Search

CN-121981694-A - Power material supply chain monitoring method and device based on WBS-BOM dynamic mapping

CN121981694ACN 121981694 ACN121981694 ACN 121981694ACN-121981694-A

Abstract

The invention discloses a power material supply chain monitoring method and device based on WBS-BOM dynamic mapping, belonging to the field of power material management, wherein the method comprises the steps of obtaining a WBS data tree and a BOM data table of a target power engineering project, and performing feature conversion through a BERT model to obtain a plurality of WBS node vectors and a plurality of BOM material vectors; the method comprises the steps of respectively calculating similarity scores between WBS node vectors and BOM material vectors, establishing a WBS-BOM dynamic mapping table according to the similarity scores, obtaining first time points when target BOM materials in a BOM material set reach target material unloading points, respectively calculating dynamic lead time of the target BOM materials, correspondingly adding the first time points and the dynamic lead time to obtain a plurality of second time points, comparing the maximum value in the second time points with planned starting time of target WBS nodes, and generating a monitoring report according to comparison results. Therefore, by implementing the invention, the accuracy and the efficiency of monitoring the power supply chain can be improved.

Inventors

  • WANG YIJIE
  • JIN YATING
  • HE XIAOJUN
  • Bao Jiangxue
  • YANG ANTAO
  • YU CHENXI
  • MA JUN
  • SHEN QI
  • WENG HUIYING
  • CHEN TIANMEI
  • ZHANG GONG
  • YIN WEIBIN
  • ZHANG WEILIN
  • MENG XIAN
  • ZHANG BING
  • XIA HONGTAO
  • WANG TAO

Assignees

  • 国网浙江省电力有限公司物资分公司
  • 国网浙江浙电招标咨询有限公司
  • 国网浙江省电力有限公司

Dates

Publication Date
20260505
Application Date
20260409

Claims (10)

