CN-121979139-A - Medical implant production regulation method, device, equipment, storage medium and computer program product

CN121979139ACN 121979139 ACN121979139 ACN 121979139ACN-121979139-A

Abstract

The present application relates to the field of intelligent manufacturing technology, and in particular, to a method, an apparatus, a device, a storage medium, and a computer program product for controlling production of a medical implant. The method comprises the steps of obtaining production process data, converging the production process data, preprocessing the production process data, aligning the production process data with time, forming an aligned multi-source data set, generating fusion characteristic representation by adopting self-adaptive weighted fusion based on the multi-source data set, constructing a production line state evaluation index based on the fusion characteristic representation, generating regulation and control trigger information, when the trigger information indicates that the production process parameters need to be adjusted, establishing a mapping model between the production process parameters and quality characterization quantities based on a historical sample set, determining a target production process parameter set according to the mapping model, executing production process parameter update based on the target production process parameter set, obtaining a corresponding quality detection result, writing the target production process parameter set and the quality detection result into the historical sample set to update the mapping model, and reflecting closed loop iteration.

Inventors

LI LING
WANG YAXIAN
ZHANG YONGLIANG
LI RAN
Ren lijing
PAN LIYAN

Assignees

广东机电职业技术学院

Dates

Publication Date: 20260505
Application Date: 20260112

Claims (10)

1. A method of regulating production of a medical implant, the method comprising: acquiring production process data, preprocessing and time alignment the production process data to obtain an aligned multi-source data set, wherein the production process data comprises processing process parameters; generating a fusion feature representation by adopting self-adaptive weighted fusion processing based on the aligned multi-source data set; Constructing a production line state evaluation index based on the fusion characteristic representation, and generating regulation and control trigger information according to the production line state evaluation index; When the regulation trigger information indicates that the machining process parameters need to be regulated, a mapping model between the machining process parameters and the quality characterization quantity is established based on a historical sample set, and a target machining process parameter set is determined based on the mapping model; And based on the target machining process parameter set, updating the machining process parameter, acquiring an updated quality detection result, and writing the target machining process parameter set and the updated quality detection result into the historical sample set so as to update the mapping model and form closed loop iteration.
2. The method of claim 1, wherein the steps of obtaining production process data and pre-processing and time-aligning the production process data to obtain an aligned multi-source data set comprise: acquiring the production process data from a plurality of data sources of a medical implant production line, and converging the production process data to a preset data processing node to obtain a data set to be processed; preprocessing the data set to be processed to obtain preprocessed data, wherein the preprocessing comprises one or more of data cleaning, deletion processing, exception processing and field mapping; And based on the time mark in the preprocessed data, performing time alignment on the data from different data sources, and performing alignment and merging according to a preset alignment rule to obtain an aligned multi-source data set.
3. The method of claim 1, wherein the step of generating a fused feature representation using an adaptively weighted fusion process based on the aligned multi-source dataset comprises: Based on the aligned multi-source data set, respectively carrying out feature extraction on data of different sources, and carrying out normalization processing and dimension mapping on the multi-source features obtained by extraction to obtain a candidate feature set; based on the candidate feature set, constructing a contribution degree weight parameter, and carrying out self-adaptive updating on the contribution degree weight parameter according to the confidence degree, the integrity and the consistency information of the candidate feature set in the current production state to obtain a corresponding target weight; And carrying out weighted fusion processing on the candidate feature set based on the target weight, and carrying out vectorization coding on a fusion result to generate the fusion feature representation.
4. The method of claim 1, wherein the step of constructing a line state evaluation index based on the fused feature representation and generating regulatory trigger information from the line state evaluation index comprises: Constructing an evaluation index system for representing the running state of the production line based on the fusion characteristic representation, and inputting the fusion characteristic representation into an index mapping model corresponding to the evaluation index system to obtain at least one candidate production line state evaluation index; based on the integrity information and fluctuation information represented by the fusion characteristics, carrying out validity judgment on the candidate production line state evaluation indexes to obtain target production line state evaluation indexes; Comparing the target production line state evaluation index with a preset trigger judgment rule, judging whether a regulation and control trigger condition is met based on the comparison result, and generating regulation and control trigger information when the regulation and control trigger condition is met.
5. The method of claim 4, wherein the step of establishing a mapping model between the process parameters and quality metrics based on a set of historical samples and determining a set of target process parameters based on the mapping model when the regulatory trigger information indicates that an adjustment to the process parameters is required comprises: When the regulation trigger information indicates that the processing parameters need to be regulated, a history record corresponding to the current production stage in the history sample set is obtained, and the processing parameters in the history record and the corresponding quality characterization quantity are associated, marked and aligned with the samples to obtain a training sample pair set; establishing a mapping model between the processing parameters and the quality characterization quantity based on the training sample pair set, and training and updating model parameters of the mapping model based on the training sample pair set to obtain a target mapping model for parameter optimization; And based on the target mapping model, evaluating and screening the candidate machining process parameter combination according to a preset parameter constraint condition and a preset optimizing strategy to obtain a target machining process parameter set.
6. The method of claim 1, wherein the step of updating the process parameters based on the target process parameter set and obtaining updated quality inspection results, writing the target process parameter set and the updated quality inspection results to the historical sample set for updating the mapping model and forming a closed loop iteration comprises: Generating a parameter updating instruction corresponding to a production line control object based on the target machining process parameter set, and issuing the parameter updating instruction to the production line control object so that the production line control object can finish machining process parameter updating according to the target machining process parameter set; After the processing parameter is updated, collecting quality detection data of the production batch corresponding to the target processing parameter set, and generating an updated quality detection result based on the quality detection data; and establishing sample association based on the target machining process parameter set and the updated quality detection result, writing the sample association into the historical sample set, updating the mapping model based on the written historical sample set, and forming closed loop iteration.
7. A medical implant production regulation device, the device comprising: The data acquisition module is used for acquiring production process data, preprocessing the production process data and aligning the production process data with time to obtain an aligned multi-source data set, wherein the production process data comprises processing process parameters; the weighted fusion module is used for generating fusion characteristic representation by adopting self-adaptive weighted fusion processing based on the aligned multi-source data set; The regulation and control trigger module is used for constructing a production line state evaluation index based on the fusion characteristic representation and generating regulation and control trigger information according to the production line state evaluation index; the processing parameter module is used for establishing a mapping model between the processing parameter and the quality characterization quantity based on a historical sample set when the regulation and control trigger information indicates that the processing parameter needs to be regulated, and determining a target processing parameter set based on the mapping model; And the target module is used for updating the processing process parameters based on the target processing process parameter set, acquiring an updated quality detection result, and writing the target processing process parameter set and the updated quality detection result into the historical sample set so as to update the mapping model and form closed loop iteration.
8. A medical implant production regulation device, characterized in that the device comprises a memory, a processor and a medical implant production regulation program stored on the memory and executable on the processor, the medical implant production regulation program being configured to implement the steps of the medical implant production regulation method according to any one of claims 1 to 6.
9. A storage medium having stored thereon a medical implant production regulation program which when executed by a processor performs the steps of the medical implant production regulation method according to any one of claims 1 to 6.
10. A computer program product, characterized in that it comprises a computer program which, when executed by a processor, implements the steps of the medical implant production regulation method according to any one of claims 1 to 6.

