CN103984788A - Automatic intelligent design and optimization system for anchor bolt support of coal tunnel - Google Patents

Abstract

The invention relates to an automatic intelligent design and optimization system for coal roadway bolt support, which includes a roadway surrounding rock index for classifying and evaluating the stability of the roadway surrounding rock by processing the data input by the user or calling the parameters in the relevant knowledge base Acquisition and stability classification subsystem; based on the sample roadway training BP neural network in the internal knowledge base of the system to establish a training model to realize the bolting parameter intelligent matching subsystem of the roadway's bolting parameter intelligent matching; according to the bolting parameter intelligent matching subsystem; The bolt support parameters obtained by the protection parameter intelligent matching subsystem are theoretically checked and calculated based on suspension theory, composite beam theory, composite arch theory, and energy theory. The required support scheme is recommended to the user's support design verification and optimization subsystem as the optimal support scheme. This system effectively guarantees the scientificity and rationality of the bolt support structure and parameters, improves the reliability of the bolt support system, and promotes the healthy development of my country's coal mine roadway bolt support technology.

Description

An automatic intelligent design and optimization system for coal roadway bolt support

technical field

The invention relates to a system for the design and optimization of bolt support in coal mine roadways. the

Background technique

With the development of the coal industry, there are more and more high-yield and high-efficiency working faces in my country, and the output has increased significantly. Coal mine excavation support is a very important production link. Anchor technology is a systematic project, which involves design, construction, support materials, measurement techniques and other aspects. Bolt support design is a key technology in bolt support engineering, which is related to important issues such as the quality of bolt support engineering, whether it is safe and reliable, and whether the economy is reasonable. Due to the complexity and uncertainty of geological conditions, the selection of various parameters in support engineering design, the approximate description of various properties of rock mass and rocks, and the final evaluation of support effects all depend on the experience and personal knowledge of experts. Since quantitative or unified standards cannot be used as criteria, it is inevitable to be arbitrarily and blindly. If the bolt support form and parameter selection are unreasonable, two extremes will often result, or the roadway support strength is not enough, The deformation of the surrounding rock cannot be effectively controlled, which will lead to accidents such as roof collapse in the roadway; or the strength of the roadway support is too high, which not only wastes support materials, but also reduces the speed of roadway excavation, which seriously affects the improvement of mine economic benefits. Strengthening scientific and technological innovation, realizing the management of information resources of coal enterprises, and accelerating the construction of informatization will have far-reaching influence and great effect on the efficient and safe production of coal enterprises. With the rapid economic development and the popularization of computer and its application knowledge, applying computer knowledge to coal mine production design has extremely important theoretical and practical significance. the

Due to the particularity of production in the coal system, the application of computer technology is still relatively backward compared with other industries. In many units in the front line of mine production in our country, the technical work still adopts the traditional human sea tactics and manual labor, which not only wastes a lot of technical force, but also is not suitable for scientific management. Therefore, the application of computer technology in coal mine production design and production management system development is a problem that must be solved first in modern coal mines. the

The current bolt support design system is cumbersome to operate, and the parameters that need to be input by users are too complicated and professional. Many data are difficult to obtain in the general production site, and users are required to have a high professional level. This limits the popularization and use of such systems in production practice. the

Contents of the invention

The purpose of the invention is to provide an intelligent coal roadway bolt support design and optimization system in order to overcome the deficiencies of the prior art. the

In order to achieve the above object, the technical solution adopted in the present invention is: an automatic intelligent design and optimization system for coal roadway bolt support, which includes,

Roadway surrounding rock index acquisition and stability classification subsystem, which realizes the classification and evaluation of roadway surrounding rock stability by processing the data input by the user or calling the parameters in the relevant knowledge base;

The bolt support parameter intelligent matching subsystem, which uses the principle of BP neural network, trains the BP neural network based on the sample roadway in the internal knowledge base of the system, establishes the training model, and realizes the anchoring of the roadway after the specific relevant parameters of the roadway provided by the user Intelligent matching of pole support parameters;

