CN105302934A - Multi-objective intelligent optimization design method for anchoring and protecting network structure of coal mine underground roadway - Google Patents
Multi-objective intelligent optimization design method for anchoring and protecting network structure of coal mine underground roadway Download PDFInfo
- Publication number
- CN105302934A CN105302934A CN201510405763.9A CN201510405763A CN105302934A CN 105302934 A CN105302934 A CN 105302934A CN 201510405763 A CN201510405763 A CN 201510405763A CN 105302934 A CN105302934 A CN 105302934A
- Authority
- CN
- China
- Prior art keywords
- anchor
- model
- coal mine
- care network
- objective
- 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.)
- Granted
Links
Landscapes
- Lining And Supports For Tunnels (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Provided is a multi-objective intelligent optimization design method for an anchoring and protecting network structure of a coal mine underground roadway, pertaining to a design method for an anchoring and protecting network structure.The multi-objective intelligent optimization design method is characterized in that parameters for the anchoring and protecting network structure of the coal mine underground roadway are designed to model a multi-objective optimization problem; support quality and support cost are utilized as target functions; as for support quality, a support quality surrogate model is set up; multiple constraints present in distribution of anchor rods are taken into consideration; a two-objective optimization model for parameter design of anchoring and protecting network structure for the coal mine underground roadway is set up; finally, the multi-objective intelligent optimization method is adopted for solving the above-mentioned two-objective optimization model in order to obtain an optimal scheme set for parameters of the anchoring and protecting network structure; and an anchoring and protecting scheme for solving the set is adopted in order to guide design of the anchoring and protecting network structure of the roadway. The multi-objective intelligent optimization design method for the anchoring and protecting network structure of the coal mine underground roadway has following beneficial effects: the multi-objective intelligent optimization design method is used for design of the anchoring and protecting network structure of the coal mine underground roadway so that reasonability and scientificity of the anchoring and protecting network structure are effectively ensured; underground safety and economical construction of a coal mine are thus guaranteed.
Description
Technical field
The present invention relates to a kind of Anchor Care network structure method for designing, be specifically related to a kind of Multiobjective Intelligent Optimization Design of coal mine down-hole tunnel Anchor Care network structure.
Background technology
Coal is one of core energy of China, and corresponding coal seam geological condition is complicated and changeable.Particularly, the seam mining degree of depth constantly increased in recent years, caused country rock rock burst to aggravate, and after section has tunneled, only relied on country rock self stress cannot ensure that being caving does not appear in tunnel, the phenomenon of roof fall.Carry out roadway support timely and effectively most important.At present, bolt supporting, as a kind of support pattern of active, is the main development direction of roadway support.The Anchor Care network structure be made up of anchor pole improves driving stability of surrounding rock, one of most economical, effective method of reply deep fractures complex stress condition.The rationality of Anchor Care network architecture parameters design directly determines its supporting effect.The anchor pole quantity designed in network structure is more, although fully can ensure stability of surrounding rock, can cause raising production cost, slow down integral production progress; Otherwise, be difficult to maintain the stable of roadway surrounding rock, increase the danger that geologic hazard occurs.Visible, Anchor Care network structure reasonable in design is most important for adjoining rock stability.
At present, coal mine down-hole tunnel Anchor Care network structure method for designing mainly comprises: (1) engineering analog method.The method, based on engineering experience, by the alley way anchor protection structure network with existing successful experience, carries out Analogy.In recent years, researchist is by summing up these engineering experiences, and exploitation expert decision system, establishes the knowledge base for Rock Stability Classification and design of its support.Patent CN103984788A proposes a kind of coal entry anchor rod support automated intelligent design and optimization system, this system is by using BP neural network, based on the sample tunnel training BP neural network in internal system knowledge base, set up training pattern, realize designing the Bolting Parameters in tunnel, then numerical simulation can be adopted to verify supporting scheme and optimize, and under choosing the prerequisite meeting deformation of the surrounding rock in tunnel requirement, most economical supporting scheme is as Optimum Support scheme; (2) method for numerical simulation.The method is by simulating the mathematical model of tunnel environment and bolt supporting structure, complete the calculating to pressure from surrounding rock, structural internal force, stability quantitatively, and the design etc. of roadway shape and supporting construction, the situation making bolt supporting network design depend on experience is improved.A kind of coal mine roadway bolt supporting Intelligentized design method is proposed in patent CN101968825A, in the method according to Surrounding Rock Stability Classification in Tunnel result, from macroscopically several feasible Bolting Parameters of initial option, orthogonal experiment is adopted to arrange representative various testing programs, carry out numerical evaluation by numerical simulation software, determine final supporting parameter; (3) based on the optimization method of Mathematical Models.In tunnel support parameter designing, being regarded as by supporting parameter design is a single-object problem, founding mathematical models, and adopts optimized algorithm to solve mathematical model, obtains optimum tunnel support parameter.In paper " Feedbackanalysisoftunnelconstructionusingahybridarithmet icbasedonSupportVectorMachineandParticleSwarmOptimisatio n ", concrete thickness and these two supporting parameters of concrete yang type modulus are optimized, be specially: adopt orthogonal experiment and Method for Numerical to produce data sample, support vector machine is utilized to carry out identification to relation between roof-to-floor convergence and supporting parameter, set up the agent model based on support vector machine, and adopt optimized algorithm to solve model, obtain optimum supporting parameter.
