CN-121998264-A - Method for detecting time sequence abnormality of engineering material supply chain

CN121998264ACN 121998264 ACN121998264 ACN 121998264ACN-121998264-A

Abstract

The invention discloses a method for detecting time sequence abnormality of an engineering material supply chain, which relates to the technical field of power grid material management and comprises the following steps of obtaining historical operation data and constraint parameters of a supply chain, determining a node supply capacity boundary, dividing the node into a continuous node set and an intermittent node set according to communication stability, constructing a double-layer directed network structure, constructing a diffusion redistribution model in the network structure based on transport uncertainty and upstream and downstream inventory releasable degree difference, iteratively calculating effective supply capacity of the node under the constraint of the supply capacity boundary, wherein the effective supply capacity comprises the intermittent node supply capacity, correcting node opening probability according to the effective supply capacity, performing seepage simulation based on the corrected node opening probability, and judging whether the supply chain is in an abnormal state according to the coupling relation of a cross-layer maximum communication component scale and critical seepage threshold deviation.

Inventors

Zhang canglong
FENG RONGGUO
JIANG HAO
YIN HANG
Tang Cuizhi
JIANG WEI
YANG TAO
GONG GAIHUA
CHEN CHAO
CAI CHEN
WANG CHAO
ZHANG RUOFAN
HAN JUN

Assignees

安徽继远软件有限公司
国网安徽省电力有限公司芜湖供电公司

Dates

Publication Date: 20260508
Application Date: 20260410

Claims (10)

1. The method for detecting the time sequence abnormality of the engineering material supply chain is characterized by comprising the following steps of: Acquiring historical operation data and constraint parameters of a supply chain, and determining a node supply capacity boundary; dividing nodes into a continuous node set and an intermittent node set according to the communication stability, and constructing a double-layer directed network structure; in the network structure, a diffusion redistribution model is constructed based on the transport uncertainty and the upstream and downstream inventory releasability difference, the effective supply capacity of intermittent nodes is calculated in an iterative mode under the boundary constraint of the supply capacity, and the node opening probability is corrected according to the effective supply capacity; And performing seepage simulation based on the corrected node opening probability, and judging whether the supply chain is in an abnormal state or not according to the coupling relation between the cross-layer maximum connected component scale and the critical seepage threshold value deviation.
2. The engineering material supply chain timing anomaly detection method of claim 1, wherein determining a node supply capacity boundary comprises: establishing a constraint model taking the side flow as a variable under the constraint of transportation capacity, throughput capacity, production capacity and safety stock, and solving the maximum feasible outflow of the node to obtain a basic supply upper bound of the node; and constructing an attenuation coefficient for the intermittent node based on the historical shipping fluctuation coefficient and the communication stability, and correcting the basic supply upper bound to obtain a preliminary node supply capacity boundary.
3. The method for detecting a timing anomaly in an engineering material supply chain according to claim 2, further comprising, after obtaining the preliminary supply capacity boundary: Constructing an upstream redundancy correction coefficient according to the ratio of the effective upstream quantity of the nodes to the theoretical upstream quantity, and correcting the primary supply capacity boundary; re-solving the upper supply bound under the stress scene of stock reference pressure drop, critical edge removal or node throughput capacity reduction to obtain the upper supply bound under the stress scene; and determining a smaller value in the redundancy corrected supply capacity and the upper supply boundary in the stress scene as a final supply capacity boundary of the node.
4. The method for detecting abnormal time sequence of engineering material supply chain according to claim 1, wherein the dividing the nodes into a continuous node set and an intermittent node set according to the connection stability, and constructing a dual-layer directed network structure comprises: Dividing the historical time period into a plurality of time slices, and determining a communication instruction according to the material flowing condition in each time slice; Calculating a node communication stability index based on the upstream edge communication indication, comparing the node communication stability index with a preset stability threshold, and dividing the node into a continuous node or an intermittent node according to a comparison result; And determining the directed edges connecting the continuous nodes and the intermittent nodes as cross-layer edges to form a double-layer directed network structure comprising a continuous node layer, an intermittent node layer and cross-layer edges.
5. The method for detecting a timing anomaly in an engineering material supply chain according to claim 4, wherein the nodes are divided into a continuous node set and an intermittent node set according to the connection stability, further comprising: Continuous segment decomposition is carried out on a communication indication sequence of the edge in a historical time slice, and the length of the continuous communication segment is counted; when the length of the continuous segment is not smaller than a preset minimum continuous length threshold value, determining that the segment stability is achieved; Counting the number of edges with paragraph stability in all upstream edges of the node, and calculating the proportion of the edges to the total number of the upstream edges; And when the proportion is not smaller than a preset proportion threshold value and the node communication stability index is not smaller than a preset stability threshold value, dividing the node into a continuous node set.
6. The method for detecting a time-series abnormality of an engineering material supply chain according to claim 1, wherein the calculation process of the intermittent node effective supply capacity includes: Calculating node inventory releasability based on the node inventory and the safety inventory, and determining upstream and downstream inventory releasability differences and an immediate upper supply limit based on the inventory releasability of the nodes at the two ends of the edge; calculating side damping weight based on the transportation delay fluctuation index; Constructing an edge diffusion coefficient according to the inventory releasability difference, the channel scheduling response coefficient, the channel capacity item and the damping weight, wherein the channel scheduling response coefficient is estimated by the ratio of the historical actual allocation amount to the channel capacity, and the channel capacity item is determined according to the maximum transport capacity or the node throughput capacity of the edge; And iteratively calculating diffusion contribution of the upstream node to the downstream node based on the diffusion coefficient, and projecting under the boundary constraint of throughput capacity and supply capacity to obtain the effective supply capacity of the node.
7. The method for detecting engineering material supply chain time sequence abnormality according to claim 1, wherein the diffusion reassignment model further comprises a tail risk triggering mechanism: Counting a high-ranking threshold value of historical transportation delay for a cross-layer edge, marking the cross-layer edge as a tail risk edge when the tail occurrence proportion is not smaller than the threshold value, and applying conservative constraint to the tail risk edge in the process of calculating a diffusion coefficient or distributing diffusion contribution; The conservation constraint specifically comprises attenuation correction of damping weight, attenuation of channel capacity item or prohibition of the cross-layer edge from participating in diffusion contribution in the iteration.
8. The method for detecting a time-series abnormality of an engineering material supply chain according to claim 6, wherein the iterative updating of the diffusion reassignment model comprises: direction screening is performed before the diffusion contribution is calculated, and only edges from the continuous layer to the intermittent layer or edges with stability differences meeting a threshold are allowed to participate in contribution distribution; The total amount of the edge contribution of the upstream node is normalized to ensure that the total amount does not exceed the current effective supply capacity of the node.
9. The method for detecting a time-series abnormality of an engineering material supply chain according to claim 1, wherein the node opening probability correction includes: Performing power law correction based on the ratio of the effective supply capacity of the node to the supply capacity boundary; performing exponential decay correction based on the node edge-entering delay fluctuation weighted average; And identifying bridging key nodes based on the node cross-layer edge duty ratio and applying conservative attenuation correction.
10. The method of claim 9, wherein the step of determining whether the supply chain is in an abnormal state comprises: Based on the corrected node opening probability, constructing a node reservation network through repeated random simulation, and calculating the maximum connected component scale simultaneously containing continuous layer nodes and intermittent layer nodes; Estimating a critical seepage threshold value by combining with unified probability scanning; and calculating a structural abnormality index according to the degree of decline of the maximum connected component scale relative to a reference state and the offset of the critical seepage threshold value, and judging the abnormality grade of the supply chain by combining the structural abnormality index.

