CN114329945A - Dynamic cloth simulation method based on identification analysis and PLM system - Google Patents

Abstract

The invention discloses a dynamic cloth simulation method based on identification analysis and a PLM system, and belongs to the technical field of garment simulation. The invention comprises the following steps: the material manufacturer codes the cloth sample information and uploads the coded cloth sample information to an identification analysis platform; the method comprises the steps that a clothing enterprise obtains cloth sample information through an identification analysis platform, records the cloth sample information to an enterprise PLM system, and establishes a spring particle model of the cloth; during dynamic simulation, calculating the speed and displacement of mass points in real time, and detecting whether severe deformation occurs or not so as to determine whether the action of an extension spring and a bending spring is introduced or not; correcting the spring with the occurrence of the super-bounce in real time to limit the elongation of the spring within a system threshold value; and giving an identification analysis two-dimensional code to the obtained cloth dynamic simulation result, and uploading the identification analysis platform and the enterprise PLM system. The invention improves the spring mass point model, solves the problem of spring superelasticity, realizes the dynamic simulation of the cloth with high stability and high calculation efficiency, and can be used in the fields of virtual clothes fitting, personalized customization and the like.

Description

Dynamic cloth simulation method based on identification analysis and PLM system

Technical Field

The invention belongs to the technical field of garment simulation in the aspect of information processing, and relates to a cloth dynamic simulation method based on a physical modeling, identification analysis and product full life cycle management system (PLM).

Background

Personalized customization in digital economy is becoming mainstream, and how to meet the personalized demands of consumers is a major challenge in the clothing industry. The application of the garment simulation technology is promoted by the technical development of 5G, industrial internet, virtual reality and the like, and a new solution is provided for online virtual fitting and personalized garment customization. Therefore, the method has important practical significance for improving the simulation performance and physical authenticity of the cloth.

Physical-based modeling is currently the most effective method for cloth simulation, and comprises a discrete particle model and a continuum model. The continuum model has high simulation precision and can reflect the characteristics of cloth made of different materials, but the calculation amount is obviously increased in the fabric simulation involving a large amount of deformation and movement. The spring mass point model proposed by Provot, the most popular of the particle models, describes the cloth as a collection of mass points connected by ideal springs and establishes the relationship of force and deformation based on the theory of elasticity and newton's law. The introduction of the dynamic equation brings a cloth simulation model which is more real and more in line with the physical motion rule.

However, during the cloth simulation based on the spring mass point model, unreasonable overstretching is easy to occur to cause simulation distortion, namely, the phenomenon of super-elasticity, so that the development of the garment simulation technology is restricted. The existing ultra-elastic solution is difficult to meet the dynamic simulation requirements of the clothes in the aspects of simulation stability and reality, and corresponding theoretical support is lacked in the aspect of selection of correction parameters.

In addition, the existing cloth dynamic simulation method cannot acquire and identify real cloth information in real time, cannot trace simulation original data and simulation results, and cannot effectively search positioning error data. Therefore, there is a need to combine the product full lifecycle management system (PLM) and the identity resolution technique to solve this problem.

An industrial internet identification analysis system is a key hub for realizing information intercommunication of all industrial elements and all links. By giving an identifier to each fabric object, cross-region, cross-industry and cross-enterprise information query and sharing are realized. The identification technology is utilized to record and query the information of the fabric such as color, weight, thickness, manufacturer and the like, the data of the product is mastered in all directions in the whole process, the traceability of the garment simulation algorithm can be improved, and the accuracy and effectiveness of the algorithm are improved.

The PLM system effectively integrates people, processes and information related in the garment fabric production process, acts on the whole garment supply chain, traverses the whole life cycle of products from concepts to scrapping, supports information such as collaborative research and development, management, distribution and use products and the like related to the products, and establishes a real-time, systematic and traceable management system of garment fabric physical data and simulation data by combining with an identification analysis system.

Disclosure of Invention

The invention aims to provide a cloth dynamic simulation method based on identification analysis and a PLM system, which solves the problems that simulation distortion exists in the existing cloth simulation technology, real cloth information collection and identification cannot be carried out in real time and the like. The method realizes a high-stability and high-calculation-efficiency dynamic cloth simulation method by combining a physical modeling technology through an identification analysis technology and a PLM system, and can be used in the fields of virtual clothes fitting, personalized customization and the like. The method also aims at the problems of large calculated amount and low efficiency of the traditional cloth simulation method, improves the spring mass point model and solves the problem of spring super-elasticity.