Above-mentioned technological achievement is that the design of coal mine roadway Anchor Care network architecture parameters provides solution, but, it should be pointed out that existing method still exists following deficiency: (1) adopts engineering analog method too will rely on engineering experience, has larger randomness and blindness.Owing to may there is very large otherness between different mine geology condition, can not realize comprehensively collecting the alley way anchor protection structure network parameter of successful experience existing under different geological conditions, thus lack can analogy case history, design proposal may be caused the most at last to there is potential safety hazard or too conservative; (2) if simple employing method for numerical simulation, computational resource and the computing time of at substantial is needed, in limited design time, the method cannot run abundant number of times, thus is difficult to, in whole decision variable space, search the supporting parameter of global optimum.If by engineering analog method, carrying out numerical simulation from macroscopically selecting several feasible supporting parameter, there is the weak point mentioned in first equally; (3) tunnel and colliery all have larger difference in a lot.First the geologic condition residing for tunnel is relatively simple, buried depth is little, suffered terrestrial stress is little, the digging mode in tunnel is also all different from coal mine roadway with requiring simultaneously, the support pattern adopted is bolt-spray support, thus the decision variable number will optimized is few, so higher for the design complexity of coal mine down-hole tunnel Anchor Care network structure.In addition, using the support parameter optimization in tunnel as a single-object problem, be as optimization aim using supporting quality, or using support cost as optimization aim, supporting quality, as constraint condition, can not obtain the supporting scheme collection that a group considers the optimum of supporting quality and support cost.
Summary of the invention
Technical matters to be solved by this invention is: the Multiobjective Intelligent Optimization Design proposing a kind of coal mine down-hole tunnel Anchor Care network structure, is intended to the optimum Anchor Care scheme that acquisition one group takes into account supporting quality and support cost.
Technical solution of the present invention: the Model for Multi-Objective Optimization setting up the design of coal mine down-hole tunnel Anchor Care network architecture parameters, and adopt Multiobjective Intelligent optimization method to solve the multi-objective Model set up, realize the optimization to Anchor Care network structure; Method step is as follows:
1. set up the supporting quality agent model of Anchor Care network architecture parameters
The combination supporting form that Anchor Care network adopts anchor pole and anchor cable to form; Under roadway support quality is reflected in Anchor Care role of network, the extent of stability of roadway surrounding rock, measured by tunnel roof-to-floor convergence and the roadway's sides amount of shifting near, identification is carried out to the nonlinearity relation between Anchor Care network architecture parameters and rock deformation, obtains the agent model of supporting quality;
Supporting quality agent model adopts following steps to obtain:
Step1: adopt orthogonal experiment and method for numerical simulation, produces sample data collection;
Step2: adopt supervised learning method training sample data, identification is carried out to the nonlinearity relation between Anchor Care network architecture parameters and rock deformation, obtains the agent model of supporting quality;
2. take into account the multi-objective optimization design of power model of supporting quality and support cost
For meeting security and the cost-effectiveness requirement of coal mining, the corresponding evaluation index of Anchor Care network instruction plan comprises two parts; One, adopts supporting quality Quality (x) to embody security requirement; Its two, adopt support cost Cost (x) embody cost-effectiveness requirement; Take into account supporting quality and support cost two aspect requirement, set up two objective optimization model of coal mine down-hole tunnel Anchor Care network architecture parameters; Two objective optimization model are: for supporting quality, set up the agent model based on vector machine; For support cost, choose tunnel and set the total price of support material needed for anchor pole and anchor cable to represent the mathematical model of support cost, by decision variable span in actual applications, be added in mathematical model as constraint condition, finally set up two objective optimization model of coal mine down-hole tunnel Anchor Care network architecture parameters design;
Two objective optimization model: Quality (x)=SVM (x)
Wherein, x
1∈ { x
1(1), x
1(2) ..., x
1(n) }; x
2∈ { x
2(1), x
2(2) ..., x
2(n) }; x
4∈ { x
4(1), x
4(2) ..., x
4(n) }; x
6, x
7, x
8>=min; Min_x
9≤ x
9≤ max_x
9; Min_x
10≤ x
10≤ max_x
10; Min_x
11≤ x
11≤ max_x
11;
In formula, x={x
1, x
2..., x
11represent Anchor Care network architecture parameters; x
1represent bolt diameter, { x
1(1), x
1(2) ..., x
1(n) } represent its discrete span; x
2represent rock-bolt length, { x
2(1), x
2(2) ..., x
2(n) } represent its discrete span; x
3{ 1,2,3} represents anchor cable distribution form to ∈, and wherein, 1 representative often to be arranged along drift section and put an anchor cable, and 2 representatives establish two anchor cables and even rows lays an anchor cable along the arrangement of drift section odd number, and 3 representatives are often arranged along drift section and established two anchor cables; x
4represent anchor cable diameter, { x
4(1), x
4(2) ..., x
4(n) } represent its discrete span; x
5{ 0,1,2,3,4} represents the multiple proportion between anchor cable array pitch and anchor pole array pitch to ∈, and wherein, 0 representative does not lay anchor cable along drift section; x
6represent anchor pole top spacing; x
7represent anchor pole array pitch; x
8represent anchor pole side spacing; x
9represent anchor cable length, min_x
9for anchor cable length minimum value, max_x
9for anchor cable length maximal value; x
10represent bolt pretension, min_x
10for bolt pretension minimum value, max_x
10for bolt pretension maximal value; x
11represent anchor cable pretightning force, min_x
11for anchor cable pretightning force minimum value, max_x
11for anchor cable pretightning force maximal value.If C
gfor unit quality anchor pole cost, C
sfor unit quality anchor cable cost, K
dfor back being laid the range size of anchor pole, K
cfor roadway's sides being laid the range size of anchor pole, L is tunnel length, ρ
gfor anchor pole density, ρ
sfor the density of anchor cable;
3. adopt evolution algorithm to solve object and multi object mathematical model
Adopt the Model for Multi-Objective Optimization of taking into account supporting quality and support cost of multi-objective Evolutionary Algorithm to above-mentioned foundation to solve, obtain Pareto optimal solution set, realize Anchor Care network architecture parameters optimal design.
