatic mine fault identification method the most according to claim 1 and 2, it is characterised in that in described step 3) In, including for each class, calculating center and the covariance matrix of apoplexy due to endogenous wind event, its covariance matrix is:

$C=\left(\begin{array}{ccc}{\mathrm{\σ}}_x^2& \mathrm{cov}(x,y)& \mathrm{cov}(x,z)\\ \mathrm{cov}(y,x)& {\mathrm{\σ}}_y^2& \mathrm{cov}(y,z)\\ \mathrm{cov}(z,x)& \mathrm{cov}(z,y)& {\mathrm{\σ}}_z^2\end{array}\right)$

Wherein σ²For it in the variance of respective coordinates value, cov () is the covariance function of respective coordinates value.