The invention discloses a cloth dynamic simulation method based on identification analysis and PLM system, comprising the following steps:

step 1, a material manufacturer uploads attribute information such as color, thickness, material quality, glossiness, softness, transparency and the like of a cloth sample to an identification analysis platform through identification analysis technology coding;

step 2, the clothing enterprises obtain the coded cloth sample information such as the color of the cloth sample uploaded by the material suppliers through the identification analysis platform and record the coded cloth sample information to the PLM system of the enterprises; the technical department of the enterprise analyzes the clothing cloth model according to the original data, establishes a spring mass point model of the simulated cloth, and determines model data such as the number of spring mass points, the mass of the spring mass point, the coefficient of structural spring, the elastic coefficient of bending spring, the elastic coefficient of extension spring, the gravity acceleration, the time step and the like of the model;

step 3, when a spring mass point model of the cloth is established, dispersing the obtained initial cloth information into a mesh model consisting of mass points and three types of springs connecting the mass points, and obtaining the position of each mass point on the cloth and the relative position of the mass point with other mass points; the structure spring acts on left, right, upper and lower mass points, the bending spring is connected with the mass point on the diagonal line, and the extension spring is connected with the mass point which is separated by one mass point in the transverse direction and the longitudinal direction;

step 4, in the dynamic simulation of the cloth, the time is discretized into t, the time interval is h, and the spring elasticity f borne by each mass point is obtained according to the Newton's law of mechanics and the Hooke's law_sGravity f_gAnd a damping force f_dAnd corresponding acceleration a, calculating the resultant force borne by the mass point: f ═ F_g+f_s+f_d(ii) a Solving a kinetic equation set based on an implicit Euler method, and calculating the speed and displacement of each particle at the next moment;

step 5, judging whether to introduce a bending spring and an extension spring according to the deformation degree of the cloth model: when the model is severely deformed, the functions of an extension spring and a bending spring are introduced; when the cloth is in non-violent deformation, the extension spring and the bending spring are removed, and only the action of the structural spring on mass points is considered;

step 6, in the dynamic simulation of the cloth, the elongation of each spring is detected, when the elongation exceeds a system threshold value, the spring is judged to be in the state of super-bounce, and the spring in the state of super-bounce is corrected, so that the elongation is limited within the system threshold value;

step 7, constructing FDH (Fixed orientations shells, FDH) of cloth and other objects, and performing collision detection on the collided objects; when collision response occurs, detecting the position relation between the material points of the cloth model and a collision object, if penetration occurs, adjusting the penetration points to the surface of the object, and if penetration does not occur, not processing;

step 8, displaying the dynamic simulation effect of the cloth, and repeating the steps 3-8 according to the dynamic change of the cloth;

and 9, endowing the obtained cloth dynamic simulation result with a unique identification analysis two-dimensional code, and uploading the identification analysis platform and the enterprise PLM system so that cloth manufacturers and consumers can obtain effect data.

Further, in the step 5, the deformation degree of the cloth model is judged, and the functions of the extension spring and the bending spring are dynamically introduced or eliminated; the method specifically comprises the following steps: calculating dihedral angles of all adjacent triangles, judging that severe deformation occurs when the dihedral angles are smaller than 90 degrees, and simultaneously considering the acting force of 3 spring forces of a structural spring, a bending spring and an extension spring on a mass point; otherwise, judging that non-violent deformation occurs, and only calculating the acting force of the structural spring on the cloth model at the moment.

Further, in the step 6, the spring force of the spring with the occurrence of the superelasticity is corrected; the method specifically comprises the following steps: setting the original length of the spring with the threshold value of 1.2 times, traversing the elongation of all springs connected with mass points, judging whether the elongation is smaller than the set threshold value, and if the elongation exceeds the set threshold value, adjusting the elasticity of the spring; the adjusting method comprises the following steps: and updating the elastic force between the over-stretched springs by using the average value of the forces applied to the adjacent springs of the super-elastic spring as a correction value, so that the elongation of the springs is limited within a system threshold value.