Multi-objective Evolutionary Algorithm is adopted to solve the two target mathematical models set up, wherein respectively using supporting quality and support cost as two objective functions, obtain optimum Anchor Care network architecture parameters, form the Multiobjective Intelligent Optimization Design of complete coal mine down-hole tunnel Anchor Care network structure, instruct the design of coal mine roadway Anchor Care network architecture parameters.
Beneficial effect, owing to have employed such scheme, utilizes method for numerical simulation, and based on machine Learning Theory, build the supporting quality agent model of reflection stability of surrounding rock, its calculation cost significantly reduces than simple method for numerical simulation; Consider the security in coal mining and cost-effectiveness requirement, using supporting quality and support cost as objective function, take into account the multiple constraint that anchor pole layout exists, set up the Model for Multi-Objective Optimization of coal mine down-hole tunnel Anchor Care network architecture parameters design; Adopt and use the above-mentioned Model for Multi-Objective Optimization of Multiobjective Intelligent Optimization Method, obtain its Pareto optimal solution set, i.e. best Anchor Care network architecture parameters scheme collection.Concentrate in the program and not only can obtain the best Anchor Care strategy parameter of supporting quality, the minimum Anchor Care strategy parameter of support cost can also be obtained, the Anchor Care scheme taking into account supporting quality and support cost can also be obtained in addition, colliery is based on the supporting scheme of this disaggregation, instruct the practical operation of alley way anchor protecting wire net network, thus ensure underground coal mine production safety, and economic construction.
Anchor Care network architecture parameters purpose of design is under the prerequisite ensureing adjoining rock stability, the reinforcement ability of anchor pole and the self-bearing ability of country rock are all not fully exerted, realize anchor pole-country rock jointly to carry, thus not only obtain good consolidation effect, and save anchor pole quantity, reduce support material consumption and support cost, improve support speed.
Advantage: (1), in coal mine down-hole tunnel Anchor Care network, supporting parameter directly affects supporting quality and the support cost in tunnel.Supporting quality and support cost two aspect restrict again between requiring mutually simultaneously, and obtaining the Optimum Support scheme collection taking into account these two aspect performance requirements is a NP-hard problem.The present invention just with supporting quality and support cost for objective function, be a multi-objective optimization question by underground coal mine Anchor Care network parameter design setting model, and then adopt Multipurpose Optimal Method to solve, be intended to the supporting scheme collection that acquisition one group considers the optimum of supporting quality and support cost.
(2) calculation cost of evaluate alternatives is reduced by setting up supporting quality agent model.The present invention adopts orthogonal experiment and method for numerical simulation to produce set of data samples, utilizes support vector machine to establish the agent model of underground coal mine supporting quality.The foundation of this agent model, on the one hand, compared with the simple method adopting numerical simulation method to find optimum parameter combinations in a large amount of Anchor Care network architecture parameters combines, significantly reduces the calculation cost of evaluate alternatives; On the other hand, making to rely on has had the situation of experience to alley way anchor protecting wire net network structural design to be improved, significantly reduce the randomness in design process, can realize in the combination of a large amount of Anchor Care network architecture parameters, carrying out optimizing, to find optimum Anchor Care scheme.
(3) adopt Multipurpose Optimal Method to solve the multi-objective Model set up, obtain Pareto optimal solution set, realize Anchor Care network architecture parameters optimal design, obtain the optimum Anchor Care scheme that a group is taken into account supporting quality and support cost.
Accompanying drawing explanation
Fig. 1 is the Multiobjective Intelligent Optimization Design process flow diagram of coal mine down-hole tunnel Anchor Care network structure of the present invention.
Fig. 2 is Anchor Care network structure optimization design scheme collection of the present invention.
Concrete implementing measure
Below in conjunction with concrete accompanying drawing and example, the embodiment to institute of the present invention extracting method is described in detail.