The support design verification and optimization subsystem, which uses the bolt support parameters obtained by the intelligent matching subsystem of the bolt support parameters based on the suspension theory, composite beam theory, and composite arch theory for theoretical verification, and transfers the values The simulation module automatically performs FLAC3D modeling, simulation, and optimization processing, and automatically analyzes various support schemes to select the support scheme that meets the deformation requirements of the surrounding rock of the roadway as the optimal support scheme and recommend it to the user. the

Optimally, the roadway surrounding rock index acquisition and stability classification subsystem automatically completes the sample data preprocessing, standardization, weighting, calibration, and clustering based on the imported sample data and the fuzzy clustering method based on the equivalence relationship. According to the user's selection or the system's automatic judgment, the sample lanes are classified and the cluster centers are given. the

Optimally, a set of basic sample data is created inside the subsystem of roadway surrounding rock index acquisition and stability classification. the

Optimally, the roadway surrounding rock index acquisition and stability classification subsystem selects the roof strength, the strength of the two sides, the floor strength, the depth of the roadway, the initial collapse step of the direct roof, the ratio of the roof height, the width of the coal pillar, the maximum Parameters such as horizontal principal stress are used as the classification and evaluation indicators of the roadway surrounding rock stability. The user inputs the data corresponding to the above parameters of the newly excavated roadway, and the system automatically evaluates the stability of the roadway surrounding rock based on the fuzzy comprehensive evaluation method, and obtains the roadway The surrounding rock stability category. the

Optimally, the intelligent matching subsystem of the bolt support parameters will predict and calculate the roadway parameters for the parameters required for the stability of the surrounding rock of the roadway, and the operation includes firstly analyzing the sample data and the support parameters The knowledge base conducts learning and training to obtain optimized weights and thresholds, and then calculates the input parameters to be predicted with weights and thresholds to obtain the basic support parameters predicted by the system. the

Optimally, the roadway surrounding rock index acquisition and stability classification subsystem, the bolt support parameter intelligent matching subsystem and the support design verification and optimization subsystem are used as an interrelated whole, one-time and systematically The surrounding rock stability classification of the roadway, the intelligent matching of bolt support parameters, and the intelligent optimization of support schemes have been completed. the

Optimally, the roadway surrounding rock index acquisition and stability classification subsystem, the bolt support parameter intelligent matching subsystem, and the support design verification and optimization subsystem are independent according to the actual situation of the user. the

Due to the application of the above technical solutions, the present invention has the following advantages compared with the prior art: This system uses the intelligent expert system to carry out quantitative analysis and optimal design of the bolt support, effectively ensuring the scientific nature of the bolt support structure and parameters It improves the reliability of the bolt support system, realizes the maximum benefit of the bolt support, improves the safety of the support, and fundamentally changes the shortcomings of the engineering analogy method widely used in my country's bolt support. And limitations, promote the development of my country's coal mine roadway bolting technology and digital mine. The system establishes an "expert-level" knowledge base based on relevant experts and typical cases in 8 major mining areas to provide reference data for users to call, so that users can meet the data requirements of bolt design on the premise of only a small input and actually available data; The system inherits the simple interface design and the "fool-like" operation mode, which is easy to grasp and use; the intelligent scheme decision-making improves the rationality of support design and the standardization and standardization of production management; the content-rich "expert-level" roadway support The knowledge base effectively guarantees the scientificity and accuracy of support plan decision-making; the humanized and flexible original parameter input and selection mechanism truly realizes the practical, usable, usable, and easy-to-use development goals of the system; perfect instruction for use And the help system can effectively improve the design level of technical personnel. the

Description of drawings

Accompanying drawing 1 is the automatic intelligent design and optimization system flowchart of coal roadway bolt support of the present invention;

Accompanying drawing 2 is the flow chart of roadway surrounding rock index acquisition and corresponding stability classification system of the present invention;

Accompanying drawing 3 is the intelligent matching subsystem flowchart of bolting parameter of the present invention;

