Cross-border logistics scheduling abnormal event adaptive handling method based on large model

Abstract

The application relates to a large model-based cross-border logistics scheduling abnormal event adaptive disposal method. The method comprises the following steps: acquiring state information returned by any logistics node, if it is detected that there is abnormal target state information, extracting abnormal event features from the target state information; inputting the abnormal event features into a knowledge graph constructed based on a cross-border logistics scene, performing chain reasoning on the knowledge graph combined with historical abnormal disposal cases to obtain a disposal operation combination for the abnormal event features; and according to the disposal operation combination, performing abnormal event disposal on the logistics node to which the target state information belongs and the affected logistics nodes until the state information returned by the corresponding logistics node is restored to normal. The method can solve the problems of poor adaptability to complex abnormal scenes and low efficiency of abnormal event disposal.

Description

Technical Field

[0001] This application relates to the field of cross-border logistics, and in particular to an adaptive handling method for cross-border logistics scheduling anomalies based on a large model. Background Technology

[0002] In the field of cross-border logistics, the handling of abnormal events during cross-border logistics transportation is crucial for achieving abnormal scheduling and control. Common methods for handling abnormal events include manual scheduling or using fixed rules. However, these methods suffer from drawbacks such as poor adaptability to complex abnormal scenarios and low efficiency in handling abnormal events. Summary of the Invention

[0003] Therefore, it is necessary to provide an adaptive handling method, system, computer equipment, and computer-readable storage medium for cross-border logistics scheduling anomalies based on a large model to address the aforementioned technical problems, thereby solving the issues of poor adaptability to complex anomaly scenarios and low efficiency in handling anomalies.

[0004] Firstly, this application provides an adaptive handling method for cross-border logistics scheduling anomalies based on a large model, applicable to systems deploying large models, including: Obtain the status information returned by any logistics node; if abnormal target status information is detected, extract the abnormal event features from the target status information. The abnormal event characteristics are input into a knowledge graph constructed based on cross-border logistics scenarios. The knowledge graph is combined with historical abnormal handling cases to perform chain reasoning to obtain a combination of handling operations for the abnormal event characteristics. According to the aforementioned combination of handling operations, abnormal events are handled for the logistics node to which the target status information belongs and the affected logistics nodes until the status information transmitted back by the corresponding logistics node returns to normal.

[0005] Secondly, this application also provides an adaptive handling system for cross-border logistics scheduling anomalies based on a large model, including: The acquisition module is used to acquire the status information returned by any logistics node. If abnormal target status information is detected, abnormal event features are extracted from the target status information. The reasoning module is used to input the abnormal event features into a knowledge graph constructed based on cross-border logistics scenarios. The knowledge graph, combined with historical abnormal event handling cases, performs chain reasoning to obtain a combination of handling operations for the abnormal event features. The handling module is used to handle abnormal events of the logistics node to which the target status information belongs and the affected logistics nodes according to the handling operation combination, until the status information returned by the corresponding logistics node is restored to normal.

[0006] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the above steps.

[0007] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the above steps.

[0008] The aforementioned adaptive handling method, system, computer equipment, and computer-readable storage medium for cross-border logistics scheduling anomalies based on a large model first detects the target state information of anomalies based on the status information returned by logistics nodes and extracts anomaly event features from the target state information, thereby enabling the formation of corresponding anomaly description information for the current anomaly event. Secondly, it performs chain-like reasoning on the anomaly event features based on a knowledge graph and historical anomaly handling cases, thereby obtaining a combination of handling operations suitable for the current anomaly event, improving the operability of anomaly event handling. Thirdly, it handles the anomaly event at the corresponding logistics node according to the handling operation combination and judges the handling effect based on the re-returned status information, thus ensuring the complete implementation of the anomaly event handling process. Based on this, the entire technical solution achieves adaptive handling of anomalies in the cross-border logistics flow process without relying on manual scheduling or fixed rules, thereby solving the problems of poor adaptability to complex anomaly scenarios and low efficiency in anomaly event handling. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart illustrating an adaptive handling method for cross-border logistics scheduling anomalies based on a large model in one embodiment. Figure 2 for Figure 1 A flowchart illustrating the content involved in step S100; Figure 3 for Figure 1 A flowchart illustrating the content involved in step S200; Figure 4for Figure 1 A flowchart illustrating the content involved in step S300; Figure 5 This is a block diagram of an adaptive handling system for cross-border logistics scheduling anomalies based on a large model, as shown in one embodiment. Figure 6 This is an internal structure diagram of a computer device that implements an adaptive handling method for cross-border logistics scheduling anomalies based on a large model, as shown in one embodiment. Figure 7 This is an internal structure diagram of a computer device that implements an adaptive handling method for cross-border logistics scheduling anomalies based on a large model, as shown in another embodiment. Figure 8 This is an internal structure diagram of a computer-readable storage medium that implements an adaptive handling method for cross-border logistics scheduling anomalies based on a large model, as shown in one embodiment. Detailed Implementation

[0011] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0012] In one embodiment, such as Figure 1 As shown, an adaptive handling method for cross-border logistics scheduling anomalies based on a large model is provided. This embodiment illustrates the application of this method to a system deploying a large model. In this embodiment, the method includes the following steps S100 to S300.

[0013] Step S100: Obtain the status information returned by any logistics node. If abnormal target status information is detected, extract the abnormal event features from the target status information.

[0014] For example, a logistics node represents a specific logistics processing location in the cross-border logistics flow process, such as a warehouse node, port node, distribution node, or transportation handover node, which is used to undertake logistics processing tasks such as cargo storage, loading and unloading, transshipment, customs clearance, or handover. Each logistics node will continuously transmit status information corresponding to the current logistics processing task. The status information may include the current arrival status, outbound status, loading status, customs clearance status, waiting status, or transshipment status of the goods, which is used to reflect the logistics execution status corresponding to the current logistics node.