Set up the Model for Multi-Objective Optimization of coal mine down-hole tunnel Anchor Care network architecture parameters design, and adopt Multiobjective Intelligent optimization method to solve the multi-objective Model set up, realize the optimization to Anchor Care network structure; Method step is as follows:
1. set up the supporting quality agent model of Anchor Care network architecture parameters
The combination supporting form that Anchor Care network adopts anchor pole and anchor cable to form; Under roadway support quality is reflected in Anchor Care role of network, the extent of stability of roadway surrounding rock, is measured by tunnel roof-to-floor convergence and the roadway's sides amount of shifting near, and supporting quality agent model is a multiple input single output model;
Supporting quality agent model adopts following steps to obtain:
Step1: adopt orthogonal experiment and method for numerical simulation, produces sample data collection;
Step2: adopt supervised learning method training sample data, identification is carried out to the nonlinearity relation between Anchor Care network architecture parameters and rock deformation, obtains the agent model of supporting quality;
2. take into account the multi-objective optimization design of power model of supporting quality and support cost
For meeting security and the cost-effectiveness requirement of coal mining, the corresponding evaluation index of Anchor Care network instruction plan comprises two parts; One, adopts supporting quality Quality (x) to embody security requirement; Its two, adopt support cost Cost (x) embody cost-effectiveness requirement; Take into account supporting quality and support cost, two objective optimization model setting up coal mine down-hole tunnel Anchor Care network architecture parameters are: for supporting quality, set up the supporting quality agent model based on vector machine; For support cost, choose tunnel and set the total price of support material needed for anchor pole and anchor cable to represent the mathematical model of support cost, by decision variable span in actual applications, be added in mathematical model as constraint condition, finally set up two objective optimization model of coal mine down-hole tunnel Anchor Care network architecture parameters design;
Two objective optimization model: Quality (x)=SVM (x)
Wherein, x
1∈ { x
1(1), x
1(2) ..., x
1(n) }; x
2∈ { x
2(1), x
2(2) ..., x
2(n) }; x
4∈ { x
4(1), x
4(2) ..., x
4(n) }; x
6, x
7, x
8>=min; Min_x
9≤ x
9≤ max_x
9; Min_x
10≤ x
10≤ max_x
10; Min_x
11≤ x
11≤ max_x
11;
In formula, x={x
1, x
2..., x
11represent Anchor Care network architecture parameters; x
1represent bolt diameter, { x
1(1), x
1(2) ..., x
1(n) } represent its discrete span; x
2represent rock-bolt length, { x
2(1), x
2(2) ..., x
2(n) } represent its discrete span; x
3{ 1,2,3} represents anchor cable distribution form to ∈, and wherein, 1 representative often to be arranged along drift section and put an anchor cable, and 2 representatives establish two anchor cables and even rows lays an anchor cable along the arrangement of drift section odd number, and 3 representatives are often arranged along drift section and established two anchor cables; x
4represent anchor cable diameter, { x
4(1), x
4(2) ..., x
4(n) } represent its discrete span; x
5{ 0,1,2,3,4} represents the multiple proportion between anchor cable array pitch and anchor pole array pitch to ∈, and wherein, 0 representative does not lay anchor cable along drift section; x
6represent anchor pole top spacing; x
7represent anchor pole array pitch; x
8represent anchor pole side spacing; x
9represent anchor cable length, min_x
9for anchor cable length minimum value, max_x
9for anchor cable length maximal value; x
10represent bolt pretension, min_x
10for bolt pretension minimum value, max_x
10for bolt pretension maximal value; x
11represent anchor cable pretightning force, min_x
11for anchor cable pretightning force minimum value, max_x
11for anchor cable pretightning force maximal value.If C
gfor unit quality anchor pole cost (unit/kg), C
sfor unit quality anchor cable cost (unit/kg), K
dfor back being laid the range size of anchor pole, K
cfor roadway's sides being laid the range size of anchor pole, L is tunnel length, ρ
gfor anchor pole density, ρ
sfor the density of anchor cable;
3. adopt evolution algorithm to solve object and multi object mathematical model
Multi-target evolution optimization method can realize many aspects to require as target, find global optimum, the Model for Multi-Objective Optimization of taking into account supporting quality and support cost of multi-objective Evolutionary Algorithm to above-mentioned foundation is adopted to solve, obtain Pareto optimal solution set, realize Anchor Care network architecture parameters optimal design.
In the design of coal mine down-hole tunnel Anchor Care network structure, for the foundation of supporting quality agent model, adopt orthogonal experiment and method for numerical simulation to combine and produce sample data collection, and set of data samples is normalized, then based on above-mentioned sample data collection, adopting support vector machine to carry out identification to having nonlinearity relation between supporting parameter and rock deformation, obtaining the agent model of supporting quality.For the foundation of support cost mathematical model, the total price adopting tunnel to set support material needed for anchor pole and anchor cable represents.
Be two objective optimisation problems by underground coal mine Anchor Care network parameter design setting model, with supporting quality and support cost for objective function, take into account the multiple constraint that anchor pole layout exists simultaneously, set up two objective optimization model of coal mine down-hole tunnel Anchor Care network architecture parameters design.
Multi-objective Evolutionary Algorithm is adopted to solve the two target mathematical models set up, wherein respectively using supporting quality and support cost as two objective functions, obtain optimum Anchor Care network architecture parameters, form the Multiobjective Intelligent Optimization Design of complete coal mine down-hole tunnel Anchor Care network structure, instruct the design of coal mine roadway Anchor Care network architecture parameters.
Concrete:
For in the assistant conveyance gate road digging process of in coal seam, the optimization of its Anchor Care scheme is as carried example.The geologic parameter in this tunnel known is: tunnel adopts rectangular cross section, and wide 5.2m, high 3.7m, long 1000m are designed in tunnel.Tunnel buried depth 180m.Intend each rock layer mechanics parameter in digging upper and lower, tunnel as shown in Table 1 and Table 2, below according to the geologic condition parameter in this tunnel to be designed, point five aspects are explained in detail the concrete steps adopting method of the present invention to obtain optimum supporting scheme.
The each rock layer mechanics parameter in top, table 1 tunnel
The each rock layer mechanics parameter in bottom, table 2 tunnel
1. choose decision variable
The combination supporting form that the Anchor Care network that the present invention considers adopts anchor pole and anchor cable to form, in this kind of support form, the Anchor Care network architecture parameters determined comprises: (1) anchor pole and anchorage cable materials.Wherein, anchor pole material is screw-thread steel, and anchorage cable materials is steel strand wires.(2) anchor pole and anchor cable set angle.Wherein, anchor pole is inclined upwardly 45 degree at rectangular shaped roadways shoulder nest place surface level and squeezes into shoulder nest anchor pole, and downward-sloping 30 degree of base angle place surface level squeezes into FLOOR ANCHOR, and other position anchor poles are to be squeezed into perpendicular to surface, tunnel, and anchor cable is all squeezed into perpendicular to surface, tunnel.
Anchor Care network architecture parameters as decision variable comprises: (1) bolt diameter x
1mm; (2) rock-bolt length x
2m; (3) anchor cable form x
3; (4) anchor cable diameter x
4mm; (5) anchor cable array pitch x
5; (6) anchor pole top spacing x
6m; (7) anchor pole array pitch x
7m; (8) anchor pole side spacing x
8m; (9) anchor cable length x
9m; (10) bolt pretension x
10kN; (11) anchor cable pretightning force x
11kN.