Compared with the prior art, the invention has the following advantages and positive effects:

(1) compared with the traditional hyperelastic correction method, the method has the advantages that the value of the correction value is more in line with the real physical significance, the hyperelastic spring can be well corrected, and the stability and the simulation efficiency are higher.

(2) According to the invention, the extension spring and the bending spring are dynamically introduced according to the deformation degree of the model, the system calculation amount can be reduced on the premise of keeping fold details and simulation precision, and the simulation efficiency is improved.

(3) The invention utilizes PLM system and identification analysis technology, cloth original data and simulation data can be traced and fed back, and cloth simulation information can be better obtained and returned.

(3) The invention has better application value for the clothing retail industry and the personalized customization industry.

Drawings

FIG. 1 is a flow chart of the implementation of the dynamic cloth simulation method based on PLM and identification analysis according to the present invention;

FIG. 2 is an exemplary diagram of information after a sample cloth is coded in the embodiment of the present invention;

FIG. 3 is a schematic diagram of a cloth model showing a phenomenon of spring superelasticity deformation;

FIG. 4 is a schematic view of the force analysis of a spring near the superelastic region in the method of the present invention;

FIG. 5 is a graph comparing the treatment of the wrinkle effect in the method of the present invention;

FIG. 6 is a graph comparing the effect of the method of the present invention on the superelastic correction.

Detailed Description

In order to make the objects, technical means and advantages of the present invention more comprehensible, embodiments of the present invention are described in detail below with reference to the accompanying drawings.

The cloth dynamic simulation method based on the identification analysis and PLM system is realized based on the following software and hardware configuration.

Hardware environment: a computer;

software configuration: windows 2000/XP; python et al, any language environment, tachi software.

As shown in fig. 1, the method for dynamically simulating a fabric based on an identifier resolution and PLM system according to the embodiment of the present invention includes the following 9 steps, and the implementation of each step is described below.

Step 1, a material manufacturer needing to obtain cloth simulation data assigns codes to cloth sample information through an identification analysis technology according to the characteristics of the cloth sample such as color, thickness, material, glossiness, softness and transparency, and uploads the assigned codes to an identification analysis platform.

As shown in fig. 2, the coded cloth sample information includes an identification ID, cloth attribute information, a network address and an interoperation interface; the ID is obtained by an identification analysis technology; the attribute information comprises information such as a cloth producing area and material quality; the network address is the network address where the cloth sample is uploaded.

Step 2, the garment enterprises acquire the code assigning data such as the color of the cloth sample uploaded by the material suppliers through the identification analysis platform and record the data to the PLM system of the enterprises; the enterprise technical department carries out modeling analysis on the clothes and cloth according to the original data, establishes a cloth spring mass point model, and determines model data such as the number of spring mass points, the mass of the spring mass point, the elastic coefficient of a structural spring, the elastic coefficient of a bending spring, the elastic coefficient of a tension spring, the gravity acceleration, the time step and the like.

And 3, constructing a cloth physical structure analysis model by using the spring mass point model. The cloth is discretized into the set of mass point and spring in the spring mass point model modeling process, and each mass point is connected by the spring with different functional actions, and 3 kinds of springs are mainly used: structural springs, bending springs and extension springs were used to simulate the restraint in the cloth. The structural spring acts on the left, right, upper and lower mass points to fix the model structure. The bending spring connects mass points on the diagonal to prevent the model from being distorted. The extension spring is connected with mass points which are separated by mass points in the transverse direction and the longitudinal direction, so that the edge smoothness of the model is ensured when the model is greatly deformed.

And acquiring the position of each particle on the cloth and the relative position of each particle to other particles in real time.

Step 4, in the dynamic simulation of the cloth, the time is discretized into t, the time interval is h, and the time interval is [ t, t + h ] at any time interval]And analyzing the stress condition of the mass points of the cloth model. The acceleration a of each mass point can be obtained by analyzing the stress condition of each mass point and according to Newton's second law F ═ ma, and m is the mass of the spring mass point. The resultant force of mass points in the system includes the spring force f_sGravity f_gAnd a damping force f_dCalculating the resultant force borne by the particles: f ═ F_g+f_s+f_d. And solving a kinetic equation set based on an implicit Euler method, and calculating the speed and displacement of each particle at the next moment:

v(t+h)＝v(t)+a(t+h)h

x(t+h)＝x(t)+v(t+h)h (1)

in the formula, a (t + h) and v (t + h) are the acceleration and velocity of the particle at the time t + h. v (t), x (t) represents the velocity and position of the particles at time t. The position x (t + h) and the velocity v (t + h) of the particle at the next time can be calculated from equation (1).

