CN-121994795-A - Photoelastic sample crystallinity multi-mode data fusion analysis method and system

CN121994795ACN 121994795 ACN121994795 ACN 121994795ACN-121994795-A

Abstract

The invention discloses a photoelastic sample crystallinity multi-mode data fusion analysis method and system, which comprise the steps of identifying photoelastic particles from a two-dimensional image of a photoelastic sample, extracting mass center coordinates and a strong chain network of each photoelastic particle, screening out the photoelastic particles with local area fractions and key orientation order parameters meeting corresponding thresholds, jointly defining the screened photoelastic particles and all the photoelastic particles in direct contact with the screened photoelastic particles as a structural crystal particle set, connecting mass centers of adjacent photoelastic particles in the strong chain network as line segments, constructing a strong chain skeleton, screening out a long-force chain, defining the photoelastic particles covered by the long-force chain belonging to the same cluster as a strong chain crystal particle set, and calculating the fusion crystallinity of the photoelastic sample based on the proportion of the two sets to the total number of the photoelastic particles. The invention overcomes the limitation of the traditional single analysis method and provides a precise and reliable technical means for quantitative analysis of the crystallinity of photoelastic samples and discrete element particle samples.

Inventors

SHI XIUSONG
CHEN MINGYU
CHEN XIANGSHENG
LI AIGUO
LIU KAI
XU JIN
SONG YONGPING

Assignees

深圳大学

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (10)

1. A photoelastic sample crystallinity multimode data fusion analysis method is characterized by comprising the following steps: Acquiring a two-dimensional image of a photoelastic sample, identifying photoelastic particles from the two-dimensional image, and extracting centroid coordinates and a strong chain network of each photoelastic particle; dividing Voronoi cells based on the centroid coordinates, and calculating local area fractions and key orientation order parameters of each photoelastic particle; Screening photoelastic particles with local area fractions and key orientation order parameters meeting corresponding screening thresholds, and defining the screened photoelastic particles and all photoelastic particles in direct contact with the screened photoelastic particles as a structural crystal particle set; Connecting the centroids of adjacent photoelastic particles in the strong chain network into line segments to construct a strong chain framework; Merging adjacent line segments meeting merging conditions into force chains based on directional sequence parameters between adjacent line segments in the strong chain skeleton, and screening out long force chains based on the average value of the number of photoelastic particles forming each force chain; Clustering the direction vectors of the long force chains, and defining photoelastic particles covered by the long force chains belonging to the same cluster as a force chain crystallization particle set; and calculating the fusion crystallinity of the photoelastic sample based on the proportion of the structural crystal particle set and the force chain crystal particle set to the total number of photoelastic particles.
2. The method for multi-modal data fusion analysis of photoelastic sample crystallinity according to claim 1, wherein the strong chain network of each photoelastic particle is extracted based on preset conditions; The preset conditions comprise that the contact force among the photoelastic particles is higher than the average value, the included angle between the main contact force direction and the connecting line of the mass center is smaller than 45 degrees, and the number of continuous photoelastic particles is larger than 3.
3. The method for multi-modal data fusion analysis of photoelastic sample crystallinity according to claim 1, wherein the local area fraction corresponds to a screening threshold greater than 0.8.
4. The method for multi-modal data fusion analysis of photoelastic sample crystallinity according to claim 1, wherein the screening threshold corresponding to the key orientation order parameter is greater than 0.85.
5. The method for multi-modal data fusion analysis of photoelastic sample crystallinity according to claim 1, wherein the steps of merging line segments into a plurality of force chains based on directional order parameters between adjacent line segments in the strong chain skeleton, and screening out long force chains based on average values of number of photoelastic particles constituting each force chain, specifically comprise: calculating the directional sequence parameters between adjacent line segments in the strong chain skeleton; Judging the direction sequence parameter, namely judging that the adjacent line segments belong to the same force chain when the direction sequence parameter is larger than or equal to a preset threshold value, and recording the number of photoelastic particles forming the force chain; the average value of the number of photoelastic particles constituting each force chain is calculated, and a force chain larger than the average value of the number of photoelastic particles is defined as a long force chain.
6. The method for analyzing the multi-modal data fusion of the crystallinity of the photoelastic sample according to claim 1, wherein the clustering of the direction vectors of the long force chain specifically comprises: carrying out weighted average on the included angles of each long force chain and the positive direction of the x axis to obtain a corresponding long force chain direction angle; carrying out standardization processing on all the long force chain direction angles and converting the long force chain direction angles into two-dimensional coordinate vectors; and setting a neighborhood radius and a minimum sample number by taking the two-dimensional coordinate vector as input, and clustering by adopting a DBSCAN clustering algorithm.
7. The method for multi-modal data fusion analysis of photoelastic sample crystallinity according to claim 1, wherein the relationship between the fused crystalline particle set composed of the fused crystallinity and the structural crystalline particle set and the force chain crystalline particle set is as follows: Wherein T represents a fused crystalline particle set; the fusion crystal particles belong to the structural crystal particle set C or the force chain crystal particle set F.
8. The method for fusion analysis of photoelastic sample crystallinity multi-modal data as set forth in claim 1, wherein the fused crystallinity is expressed as Wherein, the Represents the degree of fusion crystallinity; Representing the number of photoelastic particles in the collection of structured crystalline particles; Representing the number of photoelastic particles in the collection of force-chain crystallized particles; Indicating the total number of photoelastic particles; A photoelastic particle in which both the collection of structural crystalline particles and the collection of force-chain crystalline particles coincide is represented.
9. A photoelastic sample crystallinity multimode data fusion analysis system, characterized in that the method of any one of claims 1-8 is applied, the system comprises: the image acquisition module is used for acquiring a two-dimensional image of the photoelastic sample; the image processing module is used for identifying photoelastic particles from the two-dimensional image and extracting centroid coordinates and a strong chain network of each photoelastic particle; The structure analysis module is used for carrying out Voronoi cell division based on the centroid coordinates and calculating local area fraction and key orientation order parameters of each photoelastic particle, screening the photoelastic particles with the local area fraction and the key orientation order parameters meeting the corresponding screening threshold, and defining the screened photoelastic particles and all photoelastic particles in direct contact with the screened photoelastic particles as a structure crystallization particle set; The system comprises a force chain analysis module, a force chain crystallization particle collection module, a force chain analysis module, a force chain crystallization particle collection module and a force chain analysis module, wherein the force chain analysis module is used for connecting the centroids of adjacent photoelastic particles in a force chain network into line segments to construct a force chain skeleton; And the fusion calculation module is used for calculating the fusion crystallinity of the photoelastic sample based on the proportion of the structural crystal particle set and the force chain crystal particle set to the total number of photoelastic particles.
10. A terminal, characterized by comprising a processor and a memory, wherein the memory stores a computer program, and the processor implements the photoelastic sample crystallinity multi-mode data fusion analysis method according to any one of claims 1-8 when executing the computer program.