In above-mentioned decision variable, x
1, x
2..., x
5for discrete variable, x
6, x
7..., x
11for continuous variable.It should be noted that, discrete variable x
3{ 1,2,3} represents anchor cable distribution form to ∈, and wherein, 1 representative often to be arranged along drift section and put an anchor cable, and 2 representatives establish two anchor cables and even rows lays an anchor cable along the arrangement of drift section odd number, and 3 representatives are often arranged along drift section and established two anchor cables.Discrete variable x
5{ 0,1,2,3,4} represents the multiple proportion between anchor cable array pitch and anchor pole array pitch to ∈, and wherein, 0 representative does not lay anchor cable along drift section.
In addition, determine that the decision variable span in actual applications in the present invention is very necessary, this span is using the constraint condition as institute's founding mathematical models.In carried example, the span of decision variable comprises following four aspects:
(1) span of bolt diameter and length.Bolt diameter x is got in carried example
1{ 16mm, 18mm, 20mm, 22mm, 25mm}, consider the convenience of underground construction to ∈, requires that rock-bolt length is no more than 2.4m, therefore rock-bolt length x
2∈ { 1.6m, 1.8m, 2m, 2.2m, 2.4m}.
(2) span of anchor cable length and diameter.In carried example, choose the span of length range as this decision variable of anchor cable length of conventional anchor cable.Therefore, getting 5m is anchor cable length minimum value, and 10m is maximal value, anchor cable diameter x
4∈ { 15.24mm, 17.78mm, 18mm, 20mm, 22mm}.
(3) span of the pretightning force of anchor pole and anchor cable.For anchor pole, require in known specifications that bolt pretension square must not be less than 100KN/m, and the maximal value of pretightning force should be impact wrench can apply pretightning force maximal value to anchor nut, so bolt pretension minimum value is set to 15KN, maximal value is set to 75KN.For anchor cable, pretightning force is break the 40%-50% of load, so anchor cable pretightning force minimum value is set to 60KN, maximal value is set to 300KN.
(4) span of bolt interval.If the too small meeting of the spacing of two anchor poles causes Rock Mass broken serious, accelerate country rock and be caving.Therefore, in carried example, array pitch minimum value between anchor pole is set to 0.5m.
It should be noted that, the span of above-mentioned decision variable is only the span of put forward example, when the inventive method is applied to other examples, can adjust according to the span of difference to decision variable of colliery actual conditions in different instances.
2. obtain sample data collection
Obtaining set of data samples is set up the most important condition of supporting quality agent model.Make again sample have typical representative to reduce calculated amount as far as possible, the present invention adopts orthogonal experiment and method for numerical simulation, produces sample data collection.Concrete steps are set forth as follows in conjunction with carried example:
(1) first above-mentioned decision variable is removed anchor cable form x
3ten outer decision variables, as 10 factors, in the span that each factor is above-mentioned, have chosen five levels, construct the orthogonal arrage of ten factor five levels.Wherein, a upper joint is not provided to the decision variable x of concrete span
6, x
7, x
8, according to its common span, all getting minimum value is 0.6m, and maximal value is 1.4m.Therefore, in the orthogonal arrage built, have the various combination mode of 50 kinds of Anchor Care network architecture parameters, but do not comprise discrete variable anchor cable form x
3.For anchor cable form x
3, respectively at x
3three kinds of different values under, adopt the above-mentioned orthogonal arrage constructed to combine Anchor Care network architecture parameters, form totally 150 kinds of Anchor Care schemes with this.But, as anchor cable array pitch x
5when value is 0, represents and do not lay anchor cable along drift section, also just without the need to discussing for different anchor cable forms, therefore finally having the array mode of 130 kinds of different Anchor Care network architecture parameters, representing 130 kinds of Anchor Care schemes.
(2) in numerical simulation software FLAC, numerical simulation is carried out to the 130 kinds of Anchor Care schemes adopting said method to be formed, obtain the rock deformation in tunnel under each Anchor Care scheme.For reducing the calculated amount of numerical simulation, save computational resource and computing time, choose the 10m of tunnel to be designed length as the tunnel initial model in numerical simulation, according to known geologic condition, tunnel initial environment is simulated, obtain tunnel initial balance model, then, respectively existing 130 kinds of Anchor Care schemes are applied in constructed tunnel initial balance model, the alley way anchor obtained under each Anchor Care scheme protects network structure numerical simulator, protect network structure numerical simulator to above-mentioned 130 kinds of alley way anchors to simulate and calculate, under finally obtaining each Anchor Care scheme, tunnel roof-to-floor convergence and two helps the amount of shifting near, the cumulative sum of the amount of shifting near roof-to-floor convergence and two is helped to be combined to form sample data collection with corresponding Anchor Care network instruction plan.
3. set up supporting quality agent model
For supporting quality, measured mainly through tunnel roof-to-floor convergence and the roadway's sides amount of shifting near, roof and floor displacement and two are helped the quantizating index of cumulative sum as supporting quality of displacement by the present invention.And being related to this difficult point for having nonlinear between Anchor Care network parameter and rock deformation, the present invention adopts support vector machine to carry out identification to above-mentioned nonlinear relationship, thus sets up the agent model of supporting quality.Be specially: first the set of data samples obtained by above-mentioned work being loaded in support vector as the data sample of support vector machine, in carried example, in order to obtain better fitting effect, normalized having been carried out to set of data samples.Then, random selecting sample data concentrates 100 samples as support vector machine training sample, trains, remain 30 samples and test the agent model trained as support vector machine test sample book agent model.When error of fitting is in the scope of acceptable error, the foundation of the supporting quality agent model achieved based on support vector machine is described.
