CN101059872A - Real-time visualized emulation method facing to cone crusher performance optimization - Google Patents

Abstract

The invention relates to a real-time visible simulating method for optimizing property of cone breaker in mine machines belonging to computer application technical field, comprising 1, building a stone geometry model construction method, 2, building a stone quick texture projection method facing real-time display, 3, building the projection between the optimization calculation of breaker and the stone model generation. The invention can avoid or reduce tests, to save cost, accelerate optimization, and save time, with the application in mountain modeling, controllable result and high modeling efficiency.

Description

Real-time visualized emulation method towards cone crusher performance optimization

Technical field

What the present invention relates to is the emulation mode in a kind of Computer Applied Technology field, specifically, is a kind of real-time visualized emulation method towards cone crusher performance optimization that is used on the mining machinery.

Background technology

Gyratory crusher is a major equipment in small, broken bits in the brittle material, be widely used in industrial sectors such as mine, smelting, water conservancy, highway, railway and chemical industry, and the material shattering process is difficult to owing to rule complex effects factor is numerous analyze, and the big work under bad environment of the size of gyratory crusher own is difficult to carry out actual loading test research, and therefore the research in corresponding field brings great difficulty.

Because the complicacy of prose style free from parallelism stone shattering process, shattering process still can't be carried out visual simulating when the design disintegrating machine, the disintegrating machine computation optimization only can obtain a small amount of statistical parameter of broken results of a small amount of structural parameters, running parameter and these parameter correspondences, the analysis of optimizing the result is undertaken by test usually, and this expends a large amount of time and cost.The virtual emulation technology provides new tool for the optimization and the simulation study of the gyratory crusher course of work, and it can comprehensive considering various effects, overcomes space, time and other condition the course of work is carried out simulation study.Present virtual emulation technology generally can only be used for the real-time or non real-time modeling of regular geometric object, and can't describe for nonlinear complex geometry process owing to the restriction on modeling method.The simulation calculation because the geometric object of simulation process places one's entire reliance upon in addition, therefore not only modeling efficiency is comparatively low, also because the limitation that emulated physics calculates causes the modeling of complex geometry process (as the fragmentation that is squeezed of a large amount of prose style free from parallelism stones) to realize.

Find through literature search prior art, " the A new fractal-based approach for 3D visualization of mountains in VRMLstandard " that Mohsen Sharifi etc. deliver on " Proceedings of the2nd international conference on Computer graphics and interactivetechniques in Australasia and South East Asia " (Australia and second computer graphics in Southeast Asia and interaction technique international conference collection of thesis) (2004) 100-105 page or leaf is (a kind of visual towards 3 dimensions, based on fractal mountain peak VRML model construction method), propose a kind of fractal method in this article and construct and describe the mountain peak model, this method is visual for 3 dimensions on the mountain peak of realizing the sense of reality, use a kind of fractal geometry method that is derived from terrain configuration, and use a kind of erosion algorithm to make that the mountain peak model is truer.Because the mountain peak is irregular polyhedrons, can be similar to the stone that has amplified is regarded as in the mountain peak, so this method also can be used to construct stone by regulating parameter, and owing to used fractal method, the surface detail model of being constructed is very true to nature.Its deficiency is: this method only is to use fractally constructs sense of reality mountain peak, and the mountain peak of the structure that do not combine with physical computing just " looking like " and can't approach certain specific true mountain peak; On the other hand, the mountain peak model belongs to the model of 2 dimensions half in essence, therefore uses the improvement of terrain configuration method to realize, and the fragmentation of the prose style free from parallelism stone during disintegrating machine work, each stone all is typical three-dimensional model, uses this method can't description true to nature on global shape.The visual no real-time requirement in mountain peak in the document in addition, so the model formation efficiency also can't satisfy disintegrating machine a large amount of prose style free from parallelism stone model construction needs in when work.

Summary of the invention

The object of the invention is at the deficiencies in the prior art, a kind of real-time visualized emulation method towards cone crusher performance optimization (being called for short SGRCS) is proposed, by adopting physical behavio(u)r computation model and virtual phenomenon generation systems, make it solve computation optimization and only can obtain the few results data, and a large amount of stones broken complex process can't real-time visual causes in the disintegrating machine chamber optimizing process and result is not directly perceived, a difficult problem that is difficult to the confidence level of analysis optimization calculating, realized with the result of gyratory crusher structural parameters and running parameter computation optimization and this as a result the crushing effect of correspondence show in real time, thereby make the optimization of disintegrating machine and emulation combine, analysis and the observation of optimizing the result are directly perceived more and simple, increase optimization result's confidence level.