Description

Method for detecting time sequence abnormality of engineering material supply chain Technical Field The invention relates to the technical field of power grid material management, in particular to a method for detecting time sequence abnormality of an engineering material supply chain. Background The existing engineering material supply chain monitoring method generally performs statistical analysis on running indexes such as inventory level, supply period, transportation delay time or delay times and the like, and triggers early warning when relevant indexes exceed a preset threshold. However, in an actual engineering material supply network, there often exists a complex material flow relationship between different nodes, and a plurality of nodes form an interdependent supply structure through a plurality of transportation paths. When the operational status of certain nodes in the supply chain changes, the effect thereof tends to be progressively transmitted in the network along the supply path. Therefore, the abnormality judgment is performed only by relying on a single node or a local index, and the change condition of the overall structure of the supply network is difficult to comprehensively reflect, so that the abnormality recognition result may be inaccurate. Furthermore, in engineering material supply networks, there is often a significant difference in the operational stability of the different nodes. Some key supply nodes or core storage nodes can keep stable supply for a long time, while part of the nodes are influenced by production capacity, transportation conditions or staged requirements, participate in material circulation only in a specific time period, and show intermittent supply characteristics. If node communication stability is not distinguished in the supply network modeling process, a new communication path may be formed by the temporary transportation channel or the intermittent node in a short time, so that the real influence caused by the failure of a key node or the blockage of the channel in the supply chain structure is covered, and the recognition capability of the abnormal state of the supply chain is reduced. On the other hand, the prior art generally lacks a method for performing system analysis on the overall structural state of the supply chain by comprehensively considering factors such as node stability difference, inventory releasability, transport uncertainty and the like. These factors are transferred step by step along the supply network along the material flow path and create a coupling effect such that the actual supply capacity of the node exhibits a dynamically changing characteristic. As transport delays increase or inventory releasability decreases, the connectivity of the supply network as a whole may change, even causing the supply chain structure to approach a destabilized state. Disclosure of Invention In order to overcome the above-mentioned drawbacks of the prior art, embodiments of the present invention provide a method for detecting a timing anomaly of a supply chain of an engineering material, so as to solve the problem that it is difficult to uniformly reflect a coupling relationship among a node supply capability change, a node connection stability difference and a supply network connection structure evolution in a timing operation process in the existing method for detecting a timing anomaly of a supply chain of an engineering material, so that it is difficult to accurately identify a supply chain anomaly. In order to achieve the above purpose, the present invention provides the following technical solutions: The method comprises the steps of obtaining historical operation data and constraint parameters of a supply chain, determining a node supply capacity boundary, dividing the nodes into a continuous node set and an intermittent node set according to communication stability, constructing a double-layer directed network structure, constructing a diffusion redistribution model in the network structure based on transport uncertainty and upstream and downstream inventory releasable degree difference, iteratively calculating the effective supply capacity of the intermittent node under the constraint of the supply capacity boundary, correcting node opening probability according to the effective supply capacity, performing seepage simulation based on the corrected node opening probability, and judging whether the supply chain is in an abnormal state according to the coupling relation of the cross-layer maximum communication component scale and critical seepage threshold deviation. The engineering material supply chain time sequence abnormality detection method comprises the following steps: In a preferred embodiment, the determining the node supply capacity boundary comprises establishing a constraint model taking the side flow as a variable under the constraint of transportation capacity, throughput capacity, production capacity and safety stock, solvi