4. set up Model for Multi-Objective Optimization
Embody this objective function of supporting quality with the supporting quality agent model of above-mentioned foundation, set support material total price needed for anchor pole and anchor cable to embody this objective function of support cost with tunnel.The concrete account form of required support material total price is that the product of quality by calculating required anchor pole and anchor cable and corresponding unit mass anchor pole price and unit mass anchor cable price realizes, wherein, determine, so unit mass anchor pole and anchor cable price are definite value owing to using the material of anchor pole and anchor cable.So, using function model Cost (x) of support cost with supporting quality agent model Quality (x) to have set up as two optimization object function embodying coal mining economy and security.
In carried example, choose unit mass anchor pole cost C
gbe 15 yuan/kg, unit mass anchor cable cost C
sbe 20 yuan/kg, back lays the scope K of anchor pole
dfor 5.2m, one, tunnel side lays the scope K of anchor pole
cfor 3.7m, for reducing calculated amount, tunnel length L chooses 10m, anchor pole ρ
gwith anchor cable ρ
sdensity all choose the density 7850kg/m of steel
3.The constraint condition of span as mathematical model of decision-making will be provided in first segment, and constraint condition be added in objective function, obtain final optimization aim:
Quality(x)＝SVM(x)
Wherein, x
1∈ { 16mm, 18mm, 20mm, 22mm, 25mm}; x
2∈ { 1.6m, 1.8m, 2m, 2.2m, 2.4m}; x
4∈ { 15.24mm, 17.78mm, 18mm, 20mm, 22mm}; x
6, x
7, x
8>=0.5; 5≤x
9≤ 10; 15≤x
10≤ 75; 60≤x
11≤ 300.
5. adopt evolution algorithm to solve object and multi object mathematical model
The object and multi object mathematical model of multi-target evolution optimized algorithm to the coal mine down-hole tunnel Anchor Care network structure problem of above-mentioned foundation is adopted to solve, the process flow diagram providing the concrete process of establishing of model in accompanying drawing 1 and adopt multi-objective optimization algorithm to solve model.
Finally, obtain after being solved by the mathematical model of multi-objective optimization algorithm to multi-objective optimization question, meeting under the minimum condition of rock deformation, optimum Anchor Care strategy parameter is: (1) bolt diameter 20mm; (2) rock-bolt length 1.8m; (3) anchor cable form is 2; (4) anchor cable diameter 15.28mm; (5) anchor cable array pitch 3 times; (6) anchor pole top spacing 0.518m; (7) anchor pole array pitch 0.5m; (8) anchor pole side spacing 0.5m; (9) anchor cable length 5m; (10) bolt pretension 37.189KN; (11) anchor cable pretightning force 217.792KN.Numerical simulation software is adopted to carry out analog computation to the above-mentioned optimum Anchor Care scheme obtained, the end distension amount obtained under optimum Anchor Care scheme is 0.239m, this is because the inventive method is only helped to carry out supporting to the top board and two in tunnel, base plate does not implement any supporting measure, so floor lift in gallery amount is larger.But the amount of crushing and two under above-mentioned optimum Anchor Care scheme helps the amount of shifting near all very little, and occurrence is: the amount of crushing 0.030m, and two help the amount of shifting near 0.043m, this illustrates, the supporting quality under this kind of Anchor Care scheme is fine.
Based in the supporting agent model of support vector machine, the displacement of wall rock of optimum Anchor Care scheme is 0.2048m, and numerical simulation obtains surrouding rock deformation total amount is 0.312m, this is because support vector machine has certain error of fitting when Function Fitting, but in the data sample of numerical simulation generation, rock deformation minimum value is 0.344m.As can be seen here, little obviously than in data sample of the displacement of wall rock of Anchor Care strategy parameter adopting the inventive method to optimize out, this illustrates that optimization method of the present invention is feasible.
Optimize the Pareto optimal solution set obtained and see accompanying drawing 2, the best Anchor Care strategy parameter of supporting quality not only can be obtained in Pareto optimal solution set, the minimum Anchor Care strategy parameter of support cost can also be obtained, the Anchor Care scheme taking into account supporting quality and support cost can also be obtained in addition, colliery can be chosen the Anchor Care scheme in Pareto optimal solution set according to self actual demand, adopt the Anchor Care scheme of this disaggregation, instruct this alley way anchor to protect the design of network structure, thus ensure coal mine downhole safety, economic construction.
Claims (4)
1. the Multiobjective Intelligent Optimization Design of a coal mine down-hole tunnel Anchor Care network structure, it is characterized in that: the Model for Multi-Objective Optimization setting up the design of coal mine down-hole tunnel Anchor Care network architecture parameters, and adopt Multiobjective Intelligent optimization method to solve the multi-objective Model set up, realize the optimization to Anchor Care network structure; Method step is as follows:
(1). set up the supporting quality agent model of Anchor Care network architecture parameters
The combination supporting form that Anchor Care network adopts anchor pole and anchor cable to form; Under roadway support quality is reflected in Anchor Care role of network, the extent of stability of roadway surrounding rock, measured by tunnel roof-to-floor convergence and the roadway's sides amount of shifting near, identification is carried out to the nonlinearity relation between Anchor Care network architecture parameters and rock deformation, obtains the agent model of supporting quality;
(2). take into account the multi-objective optimization design of power model of supporting quality and support cost
For meeting security and the cost-effectiveness requirement of coal mining, the corresponding evaluation index of Anchor Care network instruction plan comprises two parts; One, adopts supporting quality Quality (x) to embody security requirement; Its two, adopt support cost
embody cost-effectiveness requirement; Take into account supporting quality and support cost two aspect requirement, set up two objective optimization model of coal mine down-hole tunnel Anchor Care network architecture parameters; Two objective optimization model are: for supporting quality, set up supporting quality agent model; For support cost, choose tunnel and set the total price of support material needed for anchor pole and anchor cable to represent the mathematical model of support cost, by decision variable span in actual applications, be added in mathematical model as constraint condition, finally set up two objective optimization model of coal mine down-hole tunnel Anchor Care network architecture parameters design;
(3). adopt evolution algorithm to solve object and multi object mathematical model
Adopt the Model for Multi-Objective Optimization of taking into account supporting quality and support cost of multi-objective Evolutionary Algorithm to above-mentioned foundation to solve, obtain Pareto optimal solution set, realize Anchor Care network architecture parameters optimal design.