[0015] Next, the status information returned by each logistics node is identified. This involves associating the current status information with the cross-border logistics flow process of the current logistics node to determine if any abnormal target status information exists. For example, if a logistics node remains in a waiting state, goods fail to be shipped out for an extended period, the status sequence of logistics nodes changes abruptly, or there is no normal connection between the current and previous logistics node statuses, then the current status information is deemed abnormal, and the corresponding status information is identified as the target status information. Furthermore, the specific abnormal events reflected in the cross-border logistics flow process by the target status information are extracted as abnormal event features. For example, the abnormal event's occurrence time, abnormal logistics node location, abnormal type (such as node delay, status change, or logistics chain break), abnormal impact range, associated cargo information (such as cargo identification, cargo batch, or cargo order), logistics flow stage (such as warehousing, sorting, transportation, port clearance, transshipment, handover, and receipt), and remaining processing time are used as the corresponding abnormal event features.

[0016] Step S200: Input the abnormal event features into the knowledge graph constructed based on the cross-border logistics scenario. The knowledge graph, combined with historical abnormal handling cases, performs chain reasoning to obtain the combination of handling operations for the abnormal event features.

[0017] For example, a knowledge graph constructed based on a cross-border logistics scenario is used to represent the correspondence between various logistics nodes and their states during the cross-border logistics flow process. For instance, in the cross-border logistics flow process where goods are transferred from a warehouse node to a port node for customs clearance and then to a transportation transfer node for transshipment, the knowledge graph records the logistics flow relationships between the warehouse node, port node, and transportation transfer node, as well as the state change relationships between the outbound state, customs clearance state, and transshipment state, thereby forming a logistics flow association structure corresponding to the actual logistics processing task.

[0018] Based on this, the characteristics of abnormal events are input into the aforementioned knowledge graph. On the one hand, historical abnormal event handling cases related to the current abnormal event are retrieved. On the other hand, the knowledge graph combines the retrieved target cases most relevant to the current abnormal event to perform chain reasoning. Among them, historical abnormal event handling cases represent abnormal events that have occurred in the cross-border logistics flow and their corresponding abnormal handling processes, such as the transshipment process after node delays, the supplementary recording process after node chain breaks, or the correction process after state jumps. These are used to represent the handling methods of different abnormal events in the actual cross-border logistics flow process.

[0019] Specifically, in the chain reasoning process, the characteristics of abnormal events are associated with the most relevant target cases at the link level. Based on the association results, the knowledge graph is used to perform hierarchical association deduction along the logistics flow relationship between logistics nodes and the state change relationship between logistics node states. This determines the location of the abnormal event and the scope of its impact in the knowledge graph. The abnormal handling process in the target case is then mapped to the corresponding logistics node and logistics node state, thereby obtaining a handling operation combination. This handling operation combination represents the operation method for handling abnormal events on the logistics node to which the target state information belongs and the affected logistics nodes.

[0020] Optionally, the handling operation combination includes at least one of the following: in-node processing operation, capacity adjustment operation, route switching operation, and anomaly notification operation. In-node processing operation refers to correcting anomalies in the logistics processing flow within a logistics node, such as re-verifying the status of the logistics node, supplementing incomplete logistics processing tasks, or reconfirming logistics records corresponding to anomalies. Capacity adjustment operation refers to adjusting the transportation resources (e.g., transport vehicles, flights, ships, or transshipment space) corresponding to the current cross-border logistics flow to change the transportation execution method of goods in subsequent cross-border logistics flows. Route switching operation refers to adjusting the logistics flow path corresponding to the current goods so that the goods continue to execute the subsequent cross-border logistics flow process according to the new logistics flow relationship. Anomaly notification operation refers to sending the anomaly description information corresponding to the current anomaly event to the processing objects related to the current cross-border logistics flow process (e.g., the processing personnel corresponding to the logistics node, the consignee, logistics dispatcher, or logistics tracker corresponding to the goods), so that the corresponding processing objects can perform subsequent confirmation, waiting, or termination processing based on the anomaly description information.

[0021] Furthermore, the node-based processing operations are applicable to various logistics nodes such as warehouse nodes, port nodes, distribution nodes, and transportation handover nodes; the exception notification operations are also applicable to various logistics nodes; while the capacity adjustment operations and route switching operations are mainly applicable to logistics nodes involving cargo flow scheduling, including distribution nodes, port nodes, and transportation handover nodes.

[0022] Step S300: Based on the handling operation combination, perform abnormal event handling on the logistics node to which the target status information belongs and the affected logistics nodes until the status information returned by the corresponding logistics node is restored to normal.

[0023] For example, the logistics node to which the target status information belongs represents the logistics node where the abnormal event is currently occurring, while the affected logistics nodes represent other logistics nodes that have a logistics relationship with the current abnormal event during the cross-border logistics flow, such as logistics nodes that have a logistics flow relationship or a status change relationship with the current logistics node. Since abnormal events can affect multiple logistics nodes during cross-border logistics flow, it is necessary to process both the target logistics node and the affected logistics nodes simultaneously when handling abnormal events. Furthermore, since the association between different logistics nodes in the current abnormal event is different, the handling operations in the handling operation combination are not executed uniformly on all logistics nodes, but rather the corresponding handling content is executed separately for different logistics nodes according to their association in the current abnormal event.

[0024] For example, when a port node experiences a clearance anomaly, intra-node processing operations can be performed on that port node to re-verify the abnormal clearance status or reconfirm the abnormal logistics records. Simultaneously, since subsequent distribution nodes and transport handover nodes with logistics connections to that port node still need to complete the subsequent cross-border logistics process, capacity adjustment or route switching operations can be performed on the corresponding distribution nodes or transport handover nodes to readjust the corresponding transport vehicles, flights, ships, transshipment warehouses, or logistics routes in the subsequent cargo flow process. Furthermore, for logistics nodes with logistics flow and status change relationships with the current anomaly, anomaly notification operations can be performed so that the corresponding processing objects can perform subsequent confirmation, waiting, or termination processing based on the anomaly description information.