Accompanying drawing 4 is support design verification and optimization subsystem flowchart of the present invention;

Detailed ways

Preferred embodiment of the present invention will be described in detail below in conjunction with accompanying drawing:

The automatic intelligent design and optimization system of coal roadway bolt support of the present invention, as shown in Figure 1, includes the roadway surrounding rock index acquisition and stability subcategory system, bolt support parameter intelligent matching subsystem, support design verification and optimization subsystems. details as follows:

1. Roadway Surrounding Rock Index Acquisition and Stability Classification Subsystem

As shown in Figure 2, in the roadway surrounding rock index acquisition and stability classification subsystem, the system selects roof strength _σtop , two sides strength _σside , floor strength _σbottom , roadway burial depth H, direct roof initial collapse Eight parameters, including step distance L, top-height ratio N, coal pillar width B, and maximum horizontal principal stress σ _h , are used as the classification and evaluation indexes of roadway surrounding rock stability. Users can input relevant data through the operation interface. At the same time, the system also collects relevant data through investigation, analysis, and collation. Choose.

When the subsystem is initially used, a certain number of typical roadway samples need to be imported to complete the classification of surrounding rock stability. The system can automatically complete the preprocessing, standardization, weighting, calibration, clustering and other processing of the sample data according to the imported sample data and the fuzzy clustering method based on the equivalence relationship, and can automatically determine the sample roadway according to the user's selection or the system's automatic judgment. Classify and give the cluster center. There is also a set of basic samples inside the system, so that users can call when it is difficult to obtain samples. the

After completing the classification work, the user inputs 8 relevant data of the newly excavated roadway, and the system automatically judges the stability of the surrounding rock of the roadway based on the fuzzy comprehensive evaluation method, and obtains the stability category of the surrounding rock of the roadway. the

The specific processing method is:

(1) Stability clustering of roadway surrounding rock

1. Input data preprocessing

The information of n roadway samples will be input, and each roadway has 8 indicators, which are the roof strength _σtop , the strength of the two sides _σside , the floor strength _σbottom , the buried depth of the roadway H, the initial collapse step of the direct roof L, and the roof height Ratio N, coal pillar width B, maximum horizontal principal stress σ _h . Then it is an n×8 matrix.

Data preprocessing:

When the coal is soft coal ( _σ <10MPa), the value of W' is as follows:

${\mathrm{<span\; class="notranslate">\; W</span>}}^{<span\; class="notranslate">\; ,</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2.6</span>}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; -</span>\frac{{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}{\mathrm{<span\; class="notranslate">\; 3</span>}}}{{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>$

When the coal is medium-hard (10MPa< _σb <20MPa), the value of W' is as follows:

${\mathrm{<span\; class="notranslate">\; W</span>}}^{<span\; class="notranslate">\; ,</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3.6</span>}\frac{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; -</span>\frac{{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}{\mathrm{<span\; class="notranslate">\; 4</span>}}}{{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ]</span>$

When the coal is hard (20MPa< _σ ), the value of W' is as follows:

${\mathrm{<span\; class="notranslate">\; W</span>}}^{<span\; class="notranslate">\; ,</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0.3</span>}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 3.6</span>}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; -</span>\frac{{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}{\mathrm{<span\; class="notranslate">\; 4</span>}}}{{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$

Select the B ₀ value according to the table below

The buried depth here is the H value entered above,

If N>4, take 4, and H, L, and σ _h will not be processed.

2. Data standardization

Transform the i-th index x _ij of the j-th sample into x′ _ij , namely

${{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; -</span>{\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}{{\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}$

(2) Poor regularization

Transform the i-th index x′ _ij of the j-th sample after standard deviation normalization into x″ _ij , namely

${\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; \{</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{<span\; class="notranslate">\; \}</span>}_{\mathrm{<span\; class="notranslate">\; min</span>}}}{<span\; class="notranslate">\; \{</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{<span\; class="notranslate">\; \}</span>}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; \{</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}^{<span\; class="notranslate">\; \&prime;</span>}{<span\; class="notranslate">\; \}</span>}_{\mathrm{<span\; class="notranslate">\; min</span>}}}$