2. the Multiobjective Intelligent Optimization Design of coal mine down-hole tunnel Anchor Care network structure according to claim 1, is characterized in that: supporting quality agent model adopts following steps to obtain:
Step1: adopt orthogonal experiment and method for numerical simulation, produces sample data collection;
Step2: adopt supervised learning method training sample data, identification is carried out to the nonlinearity relation between Anchor Care network architecture parameters and rock deformation, obtains the agent model of supporting quality.
3. the Multiobjective Intelligent Optimization Design of coal mine down-hole tunnel Anchor Care network structure according to claim 1, is characterized in that: two objective optimization model: Quality (x)=SVM (x)
Wherein, x
1∈ { x
1(1), x
1(2) ..., x
1(n) }; x
2∈ { x
2(1), x
2(2) ..., x
2(n) }; x
4∈ { x
4(1), x
4(2) ..., x
4(n) }; x
6, x
7, x
8>=min; Min_x
9≤ x
9≤ max_x
9; Min_x
10≤ x
10≤ max_x
10; Min_x
11≤ x
11≤ max_x
11;
In formula, x={x
1, x
2..., x
11represent Anchor Care network architecture parameters; x
1represent bolt diameter, { x
1(1), x
1(2) ..., x
1(n) } represent its discrete span; x
2represent rock-bolt length, { x
2(1), x
2(2) ..., x
2(n) } represent its discrete span; x
3{ 1,2,3} represents anchor cable distribution form to ∈, and wherein, 1 representative often to be arranged along drift section and put an anchor cable, and 2 representatives establish two anchor cables and even rows lays an anchor cable along the arrangement of drift section odd number, and 3 representatives are often arranged along drift section and established two anchor cables; x
4represent anchor cable diameter, { x
4(1), x
4(2) ..., x
4(n) } represent its discrete span; x
5{ 0,1,2,3,4} represents the multiple proportion between anchor cable array pitch and anchor pole array pitch to ∈, and wherein, 0 representative does not lay anchor cable along drift section; x
6represent anchor pole top spacing; x
7represent anchor pole array pitch; x
8represent anchor pole side spacing; x
9represent anchor cable length, min_x
9for anchor cable length minimum value, max_x
9for anchor cable length maximal value; x
10represent bolt pretension, min_x
10for bolt pretension minimum value, max_x
10for bolt pretension maximal value; x
11represent anchor cable pretightning force, min_x
11for anchor cable pretightning force minimum value, max_x
11for anchor cable pretightning force maximal value.If C
gfor unit quality anchor pole cost, C
sfor unit quality anchor cable cost, K
dfor back being laid the range size of anchor pole, K
cfor roadway's sides being laid the range size of anchor pole, L is tunnel length, ρ
gfor anchor pole density, ρ
sfor the density of anchor cable.
4. the Multiobjective Intelligent Optimization Design of coal mine down-hole tunnel Anchor Care network structure according to claim 1, it is characterized in that: adopt multi-objective Evolutionary Algorithm to solve the two target mathematical models set up, wherein respectively using supporting quality and support cost as two objective functions, obtain optimum Anchor Care network architecture parameters, form the Multiobjective Intelligent Optimization Design of complete coal mine down-hole tunnel Anchor Care network structure, instruct the design of coal mine roadway Anchor Care network architecture parameters.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510405763.9A CN105302934B (en) | 2015-07-10 | 2015-07-10 | The Multiobjective Intelligent optimum design method of coal mine down-hole tunnel Anchor Care network structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510405763.9A CN105302934B (en) | 2015-07-10 | 2015-07-10 | The Multiobjective Intelligent optimum design method of coal mine down-hole tunnel Anchor Care network structure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105302934A true CN105302934A (en) | 2016-02-03 |
| CN105302934B CN105302934B (en) | 2018-05-15 |
Family
ID=55200202
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510405763.9A Expired - Fee Related CN105302934B (en) | 2015-07-10 | 2015-07-10 | The Multiobjective Intelligent optimum design method of coal mine down-hole tunnel Anchor Care network structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105302934B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787645A (en) * | 2016-02-19 | 2016-07-20 | 龙口矿业集团有限公司 | Soft rock tunnel bolting-grouting support system quantitative evaluation method |
| CN108549776A (en) * | 2018-04-17 | 2018-09-18 | 山东科技大学 | A kind of finite element optimization method of base plate tunnel protection coal pillar width |
| CN109190297A (en) * | 2018-09-30 | 2019-01-11 | 山东科技大学 | A kind of Large Underground two-ended cables Anchorage Parameters optimization method |
| CN109854252A (en) * | 2019-02-01 | 2019-06-07 | 王�琦 | Underground engineering complete control method stage by stage |
| CN111980736A (en) * | 2020-08-31 | 2020-11-24 | 常州机电职业技术学院 | Danger degree prediction and roof fall early warning method for mine anchor bolt supporting roadway |
| CN115146420A (en) * | 2022-09-05 | 2022-10-04 | 中南大学 | Method for establishing railway tunnel anchor rod refined model |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101694163A (en) * | 2009-10-14 | 2010-04-14 | 山东科技大学 | Determination method of deep tunnel roof support forms and support depth |