Description

Medical implant production regulation method, device, equipment, storage medium and computer program product Technical Field The present application relates to the field of intelligent manufacturing technology, and in particular, to a method, an apparatus, a device, a storage medium, and a computer program product for controlling production of a medical implant. Background Manufacturing of medical implants often requires strict control over the stability of the process, consistency of product quality, and traceability of the manufacturing process. The existing production line is generally formed by cooperation of multiple types of processing equipment, detection equipment and an informatization system, and multi-source data including equipment running state information, processing parameter information and quality detection information can be generated in the production process. Because the sampling frequency, the time reference and the data format of each data source are different, the data in the production process often has the characteristics of isomerism, asynchronism and quality spread, and the subsequent analysis and decision are easy to develop based on the same time sequence context. In the prior art, relatively independent strategies are adopted for production regulation and quality control, namely, on one hand, the setting and adjustment of processing parameters usually depend on process experience, off-line test or fixed rules, time delay exists between parameter updating and quality detection, and on the other hand, quality detection results are mostly used for post-decision and retrospection, and associated modeling which can be used for on-line regulation and control is difficult to form with the processing parameters. Particularly, in the multi-type mixed line production scene, the sensitivity of different products to process windows is different, the existing strategy is difficult to dynamically adjust the weight of a data source and a decision basis when the production state changes, and therefore, a stable unified state evaluation caliber and an executable regulation and control trigger mechanism are difficult to form, and the production regulation and control effect of the medical implant is poor. Therefore, how to improve the effect of the production control of the medical implant is a technical problem to be solved. Disclosure of Invention The application mainly aims to provide a medical implant production regulation and control method, a device, equipment, a storage medium and a computer program product, which aim to solve the technical problem of how to improve the effect of medical implant production regulation and control. To achieve the above object, the present application provides a method for controlling the production of a medical implant, comprising the steps of: acquiring production process data, preprocessing and time alignment the production process data to obtain an aligned multi-source data set, wherein the production process data comprises processing process parameters; generating a fusion feature representation by adopting self-adaptive weighted fusion processing based on the aligned multi-source data set; Constructing a production line state evaluation index based on the fusion characteristic representation, and generating regulation and control trigger information according to the production line state evaluation index; When the regulation trigger information indicates that the machining process parameters need to be regulated, a mapping model between the machining process parameters and the quality characterization quantity is established based on a historical sample set, and a target machining process parameter set is determined based on the mapping model; And based on the target machining process parameter set, updating the machining process parameter, acquiring an updated quality detection result, and writing the target machining process parameter set and the updated quality detection result into the historical sample set so as to update the mapping model and form closed loop iteration. In one embodiment, the step of obtaining production process data, and preprocessing and time-aligning the production process data to obtain an aligned multisource data set includes: acquiring the production process data from a plurality of data sources of a medical implant production line, and converging the production process data to a preset data processing node to obtain a data set to be processed; preprocessing the data set to be processed to obtain preprocessed data, wherein the preprocessing comprises one or more of data cleaning, deletion processing, exception processing and field mapping; And based on the time mark in the preprocessed data, performing time alignment on the data from different data sources, and performing alignment and merging according to a preset alignment rule to obtain an aligned multi-source data set. In an embodiment, the step of generating the fusion