CN-122022847-A - Intelligent full-flow quality tracing method and system

CN122022847ACN 122022847 ACN122022847 ACN 122022847ACN-122022847-A

Abstract

The invention relates to the technical field of product quality tracing, in particular to an intelligent full-flow quality tracing method and system, which are implemented by collecting full-flow monitoring data covering the life cycle of a product; the method comprises the steps of calculating first risk indexes of a product in a plurality of production links based on full-flow monitoring data and preset historical normal data, obtaining link importance of each production link, correcting the first risk indexes according to the link importance to obtain second risk indexes, constructing a full-flow risk table based on the second risk indexes, tracing according to the full-flow risk table when quality problems occur in the product, screening target risk links, tracing front-end influence links based on the target risk links, locking abnormal data dimensions corresponding to the front-end influence links, and generating monitoring and early warning information. Therefore, the data integration and the risk dynamic quantification of the whole life cycle of the product are realized, the accuracy and the efficiency of quality tracing are improved, and the potential risks can be identified in a prospective manner.

Inventors

CHEN FENGQIN
WU CHUNBAO

Assignees

苏州数琨创享信息技术有限公司

Dates

Publication Date: 20260512
Application Date: 20260204

Claims (10)

1. An intelligent full-flow quality tracing method is characterized in that the method is applied to an intelligent full-flow quality tracing system, and the method comprises the following steps: Collecting data covering a raw material end, a production end, a storage end, a logistics distribution end, a terminal consumption end and an after-sale feedback end in the life cycle of a product to obtain full-flow monitoring data; Calculating first risk indexes of the product in a plurality of production links based on the whole process monitoring data and preset historical normal data, wherein the first risk indexes are at least obtained by fusing the deviation degree of the monitoring data in each link with link abnormality; acquiring link importance of each production link, wherein the link importance is determined based on influence weight of each production link on product quality, historical quality problem occurrence rate and economic influence information; correcting the first risk index according to the link importance degree to obtain a second risk index; constructing a full-flow risk table based on the second risk index, and tracing according to the full-flow risk table when the quality problem of the product occurs, and screening target risk links; and tracing the preamble influencing link based on the target risk link, locking the abnormal data dimension corresponding to the preamble influencing link, and generating monitoring and early warning information.
2. The intelligent full-process quality tracing method according to claim 1, wherein before collecting data covering a raw material end, a production end, a storage end, a logistics distribution end, a terminal consumption end and an after-sales feedback end in a product life cycle to obtain full-process monitoring data, the method further comprises: setting data acquisition nodes for a plurality of production links in a product life cycle, and configuring corresponding data acquisition interfaces for each data acquisition node; And writing the mapping relation between the data acquisition interface of the data acquisition node and the corresponding production link into a system configuration table, and completing loading and verification of the mapping relation between the data acquisition interface and the production link in the system configuration table in a system initialization stage.
3. The intelligent full-process quality tracing method of claim 1, wherein said step of calculating a first risk indicator of a product in a plurality of production links based on said full-process monitoring data and a preset historical normal data comprises: Extracting real-time monitoring data corresponding to a target production link, and comparing the real-time monitoring data with historical normal data corresponding to the target production link, wherein the target production link is any one of a plurality of production links in the product life cycle; Calculating the deviation degree of each item of monitoring data in the target production link, wherein the deviation degree is used for reflecting the deviation condition between the real-time data and the historical base line; And calculating link abnormality of the target production link based on the deviation degree of the monitoring data corresponding to the target production link to obtain a first risk index.
4. The intelligent full-process quality tracing method of claim 3, wherein said step of calculating the deviation degree of each item of monitoring data in said target production link comprises: Acquiring reference sequences of all monitoring data of the target production link in a plurality of historical batches; Calculating the correlation index coefficient and the credibility difference between the real-time monitoring data and the reference sequence; and calculating the deviation degree of the monitoring data based on the correlation index coefficient and the reliability difference.
5. The intelligent full-process quality tracing method according to claim 1, wherein said step of obtaining link importance of each of said production links comprises: extracting the times and processing cost of quality problems of a target production link in the past based on a historical quality event database, wherein the target production link is any one of a plurality of production links in the product life cycle; Calculating the occurrence rate of historical problems and economic influence information according to the times of quality problems in the past of the target production link and the processing cost; Determining the influence weight corresponding to the target production link by combining the position, the process complexity and the repair difficulty of the target production link in the product structure; and integrating the influence weight, the historical problem occurrence rate and the economic influence information, and calculating to obtain the link importance.
6. The intelligent full-process quality tracing method of claim 1, wherein said step of correcting said first risk indicator according to said link importance level to obtain a second risk indicator comprises: if the first risk index of the current production link is higher than a preset risk threshold, tracing the risk state of the preamble production link of the current production link; calculating an influence factor of the preamble production link on the current production link, wherein the influence factor is determined based on the correlation between the monitoring data deviation sequence of the preamble production link and the current production link; and if the influence factor is larger than a preset influence threshold, carrying out weighted correction on the first risk index based on the influence factor and the link importance degree to obtain a second risk index.
7. The intelligent full-process quality traceability method according to claim 1, wherein said step of constructing a full-process risk table based on said second risk index comprises: arranging the second risk indexes of each production link according to the process flow sequence to form a structured risk table; Labeling each production link with a corresponding link importance and an identification of the production link affecting the preamble; And associating the structured risk table with the product batch information to obtain a whole-flow risk table, wherein the whole-flow risk table can be used for inquiring and screening production links according to batches, time ranges or link types.
8. The intelligent full-process quality traceability method according to claim 7, further comprising performing a multi-stage risk potential analysis on said intelligent full-process quality, comprising: Detecting minor parameter deviations of the product in a plurality of production links of production, storage and logistics; based on a cross-stage parameter interaction model, calculating an accumulated potential risk index according to the tiny parameter deviation; and generating potential defect early warning information when the accumulated potential risk index exceeds a preset early warning threshold value.
9. The intelligent full-process quality traceability method according to claim 8, wherein said step of calculating an accumulated risk potential index based on said micro-parameter bias based on a cross-phase parametric interaction model comprises: When the micro parameter deviation is monitored, responding to the micro parameter deviation, and extracting micro deviation trend, intra-batch variability and inter-stage correlation characteristics of key parameters of the product in each production link; Based on a historical failure mode and a preset expert rule, constructing the cross-stage parameter interaction model; and carrying out weighted cumulative calculation on the small deviation trend, the intra-batch variability and the inter-stage correlation characteristic of the key parameter corresponding to the small parameter deviation to obtain the cumulative potential risk index.
10. An intelligent full-process quality traceability system, the system comprising: The full-link acquisition module is used for acquiring data of a raw material end, a production end, a storage end, a logistics distribution end, a terminal consumption end and an after-sale feedback end in the life cycle of the covered product to obtain full-flow monitoring data; The monitoring data calculation module is used for calculating first risk indexes of the product in a plurality of production links based on the whole-flow monitoring data and preset historical normal data, wherein the first risk indexes are at least obtained by fusing the deviation degree of the monitoring data in each link with link abnormality; the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring link importance of each production link, and the link importance is determined based on influence weight of each production link on product quality, historical quality problem occurrence rate and economic influence information; The correction module is used for correcting the first risk index according to the link importance degree to obtain a second risk index; the tracing module is used for constructing a full-flow risk table based on the second risk index, tracing according to the full-flow risk table when the quality problem occurs in the product, and screening target risk links; and the early warning module is used for tracing the preamble influencing link based on the target risk link, locking the abnormal data dimension corresponding to the preamble influencing link and generating monitoring early warning information.