[0025] Furthermore, in the abnormal event handling process, the status information re-transmitted by the corresponding logistics nodes is continuously acquired, and the handling status of the current abnormal event is determined based on the re-transmitted status information. That is, it is equivalent to making a correlation judgment based on the correspondence between the current status information and the cross-border logistics flow process of the current logistics node, so as to identify whether the re-transmitted status information is still abnormal. If the re-transmitted status information is still abnormal, the subsequent abnormal event handling is carried out according to the corresponding handling content in the handling operation combination. When the status information re-transmitted by the corresponding logistics node returns to normal, the handling process corresponding to the current abnormal event ends.

[0026] For example, the above technical solution is applied to a system deploying a large model, where the large model analyzes the status information returned by logistics nodes and, in conjunction with a knowledge graph and historical anomaly handling cases, performs chain-like reasoning on the characteristics of abnormal events to generate corresponding handling operation combinations. During system operation, the large model reasones and analyzes the anomaly handling process corresponding to the target logistics node and the affected logistics node based on the correlation between the characteristics of abnormal events, the knowledge graph, and historical anomaly handling cases; for example, the large model may include the DeepSeek large model, the Tongyi Qianwen large model, or the Wenxin Yiyan large model, etc.

[0027] In the aforementioned adaptive handling method for cross-border logistics scheduling anomalies based on a large model, in step S100, the target state information of the anomaly is detected based on the status information returned by the logistics nodes, and the anomaly event features are extracted from the target state information, thereby enabling the current anomaly event to form corresponding anomaly description information; in step S200, the anomaly event features are subjected to chain reasoning based on the knowledge graph and historical anomaly handling cases, thereby obtaining a combination of handling operations suitable for the current anomaly event, improving the operability of anomaly event handling; in step S300, the anomaly event is handled on the corresponding logistics nodes according to the handling operation combination, and the handling effect is judged based on the re-returned status information, thereby ensuring the complete implementation of the anomaly event handling process; based on this, the entire technical solution realizes adaptive handling of anomalies in the cross-border logistics flow process without relying on manual scheduling or fixed rules, thereby solving the problems of poor adaptability to complex anomaly scenarios and low efficiency in anomaly event handling.

[0028] In one exemplary embodiment, such as Figure 2 As shown, step S100, which involves "extracting abnormal event features from the target state information", includes steps S101 to S102.

[0029] Step S101: Parse the target state information into abnormal event elements corresponding to each event dimension.

[0030] For example, the target status information is divided according to different event dimensions, including time, location, type, cargo, and circulation. The time dimension represents the temporal correlation of anomalies in the cross-border logistics flow, with corresponding anomaly elements including anomaly occurrence time, anomaly duration, and remaining processing time. The location dimension represents the location correlation of anomalies in the cross-border logistics flow, with corresponding anomaly elements including anomaly logistics node location, anomaly transportation resources, and anomaly logistics flow path. The type dimension represents the anomaly changes corresponding to the current anomaly, with corresponding anomaly elements including node delays, status jumps, and logistics chain breaks. The cargo dimension represents the cargo correlation with the current anomaly, with corresponding anomaly elements including cargo identification, cargo batch, and cargo order. The circulation dimension represents the circulation stage relationship of the current anomaly in the cross-border logistics flow, with corresponding anomaly elements including the direction and sequence of anomaly changes in the circulation stages.

[0031] Step S102: Based on the correlation description relationship between each event dimension under abnormal situations, the abnormal event elements corresponding to each event dimension are vectorized and organized to obtain abnormal event features represented in vector form and used for abnormal situation description, which are then used as input to the knowledge graph.

[0032] For example, the relational descriptions between event dimensions in anomaly scenarios are used to represent the association between different event dimensions in describing the same anomaly event. This is equivalent to: prioritizing the description of the type and location dimensions in anomaly scenarios to determine the anomaly content and location of the current event; jointly describing the flow and time dimensions to determine the sequence, direction, and duration of anomalies in the cross-border logistics flow process; and providing supplementary descriptions of the cargo dimension to determine the scope of related cargo corresponding to the current anomaly event. In other words, the relational descriptions are used to define the description priority, combination method, and organization order of different event dimensions in the anomaly scenario description process, so that different event dimensions can form a unified anomaly scenario description structure around the same anomaly event.

[0033] Based on the above-mentioned correlation description relationship, the abnormal event elements corresponding to each event dimension are vectorized and organized, which is equivalent to sequentially encoding and arranging the abnormal event elements corresponding to each event dimension. For example, the "node delay" of the type dimension is encoded as "TYPE_DETAIN", the "port node" of the location dimension is encoded as "PORT_NODE", the "interruption of port clearance stage → transshipment stage" of the flow dimension is encoded as "FLOW_CUSTOMS_TO_TRANSFER_BREAK", and the "abnormality lasting 48 hours" of the time dimension corresponding to the abnormal event element of the flow dimension is encoded as "TIME_48H", so that the flow dimension and the time dimension form a corresponding relationship around the same abnormal flow process; at the same time, the "cargo batch A001" of the cargo dimension is encoded as "CARGO_BATCH_A001". Based on the correlation between the event dimensions, the above-mentioned encoded content is organized and associated. The type and position codes, which are prioritized for description, are placed at the beginning of the vector; the flow and time codes, which are jointly described, are arranged in the middle of the vector according to their correspondence; and the cargo codes, which are supplementary descriptions, are placed at the end of the vector. This forms a vectorized data structure, such as "[TYPE_DETAIN, PORT_NODE, FLOW_CUSTOMS_TO_TRANSFER_BREAK, TIME_48H, CARGO_BATCH_A001]", which serves as the abnormal event feature describing the current abnormal situation and as input to the knowledge graph. In other words, the abnormal event feature is represented as an abnormal description vector formed by sequentially encoding, arranging, and uniformly organizing the abnormal event elements corresponding to each event dimension according to their correlation, serving as the abnormal situation description input for chain-like reasoning in the knowledge graph.