  1. 1. A power supply chain monitoring method based on WBS-BOM dynamic mapping, comprising: Acquiring a WBS data tree and a BOM data table of a target power engineering project, and respectively inputting the WBS data tree and the BOM data table into a pre-trained BERT model so that the BERT model converts the WBS data tree into a plurality of WBS node vectors and converts the BOM data table into a plurality of BOM material vectors; Based on cosine similarity and a preset expert rule base, similarity scores between the WBS node vectors and the BOM material vectors are calculated respectively, and a WBS-BOM dynamic mapping table is established according to numerical comparison results between the similarity scores and preset thresholds; Extracting a BOM material set corresponding to a target WBS node according to the WBS-BOM dynamic mapping table, obtaining a first time point when each target BOM material in the BOM material set reaches a target material unloading point, respectively calculating the dynamic lead time of each target BOM material based on the material attribute of each target BOM material and a preset risk buffering period, and correspondingly adding each first time point and each dynamic lead time to obtain a plurality of second time points; Comparing the maximum value in each second time point with the planned starting time of the target WBS node, and generating a power supply chain monitoring report according to the comparison result.
  2. 2. The WBS-BOM dynamic mapping-based power supply chain monitoring method of claim 1, wherein the obtaining WBS data tree and BOM data table of the target power engineering project and inputting the WBS data tree and BOM data table into a pre-trained BERT model, respectively, so that the BERT model converts the WBS data tree into a plurality of WBS node vectors and the BOM data table into a plurality of BOM material vectors, comprises: Acquiring a WBS data tree from an engineering progress management system through a data interface, and acquiring a BOM data table from a material supply chain system; Word segmentation processing is respectively carried out on the WBS data tree and the BOM data table, and a WBS node description text set and a BOM material description text set are correspondingly obtained; and respectively converting each WBS node description text in the WBS node description text set into WBS node vectors through a pre-trained BERT model, and respectively converting each BOM material description text in the BOM material description text set into BOM material vectors.
  3. 3. The WBS-BOM dynamic mapping-based power supply chain monitoring method according to claim 1, wherein before the obtaining WBS data tree and BOM data table of the target power engineering project and inputting the WBS data tree and BOM data table into the pre-trained BERT model, respectively, further comprising: acquiring a corpus sample set in the field of power engineering; And carrying out iterative training on the initial BERT model based on the corpus sample set in the electric power engineering field until the BERT model reaches a preset convergence condition to obtain an optimal BERT model.
  4. 4. The WBS-BOM dynamic mapping-based power supply chain monitoring method of claim 1, wherein the calculating similarity scores between the WBS node vectors and the BOM material vectors based on cosine similarity and a preset expert rule base, respectively, comprises: Calculating cosine similarity between each WBS node vector and each BOM material vector, and taking a calculation result as a first similarity score; Judging the logic relation between each WBS node vector and each BOM material vector according to a preset expert rule base, and mapping the judgment result into a second similarity score; And respectively carrying out weighted summation on each first similarity score and the corresponding second similarity score to obtain a plurality of similarity scores.
  5. 5. The method for monitoring the power supply chain based on the WBS-BOM dynamic mapping according to claim 1, wherein the establishing the WBS-BOM dynamic mapping table according to the numerical comparison result between each similarity score and the preset threshold value includes: If any similarity score is larger than a preset threshold, determining that a mapping relationship exists between the WBS node vector corresponding to the similarity score and the BOM material vector, and generating a mapping record between the WBS node vector and the BOM material vector; And storing each mapping record to obtain a WBS-BOM dynamic mapping table.
  6. 6. The method for monitoring the power supply chain based on the WBS-BOM dynamic mapping according to claim 1, wherein the extracting the BOM material set corresponding to the target WBS node according to the WBS-BOM dynamic mapping table, and obtaining the first time point when each target BOM material in the BOM material set reaches the target material unloading point, includes: Acquiring WBS node vectors of target WBS nodes, extracting BOM material vectors with mapping relation with the WBS node vectors through the WBS-BOM dynamic mapping table, and taking BOM materials corresponding to the BOM material vectors as target BOM materials; And for any target BOM material, acquiring a first GPS coordinate of the target BOM material and a second GPS coordinate of a target material unloading point, and calculating to obtain a first time point when the target BOM material reaches the target material unloading point through a road condition API based on the first GPS coordinate and the second GPS coordinate.
  7. 7. The WBS-BOM dynamic mapping-based power supply chain monitoring method of claim 6, wherein the calculating the dynamic lead time of each target BOM material based on the material property of each target BOM material and a preset risk buffer period, respectively, includes: For any target BOM material, calling historical average unpacking inspection time consumption of the target BOM material from a material management database based on material properties of the target BOM material; acquiring BIM coordinates of a target equipment installation point through a building information model, and calculating the in-site transportation time consumption of the target BOM material based on the second GPS coordinates and the BIM coordinates; and adding the historical average unpacking inspection time consumption, the in-field transportation time consumption and the risk buffering period preset by the target power engineering project to obtain the dynamic lead time of the target BOM material.
  8. 8. The WBS-BOM dynamic mapping-based power supply chain monitoring method of claim 1, wherein comparing the maximum value of each of the second time points with the planned starting time of the target WBS node, and generating a power supply chain monitoring report according to the comparison result, comprises: Analyzing the material lag risk condition of the target WBS node and the construction period delay risk condition of the target power engineering project according to the comparison result and the total floating time length corresponding to the target WBS node, wherein the total floating time length is determined through the path position of the target WBS node in the target power engineering project; If the target WBS node does not have the material lag risk, generating a first prediction result, and packaging the first prediction result into a power material supply chain monitoring report; if the target WBS node has a material lag risk and the target power engineering project does not have a construction period lag risk, generating a second prediction result, and packaging the second prediction result and a material lag time length into a power material supply chain monitoring report, wherein the material lag time length is the difference value between the maximum value and the planning starting time; If the target power engineering project has a construction period delay risk, a third prediction result is generated, and the third prediction result, the material lag time and the construction period delay time are packaged into a power material supply chain monitoring report, wherein the construction period delay time is the difference value between the material lag time and the total floating time.
  9. 9. The method for monitoring the power supply chain based on the WBS-BOM dynamic mapping according to claim 8, wherein the analyzing the material lag risk condition of the target WBS node and the construction period lag risk condition of the target power engineering project according to the comparison result and the total floating time length corresponding to the target WBS node comprises: if the maximum value is smaller than or equal to the planned starting time, judging that the target WBS node does not have the material hysteresis risk; if the maximum value is greater than the planned starting time, judging that the target WBS node has a material hysteresis risk; when the target WBS node has a material lag risk, comparing the material lag time with the total floating time, and if the material lag time is less than or equal to the total floating time, judging that the target power engineering project has no construction period lag risk; And if the material lag time is longer than the total floating time, judging that the target power engineering project has a construction period delay risk.
  10. 10. The power material supply chain monitoring device based on WBS-BOM dynamic mapping is characterized by comprising a feature conversion module, a mapping construction module, a time prediction module and a result generation module; the feature conversion module is used for acquiring a WBS data tree and a BOM data table of a target power engineering project, and respectively inputting the WBS data tree and the BOM data table into a pre-trained BERT model so that the BERT model converts the WBS data tree into a plurality of WBS node vectors and converts the BOM data table into a plurality of BOM material vectors; the mapping construction module is used for respectively calculating similarity scores between the WBS node vectors and the BOM material vectors based on cosine similarity and a preset expert rule base, and establishing a WBS-BOM dynamic mapping table according to numerical comparison results between the similarity scores and preset thresholds; The time prediction module is configured to extract a BOM material set corresponding to a target WBS node according to the WBS-BOM dynamic mapping table, obtain a first time point when each target BOM material in the BOM material set reaches a target material unloading point, respectively calculate a dynamic lead time of each target BOM material based on a material attribute of each target BOM material and a preset risk buffering period, and correspondingly add each first time point and each dynamic lead time to obtain a plurality of second time points; and the result generation module is used for comparing the maximum value in each second time point with the planned starting time of the target WBS node and generating a power supply chain monitoring report according to the comparison result.