And 5, judging the deformation degree of the cloth model to determine whether acting forces of a bending spring and an extension spring are introduced to the mass point. The invention provides a method for calculating dihedral angles of adjacent triangles in real time, judging local deformation degree and dynamically introducing extension springs and bending springs, and solves the problem that the function of spring force with 3 different functions is solved in real time by considering a traditional model, so that system calculation amount is increased. In the embodiment of the invention, a deformation threshold is firstly set, preferably the deformation threshold is set to be 90 degrees, when the dihedral angle of adjacent triangles is smaller than 90 degrees, the area is judged to be severely deformed, and at the moment, the acting force of three springs on mass points is considered at the same time. When the dihedral angle is greater than or equal to 90 degrees, the cloth is in a non-severe deformation state (such as suspension), and only the acting force of the structural spring to the mass point is calculated, and the acting forces of the extension spring and the bending spring are ignored.

And 6, performing the superelastic correction on the spring excessively stretched in the step 4.

The reason why the super-elasticity occurs is that in the simulation process, for convenience of processing, the elastic force and the deformation are generally in an ideal linear relationship, the deformation of the spring is increased when the elastic force applied to the vertexes is increased, and the phenomenon that the spring is excessively stretched between the vertexes due to the adoption of the ideal linear spring is called as the super-elasticity phenomenon. As shown in FIG. 3, the spring superelasticity occurs, and the numbers 1-16 are spring numbers.

In the present invention, first, the threshold value of the spring elongation is set to 1.2 times the original spring length. And detecting the elongation of each spring in the system in real time, when the elongation of a certain spring exceeds a system threshold value, the spring generates the overstretch, and for the spring generating the overstretch, calculating the average value of the tensile force borne by the surrounding springs as a correction value to update the elastic force among the springs, so that the elongation of the spring is recovered to a normal state. The method comprises the following specific steps:

firstly, intercepting the area with serious deformation but without super-elasticity in the cloth simulation diagram for amplification analysis, as shown in fig. 3, observing that the deformation trends in adjacent areas are approximately the same, so that it can be assumed that: the stress condition of any spring is similar to that of the adjacent spring.

Next, the area in fig. 3 where the super-bounce occurred was subjected to mechanical analysis, as shown in fig. 4. In the simulation, the vertex a is subjected to the tension f of the vertices b and d_b→a，f_d→aAnd the dots b and d are located to the left and below the dot a, respectively. In order to simplify the calculation amount, only the force action of the structural spring on the mass point is considered, and the mechanical analysis is further performed on the point b, and the pulling forces from the vertexes a, c and e are respectively as follows: f. of_a→b、f_c→bAnd f_e→b. Vertex c is further analyzed and the tension from vertices b, d, f, h is: f. of_b→c、f_d→c、f_f→c、f_h→c. And the force experienced by vertex d: f. of_a→d、f_c→d、f_i→dSimilarly, there is a vertex e which is pulled by the tension f of the vertexes f and b_f→e、f_b→eThe vertex f is subjected to the tension force f of the vertices g, e and c_g→f、f_e→f、f_c→fThe vertex i is pulled by the vertices d and h_d→i、f_h→iThe vertex h is subjected to the tension f of the vertices g, i and c_g→h、f_i→h、f_c→h. According to the method of the invention, the springs ab and ad are overstrained, and the left tension f is applied to the springs around the springs ab and the springs ab_b→aThe left-side tensile force with the same action has f_e→b、f_c→d、f_f→c、f_h→iAnd f_g→h(ii) a Around the spring adLower tension f on the surrounding spring and applied to the spring ad_d→aThe same downward force has f_c→b、f_i→d、f_h→c、f_f→eAnd f_g→f(ii) a The correction values at the springs ab, ad where the super bounce occurs are respectively:

using correction values

And

as new elasticity values of ab edge and ad edge, FIG. 6(f) is a result obtained by correcting the super-elasticity by the method of the present invention, and simulation results show that the method can well inhibit the occurrence of the super-elasticity phenomenon.