The present invention is achieved through the following technical solutions, and concrete steps are as follows:

Step 1, the foundation of stone geometric model structure unified approach

To disintegrating machine broken various stones adopt and dynamically construct based on the unified approach of fractal geometry, realize unified structure to various rocks, ore, the controlled variable of regulating construction process can obtain the stone model of difformity and surface details.Thereby can construct the multiple stone of disintegrating machine with fragmentation under the situation that does not rely on physical computing fast, the stone shape and the surface details of being constructed are true to nature, and as required control accuracy and speed to satisfy real-time simulation to disintegrating machine work.

Described stone geometric model structure unified approach, it generates new node by iterative formula, thereby obtains new polygon model on simple polygon model (tetrahedron, hexahedron, the octahedron) basis of generator.The global shape that controlled variable (fractal dimension, random factor, scale factor, iterations) controls the stone of constructing, local smooth and degree of roughness are provided during iteration.The principal element that influences formation speed is an iterations, and the stone during for cone crushing dynamically generates emulation, and the iterations when generating stone gets 5～7 usually.

Step 2 is set up towards the quick texture mapping method of the stone of real-time demonstration

Conventional texture generation method need be calculated the texture coordinate, and stone is because there is the face in a plurality of different sizes, orientation in the surface details complexity, use conventional method must cause the texture computing velocity to slow down, thereby can't satisfy the real-time simulation requirement of crusher in crushing process.

Described quick texture mapping method, be meant: on stone geometric model basis, be used for the texture picture fast mapping of stone is arrived each surface of stone geometric model, carrying out texture between 6 faces of picture and bounding box calculates, employing is shone upon texture picture on the bounding box face, again will be on the model point to the bounding box projection, the texture information at acquisition point place, the texture that uses the interpolation calculation of point texture information to finish model at last generates.The stone dynamic modelling method

Because follow-up projection and interpolation calculation is simple computation, has faster speed so the method is compared with conventional method.The selection of putting on the stone geometric model also can be simplified or be increased as required in addition, when growing up in 5mm, a tri patch minimum edge on the model surface can select whole nodes to add 2～5 sampled points, if lip-deep plane polygon is not a triangle, then always can be decomposed into the leg-of-mutton combination more than 2; And can only use node to carry out interpolation during less than 5mm when the minimum length of side of tri patch on the model surface, if the average length of side then can only select the part node to carry out interpolation less than 3mm, remaining node according to and sampled point between the minimum principle of the distance texture information value of directly getting sampled point.Can control the precision and the formation speed of texture in view of the above.

Step 3, the mapping relations between disintegrating machine computation optimization and stone model generate are set up;

Mapping relations are that stones such as the generator that generates of disintegrating machine computation optimization model and stone model, fractal dimension, iterations, the random deformation factor dynamically generate the relation between the controlled variable, though dynamic generation and the extinction broken phenomenon that can simulate stone of stone in the disintegrating machine chamber, but can not reflect the actual physical process under the specific disintegrating machine parameter condition, because this simulation does not combine with the physical computing of disintegrating machine.

Dynamically generate at described stone on the basis of modeling, broken results statistical parameter and stone modeling process that the disintegrating machine computation optimization is obtained combine, set up the mapping relations between the two, make the dynamic generation and the extinction of stone model meet the result of disintegrating machine computation optimization, thereby reach emulation the stone shattering process.

It is generator, fractal dimension, iterations, random factor (scale factor and probability factor) that stone dynamically generates controlled variable, and the stone broken results parameter that the disintegrating machine computation optimization obtains is broken hierarchy number, every layer of grain-size distribution curve in throughput rate (T/h), particle shape coefficient, the chamber, set up mapping relations between them and just can make the stone broken results parameter that reflection physics calculates in the broken emulation of stone, therefore can carry out visual simulating the physical computing result.Parameter when changing computation optimization model or optimization causes that stone broken results parameter changes, and this changes the variation of the controlled variable of stone modeling when causing emulation, thereby makes computation optimization can control the emulation phenomenon.

The present invention is divided into two relatively independent subsystems with the stone structure with the disintegrating machine computation optimization, they interrelate by mapping relations, stone structure generates the stone model information and transfers to virtual reality system and play up and show, thus the broken emulation of stone when finishing gyratory crusher work.After using the present invention, disintegrating machine computation optimization model can obtain and existence in a variety of forms by multiple means, as numerical value method solver, experimental formula, test findings data etc.It does not bear work such as stone model construction and color and vein generation.May not there be contacting directly on natural, the physical significance between it and stone model generate, but can be by setting up the mathematics mapping relations between them, thereby realize control that shattering process stone model is changed.

The method that the present invention proposes can be used for the parameter optimization of gyratory crusher design phase.Present lamination gyratory crusher computation optimization can access the structural parameters of main broken machine mechanism of disintegrating machine and the running parameter of disintegrating machine, and the broken results of stone can only draw a small amount of statistical parameter, only the broken situation of stone in crusher chamber can't be described with these statistics parameters, in order to verify the computation optimization result, usually need carry out test in kind, adjust the computation optimization parameter according to actual loading test and be optimized calculating again, so repeatedly up to definite final disintegrating machine parameter.Actual loading test requires a great deal of time and cost, and particularly in the time of the needs repetition test, this problem is more outstanding.