3. Weighting

After the data has been processed above, each indicator must be multiplied by the corresponding weight. the

4. Calibration

Calculate the similarity coefficient γ _ij

${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ik</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; jk</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; )</span>}{\sqrt{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; ik</span>}}<span\; class="notranslate">\; -</span>\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}\sqrt{{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}^{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\text{<span class="notranslate"> jk</span>}}\text{<span class="notranslate"> -</span>}\stackrel{<span\; class="notranslate">\; \&OverBar;</span>}{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}$

After calibration, it is obtained: 0≤γ _ij ≤1, (i=1, 2,..., m; j=1, 2,..., m); then the fuzzy relationship matrix can be determined

5. Cluster analysis

The similarity matrix obtained by calibration The transitive package R ^* obtained by the square method contains The minimum fuzzy equivalent matrix, and then according to the λ-cut relationship of R ^* , perform dynamic clustering analysis on X, the ones smaller than λ are recorded as 0, and the ones larger than λ are recorded as 1, and finally the identical rows are classified into one category,

6. Determine the optimal number of categories

F-statistics method: Let r be the number of categories corresponding to λ, the number of samples of the jth class is n _j , and the average value of the kth feature of the jth class is Make F-statistics: $f=\frac{\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}{\mathrm{no}}_j\&Center\; Dot;\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}\frac{{(\stackrel{\&OverBar;}{x_k^{\left(j\right)}}-\stackrel{\&OverBar;}{x_k})}^2}{r-1}}{\underset{j=1}{\overset{r}{\mathrm{\&Sigma;}}}\underset{i=1}{\overset{{\mathrm{no}}_j}{\mathrm{\&Sigma;}}}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}\frac{{(x_{\mathrm{ik}}^{\left(j\right)}-\stackrel{\&OverBar;}{x_k^{\left(j\right)}})}^2}{\mathrm{no}-r}}$

The F-statistic is an F-distribution with (r-1, n-1) degrees of freedom. Its numerator represents the distance between classes and the denominator represents the distance between samples within a class. Therefore, the larger the F value, the greater the distance between classes, that is, the greater the difference between classes, the better the classification effect. the

If F>F _1-λ (r-1, nr) (α=0.05), according to the variance analysis theory of mathematical statistics, it can be known that the difference between classes is significant, indicating that the classification is optimal.

If there is more than one F value that satisfies F>F _1-λ (r-1, nr), the size of FF _α can be further investigated, and a satisfactory F value can be selected from the larger one as the optimal classification.

After classification, the average value of each index of each sample roadway is taken to obtain the (preliminary) cluster center. the

(2) Fuzzy comprehensive evaluation of roadway surrounding rock stability

1. Construct fuzzy relationship matrix

Construct the fuzzy relationship matrix R={r _ij }9*5, where r _ij represents the degree of membership of the j-th category from the perspective of factor i. The r _ij calculation method can be selected from the following methods:

Method 1. Normal membership function method:

${\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; (</span>\frac{{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}}{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$

2. Fuzzy transformation

The set C={c ₁ , c ₂ , c ₃ , c ₄ , c ₅ , c ₆ , c ₇ , c ₈ , c ₉ } is the influence weight of 9 factors (same as the previous procedure). Do fuzzy transformation B=C ^o R. Fuzzy vector B={b ₁ , b ₂ , b ₃ , b ₄ , b ₅ }T={b j}, where b ₁ , b ₂ , b ₃ , b ₄ , b ₅ represent the clusters of the roadway to be predicted The degree of membership of the center, the category of the roadway to be predicted is determined by the samples with a high degree of membership between the roadway and the clustering center.

The algorithm of b _j has the following four methods:

Method 1. Determinant type of main factor: Among them, ∧ is the smaller symbol of the two, and ∨ is the larger symbol of the two;

Method 2. Main factor prominent type:

Method 3. Main factor prominent type:

Method 4. Weighted average type:

Then the evaluation matrix B={b ₁ , b ₂ , b ₃ , b ₄ , b ₅ }T={b _j } can be obtained

3. Evaluation index processing:

Method 1, the maximum degree of membership method:

The subscript j corresponding to the maximum value among b ₁ to b ₅ is the type of roadway to be tested.