[0034] In this embodiment, in step S101, the target state information is parsed into abnormal event elements corresponding to each event dimension, thereby forming a classification correspondence between different abnormal description information in the same abnormal event; in step S102, according to the association description relationship between each event dimension in the abnormal situation, the abnormal event elements corresponding to each event dimension are vectorized and organized, thereby forming a unified vector data in sequence for the abnormal event elements corresponding to different event dimensions; based on this, in the entire technical solution, the abnormal event elements corresponding to each event dimension in the cross-border logistics flow process can be associated and expressed in a unified vector form, so as to serve as the feature input of the knowledge graph.

[0035] In one exemplary embodiment, such as Figure 3 As shown, step S200 involves “using the knowledge graph in conjunction with historical anomaly handling cases to perform chain reasoning to obtain a combination of handling operations for the characteristics of anomaly events”, including steps S201 to S203.

[0036] Step S201: Compare the abnormal event characteristics with each historical abnormal event handling case to obtain the target case that best matches the abnormal event characteristics.

[0037] For example, firstly, the case features corresponding to each historical anomaly handling case are obtained, and the anomaly event features are matched with the case features. Both case features and anomaly event features are anomaly description vectors represented according to event dimensions such as time, location, type, goods, and flow, describing the occurrence and handling of the anomalies in the corresponding historical anomaly handling cases. Based on this, feature comparison can be performed on the vector dimension between the anomaly event features of the current anomaly event and the case features corresponding to the historical anomaly handling cases. This is equivalent to performing a dimension-by-dimensional matching analysis on the vector content corresponding to different event dimensions based on the vector correspondence between the anomaly event features and the case features in each event dimension, and determining the feature correlation degree between the anomaly event features and the case features based on the vector deviation between each event dimension, thus obtaining the comparison results for each historical anomaly handling case. Further, based on the comparison results for each historical anomaly handling case, the historical anomaly handling cases are ranked according to their correlation degree, and the historical anomaly handling case with the highest correlation degree with the anomaly event features is determined as the target case.

[0038] For example, for a specific historical anomaly handling case, the characteristics of the current anomaly are compared with the characteristics of a historical anomaly handling case in terms of time, location, type, cargo, and flow. If the duration of the current anomaly is "48 hours," while the duration of the corresponding historical anomaly handling case is "52 hours," then the degree of feature correlation in the time dimension is determined based on the proportion of the time difference between the two, calculated as "1 - |48 - 52| ÷ 52," resulting in a correlation of approximately "0.92." If both the current anomaly and the corresponding historical anomaly handling case occur at a "port node," and... If the abnormal logistics flow path is consistent, the feature correlation degree in the location dimension is determined to be "1.00"; if the current abnormal event and the corresponding historical abnormal handling case both correspond to the "node delay" abnormal type, the feature correlation degree in the type dimension is determined to be "1.00"; if the cargo batch corresponding to the current abnormal event and the cargo batch in the corresponding historical abnormal handling case belong to the same cargo order but are different cargo batches, the feature correlation degree in the cargo dimension is determined to be "0.80"; if the current abnormal event and the corresponding historical abnormal handling case both occur in the abnormal change process of "port clearance stage → transshipment stage", the feature correlation degree in the flow dimension is determined to be "1.00".

[0039] Then, a comprehensive calculation is performed based on the degree of correlation of features corresponding to each event dimension. For example, the degree of correlation of features corresponding to time dimension, location dimension, type dimension, goods dimension and circulation dimension is averaged and calculated. That is, the overall degree of correlation of features is approximately "0.94" according to "(0.92+1.00+1.00+0.80+1.00)÷5", which is used as the comparison result between the current abnormal event and the corresponding historical abnormal handling cases.

[0040] Step S202: Based on the entity association links of the target case in the knowledge graph, perform link mapping between the abnormal event features and the target case, and perform chain reasoning based on the link mapping results to obtain the handling reasoning results for the abnormal event features.

[0041] For example, the entity association links of the target case in the knowledge graph are used to represent the entity association paths of the target case in the cross-border logistics flow process. This includes the logistics flow process of goods moving from a warehouse node to a port node, and then to a transportation handover node for subsequent transfer, as well as the state changes of the corresponding logistics nodes at each stage of the flow. Since the target case already has a high degree of feature correlation with the current abnormal event, the entity association links of the target case in the knowledge graph can be used to deduce the corresponding abnormal situations of the current abnormal event in the cross-border logistics flow process.

[0042] Specifically, based on the entity association links of the target case in the knowledge graph, link mapping is performed between the abnormal event features and the target case. Link mapping means associating features and replacing entities with the entity association links of the target case in the knowledge graph to establish a corresponding inference link for the abnormal event features based on the existing entity association links. Feature association means matching the abnormal description content in the abnormal event features with the historical abnormal description content in the corresponding entity association links of the target case to determine the feature correspondence between the current abnormal event and the target case. Entity replacement means replacing the historical logistics flow content in the corresponding entity association links of the target case with the logistics flow content corresponding to the current abnormal event according to the feature correspondence, so as to form a corresponding inference link applicable to the current abnormal event.

[0043] Specifically, firstly, based on the abnormal logistics node locations, abnormal logistics flow paths, abnormal types, and abnormal flow stages in the abnormal event characteristics, feature association is performed with the historical logistics node locations, historical logistics flow paths, historical abnormal types, and historical flow stages in the corresponding entity association links of the target case to determine the feature correspondence between the current abnormal event and the target case in the abnormal handling process; then, based on this feature correspondence, the historical logistics nodes, historical logistics flow paths, and historical status change content in the corresponding entity association links of the target case are replaced with the logistics nodes, logistics flow paths, and status change content corresponding to the current abnormal event, thereby forming a reasoning link corresponding to the current abnormal event.