Description

Power material supply chain monitoring method and device based on WBS-BOM dynamic mapping Technical Field The invention relates to the field of power material management, in particular to a power material supply chain monitoring method and device based on WBS-BOM dynamic mapping. Background In the existing power material management system, engineering Project Management (PMS) and material supply chain management (ERP/WMS) are two core systems which operate independently. The PMS system focuses on engineering progress management and control of WBS (Work Breakdown Structure, work decomposition structure) layers, and has the core function of decomposing the whole electric power engineering project into a tree structure according to the logic of the branches, the branches and the working procedures, and a manager tracks the advancing condition of each working procedure through the system. For example, the "110kV substation construction project" may be disassembled into the partitions such as "civil engineering," "equipment installation," and "debug test," and the "equipment installation" may be further disassembled into the partitions such as "main transformer installation" and "switchgear installation," and finally refined to a specific operation process node. The ERP/WMS system is focused on logistics management of BOM (Bill of Materials ) layers and is mainly responsible for information recording and circulation of links such as purchasing, storage and transportation of electric power materials. For example, for specific materials such as "110kV main transformer" and "GIS switchgear", state data such as purchase order generation, manufacturer production, logistics transportation, warehouse entry and ex-warehouse delivery are tracked. However, the existing power material management methods have the following defects that firstly, the prior art lacks a dynamic and automatic association mechanism between engineering progress and material states. The existing association mechanism is mostly dependent on a static comparison table or simple rules maintained manually, cannot adapt to engineering change and real-time change of material states, and is easy to form a data island. The manager is difficult to quickly locate the influence of material delay on a specific process, and delay time cannot be quantized. Second, existing monitoring approaches are based on two-dimensional reports, graphs, or independent GIS (geographic information system) track displays, lacking deep fusion with building information models (Building Information Modeling, BIM) and digital twinning (DIGITAL TWIN) techniques. Visual views from macroscopic logistics networks to microscopic engineering components cannot be provided, and real-time states of materials in concealed engineering or complex installation areas are particularly difficult to intuitively display. Third, existing systems focus on a certain link of the supply chain (such as transportation tracking or warehouse management) and do not build a state closed-loop management model covering the complete life cycle of the resource "production-transportation-warehouse-installation". The lack of historical playback and accident traceability based on space-time trajectories results in difficulty in problem responsibility determination, and the lack of data support for process optimization. Disclosure of Invention The invention provides a power supply chain monitoring method and device based on WBS-BOM dynamic mapping, which can improve the accuracy and efficiency of power supply chain monitoring. The embodiment of the invention provides a power supply chain monitoring method based on WBS-BOM dynamic mapping, which comprises the following steps: Acquiring a WBS data tree and a BOM data table of a target power engineering project, and respectively inputting the WBS data tree and the BOM data table into a pre-trained BERT model so that the BERT model converts the WBS data tree into a plurality of WBS node vectors and converts the BOM data table into a plurality of BOM material vectors; Based on cosine similarity and a preset expert rule base, similarity scores between the WBS node vectors and the BOM material vectors are calculated respectively, and a WBS-BOM dynamic mapping table is established according to numerical comparison results between the similarity scores and preset thresholds; Extracting a BOM material set corresponding to a target WBS node according to the WBS-BOM dynamic mapping table, obtaining a first time point when each target BOM material in the BOM material set reaches a target material unloading point, respectively calculating the dynamic lead time of each target BOM material based on the material attribute of each target BOM material and a preset risk buffering period, and correspondingly adding each first time point and each dynamic lead time to obtain a plurality of second time points; Comparing the maximum value in each second time point with the planned starting time