Method 2: Weighted average method:

Let v'=(b ₁ +2b ₂ +3b ₃ +4b ₄ +5b ₅ )/(b ₁ +b ₂ +b ₃ +b ₄ +b ₅ ), then v' is close to which category the roadway belongs to 】

2. Intelligent matching subsystem of bolt support parameters

In this subsystem, as shown in Figure 3, the required parameters, such as the initial caving step of the direct roof, the strength of the roof, the strength of the floor, the buried depth of the roadway, the clear width of the roadway, and the clear height of the roadway, are input to the interface , and then predict the roadway parameters and perform calculations. The calculation is mainly carried out in the background of the system, that is, the system first learns and trains the samples and the knowledge base of support parameters, obtains the optimized weights and thresholds, and then conducts a series of calculations on the input parameters to be predicted, weights, and thresholds. Finally, the system predicted support basic parameter values are obtained. the

The first step is to initialize the network and normalize the input and output. the

Assign a random number in the interval (-1, 1) to each connection weight, set the error function ε, and give the calculation accuracy value and the maximum number of learning times. the

The second step is to randomly select the kth input sample and the corresponding expected output

d _o (k) = (d ₁ (k), d ₂ (k), ..., d _q (k))

x(k)=(x ₁ (k), x ₂ (k), . . . , x _n (k))

The third step is to calculate the input and output of each neuron in the hidden layer. the

The fourth step is to use the expected output and the actual output of the network to calculate the partial derivative δ _o (k) of the error function to each neuron in the output layer.

The fifth step is to use the connection weights from the hidden layer to the output layer, δ _o (k) of the output layer and the output of the hidden layer to calculate the partial derivative δ _h (k) of the error function to each neuron in the hidden layer.

The sixth step is to use the δ _o (k) of each neuron in the output layer and the output of each neuron in the hidden layer to correct the connection weight who _ho (k)

The seventh step is to use the δ _h (k) of each neuron in the hidden layer and the input of each neuron in the input layer to modify the connection weight.

The eighth step, calculate the global error $\mathrm{E.}=\frac{1}{2m}\underset{k=1}{\overset{m}{\mathrm{\&Sigma;}}}\underset{o=1}{\overset{q}{\mathrm{\&Sigma;}}}{(d_o\left(k\right)-{\mathrm{the\; y}}_o\left(k\right))}^2.$

The ninth step is to judge whether the network error meets the requirements. When the error reaches the preset accuracy or the number of learning times is greater than the set maximum number of times, the algorithm ends. Otherwise, select the next learning sample and the corresponding expected output, return to the third step, and enter the next round of learning. the

3. Support design verification and optimization subsystem

As shown in Figure 4, the bolt support parameters obtained above are automatically transferred into the subsystem. Firstly, according to the theoretical calculation method, the design scheme is automatically imported and carried out based on suspension theory, composite beam theory, and composite arch theory. , Energy theory calculation. After the experience calculation of the design scheme meets the requirements, the background is called to obtain a series of support design schemes based on the orthogonal test module, and these schemes are transferred into the numerical simulation module to automatically perform FLAC3D modeling, simulation, optimization, etc. According to the analysis of the protection scheme, the most economical support scheme that meets the deformation requirements of the roadway surrounding rock is selected as the optimal support scheme and recommended to the user. the

The system includes three subsystems: the intelligent classification subsystem of surrounding rock stability, the intelligent matching subsystem of bolt support parameters, and the intelligent optimization subsystem of support schemes. The three subsystems can be used as an interrelated whole to complete the stability classification of the roadway surrounding rock, the intelligent matching of bolt support parameters, and the intelligent optimization of support schemes in a one-time system, or according to the actual situation of the user , use one of the modules independently to realize related functions, so that the system has stronger applicability and selectivity. It is conducive to the stable, efficient and safe functions of the system, and at the same time ensures the coordination, consistency and sustainability of the subsystems. the