[0044] For example, when the target case corresponds to the entity association link of "the Shenzhen port node is delayed during the customs clearance stage and then switches to the backup transportation route", and the current abnormal event corresponds to "the Guangzhou port node is delayed during the customs clearance stage", then feature association is first established based on the "node delay" anomaly type, the "port node" type, and the "customs clearance stage" processing procedure. Then, based on the feature association, the "Shenzhen port node" and the corresponding "backup transportation route" in the target case are replaced with the "Guangzhou port node" and the corresponding "backup transportation route" corresponding to the current abnormal event, thereby forming a corresponding inference link applicable to the current abnormal event.

[0045] Furthermore, after completing the link mapping, based on the formed inference link, the abnormal propagation and subsequent flow of the current abnormal event in the cross-border logistics flow process are continuously deduced; for example, based on the subsequent nodes corresponding to the current abnormal logistics node, the subsequent state changes corresponding to the current abnormal state, and the subsequent logistics processing corresponding to the current logistics flow path, the impact of the current abnormal event in the subsequent cross-border logistics flow process is analyzed step by step, and the corresponding disposal content and execution relationship are determined based on the analysis results, thereby obtaining the disposal inference result for the current abnormal event.

[0046] In other words, both entity association links and reasoning links are used to represent entity association paths in a knowledge graph. Both are based on logistics nodes, logistics flow paths, and state change content to form an association structure. The difference is that entity association links are used to represent the original entity association relationships in a knowledge graph, while reasoning links are corresponding entity association paths formed on the basis of link mapping and oriented towards the reasoning process of the current abnormal event.

[0047] Step S203: Convert the handling reasoning results into a combination of handling operations based on the characteristics of the abnormal event.

[0048] For example, based on the handling content and execution relationship in the handling reasoning result, the corresponding abnormal handling behaviors are classified and organized in sequence. Specifically, the handling content for correcting anomalies within logistics nodes is determined as node-internal processing operation, the handling content for adjusting transportation resources is determined as capacity adjustment operation, the handling content for adjusting logistics flow paths is determined as path switching operation, and the handling content for transmitting abnormal information is determined as abnormal notification operation. Then, based on the execution relationship between each handling content in the handling reasoning result, the corresponding handling operations are arranged sequentially to form a combination of handling operations for the characteristics of the current abnormal event. For example, when the handling reasoning result corresponds to "first verify the status of the abnormal logistics node, then switch to the backup transportation path, and send an abnormal notification to the subsequent logistics node", then the handling operation combination of "node-internal processing operation - path switching operation - abnormal notification operation" is generated.

[0049] In this embodiment, in step S201, the abnormal event features are compared with each historical abnormal event handling case, and the target case that best matches the abnormal event features is determined, so that the current abnormal event can establish a correspondence with the existing historical abnormal event handling process; in step S202, based on the entity association links of the target case in the knowledge graph, the abnormal event features and the target case are mapped and chain reasoning is performed, so that the current abnormal event can form a corresponding reasoning link based on the existing entity association links; in step S203, the corresponding abnormal event handling behavior is classified and organized in sequence according to the handling reasoning results, so as to form a combination of handling operations for the abnormal event features; based on this, in the whole technical solution, the current abnormal event is associated with historical abnormal event handling cases and knowledge graph, so as to realize an abnormal event handling process adapted to the current abnormal event.

[0050] In an exemplary embodiment, step S202 involves “performing chain reasoning based on the results of the link mapping to obtain the reasoning results for handling the characteristics of the abnormal event”, including steps S2021 to S2022.

[0051] Step S2021: Based on the link mapping results, determine the starting inference entity and each candidate disposal entity corresponding to the abnormal event characteristics in the corresponding inference link.

[0052] For example, after the link mapping is completed, the starting inference entity and each candidate disposal entity corresponding to the abnormal event feature in the inference link are determined according to the inference link formed by the link mapping result; wherein, the starting inference entity is used to represent the starting abnormal position of the current abnormal event in the inference link, and each candidate disposal entity is used to represent each inference entity corresponding to the subsequent abnormal disposal content in the inference process corresponding to the current abnormal event.

[0053] Specifically, based on the abnormal logistics node location, abnormal flow stage, and abnormal type in the characteristics of the abnormal event, the starting position of the current abnormal event is determined in the inference link, and the inference entity corresponding to the starting position is determined as the starting inference entity. For example, when the current abnormal event corresponds to "the port node experiencing node delay during the customs clearance stage," the "port node customs clearance abnormal entity" in the inference link is determined as the starting inference entity. Then, following the subsequent inference direction in the inference link, the subsequent inference entities corresponding to the current abnormal event are identified level by level, and the corresponding candidate disposal entities are determined based on the participation of each subsequent inference entity in the current inference link. For example, when there are "alternate transportation route entities," "subsequent transshipment node entities," and "abnormal notification object entities" in the subsequent inference link, they are determined as the corresponding candidate disposal entities.

[0054] Step S2022: Taking the initial inference entity as the starting point of inference, generate inference sub-results step by step according to the association order of each candidate disposal entity in the inference chain, and connect each inference sub-result to form a disposal inference result for the characteristics of the abnormal event.

[0055] For example, since the candidate disposal entities have a sequential relationship in the inference chain, the inference process needs to be executed step by step according to the relationship in the inference chain to obtain the inference sub-results corresponding to each inference stage, and then each inference sub-result is integrated into the disposal inference result; that is, the inference sub-result is used to represent the local disposal content corresponding to a certain inference stage, while the disposal inference result is used to represent the overall disposal content for the current abnormal event.