Method of the present invention is calculated with the disintegrating machine parameter optimization and is combined, can carry out visual simulating to the result of computation optimization, make one to optimize the broken situation of the stone observation analysis and the analysis intuitively on computers of correspondence as a result, in the chamber, whether cause treating material, stop up because of particle shape, size-grade distribution as observing stone granularity, stone.Calculate again and emulation thereby can make amendment to the correlation computations parameter of optimizing process according to the result of emulation.Therefore can avoid or reduce actual loading test after using method of the present invention, provide cost savings greatly on the one hand, accelerate disintegrating machine Parameter Optimization process on the one hand, save a large amount of time.

When present novel lamination gyratory crusher designs, need carry out actual loading test 3～5 times usually in the disintegrating machine principal parameter optimizing phase, the actual loading test time was approximately for 8～30 weeks, spent 10～500,000 dollars.If use method of the present invention, then actual loading test can be avoided or reduce to 1～2 time, and then the actual loading test time can not consider or only needed for 4～5 weeks that cost is no more than 50,000 dollars under the equal conditions.

On the other hand, the modeling method that the present invention proposes not only is used for the modeling of stone, by changing the modeling controlled variable, can construct multiple polyhedral model, thereby can become the universal method that multiple geometric object is constructed in the virtual reality field.Compare with the method for the Mohsen Sharifi that mentions in the background technology, modeling method of the present invention also has the characteristics that controllability is strong as a result, modeling efficiency is high except can be used for the mountain peak modeling equally.

Description of drawings

The system architecture synoptic diagram that Fig. 1 adopts for the embodiment of the invention

Fig. 2 is for using the software systems composition diagram that the embodiment of the invention adopts

Fig. 3 is stone geometric model structure design sketch of the present invention

Fig. 4 is the approximate fast texture generating algorithm key diagram of stone of the present invention

Fig. 5 generates design sketch for the approximate fast texture of stone of the present invention

Fig. 6 is the result of calculation curve map of the embodiment of the invention

Embodiment

Below in conjunction with accompanying drawing embodiments of the invention are elaborated: present embodiment is being to implement under the prerequisite with the technical solution of the present invention, provided detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.

1. the quick structure of stone geometric object

The use fractal geometry are come the geometric object in the construction process emulation in the present embodiment.Different with the Euclidean geometry modeling of routine, one also has limited analog functuion to a class phenomenon based on fractal Geometric Modeling itself, and promptly the processing procedure of geometric model has reflected the characteristic that such phenomenon takes place to a certain extent.

The basic process of stone dynamic modelling method:, determine key elements such as fractal iterative model, fractal dimension, initial generator at first according to the three-dimensional fractal geometric theory; Consider randomness when broken phenomenon takes place and the solid yardstick difference of different phase, therefore when the constructive geometry model, need consider the influence of the random deformation factor; That determines iterations, fractal model and fractal dimension determines to depend on concrete phenomenon, need determine by theoretical analysis even analysis of experiments, and unusual between 2.4～2.7 with material of the fractal dimension during the stone modeling.

Below provide the quick structure unified approach of stone geometric object.Stone is as polyhedron, and once fractal iteration generates and is expressed as

${\left\{S_i^j\right\}}_{i=1,K,n}\⇒{\left\{S_k^{j+1}\right\}}_{k=1,K,m}---\left(1\right)$

s.t.

$(E_{s_j}^j=E_{s_{j+1}}^j)\⇒(E_{s_j}^{j+1}=E_{s_{j+1}}^{j+1})---\left(2\right)$

In the formula:

s _iBe the polyhedral face of stone, j is an iterations, the face sum when n is the j time iteration, the face sum when m is the j+1 time iteration; (2) represent respectively in the formula that the limit of original adjacent surface before and after the iteration is still consistent, promptly the polyhedron topology is correct after the iteration.

$S_i^j\≅{\left\{E_i^j\right\}}_{i=\mathrm{1,2,3},},\underset{i=1}{\overset{n}{U}}{S}_i^j\≅U{\left\{E_i^j\right\}}_{i=\mathrm{1,2,3}}={\left\{E_i^j\right\}}_{i=1...n}$

S again _i { E _k} _{K=1,2, K, K}(3)

In the formula: E _kBe planar S _iSide ring, K is the limit sum of this face

Suppose that face is triangle entirely, so have:

${\left\{S_i^j\right\}}_{i=1,K,n}\⇔Y{\left\{E_k^j\right\}}_{k=\mathrm{1,2,3}}\≅{\left\{E_k^j\right\}}_{k=\mathrm{1,2},K,M}---\left(4\right)$

In the formula: M is the total limit of a polyhedron number, and j is fractal iterations.Then formula (1) can turn to

${\left\{E_i^j\right\}}_{i=1,K,N}\⇒{\left\{E_k^{j+1}\right\}}_{k=1,K,M}---\left(5\right)$

In the formula:

J is fractal iterations; N, M are respectively the limit number of adjacent twice iteration, for the triangle polyhedron M=2N are arranged.