The above-mentioned embodiments are only to illustrate the technical concept and characteristics of the present invention. Equivalent changes or modifications made in the spirit shall fall within the protection scope of the present invention. the

Claims (7)

1. a coal entry anchor rod support automated intelligent design and optimization system, is characterized in that: it comprises,

Roadway surrounding rock index selection and Stability Classification subsystem, it,, by processing the data of user's input or calling the parameter in relevant knowledge storehouse, is realized the classification of improving stability of surrounding rocks in roadway and assessment;

Bolting Parameters Intelligent Matching subsystem, it is by using BP neural networks principles, BP neural network is practiced in sample tunnel training based in internal system knowledge base, sets up training pattern, realizes the Bolting Parameters Intelligent Matching in this tunnel after the concrete correlation parameter in the tunnel that user provides;

Design of its support checking and optimization subsystem, its Bolting Parameters obtaining according to described Bolting Parameters Intelligent Matching subsystem is theoretical based on suspention, give close beam theory, built-up arch theory is carried out theory and is checked, and call in numerical simulation module, automatically carry out FLAC3D modeling, simulation, optimization process, automatically various supporting schemes are analyzed, usingd and choose the supporting scheme that meets deformation of the surrounding rock in tunnel requirement and recommend user as Optimum Support scheme.

2. coal entry anchor rod support automated intelligent design and optimization system according to claim 1, it is characterized in that: described roadway surrounding rock index selection and Stability Classification subsystem are automatically according to the sample data importing, pre-service, pushing away of mark, weighting, demarcation, clustering processing according to the automatic completed sample certificate of fuzzy clustering algorithm based on relation of equivalence, and according to user's selection or system automatic decision, classified in sample tunnel, cheat out cluster centre.

3. coal entry anchor rod support automated intelligent design and optimization system according to claim 2, is characterized in that: the inner establishment of described system roadway surrounding rock index selection and Stability Classification subsystem has a set of basic sample data.

4. coal entry anchor rod support automated intelligent design and optimization system according to claim 2, it is characterized in that: described roadway surrounding rock index selection and Stability Classification subsystem are selected strength of roof, two help intensity, base plate strength, tunnel buried depth, first roof caving step pitch, rise ratio, coal pillar width, the parameters such as maximum horizontal principal stress are as Surrounding Rock Stability Classification in Tunnel and evaluation index, user inputs the data corresponding with above-mentioned parameter in new pick tunnel, system is passed judgment on this improving stability of surrounding rocks in roadway based on fuzzy comprehensive evaluation method automatically, obtain the stability of surrounding rock classification in this tunnel.

5. coal entry anchor rod support automated intelligent design and optimization system according to claim 1, it is characterized in that: described Bolting Parameters intelligence can be mated subsystem, it will carry out the prediction union of tunnel parameter for the required parameter of improving stability of surrounding rocks in roadway, described computing comprises first carries out learning training to sample data and supporting parameter knowledge base, weights after being optimized and threshold values, then by the parameter to be predicted of having inputted, calculate with weights, threshold values, obtain the supporting basic parameter value of system prediction.

6. according to the coal entry anchor rod support automated intelligent design and optimization system described in claim l, it is characterized in that: the checking of described roadway surrounding rock index selection and Stability Classification subsystem and Bolting Parameters Intelligent Matching subsystem and design of its support is used as the integral body that is mutually related with optimizing subsystem, disposable and systematically complete surrounding rock mass stability classification, Bolting Parameters Intelligent Matching, the work of supporting scheme intelligent optimization in tunnel.

7. coal entry anchor rod support automated intelligent design and optimization system according to claim 1, is characterized in that: described roadway surrounding rock index selection and Stability Classification subsystem and Bolting Parameters Intelligent Matching subsystem and design of its support checking are independent respectively according to user's actual conditions with optimization subsystem.