Search

CN-121786044-B - Data maintenance system and method based on distributed material boxing coding system

CN121786044BCN 121786044 BCN121786044 BCN 121786044BCN-121786044-B

Abstract

The invention provides a data maintenance method of a distributed material boxing coding system, which comprises the steps of obtaining a single coding binding record, calculating a credibility metric value for each coding node, constructing a coding behavior causal relation structure, identifying conflict paths formed by multiple binding of the same material or the same bag code in a period, calculating path reliability metrics of all the conflict paths, carrying out abnormal path identification on the conflict paths, screening out a conflict path with the highest path reliability metrics as a reserved path, dynamically determining a verification strategy of the binding record in the current boxing period based on the reserved path, generating a coding binding record passing verification, and carrying out redundancy persistence storage according to the corresponding connection degree in the coding behavior causal relation structure based on the coding binding record passing verification. The invention ensures that the data consistency, the system efficiency and the traceability are uniformly ensured under the same framework.

Inventors

  • LI YONG
  • YANG ZHONG

Assignees

  • 广州慧翼智能科技有限公司

Dates

Publication Date
20260508
Application Date
20260304

Claims (9)

  1. 1. Data maintenance system based on distributed material vanning coding system, its characterized in that, the system includes: The code-assigning acquisition unit is used for acquiring a single code-assigning binding record and mapping the code-assigning binding record into code-assigning behavior nodes, wherein the single code-assigning binding record at least comprises a material identifier, a bag code identifier, a code-assigning time stamp and a code-assigning node identifier, and a credibility metric value is calculated for each code-assigning node based on the single code-assigning binding record; the causal relationship construction unit is used for constructing a coding behavior causal relationship structure based on the credibility metric value and the corresponding coding node identification; The conflict and anomaly analysis unit is used for identifying conflict paths formed by multiple binding of the same material or the same bag code in a period based on the coding behavior causal relationship structure, and calculating path reliability metric values of all the conflict paths; based on the path reliability measurement value, carrying out abnormal path identification on the conflict paths, screening out conflict paths with highest path reliability measurement values as reserved paths and other conflict paths, wherein the conflict paths are used as abnormal paths; the effective path acquisition unit is used for dynamically determining a verification strategy of the binding record in the current boxing period based on the reserved path, and generating a coded binding record passing verification as an effective binding record; The coding statistics unit is used for executing redundancy persistence storage according to the corresponding connection degree in the coding behavior causal relationship structure based on the coding binding record passing the verification and recording the path index information of the redundancy persistence storage; The coding behavior causal relationship structure is constructed based on the credibility metric value and the corresponding coding node identifier, and specifically comprises the following steps: Grouping the coding binding records according to the coding node identifiers, and establishing directed association representing the sequence in the group according to the coding time mark sequence; If the time difference of the coding time marks falls within a preset same-box candidate time range between coding binding records of different coding node marks, establishing candidate association representing concurrency relation; And calculating association weights for each directional association and candidate association by combining the credibility metric value, the node operation stability mark and the time interval.
  2. 2. The data maintenance method based on the distributed material boxing coding system is characterized by comprising the following steps of: Acquiring a single code-assigning binding record, mapping the code-assigning binding record into code-assigning behavior nodes, wherein the single code-assigning binding record at least comprises a material identifier, a bag code identifier, a code-assigning time mark and a code-assigning node identifier, and calculating a credibility metric value for each code-assigning node based on the single code-assigning binding record; constructing a coding behavior causal relationship structure based on the credibility metric value and the corresponding coding node identifier; Based on the coding behavior causal relationship structure, identifying conflict paths formed by multiple binding of the same material or the same bag code in a period, and calculating path reliability metric values of all the conflict paths; based on the path reliability measurement value, carrying out abnormal path identification on the conflict paths, screening out conflict paths with highest path reliability measurement values as reserved paths and other conflict paths, wherein the conflict paths are used as abnormal paths; dynamically determining a verification strategy of the binding record in the current boxing period based on the reserved path, and generating a coded binding record passing verification as an effective binding record; based on the code-assigning binding record passing the verification, executing redundancy persistence storage according to the corresponding connection degree in the code-assigning behavior causal relation structure, and simultaneously recording path index information of the redundancy persistence storage; The coding behavior causal relationship structure is constructed based on the credibility metric value and the corresponding coding node identifier, and specifically comprises the following steps: Grouping the coding binding records according to the coding node identifiers, and establishing directed association representing the sequence in the group according to the coding time mark sequence; If the time difference of the coding time marks falls within a preset same-box candidate time range between coding binding records of different coding node marks, establishing candidate association representing concurrency relation; And calculating association weights for each directional association and candidate association by combining the credibility metric value, the node operation stability mark and the time interval.
