CN-122021191-A - Computing processing method of drug packaging bag barrier twin simulation data

Abstract

The invention relates to the field of barrier property simulation and data processing of medicine packaging materials, in particular to a calculation processing method of barrier property twin simulation data of medicine packaging bags, which comprises the steps of obtaining original grid topology data, time sequence permeation flux data, defect-free diffusion models and defect characteristic knowledge base of a multilayer medium structure; the method comprises the steps of generating an ideal blocking reference field, mapping the ideal blocking reference field into a reference flux matrix, calling a defect parameterization operator to generate a theoretical defect state, calculating a theoretical residual matrix, calculating a real residual matrix, carrying out space-time similarity calculation on the real residual matrix and the theoretical residual matrix, generating a defect confidence result, reserving real defect area data, carrying out rejection or replacement processing on the real defect area data, outputting corrected simulation data, and updating a defect parameterization operator parameter or similarity judgment threshold value group based on a plurality of continuous processing periods.

Inventors

• SHAO JIANXIN
• Yao Yongwu

Assignees

• 温州市天艺塑料有限公司

Dates

Publication Date

20260512

Application Date

20260409

Claims (10)

1. 1. The calculation processing method of the drug packaging bag barrier twin simulation data is characterized by comprising the following steps: Step one, original grid topology data, time sequence permeation flux data, a defect-free diffusion model and a defect characteristic knowledge base of a multi-layer medium structure are obtained, wherein the defect characteristic knowledge base at least comprises defect type labels and parameterized operator configuration rules corresponding to each defect type; Generating an ideal blocking reference field based on the original grid topological data and the defect-free diffusion model, and mapping the ideal blocking reference field into a reference flux matrix which corresponds to the time sequence permeation flux data one by one on a uniform space grid coordinate and a uniform time sampling node; calling a defect parameterization operator to act on the ideal blocking reference field according to the defect characteristic knowledge base to generate a theoretical defect state, and calculating a theoretical residual matrix based on the difference value of the permeation flux distribution corresponding to the theoretical defect state and the permeation flux distribution corresponding to the ideal blocking reference field; Calculating a real residual matrix based on the difference value between the time sequence permeation flux data and the reference flux matrix; and fifthly, performing space-time similarity calculation on the actual residual matrix and the theoretical residual matrix, generating a defect confidence result by combining a similarity judgment threshold value group, retaining data judged as a real defect area based on the defect confidence result, performing rejection or replacement processing on the data judged as a digital noise area, and outputting corrected simulation data, wherein the defect confidence result is used for forming a parameter update record, and updating parameters of the defect parameterization operator or the similarity judgment threshold value group based on recognition results in a plurality of continuous processing periods.
2. 2. The method for computing and processing barrier twin simulation data of a medicine packaging bag according to claim 1, wherein the defect-free diffusion model is constructed based on a preset diffusion transmission mechanism and is used for representing a standard concentration gradient field and/or a standard permeation flux field in a stable state with uniform thickness, continuous interface and boundary conditions.
3. 3. The method for computing and processing barrier twin simulation data of a medicine packaging bag according to claim 1, wherein the original grid topology data at least comprises node coordinates, unit connection relations, layer thickness information and interlayer interface position relations of each layer of a multi-layer medium structure.
4. 4. The method for computing and processing barrier twin simulation data of a medicine packaging bag according to claim 1, wherein the defect feature knowledge base comprises at least one of an interface peeling feature, a local free volume expansion feature, a microcrack communication feature and a pinhole leakage feature, and the defect parameterization operator comprises at least one of an interface permeation resistance adjustment operator, a local diffusion coefficient adjustment operator and a communication path enhancement operator.
5. 5. The method for computing and processing barrier twin simulation data of a medicine packaging bag is characterized in that the theoretical residual error matrix is a space-time residual error data matrix formed by position-by-position difference between permeation flux distribution corresponding to the theoretical defect state and permeation flux distribution corresponding to the ideal barrier reference field under a unified space grid and time sampling axis, and the real residual error matrix is a space-time residual error data matrix formed by position-by-position difference between the time sequence permeation flux data and the reference flux matrix.
6. 6. The method for computing and processing barrier twin simulation data of a medicine packaging bag according to claim 1, wherein the space-time similarity computation comprises one of performing three-dimensional cross correlation computation on the real residual matrix and the theoretical residual matrix according to a unified space grid, performing dynamic time warping distance computation on a time sequence of a corresponding position, or performing weighted fusion after normalizing the three-dimensional cross correlation computation result and the dynamic time warping distance computation result.
7. 7. The method for calculating and processing barrier twin simulation data of a medicine packaging bag according to claim 1 is characterized in that in the fifth step, a segmentation judgment process is adopted in the process of generating a defect confidence result, retaining or executing rejection and replacement processing, when the space-time similarity is larger than or equal to a preset high threshold value, a corresponding area is judged to be a real defect area, when the space-time similarity is smaller than or equal to a preset low threshold value, a corresponding area is judged to be a numerical noise area, and when the space-time similarity is larger than the low threshold value and smaller than the high threshold value, a defect parameterization operator corresponding to an additional defect type in a defect feature knowledge base is called to regenerate the theoretical residual matrix and execute re-judgment, wherein the high threshold value is larger than the low threshold value.
8. 8. The method for calculating and processing barrier twin simulation data of a medicine packaging bag according to claim 7, wherein when a judging result is a true defect area, data corresponding to the true defect area is reserved, defect position marks are output, when the judging result is a numerical noise area, rejection processing is carried out on the data corresponding to the numerical noise area, or replacement processing is carried out on reference data, neighborhood interpolation data or historical stable period data of a position corresponding to an ideal barrier reference field, and when the judging result is carried out again, the data still falls between the low threshold value and the high threshold value, marks to be rechecked are output.
9. 9. The method for computing and processing barrier twin simulation data of a medicine packaging bag according to claim 1, wherein the parameter updating record is used for representing a distribution statistical result of a real defect area and a numerical noise area in a plurality of continuous processing periods, when the deviation of the center space distance of the real defect area in the plurality of continuous processing periods is smaller than a preset position threshold value and the consistency of the defect type is higher than a preset type threshold value, the defect weight in the defect feature knowledge base is updated according to the parameter updating record, and when the numerical noise area in the plurality of continuous processing periods is concentrated in a preset space neighborhood range in space or the frequency of repeated occurrence in time is higher than a preset frequency threshold value, the similarity judgment threshold value group is updated according to the parameter updating record.
10. 10. The method for computing and processing barrier twin simulation data of a medicine packaging bag according to claim 1, wherein the multi-layer medium structure is a medicine packaging bag composite barrier material, the original grid topology data is three-dimensional grid data of a multi-layer film body of the medicine packaging bag, and the time sequence permeation flux data is oxygen permeation simulation data or water vapor permeation simulation data.