Description

Intelligent full-flow quality tracing method and system Technical Field The invention relates to the technical field of product quality tracing, in particular to an intelligent full-flow quality tracing method and system. Background In the context of global industry chain depth fusion and manufacturing industry digital transformation acceleration, product quality management poses multiple challenges. The traditional quality tracing mode has the common information island phenomenon, data of all links are stored in different systems in a scattered mode, and a unified data interaction mechanism is lacked, so that the effective integration of all-chain data from raw material purchasing, production processing and storage logistics to terminal sales is difficult to realize. In the related art, the data acquisition process is highly dependent on manual input and paper bill transmission, so that the efficiency is low, the problems of data tampering, loss or inconsistency are easily caused, the time and labor are consumed in the tracing process, and the data reliability is difficult to guarantee. When the quality defect occurs to the product, the existing system can only locate the problem to the batch level, cannot refine to the microscopic dimensions of specific working procedures, equipment running states or operator behaviors and the like, and is difficult to accurately identify the root links of the quality hidden trouble, so that the recall range is enlarged, the cost is high, and the brand reputation is damaged. More prominently, the prior art lacks prospective identification capability for potential risks, data analysis is limited to passive traceability after quality problems occur, and risk trend cannot be predicted based on real-time monitoring data. The root of the problems is that a data monitoring mechanism covering the whole life cycle of the product cannot be established, and dynamic quantification and cross-stage impact analysis on risk indexes of each link are lacking, so that a quality traceability system cannot meet the requirements of modern manufacturing industry on accurate and real-time management. Disclosure of Invention In order to break through the problem that a data monitoring mechanism of the whole life cycle of a product does not exist, the application provides an intelligent whole-flow quality tracing method and system, so that the whole life cycle of the product can be monitored, and the adaptability of the quality tracing system to accurate and real-time management is improved. In order to achieve the above objective, in a first aspect, the present application provides an intelligent full-process quality tracing method, where the intelligent full-process quality tracing method is applied to an intelligent full-process quality tracing system, the method includes: Collecting data covering a raw material end, a production end, a storage end, a logistics distribution end, a terminal consumption end and an after-sale feedback end in the life cycle of a product to obtain full-flow monitoring data; calculating first risk indexes of the product in a plurality of production links based on the whole-flow monitoring data and preset historical normal data, wherein the first risk indexes are at least obtained by fusing the deviation degree of the monitoring data in each link with the link abnormality; acquiring link importance of each production link, wherein the link importance is determined based on influence weight of each production link on product quality, historical quality problem occurrence rate and economic influence information; correcting the first risk index according to the link importance degree to obtain a second risk index; Constructing a full-flow risk table based on the second risk index, and when a quality problem occurs in a product, tracing according to the full-flow risk table, and screening a target risk link; and tracing the preamble influencing link based on the target risk link, locking the abnormal data dimension corresponding to the preamble influencing link, and generating monitoring and early warning information. In an embodiment, the method further includes, before acquiring the full-process monitoring data, acquiring data covering the raw material end, the production end, the storage end, the logistics distribution end, the terminal consumption end and the after-sales feedback end in the product life cycle: setting data acquisition nodes for a plurality of production links in a product life cycle, and configuring corresponding data acquisition interfaces for each data acquisition node; And writing the mapping relation between the data acquisition interface of the data acquisition node and the corresponding production link into a system configuration table, and completing loading and verification of the mapping relation between the data acquisition interface and the production link in the system configuration table in the system initialization stage. In one embodim