  3. 3. The data maintenance method based on a distributed material boxing and coding system according to claim 2, wherein the calculating a credibility metric value for each coding node based on the single coding binding record is specifically as follows: Analyzing the time interval distribution of the code binding records of the same code node in the sliding time window to evaluate the time stability; Detecting whether the same material identification and bag code identification binding combination exists in a short time to evaluate repeatability; Acquiring operation monitoring information of the coding node to evaluate the node operation state; and integrating the time stability, the repeatability and the evaluation result of the node running state, and calculating to obtain the credibility metric value.
  4. 4. The data maintenance method based on a distributed material boxing and coding system according to claim 3, wherein the node operation stability mark corresponds to a dimension of the coding node mark and is used for representing the operation stability degree of the current coding node mark in a current sliding time range.
  5. 5. The data maintenance method based on a distributed material boxing and coding system according to claim 2, wherein based on the coding behavior causal relationship structure, a conflict path formed by multiple binding of the same material or the same bag code in a period is identified, specifically: Based on the fact that the same material identifier only allows a corresponding bag code identifier in the same boxing period, the same bag code identifier also only allows one material identifier to be bound in the same boxing period, and similar nodes are aggregated to screen out node sets with multiple binding candidates; And constructing a plurality of candidate causal paths along the directional association backtracking by taking the trigger point as an end point based on each conflict trigger point of the node set to form conflict paths, wherein in the backtracking process, links from the same coding node are unfolded along the sequence edges of stations, links crossing the coding node are unfolded along the concurrent candidate edges, the latest record index mapping of the coding node is utilized in traversal, the latest record number of each coding node is used as a backtracking starting point, and the backtracking is carried out forward along the sequence edges of the node only to a preset backtracking depth, so that the path construction is limited in the range of the recent action chain of each node in a high concurrency period, and the unbounded expansion of the whole graph is avoided.
  6. 6. The data maintenance method based on a distributed material boxing and coding system according to claim 5, wherein the conflict path reliability measurement value accumulates the weight of each association on a path, penalizes frequent hops across coding nodes in the path, and is calculated and generated based on the number of hops of the conflict path in the dimension of the coding nodes.
  7. 7. The data maintenance method based on the distributed material boxing and coding system according to claim 2, wherein the step of dynamically determining a verification policy of the binding record in the current boxing period based on the reserved path, and generating the coding binding record passing verification specifically includes: acquiring nodes of the abnormal path and the reserved path, wherein the number of behavior nodes which are invalid and correspond to the abnormal path corresponds to the total number of behavior nodes which participate in boxing in the boxing period and correspond to the reserved path; calculating the abnormal proportion of the structure according to the number of the invalid behavior nodes and the total number of the behavior nodes participating in boxing in the boxing period; Calculating the average value of the credibility metric values of all the effective nodes in the current period; calculating a check state index of the current boxing period by combining the structural abnormality proportion and the average value of the credibility metric value; And determining a final verification strategy according to a preset verification strategy mapping table and the interval in which the verification state index is located, wherein the strategy at least comprises non-checking, spot checking and full checking.
  8. 8. The data maintenance method based on a distributed material boxing and coding system according to claim 2, wherein the performing redundancy persistence storage is specifically: acquiring the number of direct predecessor and the number of direct successor of the effective binding record in the coding behavior causal relationship structure; Calculating redundancy factors of the structural drive according to the sum of the direct precursor number and the direct subsequent number; Determining the number of redundant copies according to the redundancy factor, and writing the binding records into the corresponding number of redundant storage nodes; wherein, in the step of recording path index information: and storing the identification information of the direct predecessor node and the direct successor node of the binding record in the coding behavior causal relationship structure.
  9. 9. The method for maintaining data based on a distributed material packing and coding system according to claim 8, wherein when selecting redundant storage nodes, storage nodes located at different physical or logical positions are preferentially selected according to a decentralized principle.