| CN101968825A (en) * | 2010-11-23 | 2011-02-09 | 中国矿业大学 | Method for intelligently designing bolting of coal mine tunnels |
| CN103984788A (en) * | 2013-12-31 | 2014-08-13 | 中国矿业大学(北京) | Automatic intelligent design and optimization system for anchor bolt support of coal tunnel |
-
2015
- 2015-07-10 CN CN201510405763.9A patent/CN105302934B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101694163A (en) * | 2009-10-14 | 2010-04-14 | 山东科技大学 | Determination method of deep tunnel roof support forms and support depth |
| CN101968825A (en) * | 2010-11-23 | 2011-02-09 | 中国矿业大学 | Method for intelligently designing bolting of coal mine tunnels |
| CN103984788A (en) * | 2013-12-31 | 2014-08-13 | 中国矿业大学(北京) | Automatic intelligent design and optimization system for anchor bolt support of coal tunnel |
Non-Patent Citations (2)
| Title |
|---|
| FS HAN 等: "Study on Rock Bolt Support of Roadway of Coal Mine Using Neural Network", 《APPLIED MECHANICS & MATERIALS》 * |
| 戴俊等: "煤矿巷道锚杆支护的参数优化", 《岩土力学》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787645A (en) * | 2016-02-19 | 2016-07-20 | 龙口矿业集团有限公司 | Soft rock tunnel bolting-grouting support system quantitative evaluation method |
| CN108549776A (en) * | 2018-04-17 | 2018-09-18 | 山东科技大学 | A kind of finite element optimization method of base plate tunnel protection coal pillar width |
| CN109190297A (en) * | 2018-09-30 | 2019-01-11 | 山东科技大学 | A kind of Large Underground two-ended cables Anchorage Parameters optimization method |
| CN109190297B (en) * | 2018-09-30 | 2022-11-29 | 山东科技大学 | Anchor parameter optimization method for anchor cable penetrating through large underground plant |
| CN109854252A (en) * | 2019-02-01 | 2019-06-07 | 王�琦 | Underground engineering complete control method stage by stage |
| CN111980736A (en) * | 2020-08-31 | 2020-11-24 | 常州机电职业技术学院 | Danger degree prediction and roof fall early warning method for mine anchor bolt supporting roadway |
| CN115146420A (en) * | 2022-09-05 | 2022-10-04 | 中南大学 | Method for establishing railway tunnel anchor rod refined model |
| CN115146420B (en) * | 2022-09-05 | 2022-11-25 | 中南大学 | Method for establishing railway tunnel anchor rod refined model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105302934B (en) | 2018-05-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105302934A (en) | Multi-objective intelligent optimization design method for anchoring and protecting network structure of coal mine underground roadway | |
| CN108985340B (en) | Shield construction ground settlement prediction method based on dual-model fusion | |
| CN105911867B (en) | Ship thrust distribution method based on NSGA-II algorithm | |
| CN103984788A (en) | Automatic intelligent design and optimization system for anchor bolt support of coal tunnel | |
| CN110889588A (en) | Method for evaluating risk level of shield tunnel construction adjacent building by using factor judgment matrix | |
| CN113153334A (en) | Risk management and control system for influence of shield underpass construction on structure | |
| CN103362519B (en) | Safe and quick coal roadway tunneling method | |
| CN106049511B (en) | A kind of composite reinforcement method suitable for non-coal open-pit slope landslide control | |
| CN105241510A (en) | Determination method of tunnel surrounding rock prestressing force reinforcing anchor rod length and radial prestressing force values | |
| CN112231870B (en) | Intelligent generation method for railway line in complex mountain area | |
| CN104156782A (en) | Balancing-optimalizing method, for project time limit, quality and cost, used in concrete faced rockfill dam construction | |
| CN104727828A (en) | Tunnel pre-support management method based on deformation control | |
| CN102789516A (en) | Stability numerical analysis and optimization design method based on monitoring information in slope construction process | |
| CN106339798A (en) | Soft rock inclined shaft supporting and protecting system stability evaluation method based on fuzzy comprehensive evaluation method | |
| Zhu et al. | Search for circular and noncircular critical slip surfaces in slope stability analysis by hybrid genetic algorithm | |
| Qi et al. | Failure characteristics and control technology of surrounding rock in deep coal seam roadway with large dip angle under the influence of weak structural plane | |
| Wu et al. | Application of computer simulation method in deep underground engineering of complicated geological condition | |
| Li et al. | Anchoring parameters optimization of tunnel surrounding rock based on particle swarm optimization | |
| Wang | Effect of genetic algorithm in optimizing deep foundation pit supporting structure | |
| CN102767293B (en) | Strengthening method for power transmission tower on hazardous area of landslide | |
| Wang et al. | A novel evaluation method for the stability of construction sites on an abandoned goaf: a case study | |
| CN114810163A (en) | Deep high-stress thin-layer hard top plate bottom suction roadway supporting process | |
| Reddy | Modeling Load-Settlement Behavior Of Driven Piles Embedded In Cohesive Soils Using Artificial Neural Networks | |
| Yang | Study on large deformation prediction and control technology of carbonaceous slate tunnel in Lixiang railway | |
| Shangguan et al. | A study on soil arching effect of anti-slide pile considering different pile arrangements |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180515 Termination date: 20190710 |