(5) implication of formula is to substitute the face iteration with the limit iteration, in like manner can substitute the limit iteration with an iteration, obtains

${\left\{P_i^j\right\}}_{i=1,K,N}\⇒{\left\{P_k^{j+1}\right\}}_{k=1,K,2N}---\left(6\right)$

In the formula:

N is counting of the iteration, has N needs to regenerate in 2N the new point, and original N point only done the geometric position and changed.

The reposition of original point generates and can calculate as follows:

At first calculate polyhedral geometric center: $P_G=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{P}_i/N---\left(7\right)$

Calculate new point according to following formula then: $\stackrel{\→}{P_G{P}_i^{j+1}}=\stackrel{\→}{P_G{P}_i^j}\·(1+r_{\mathrm{\ϵ}})+\underset{k=1}{\overset{K}{\mathrm{\Σ}}}\stackrel{\→}{P_G{P}_k^j}\·r_{\mathrm{\α}}/K---\left(8\right)$

In the formula:

P _i ^J+1Be new point after the iteration; P _k ^jBe P _i ^jNeighbor point; K is the neighbor point number; r _εBe a random quantity, its obedience-r _E～r _EBetween normal distribution, r _EAs a parameter input, it is relevant with the material and the crumbling method of stone; r _αBe another random quantity, obedience-r _A～r _ABetween normal distribution, r _AGenerally desirable

Newly-generated point serves as that the basis generates with newly-generated leg-of-mutton mid point, and an Atria new mid point can calculate according to following formula:

$\left\{\begin{array}{c}\stackrel{\→}{P_G{P}_{M,1}}=\frac{1}{2}\·(\stackrel{\→}{P_G{P}_1}+\stackrel{\→}{P_G{P}_2})\·(1+r_{\mathrm{\ϵ}})+(\stackrel{\→}{P_G{P}_3^{\′}}+\stackrel{\→}{P_G{P}_3^{\′\′}})\·r_q/2\\ \stackrel{\→}{P_G{P}_{M,2}}=\frac{1}{2}\·(\stackrel{\→}{P_G{P}_2}+\stackrel{\→}{P_G{P}_3})\·(1+r_{\mathrm{\ϵ}})+(\stackrel{\→}{P_G{P}_3^{\′}}+\stackrel{\→}{P_G{P}_3^{\′\′}})\·r_q/2\\ \stackrel{\→}{P_G{P}_{M,3}}=\frac{1}{2}\·(\stackrel{\→}{P_G{P}_3}+\stackrel{\→}{P_G{P}_1})\·(1+r_{\mathrm{\ϵ}})+(\stackrel{\→}{P_G{P}_3^{\′}}+\stackrel{\→}{P_G{P}_3^{\′\′}})\·r_q/2\end{array}\right.---\left(9\right)$

In the formula:

P ₁, P ₂, P ₃Be three new summits that generate after the former vertex position iteration of triangle; P _{M, 1}, P _{M, 2}, P _{M, 3}New point for middle dot generation on the foundation three sides of a triangle; P ₃', P ₃" for using two adjacent leg-of-mutton the 3rd summits on this limit; r _εCotype (8); r _qBe another stochastic variable, obedience-r _Q～r _QBetween normal distribution, r _QAs the parameter input, it is also relevant with the material and the crumbling method of stone, rule of thumb generally can be taken as

-r _E～r _EBetween the standard deviation S of normal distribution according to following calculating: S=k2 ^-iH(10)

In the formula: k is a scale factor, and is also relevant with the stone material; I is an iterations; H=2-D, D is fractal dimension, it be one less than 3 greater than 2 real number, relevant with material and crumbling method during the structure stone, the rock of suitable cone crushing or ore fractal dimension are usually between 2.3～2.7.

Use formula (7～10) promptly can realize generating the level of detail stone polygon model relevant with iterations from a simple polygon model (tetrahedron, hexahedron, octahedron).

The stone model need be determined generator, fractal dimension, parameter r in the said method structure prose style free from parallelism rock failure process _EWith scale factor k at random.r _ECan change with iterations with k.Tetrahedron, hexahedron and octahedron are the most frequently used generators, and the length breadth ratio of initial generator is relevant with the particle shape coefficient of physical computing.Fractal dimension need be adjusted according to the result in actual stone characteristic and when structure, in case need not change after fixing.