[0056] Specifically, the process begins by using the initial inference entity as the starting point and the next candidate disposal entity directly associated with the initial inference entity as the current inference scope. This analysis examines the impact of the current anomaly on subsequent cross-border logistics flows. For example, if the initial inference entity corresponds to a "port node customs clearance anomaly entity" and the next candidate disposal entity corresponds to an "alternate transportation route entity," the analysis then examines the subsequent nodes corresponding to the current anomaly logistics node based on the inference scope between them. This determines whether there is any subsequent interruption in the transportation node to which the goods should have flowed after the port node. Simultaneously, the analysis examines the subsequent state changes corresponding to the current anomaly state, determining whether the current "customs clearance anomaly state" will prevent the subsequent "transportation state" from switching normally. Furthermore, the analysis examines the subsequent logistics processing corresponding to the current logistics flow path, determining whether the current transportation path cannot continue to complete the subsequent transportation processing. After completing the above analysis, if it is determined that the current anomaly has affected the subsequent logistics flow process, a "switch to alternative transportation route" inference sub-result is generated.

[0057] For example, on the one hand, when analyzing the subsequent nodes corresponding to the current abnormal logistics node, the knowledge graph can be used to determine whether there is any subsequent interruption in the flow of goods to the transportation node that the goods should have flowed to after the port node, based on the logistics flow sequence, the waiting time of the current transportation node, and whether the corresponding goods have completed the node handover. On the other hand, when analyzing the subsequent state changes corresponding to the current abnormal state, the knowledge graph can be used to determine whether the current "customs clearance abnormal state" will cause the subsequent "transportation state" to be unable to switch normally, based on the state switching sequence, the prerequisite state requirements corresponding to the current state, and the triggering conditions corresponding to the subsequent state. Furthermore, when analyzing the subsequent logistics processing corresponding to the current logistics flow path, the knowledge graph can be used to determine whether the current transportation path cannot continue to complete the subsequent transportation processing, based on the logistics flow path, the transportation resource usage corresponding to the current transportation path, and the path execution requirements corresponding to the subsequent logistics processing tasks.

[0058] Furthermore, the candidate disposal entities corresponding to the previous level's inference sub-results are used as new starting points for inference. Based on the disposal content corresponding to the previous level's inference sub-results, the subsequent logistics flow processes corresponding to the subsequent candidate disposal entities are analyzed step by step according to the association order in the inference chain, thereby generating corresponding inference sub-results. In other words, the generation of subsequent level inference sub-results is not independent, but rather based on the disposal content corresponding to the previous level's inference sub-results, continuing to infer the impact on the subsequent logistics flow processes. For example, after generating the inference sub-result of "switching to an alternative transportation route," the subsequent logistics flow processes after the alternative transportation route switch are further analyzed to analyze the logistics processing processes corresponding to the subsequent transfer nodes, generating corresponding subsequent inference sub-results. After completing the inference processing corresponding to each candidate disposal entity, the inference sub-results are continuously linked together according to the generation order in the inference chain, thereby forming a disposal inference result targeting the characteristics of the current abnormal event.

[0059] Furthermore, the inference link essentially represents the existing inference path of the current abnormal event in the cross-border logistics flow process. The order of entities associated with each other comes from the logistics flow process and state change process formed after the link mapping. Inference sub-results such as "switching to an alternative transportation route" are disposal content generated based on this inference link. That is, the inference link is used to determine "which entities and flow processes will be affected by the current abnormal event," while the inference sub-results are used to determine "what disposal methods should be implemented for these subsequent impacts." Therefore, even if inference sub-results such as "switching to an alternative transportation route" are generated, they affect the subsequent disposal content, not the order of association between existing entities in the inference link.

[0060] In this embodiment, in step S2021, based on the link mapping result, the starting inference entity and each candidate disposal entity corresponding to the abnormal event characteristics in the inference link are determined, thereby clarifying the starting point of subsequent chain inference and the disposal objects that can participate in the inference; in step S2022, based on the association order of the starting inference entity and each candidate disposal entity in the inference link, inference sub-results are generated step by step and connected in series, so that the disposal inference results can form a continuous expression according to the order of the inference link; based on this, in the whole technical solution, the inference link after link mapping can be transformed into a disposal inference result with a clear starting point, a step-by-step generation process and a continuous connection relationship.

[0061] In one exemplary embodiment, such as Figure 4 As shown, step S300 involves “handling abnormal events for the logistics node to which the target status information belongs and the affected logistics node according to the combination of handling operations”, including steps S301 to S302.

[0062] Step S301: Based on the urgency level of the disposal operation combination, determine the disposal execution method corresponding to the disposal operation combination, where the disposal execution method indicates automatic execution, manual approval execution, or human-machine collaborative execution.

[0063] For example, the urgency level of the current abnormal event is analyzed based on the handling content, the number of logistics nodes involved, the affected logistics flow scope, and the remaining processing time in the handling operation combination. For instance, when the handling operation combination only includes anomaly notification operations or single route switching operations, and the corresponding abnormal event only affects local logistics nodes, the corresponding urgency level is determined to be low urgency. When the handling operation combination simultaneously includes route switching operations, capacity adjustment operations, and intra-node processing operations of multiple logistics nodes, and the corresponding abnormal event has already affected the subsequent logistics flow process, the corresponding urgency level is determined to be high urgency. When the handling operation combination involves subsequent logistics flow adjustments, but the corresponding abnormal event has not yet affected the large-scale logistics flow process, the corresponding urgency level is determined to be medium urgency.

[0064] Then, based on the execution requirements corresponding to different levels of urgency, the corresponding execution methods are determined. Specifically: for the combination of execution operations corresponding to low urgency, an automatic execution method can be determined; for the combination of execution operations corresponding to high urgency, a manual approval execution method is determined; and for the combination of execution operations corresponding to medium urgency, a human-machine collaborative execution method is determined.