And 7, constructing FDH bounding boxes of other objects in the cloth and the scene, and judging whether collision behaviors occur or not by detecting whether the bounding boxes are intersected or not. Firstly, monitoring whether the FDH bounding boxes between the cloth and surrounding objects are intersected or not in real time, if not, not colliding, and rejecting the parts which are not collided. Otherwise, the distance between the cloth material point and the object surface material point is monitored for the area where the collision occurs, when the distance is negative, the penetration is indicated, and the cloth material point is adjusted to the object surface.

Step 8, displaying the dynamic simulation effect of the cloth, and repeating the steps 3-8 according to the dynamic change of the cloth;

and 9, endowing the obtained cloth dynamic simulation result with a unique identification analysis two-dimensional code, and uploading the unique identification analysis two-dimensional code to an enterprise PLM system and an identification analysis platform, so that cloth manufacturers and consumers can obtain simulation effect data.

Aiming at the improved result of the spring mass point model in the step 5, a simulation experiment of freely falling the cloth on the sphere is carried out, and the traditional method and the method for removing the stretching and bending springs are compared with the method of the invention, as shown in fig. 5 and table 1, the removal of the two springs improves the simulation speed to a certain extent, but causes the loss of precision on the simulation details. The method of the invention reserves the fold details under the condition of saving the calculation time and ensures the simulation precision.

Table 1: parameter comparison of the method of the present invention with existing methods

Method	Number of peaks	Number of springs	Run time/s
				Conventional methods	16384	2283008	9.46170
Method of removing tension bending spring	16384	2283008	7.62344
				The method of the invention	16384	2283008	8.28423

In order to verify the effectiveness of the method for solving the problem of the super-elasticity of the spring mass point model in the cloth simulation, the method is compared with a correction method based on the variable elasticity coefficient, the position, the speed and the internal force correction.

The stress and the simulation parameters of the cloth are different under different simulation environments and are difficult to express by actual units, so that the unit of the variable parameter is omitted in the simulation environment of the invention and the unit is expressed by a specific numerical value. The number of the mass points of the spring mass point model of the simulated cloth in the experiment is 14 multiplied by 14, the mass of the spring mass point is 1, the structural spring coefficient is 20, the elastic coefficient of a bending spring is-0.25, the elastic coefficient of a tension spring is 20, the gravity acceleration is-0.0221, the model is in an initial state and is parallel to the ground, the cloth starts to freely swing down under the action of gravity when the program starts to run, the mass point coordinates and the speed are updated by adopting an implicit Euler method, the time step length is 0.005, and each frame is updated for 10 times. The simulation results are shown in fig. 6.

Fig. 6 (a) shows the phenomenon of the spring mass model occurring when no correction method is adopted, and in the experiment, when the spring elongation is 1.2 times the original length of the spring, the phenomenon is judged to occur and the above correction methods are introduced. Firstly, the elastic coefficient is directly adjusted to 700 by adopting a method of hyperelastic correction, the elastic coefficient of a structural spring is adjusted to-0.5, the elastic coefficient of a bending spring is adjusted to-0.5, and the elastic coefficient of a stretching spring is adjusted to 700, and the experimental result is shown in fig. 6 (b). Fig. 6(c) shows a position-based calibration method, which can solve the problem of the overshoot but causes the model oscillation distortion by forcibly adjusting the position of the overshoot point to 0.7 times the original length of the spring when the overshoot occurs. Fig. 6(d) shows a velocity-based correction method, in which when a pop occurs, the corresponding vertex velocity is reduced to 1/2. Fig. 6(e) shows a correction method based on the internal force correction method, in which the elastic force is increased to 1.5 times the original value when the superelasticity occurs, and the results show that the two methods can suppress the occurrence of the superelasticity phenomenon to some extent, but the simulation effect is generally stable. Fig. 6(f) shows the result of the correction by the method of the present invention, which can be seen that the method of the present invention can ensure that the model is stable and the occurrence of the superelasticity phenomenon is suppressed without distortion. In conclusion, the force-based adjustment method can avoid model instability caused by forcibly adjusting the elastic coefficient, the position or the speed, and effectively solve the problem of superbounce under the condition of ensuring the stability of the system.