Use said method to carry out petrophysical model structure example, can satisfy when usually iterations is 6 times that the user describes demand to the stone details in the broken emulation, the single model rise time is about ordinary PC (CPU frequency 1GHZ) goes up to 0.005ms, promptly can generate 20,000 models approximately 1 second, iterations one is regularly irrelevant with the model size.With the PYB1750 gyratory crusher is example, throughput rate usually about 500 tons/hour, i.e. 138.9 Kilograms Per Seconds, ls density is 2.67 * 10 ³About kilograms per cubic meter, the stone cumulative volume is about 0.052 cubic metre, and promptly 5.2 * 10 ⁷Mm ³And the particle mean size of gape stone is about 60mm, supposes that stone is a spheroid, and then radius is 30mm, volume 2827.8mm ³Promptly the stone number of the required generation of per second of calculating with minimum particle size is about 18397, less than 20000 of per seconds.The turnover rate of stone is about 47% during in fact broken emulation, and the stone granularity on upper strata many (the superiors particle mean size are 110mm) bigger than the granularity of gape, so this modeling efficiency can satisfy broken real-time simulation demand.

The original-shape of stone model generator also will influence the global shape of final mask, also be elongated such as the final generation model global shape of elongated model.Adopt the method calculating parameter r of simple and iterations linear dependence _E, and when using less scale factor k, the stone of generation will be relatively smooth, and bulk deformation is little; If adopt when stone generates with the method for iterations nonlinear dependence and calculate r _E, r when then causing primary iteration _EBigger with k, global shape is irregular, if reduce fast with iterations, then can obtain comparatively smooth local shape again.

2. the quick texture processing of stone object

The stone shattering process is accompanied by the variation of color and texture, the processing of texture has occupied the very most of the time during graph rendering, solid One's name is legion not only when stone is broken, also have complicated geometry and topological characteristic, this makes that texture processing is difficult more fast, the present invention mainly pays close attention to macroscopical result, this characteristic of statistics according to the shattering process simulation process, proposes a kind of approximate texture generating algorithm based on the polyhedron node, mainly is divided into following steps:

At first calculate polyhedral bounding box, then predetermined texture is attached on the surface of bounding box, because the direction of material texture picture does not influence the texture observing effect, so do not need to consider the coordinate direction when texture picture is mapped on face of bounding box, the minimum node of numbering of a face of therefore unified regulation bounding box overlaps with the local coordinate initial point of texture picture, which bar limit of determining this face according to right-handed coordinate system is consistent with texture coordinate x direction, and then another of shared initial point limit is that unity and coherence in writing draws the y direction when shining upon;

Tie up on the bounding box face according to pass, the position between block faces node and each face of bounding box subsequently and obtain mapping point, give corresponding node with the RGBA value at mapping point place, for a node P ₁, calculate its distance to 6 faces of bounding box, its distance is top recently more as can be known, then this spot projection is obtained P on top ₁' point, texture is at P on the face ₁' the RGBA value of locating can obtain, and gives P with it ₁Point promptly obtains the RGBA value of this node;

At last utilize the RGBA value of each node RGBA information of face by each sampled point on the linear interpolation method calculating face for a face to be played up, play up on the opposite again.Providing a kind of simple interpolations method herein, is that example is illustrated with the gore:

Suppose that waiting to play up leg-of-mutton three summits for one is P ₁(x ₁, y ₁, z ₁), P ₂(x ₂, y ₂, z ₂), P ₃(x ₃, y ₃, z ₃), corresponding texture value is respectively (R ₁, G ₁, B ₁, A ₁), (R ₂, G ₂, B ₂, A ₂), (R ₃, G ₃, B ₃, A ₃), 1 P on the triangular facet (x, y, texture information z) (R, G, B, A) can calculate according to following steps:

The distance on calculation level to three summit at first:

Calculate the texture information value then

$R=\frac{{\stackrel{\‾}{\mathrm{PP}}}_1}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·R_1+\frac{{\stackrel{\‾}{\mathrm{PP}}}_2}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·R_2+\frac{{\stackrel{\‾}{\mathrm{PP}}}_3}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·R_3---\left(11\right)$

$G=\frac{{\stackrel{\‾}{\mathrm{PP}}}_1}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·G_1+\frac{{\stackrel{\‾}{\mathrm{PP}}}_1}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·G_2+\frac{{\stackrel{\‾}{\mathrm{PP}}}_3}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·G_3---\left(12\right)$

$B=\frac{{\stackrel{\‾}{\mathrm{PP}}}_1}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·B_1+\frac{{\stackrel{\‾}{\mathrm{PP}}}_2}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·B_2+\frac{{\stackrel{\‾}{\mathrm{PP}}}_3}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·B_3---\left(13\right)$

$A=\frac{{\stackrel{\‾}{\mathrm{PP}}}_1}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·A_1+\frac{{\stackrel{\‾}{\mathrm{PP}}}_2}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·A_2+\frac{{\stackrel{\‾}{\mathrm{PP}}}_3}{{\stackrel{\‾}{\mathrm{PP}}}_1+{\stackrel{\‾}{\mathrm{PP}}}_2+{\stackrel{\‾}{\mathrm{PP}}}_3}\·A_3---\left(14\right)$

There is not interative computation in said method, supposes that a model has N node, M face, and each face up-sampling point gets 2, then will carry out 6N time distance calculation, 4N assignment operation in node texture information calculation stages; Sampled point texture value calculation stages need use formula (11～14) to calculate the texture value of 2M sampled point, and final levels of complexity is (N+M).