Description

Photoelastic sample crystallinity multi-mode data fusion analysis method and system Technical Field The invention relates to the technical field of photoelastic particle tests, in particular to a photoelastic sample crystallinity multi-mode data fusion analysis method and system. Background Photoelastic effect experiments are important technical means for exploring the network behaviors of the stress chains of the granular materials, and the stress chain network evolution characteristics of the granular media under the microscopic scale are effectively revealed through the stress birefringence effect visualization force transmission paths of amorphous materials. The single graded particles are extremely easy to form local ordered crystallization arrangement under the action of gravity, so that stress is abnormally concentrated in the crystallization particles, the overall force chain mode observed in a photoelastic image is obviously disturbed, and the accurate representation of the real force chain network state of the particle material by an experiment is influenced. Therefore, the quantitative analysis of crystallinity in the photoelastic sample is an important precondition for researching the network behavior of the force chain of the granular material. At present, crystallinity analysis based on photoelastic images mostly adopts a single structural mode, the geometric structure order of single particles is identified with emphasis, and a reasonable sample crystallinity quantization index is not formed. Meanwhile, the shape of a force chain network under the same structure can be obviously changed due to the difference of loading modes, the force chain network is probably not formed in a geometrically ordered area under the action of pressure, and the crystallinity of the photoelastic sample is only analyzed from a structural mode, so that the crystallinity of the photoelastic sample is unilateral and limited. Therefore, how to build a comprehensive and accurate crystallization degree quantization index, so as to solve the problem that the crystallization state of the system is evaluated to be inaccurate due to the fact that the existing crystallization degree analysis based on the photoelastic image only depends on a single structural mode, unfused particle geometric order and force chain network order information, and the problem is needed to be solved by a person skilled in the art. Disclosure of Invention In view of the above problems, the present invention provides a method and a system for analyzing multi-modal data fusion of photoelastic sample crystallinity, which overcome or at least partially solve the above problems. In order to achieve the above purpose, the present invention adopts the following technical scheme: in a first aspect, the present invention provides a method for analyzing multi-modal data fusion of crystallinity of a photoelastic sample, including: Acquiring a two-dimensional image of a photoelastic sample, identifying photoelastic particles from the two-dimensional image, and extracting centroid coordinates and a strong chain network of each photoelastic particle; dividing Voronoi cells based on the centroid coordinates, and calculating local area fractions and key orientation order parameters of each photoelastic particle; Screening photoelastic particles with local area fractions and key orientation order parameters meeting corresponding screening thresholds, and defining the screened photoelastic particles and all photoelastic particles in direct contact with the screened photoelastic particles as a structural crystal particle set; Connecting the centroids of adjacent photoelastic particles in the strong chain network into line segments to construct a strong chain framework; Merging adjacent line segments meeting merging conditions into force chains based on directional sequence parameters between adjacent line segments in the strong chain skeleton, and screening out long force chains based on the average value of the number of photoelastic particles forming each force chain; Clustering the direction vectors of the long force chains, and defining photoelastic particles covered by the long force chains belonging to the same cluster as a force chain crystallization particle set; and calculating the fusion crystallinity of the photoelastic sample based on the proportion of the structural crystal particle set and the force chain crystal particle set to the total number of photoelastic particles. Further, extracting a strong chain network of each photoelastic particle based on preset conditions; The preset conditions comprise that the contact force among the photoelastic particles is higher than the average value, the included angle between the main contact force direction and the connecting line of the mass center is smaller than 45 degrees, and the number of continuous photoelastic particles is larger than 3. Further, the screening threshold corresponding to the local area fra