Description

Data maintenance system and method based on distributed material boxing coding system Technical Field The invention belongs to the field of data maintenance, and particularly relates to a data maintenance system and method based on a distributed material boxing coding system. Background Along with the development of the manufacturing industry to the high automation and the large scale, the material boxing and coding system gradually evolves into a distributed system form of multi-station parallel operation and cross-node cooperative processing. The coding process not only requires the establishment of a stable one-to-one correspondence between materials and bag codes, but also ensures the consistency and the usability of binding data in links such as subsequent boxing check, logistics tracing, quality management and the like under the conditions of high concurrency and asynchronous execution. In the existing boxing coding system, a plurality of coding nodes work simultaneously, and binding results are intensively written into a background system to improve the overall efficiency, but in actual operation, different coding nodes have objective differences in network state, equipment load and operation stability, so that the situations of delay, repetition, disordered sequence and the like of binding data inevitably occur in the process of generating and transmitting. The problems can be processed by simple rules when the data volume is small or the number of nodes is limited, but after the distributed scale is enlarged, the problems are easy to evolve into structural conflict, so that a system is difficult to accurately judge which binding relations are derived from a real business process, and which are only caused by acquisition conditions or concurrent execution. Meanwhile, when key nodes such as boxing completion and box arrangement processing trigger verification or rejection operations, the existing scheme is mostly dependent on fixed proportion or static threshold values, lacks the capability of describing the overall state of coding behaviors, and can only be remedied through manual intervention or full backtracking after the fact, so that the production beat is affected, and the data maintenance cost is increased. Therefore, how to systematically describe the generation process, evolution relation and running state of the coding data under the distributed boxing coding scene, and accordingly realize a data maintenance mechanism with identifiable conflict, controllable processing and traceable result has become a key problem for restricting the further development of the prior art. Disclosure of Invention The invention aims to provide a data maintenance system and method based on a distributed material boxing coding system, and solves the problems. To achieve the above object, in a first aspect of the present invention, there is provided a data maintenance system based on a distributed material boxing and coding system, the system comprising: The code-assigning acquisition unit is used for acquiring a single code-assigning binding record and mapping the code-assigning binding record into code-assigning behavior nodes, wherein the single code-assigning binding record at least comprises a material identifier, a bag code identifier, a code-assigning time stamp and a code-assigning node identifier, and a credibility metric value is calculated for each code-assigning node based on the single code-assigning binding record; the causal relationship construction unit is used for constructing a coding behavior causal relationship structure based on the credibility metric value and the corresponding coding node identification; The conflict and anomaly analysis unit is used for identifying conflict paths formed by multiple binding of the same material or the same bag code in a period based on the coding behavior causal relationship structure, and calculating path reliability metric values of all the conflict paths; based on the path reliability measurement value, carrying out abnormal path identification on the conflict paths, screening out conflict paths with highest path reliability measurement values as reserved paths and other conflict paths, wherein the conflict paths are used as abnormal paths; the effective path acquisition unit is used for dynamically determining a verification strategy of the binding record in the current boxing period based on the reserved path, and generating a coded binding record passing verification as an effective binding record; and the coding statistical unit is used for executing redundancy persistence storage according to the corresponding connection degree in the coding behavior causal relation structure based on the coding binding record passing the verification and recording the path index information. In a second aspect of the present invention, there is provided a data maintenance method based on a distributed material boxing and coding system, the method comprising: Acqui