3. the mapping relations method for building up between computation optimization result and the stone modeling

The foundation of Optimization Model depends on theoretical analysis and engineer testing, at document [Huang Dongming, Wu Dianliang, Fan Xiumin. the design of the multiple objective programming of gyratory crusher. mechanical engineering journal, Vol.43 (3) 2007.3:204～211] in provide the computation optimization model of a kind of ls in the circular cone shattering process.Because the present invention does not relate to the method for building up of computation optimization model, Optimization Model is not within the scope of the present invention yet, only the output parameter of computation optimization and the mapping relations method for building up between the stone are listed, what the broken Optimization Model of user's lamination was exported also mainly is these parameters herein.

The result parameter of computation optimization model output is: gyratory crusher chamber type curve; Stone material parameter (density p, material texture image); Broken hierarchy number n in the crusher chamber; The particle size after cracking distribution curve η of each crushable layer _i=F _i(Φ) (every kind of shared percentage of granularity, i is a level number, and i=1,2, K, n, Φ ∈ [Φ ₁, Φ ₂]); The particle shape coefficient lambda of stone in the chamber (stone bounding box length breadth ratio); Broken throughput rate P.The parameter of its lumen type curve decision crusher chamber itself no longer influences stone and generates during emulation.Other parameter must guarantee when generating stone.

The key element of structure stone is during emulation: unit interval structure stone volume V; Total stone number N of every layer of generation _i(i is a level number); The every layer of initial generator of stone (topological structure numbering (0-tetrahedron; The 1-hexahedron; The 2-octahedron), shape coefficient λ, volume V _jFractal alternative manner is formula (7)-Shi (10).

To η _iSampling obtains total K _iPlant granularity, then each granularity is

${\mathrm{\Φ}}_j={\mathrm{\Φ}}_1+\frac{{\mathrm{\Φ}}_2-{\mathrm{\Φ}}_1}{K_1}---\left(15\right)$

In the formula: j=1,2, K, K _i

Every kind of shared number percent is η _{I, j}=F _i(Φ _j), and have $\left\{\begin{array}{c}\underset{j=1}{\overset{K_i}{\mathrm{\Σ}}}{\mathrm{\η}}_{i,j}=1\\ \underset{j=1}{\overset{K_i}{\mathrm{\Σ}}}{\mathrm{\η}}_{i,j}\·N_i\·V_j=V\end{array}\right.---\left(16\right)$

In the formula: $V_j={\mathrm{\Φ}}_j^3\·{\mathrm{\λ}}^2,$ Be the volume of every layer various granularity stones, V=P/ ρ.Through type (16) can calculate N _i

Because the size-grade distribution difference of adjacent layer, so their contained stone number difference, the number wall upper strata of lower floor is many, establishes every layer of K _iEquate (to work as K _iScope can satisfy actual conditions when enough big), can set up N by (15), (16) _iWith N _I+1Fixed proportion relation, thereby can access stone number after each stone fragmentation of upper strata, though the constancy of volume can not satisfy every block of stone and break the time, the cumulative volume of adjacent layer equates, meets the constancy of volume criterion.

The stone that volume equates can be got tetrahedron, hexahedron, octahedra generator, gets according to equiprobability usually.

The system architecture synoptic diagram that Fig. 1 adopts for present embodiment.But the physical computing model is as user's edit methods of bottom among the figure, and they are used for physical parameter is calculated; Geometric model is configured to construct the geometric model of the needed stone of emulation, and Real-time and Dynamic texture Core Generator then can generate texture according to the stone texture picture of user's appointment on the stone geometric model; The mapping relations module is used to control the geometric model generation with the result of physical computing between computation optimization model and the stone modeling, thereby can make the stone that constructs meet physics law.Virtual reality system is finished playing up, showing of disintegrating machine and stone model, and the interactive signal between process user and the computer system (as the input and output signal of equipment such as keyboard, mouse, data glove).

Fig. 2 is the composition diagram of the software systems of present embodiment employing, comprise demand management, realistic model management, the calculating of gyratory crusher parameter optimization, gyratory crusher mechanism simulation, gyratory crusher Process Simulation totally 5 one-level modules, Process Simulation also needs to call geometry maker and two submodules of dynamics calculation, and wherein how much makers comprise geometric model structure, texture generation function.