Description

Computing processing method of drug packaging bag barrier twin simulation data Technical Field The invention relates to the field of barrier property simulation and data processing of medicine packaging materials, in particular to a calculation processing method of barrier property twin simulation data of medicine packaging bags. Background In the existing barrier simulation process of the medicine packaging bag, solving an oxygen permeation process or a water vapor permeation process based on a finite element or finite volume method aiming at a multilayer composite film body formed by an outer protective layer, a barrier layer and an inner heat sealing layer, and outputting three-dimensional grid data of the multilayer film body and local permeation flux distribution data under a plurality of time nodes; In the existing data processing mode, the local abnormal high-flux area in the simulation result is generally processed by adopting modes such as fixed threshold screening, local abnormal value elimination, global smoothing or simple interpolation correction and the like so as to reduce numerical fluctuation influence caused by grid distortion, time step truncation and local solving instability; However, by adopting a fixed threshold, direct smoothing or simple replacement mode, local flux sudden increase caused by real defects and abnormal fluctuation caused by numerical noise are easily mixed, so that real tiny leakage points are deleted by mistake, or numerical noise is misjudged as a material defect, and the reliability, consistency and usability of blocking simulation data of a medicine packaging bag are reduced. Disclosure of Invention In order to solve the technical problems, the invention provides a calculation processing method of barrier twin simulation data of a medicine packaging bag, and specifically, the technical scheme of the invention comprises the following steps: The calculation processing method of the drug packaging bag barrier twin simulation data comprises the following steps: step one, original grid topology data, time sequence permeation flux data, a defect-free diffusion model and a defect feature knowledge base of a multi-layer medium structure are obtained, wherein the defect feature knowledge base at least comprises defect type labels and parameterized operator configuration rules corresponding to each defect type; Generating an ideal blocking reference field based on original grid topological data and a defect-free diffusion model, and mapping the ideal blocking reference field into a reference flux matrix which corresponds to the time sequence permeation flux data on a uniform space grid coordinate and a uniform time sampling node one by one; Calling a defect parameterization operator to act on an ideal blocking reference field according to a defect characteristic knowledge base to generate a theoretical defect state, and calculating a theoretical residual error matrix based on the difference value of permeation flux distribution corresponding to the theoretical defect state and permeation flux distribution corresponding to the ideal blocking reference field; Calculating a real residual matrix based on the difference value between the time sequence permeation flux data and the reference flux matrix; based on the defect confidence result, reserving the data judged as a real defect area, performing rejecting or replacing processing on the data judged as a numerical noise area, and outputting corrected simulation data; the defect confidence results are used for forming parameter updating records, and updating parameters or similarity judgment threshold value groups of the defect parameterization operator based on the identification results in a plurality of continuous processing periods. Preferably, the defect-free diffusion model is constructed based on a preset diffusion transport mechanism for characterizing a standard concentration gradient field and/or a standard permeation flux field at uniform thickness, continuous interface and steady state boundary conditions. Preferably, the original grid topology data at least comprises node coordinates, unit connection relations, layer thickness information and interlayer interface position relations of each layer of the multi-layer medium structure. Preferably, the defect feature knowledge base comprises at least one of an interface peeling feature, a local free volume expansion feature, a microcrack communication feature and a pinhole leakage feature, and the defect parameterization operator comprises at least one of an interface penetration resistance adjustment operator, a local diffusion coefficient adjustment operator and a communication path enhancement operator. Preferably, the theoretical residual error matrix is a space-time residual error data matrix formed by position-by-position difference between permeation flux distribution corresponding to the theoretical defect state and permeation flux distribution corresponding to