[0065] Specifically, automatic execution means that the preset control system automatically executes the corresponding abnormal event handling according to the handling operation combination; manual approval execution means that the corresponding handling object completes the approval and confirmation of the handling content in the handling operation combination, and then the preset control system executes the corresponding abnormal event handling; human-machine collaborative execution means that the preset control system first automatically executes part of the handling content in the handling operation combination (such as standardized handling content with clear rules and no need for human decision-making), and then the corresponding handling object confirms the remaining key handling content before continuing to execute the subsequent abnormal event handling.

[0066] Step S302: Based on the handling execution method, perform abnormal event handling on the logistics node to which the target status information belongs and the affected logistics node, and use the abnormal event description, handling process description and corresponding handling result as new historical abnormal handling cases and synchronize them to the knowledge graph to build the corresponding entity association link in the knowledge graph.

[0067] For example, the abnormal event description is used to represent the abnormal situation corresponding to the current abnormal event, such as the location of the abnormal logistics node, the type of abnormality, the abnormal flow stage, and the scope of the abnormality's impact; the handling process description is used to represent the handling operation process performed for the current abnormal event, such as route switching, capacity adjustment, node status correction, or abnormal notification; and the handling result is used to represent the logistics flow status after the abnormal event handling is completed, such as the logistics flow returning to normal, subsequent logistics nodes resuming flow, or the transportation status successfully switching back.

[0068] Then, based on the correspondence between the abnormal event description, the handling process description, and the handling result, the association path between the corresponding entities is established in the knowledge graph. That is, based on the content in the abnormal event description, the corresponding logistics node and the corresponding logistics flow path are located in the knowledge graph, and the content in the handling process description and the handling result is recorded in the knowledge graph as the state change content of the corresponding logistics node in different flow stages. This forms a corresponding entity association link between the corresponding logistics node, the logistics flow path, and the state change content, which serves as the historical abnormal handling basis for subsequent abnormal event link mapping and chain reasoning.

[0069] In this embodiment, in step S301, the corresponding handling execution method is determined according to the urgency of the handling operation combination, so that different abnormal events can be handled according to their corresponding urgency. In step S302, abnormal events are handled for the logistics node to which the target status information belongs and the affected logistics node according to the handling execution method, and the correlation data between the abnormal event description, handling process and corresponding handling result is synchronized to the knowledge graph, so that the handling process corresponding to the current abnormal event can form a new historical abnormal handling case and corresponding entity association link. Based on this, in the whole technical solution, the corresponding handling execution of abnormal events is realized, and the handling process corresponding to the current abnormal event can be continuously synchronized to the knowledge graph to update the corresponding historical abnormal handling basis.

[0070] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0071] Based on the same inventive concept, this application also provides a model-based adaptive handling system for cross-border logistics scheduling anomaly events, used to implement the aforementioned model-based adaptive handling method for cross-border logistics scheduling anomaly events. The solution provided by this system is similar to the implementation scheme described in the above method. Therefore, the specific limitations of one or more model-based adaptive handling system embodiments for cross-border logistics scheduling anomaly events provided below can be found in the limitations of the model-based adaptive handling method for cross-border logistics scheduling anomaly events described above, and will not be repeated here.

[0072] In one exemplary embodiment, such as Figure 5 As shown, an adaptive handling system for cross-border logistics scheduling anomalies based on a large model is provided, including: an acquisition module 100, an inference module 200, and a handling module 300, wherein: The acquisition module 100 is used to acquire the status information returned by any logistics node. If abnormal target status information is detected, abnormal event features are extracted from the target status information. The reasoning module 200 is used to input the abnormal event characteristics into the knowledge graph constructed based on the cross-border logistics scenario. The knowledge graph is combined with historical abnormal handling cases to perform chain reasoning to obtain the combination of handling operations for the abnormal event characteristics. The handling module 300 is used to handle abnormal events of the logistics node to which the target status information belongs and the affected logistics nodes according to the handling operation combination, until the status information returned by the corresponding logistics node is restored to normal.

[0073] In an exemplary embodiment, the acquisition module 100 is further configured to: parse the target state information into abnormal event elements corresponding to each event dimension; and, based on the correlation description relationship between each event dimension under abnormal circumstances, vectorize and organize the abnormal event elements corresponding to each event dimension to obtain abnormal event features represented in vector form and used for describing abnormal circumstances, as input to the knowledge graph.

[0074] In an exemplary embodiment, the reasoning module 200 is further configured to: compare the abnormal event features with each historical abnormal handling case to obtain the target case that best matches the abnormal event features; perform link mapping between the abnormal event features and the target case based on the entity association links of the target case in the knowledge graph, and perform chain reasoning based on the link mapping results to obtain the handling reasoning result for the abnormal event features; and convert the handling reasoning result into a combination of handling operations for the abnormal event features.

[0075] In an exemplary embodiment, the inference module 200 is further configured to: determine the starting inference entity and each candidate disposal entity corresponding to the abnormal event feature in the corresponding inference link based on the link mapping result; generate inference sub-results step by step according to the association order of each candidate disposal entity in the inference link, taking the starting inference entity as the starting point of inference, and connect each inference sub-result to form a disposal inference result for the abnormal event feature.

[0076] In an exemplary embodiment, link mapping refers to: associating the entity association links of the target case in the knowledge graph with the abnormal event features and performing feature replacement, so as to establish a corresponding inference link for the abnormal event features based on the existing entity association links. Specifically, feature association involves matching the abnormal description content in the abnormal event features with the historical abnormal description content in the corresponding entity association links of the target case to determine the feature correspondence between the current abnormal event and the target case; entity replacement involves replacing the historical logistics flow content in the corresponding entity association links of the target case with the logistics flow content corresponding to the current abnormal event, based on the feature correspondence, to form a corresponding inference link suitable for the current abnormal event.