Claims (7)

1. A cloth dynamic simulation method based on identification analysis and PLM system is characterized by comprising the following steps:

step 1, uploading attribute information of a cloth sample by a material provider, coding the cloth sample information by an identification analysis technology, and uploading the cloth sample information to an identification analysis platform; the cloth sample information comprises an identification code and cloth attributes;

step 2, the clothing enterprise acquires the cloth sample information uploaded by the material supplier through the identification analysis platform and records the cloth sample information to an enterprise PLM system; when an enterprise technical department analyzes the clothing and cloth model, acquiring cloth attributes and establishing a spring mass point model of the cloth; PLM represents a product full lifecycle management system;

step 3, when a spring mass point model of the cloth is established, dispersing the cloth into a mesh model consisting of mass points and three types of springs connecting the mass points, and acquiring the position of each mass point on the cloth and the relative position of the mass point with other mass points in real time; the three types of springs comprise: structural springs acting on left, right, upper and lower mass points, bending springs connecting mass points on diagonal lines, and extension springs connecting mass points separated by one mass point in the transverse direction and the longitudinal direction;

step 4, in the dynamic simulation of the cloth, dispersing time into t, and analyzing the stress condition of the mass points in real time, calculating resultant force and acceleration borne by the mass points, and calculating the speed and displacement of each mass point at the next moment;

step 5, judging whether acting forces of a bending spring and an extension spring are introduced or not according to the deformation degree of the cloth model; when the cloth model is judged to be severely deformed, the acting forces of the extension spring and the bending spring are introduced; otherwise, only the acting force of the structural spring on the mass point is considered;

step 6, in the dynamic simulation of the cloth, detecting the elongation of each spring in real time, judging that the spring is overstroked when the elongation exceeds a system threshold value, and correcting the overstroked spring to limit the elongation of the spring within the system threshold value;

step 7, building FDH (fully drawn) surrounding bodies of other objects in the cloth and the scene, and performing collision detection on the collided objects; when collision response occurs, detecting the position relation between the material points of the cloth model and a collision object, if penetration occurs, adjusting the penetration points to the surface of the object, and if penetration does not occur, not processing;

step 8, displaying the dynamic simulation effect of the cloth, and repeating the steps 3-8 according to the dynamic change of the cloth;

and 9, endowing the obtained cloth dynamic simulation result with a unique identification analysis two-dimensional code, and uploading the identification analysis platform and the enterprise PLM system.

2. The method according to claim 1, wherein in the step 5, the deformation degree of the cloth model is determined, specifically: and setting a deformation threshold value L, calculating dihedral angles of all adjacent triangles in the cloth model, judging that severe deformation occurs when the dihedral angles are smaller than L, and otherwise, judging that non-severe deformation occurs.

3. The method of claim 2, wherein the deformation threshold L is 90 degrees.

4. The method of claim 1, wherein in step 6, the system threshold for spring elongation is set to 1.2 spring original lengths.

5. The method according to claim 1 or 4, wherein in step 6, the elongation of each spring is detected in real time, and the springs with the occurrence of the superelasticity are corrected, specifically: and traversing the elongation of all the springs connected with the mass points, judging whether the elongation exceeds a system threshold value, if so, performing the spring overstroke, and updating the elasticity of the spring by using the average value of the stress of the springs adjacent to the spring as a correction value to limit the elongation of the spring within the system threshold value.

6. The method of claim 5, wherein in step 6, when the spring is overstrained and corrected, considering only the force of the structural spring on the mass point, let ab and ad be overstrained, a, b and d be mass points, b be the left mass point of mass point a, d be the lower mass point of mass point a, and f be the pulling force of mass points b and d on mass point a_b→a，f_d→a(ii) a Find the pulling force of the left mass point on the spring around the spring ab, f_b→aAveraging the tension of the mass points on the left side found out to obtain a new elastic force value of the spring ab; find the pulling force of the lower mass point on the spring around the spring ad, f_d→aAnd averaging the tension of the found lower mass point to obtain a new elastic force value of the spring ad.

7. The method of claim 1, wherein in step 4, the spring force f experienced by each mass point is obtained_sGravity f_gAnd a damping force f_dThe resultant force F ═ F borne by the calculated mass point_g+f_s+f_d(ii) a And then calculating the acceleration of each mass point according to Newton's second law and the mass of the spring mass point.

Images