The physical computing model belongs to concrete application in these software systems, should provide by the user, present embodiment uses document [Huang Dongming, Wu Dianliang, Fan Xiumin. the design of the multiple objective programming of gyratory crusher. mechanical engineering journal, Vol.43 (3) 2007.3:204～211] in the ls that provides at the computation optimization model of circular cone shattering process.

In this system, geometric model structure instrument, Real-time and Dynamic texture Core Generator, mapping relations are methods that present embodiment proposes between computation optimization model and the stone modeling.Below the application process of present embodiment in this system is described in detail.

The concrete implementation step of present embodiment is as follows:

1, using the present invention during the Application and Development software systems finishes the stone model and dynamically constructs

(1) determines initial parameter: determine generator, fractal dimension, parameter r _EWith scale factor k.Generator is appointed as 3 kinds: tetrahedron (4 triangles), topological hexahedron (12 triangles), topology octahedra (8 triangles); The present embodiment fractal dimension is 2.3 (structure that satisfies ls, grouan, iron ore requires); Scale factor k is 0.82; Parameter r _EGet 0.7.Initial particle shape coefficient is made as adjustable, and the user can be according to the stone shape adjustments, and present embodiment is taken as 1.32.

(2) polygon models are expressed as { P _i ^j} _i=1, K, N comes the new iteration polygon model of The deformation calculation according to the formula (7,8,9,10) of summary of the invention, and iterations is set to adjustable, and promptly the user can fine degree as required be provided with.

Accompanying drawing 3 is stone models that present embodiment uses the geometric model generation module to obtain, and the generation number average is 6 times.The lower right corner tetrahedral short transverse of stone model generator is bigger, and the stone of generation is also relatively elongated.And adopt the method calculating parameter r of simple and iterations linear dependence _E, use less scale factor k, so the stone that generates is comparatively smooth, bulk deformation is little.

2, using the present invention during the Application and Development software systems finishes the stone dynamic texture and generates

(1) texture tile is affixed on method on the bounding box face: texture picture is quadrilateral, the bounding box face also is a quadrilateral, because picture is the isotropy texture, coordinate direction does not influence the pinup picture effect, as long as the simple textures mapping method of the graphics that uses a computer can be realized.

(2) the RGBA value of polygon node is obtained: for any one node, the point of the graphics that uses a computer calculates the distance of six faces of bounding box to the plan range computing method, selection is from the nearest face of node, node is projected on the face, obtain a subpoint, obtain the RGBA value according to the position of subpoint on this face.

Accompanying drawing 4 shows said process: for the node P among the figure ₁, calculate its distance to 6 faces of bounding box, its distance is top recently more as can be known, then this spot projection is obtained P on top ₁' point, texture is at P on the face ₁' the RGBA value of locating can obtain, and gives P with it ₁Point promptly obtains the RGBA value of this node.

(3) the polygon texture generates: present embodiment is selected 4 sampled points on each triangle, the texture that uses node texture value and sampled point texture value to carry out on the polygon facet generates, sampled point is location determining method on triangle: calculate the triangle centre of form, getting the centre of form respectively is sampled point to the intermediate point of three limit mid point lines, and the centre of form is also as sampled point; Calculate the texture value of the sample point on the polygon facet according to the formula (11,12,13,14) of summary of the invention.

Accompanying drawing 5 is to use the approximate texture mapping of two stone models of this module acquisition, and each triangle has been got 4 sampled points during calculating.Left side stone has 676 nodes, 774 triangles, and the right stone model has 892 nodes, 1076 triangles.

3, use present embodiment during the Application and Development software systems and finish computation optimization model and the dynamically foundation of the mapping relations between the structure of stone model

Mapping relations are the relations between the generator kind, number, volume, shape coefficient of throughput rate, every layer of size-grade distribution, particle shape coefficient and every layer, determine the latter with the former exactly when implementing.Use the formula (15,16) of " gordian technique solution " " 3. mapping relations foundation side between computation optimization result and the stone modeling " of present embodiment, present embodiment K _iGet 18, promptly each layer is all got 18 kinds of granularities.

The implementation result explanation:

Used computing machine is 1 HP9300PC workstation, 6 PCs (wherein 2 are used as the stone dynamics calculation, and 4 as Visual calculation); The disintegrating machine model of institute's emulation is PYB1750,305 rev/mins of the speeds of mainshaft; Broken stone material be ls (density 2.67 * 10 ³Kg/m ³, take from Zhenjiang, Jiangsu), original particle size is φ 105～φ 210, original average particle shape coefficient is 1.18.

During emulation in the chamber stone model mostly be 55263 most, minimum is 15248, the stone model is 31% in the average turnover rate of each emulation frame.The structure of stone model and texture calculate and play up calculating on same computing machine, promptly have 4 computing machines and participate in structure and texture calculating, and the HP9300 workstation is responsible for image output as main control computer; Administrative model structure, texture generate, graph rendering.