[0077] In one exemplary embodiment, the handling operation combination includes at least one of the following: in-node processing operation, capacity adjustment operation, route switching operation, and exception notification operation.

[0078] In an exemplary embodiment, the handling module 300 is further configured to: determine the handling execution method corresponding to the handling operation combination according to the urgency level of the handling operation combination, wherein the handling execution method represents automatic execution, manual approval execution, or human-machine collaborative execution; and handle abnormal events for the logistics node to which the target status information belongs and the affected logistics node according to the handling execution method, and synchronize the abnormal event description, handling process description, and corresponding handling result as new historical abnormal handling cases to the knowledge graph, so as to construct the corresponding entity association link in the knowledge graph.

[0079] The modules in the aforementioned adaptive handling system for cross-border logistics scheduling anomalies based on a large model can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, allowing the processor to invoke and execute the corresponding operations of each module.

[0080] In one exemplary embodiment, a computer device is provided, the computer device including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps in the above-described method embodiments.

[0081] This computer device can be a server, and its internal structure diagram can be as follows: Figure 6 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores relevant data for adaptive handling of cross-border logistics anomalies. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network.

[0082] The computer device can be a terminal, and its internal structure diagram can be as follows: Figure 7As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements the steps in the aforementioned adaptive handling method for cross-border logistics scheduling anomalies based on a large model. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen; the input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs or touchpads set on the casing of the computer device, or external keyboards, touchpads or mice, etc.

[0083] Those skilled in the art will understand that Figure 6 or Figure 7 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0084] In one exemplary embodiment, such as Figure 8 The diagram shows the internal structure of a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the above-described method embodiments.

[0085] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods.

[0086] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0087] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. An adaptive handling method for cross-border logistics scheduling anomalies based on a large model, characterized in that, The method, applied to systems deploying large models, includes: Obtain status information returned by any logistics node; if abnormal target status information is detected, extract abnormal event features from the target status information. The abnormal event characteristics are input into a knowledge graph constructed based on cross-border logistics scenarios. The knowledge graph is combined with historical abnormal handling cases to perform chain reasoning to obtain a combination of handling operations for the abnormal event characteristics. According to the aforementioned handling operation combination, abnormal event handling is performed on the logistics node to which the target status information belongs and the affected logistics nodes until the status information transmitted back by the corresponding logistics nodes returns to normal.

2. The method according to claim 1, characterized in that, Extracting abnormal event features from the target state information includes: The target state information is parsed into abnormal event elements corresponding to each event dimension; Based on the correlation description relationship between each event dimension under abnormal situations, the abnormal event elements corresponding to each event dimension are vectorized and organized to obtain abnormal event features represented in vector form and used for abnormal situation description, which can be used as input for the knowledge graph.

3. The method according to claim 1, characterized in that, The chain reasoning based on the knowledge graph and historical anomaly handling cases yields a combination of handling operations for the characteristics of the anomaly event, including: The abnormal event features are compared with the features of each historical abnormal event handling case to obtain the target case that best matches the abnormal event features. Based on the entity association links of the target case in the knowledge graph, the abnormal event features are mapped to the target case, and chain reasoning is performed based on the results of the link mapping to obtain the handling reasoning results for the abnormal event features. The reasoning results are then converted into a combination of handling operations tailored to the characteristics of the anomalous event.

4. The method according to claim 3, characterized in that, The step of performing chain-like reasoning based on the link mapping results to obtain the handling reasoning results for the abnormal event characteristics includes: Based on the link mapping results, determine the starting inference entity and each candidate disposal entity corresponding to the abnormal event characteristics in the corresponding inference link; Starting with the initial inference entity, inference sub-results are generated step by step according to the association order of each candidate disposal entity in the inference chain, and the inference sub-results are concatenated to form a disposal inference result for the characteristics of the abnormal event.

5. The method according to claim 3 or 4, characterized in that, Link mapping means: associating the entity association links of the target case in the knowledge graph with the abnormal event features and replacing the entities, so as to establish the corresponding reasoning links of the abnormal event features based on the existing entity association links. In this context, feature association means matching the abnormal description content in the abnormal event features with the historical abnormal description content in the entity association link corresponding to the target case, so as to determine the feature correspondence between the current abnormal event and the target case; Entity replacement means replacing the historical logistics flow content in the entity association link corresponding to the target case with the logistics flow content corresponding to the current abnormal event, based on the feature correspondence, so as to form a corresponding inference link applicable to the current abnormal event.

6. The method according to claim 1 or 3, characterized in that, The handling operation combination includes at least one of the following: in-node processing operation, capacity adjustment operation, route switching operation, and anomaly notification operation.

7. The method according to claim 1, characterized in that, The step of handling abnormal events for the logistics node to which the target status information belongs and the affected logistics nodes according to the combination of handling operations includes: Based on the urgency level of the combination of handling operations, the corresponding execution method is determined, wherein the execution method represents automatic execution, manual approval execution, or human-machine collaborative execution. According to the aforementioned handling execution method, abnormal events are handled for the logistics nodes to which the target status information belongs and the affected logistics nodes. The abnormal event description, handling process description, and corresponding handling results are used as new historical abnormal handling cases and synchronized to the knowledge graph to construct corresponding entity association links in the knowledge graph.

8. An adaptive handling system for cross-border logistics scheduling anomalies based on a large model, characterized in that, The system includes: The acquisition module is used to acquire status information returned by any logistics node. If abnormal target status information is detected, abnormal event features are extracted from the target status information. The reasoning module is used to input the abnormal event features into a knowledge graph constructed based on cross-border logistics scenarios. The knowledge graph, combined with historical abnormal event handling cases, performs chain reasoning to obtain a combination of handling operations for the abnormal event features. The handling module is used to handle abnormal events of the logistics node to which the target status information belongs and the affected logistics nodes according to the handling operation combination, until the status information returned by the corresponding logistics node is restored to normal.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images