Output image resolution is 1024 * 768 during visual simulating, and refresh rate was 49 frame/seconds, is 24 frame/seconds (passive stereo displays) under the stereo display pattern, can satisfy the visual real-time continuous requirement of shattering process.By analysis, stone is modeled under the non-stereo display pattern and accounts for 8% in each emulation frame, and promptly 0.0016 second, texture calculated and accounts for 6%, promptly 0.0016 second.Computation optimization result can ignore in a frame to the mapping time of modeling starting condition.

Use present embodiment and carry out parameter optimization and shattering process emulation, 4 circulations have been carried out altogether, be computation optimization 4 times, broken emulation 4 times, disintegrating machine parameter and corresponding broken statistical parameter after finally having obtained optimizing, accompanying drawing 6 (a) is a chamber type curve, and 6 (b) are grain-size distribution curve (the broken number of plies is 13).The parameter optimization result that this embodiment obtains is summarized as follows: under the situation of the mantle diameter of gyratory crusher, mantle base angle, angle of precession, parallel zone length, minimum gape fixed size, the broken number of plies of lamination is 13; Rotating speed after the optimization is 305 rev/mins; Ip angle is 23.02 degree in the average cavity; The mantle oscillating stroke is 79.74mm; Eccentric throw is 25.63mm; Minimum production rate under the subparameter is 442.62 tons/hour; Peak performance is 728.57 tons/hour; Stone less than the gape size accounts for 82.48% of total stone amount.At present be about 280～480 tons/hour for the broken throughput rate of the used ls of embodiment in the PYB1750 spring cone crusher of using.

Claims (7)

1. the real-time visualized emulation method towards cone crusher performance optimization is characterized in that, comprises that concrete steps are as follows:

Step 1, the foundation of stone geometric model structure unified approach;

Step 2 is set up towards the quick texture mapping method of the stone of real-time demonstration;

Step 3, the mapping relations between disintegrating machine computation optimization and stone model generate are set up.

2. the real-time visualized emulation method towards cone crusher performance optimization according to claim 1, it is characterized in that, described stone geometric model structure unified approach, it is on the polygon model basis of generator, generate new node by iterative formula, thereby obtain new polygon model, the global shape that fractal dimension, random factor, scale factor, iterations controlled variable control the stone of constructing, local smooth and degree of roughness are provided during iteration.

3. the real-time visualized emulation method towards cone crusher performance optimization according to claim 2 is characterized in that, described iterations, and the stone during for cone crushing dynamically generates emulation, and the iterations when generating stone gets 5～7.

4. the real-time visualized emulation method towards cone crusher performance optimization according to claim 1, it is characterized in that, described quick texture mapping method, be meant: on stone geometric model basis, be used for the texture picture fast mapping of stone is arrived each surface of stone geometric model, carrying out texture between 6 faces of picture and bounding box calculates, employing is shone upon texture picture on the bounding box face, to put on the model again to the bounding box projection, the texture information at acquisition point place, the texture that uses the interpolation calculation of point texture information to finish model at last generates.

5. according to claim 1 or 2 or 4 described real-time visualized emulation methods towards cone crusher performance optimization, it is characterized in that, described stone geometric model, the selection of its upper point: when a tri patch minimum edge on the model surface is grown up in 5mm, select whole nodes to add 2～5 sampled points; If lip-deep plane polygon is not a triangle, then be decomposed into the leg-of-mutton combination more than 2; When the minimum length of side of tri patch on the model surface during less than 5mm, use node to carry out interpolation, if the average length of side less than 3mm, selects the part node to carry out interpolation, remaining node according to and sampled point between the minimum principle of the distance texture information value of directly getting sampled point.

6. the real-time visualized emulation method towards cone crusher performance optimization according to claim 2, it is characterized in that, dynamically generate at described stone on the basis of modeling, broken results statistical parameter and stone modeling process that the disintegrating machine computation optimization is obtained combine, set up the mapping relations between the two, make the dynamic generation and the extinction of stone model meet the result of disintegrating machine computation optimization, thereby reach emulation the stone shattering process.

7. the real-time visualized emulation method towards cone crusher performance optimization according to claim 1, it is characterized in that, described stone dynamically generates, its controlled variable is a generator, fractal dimension, iterations, the random factor of scale factor and probability factor, and the stone broken results parameter T/h that the disintegrating machine computation optimization obtains is a throughput rate, the particle shape coefficient, broken hierarchy number in the chamber, every layer of grain-size distribution curve, set up mapping relations between them and just can make the stone broken results parameter that reflection physics calculates in the broken emulation of stone, therefore the physical computing result is carried out visual simulating: the parameter when changing computation optimization model or optimization causes that stone broken results parameter changes, and this changes the variation of the controlled variable of stone modeling when causing emulation, thereby makes computation optimization can control the emulation phenomenon.

Images