A method for constructing a smart logistics platform based on a multi-layer network cascading failure model

Abstract

The present invention discloses a method for constructing a smart logistics platform based on a multi-layer network cascading failure model, comprising: constructing a "function-project-third party" multi-layer network for each sub-module construction project combination in a plurality of sub-module construction project combinations; performing failure analysis on the "function-project-third party" multi-layer network corresponding to each sub-module construction project combination; during the failure analysis, first determining the initial load and capacity of each node in the project layer network and each node in the third-party layer network, and setting a load distribution strategy when a node fails; based on the initial load, capacity and load distribution strategy, performing failure analysis on the nodes in the project layer network or the nodes in the third-party network layer, calculating the capability value of the smart logistics platform function realized by the project layer network and the capability value of the smart logistics platform function realized by the third-party network, and finally selecting the optimal sub-module construction project combination through the resilience measurement index of the sub-module construction project combination.

Description

Technical Field

[0001] The present invention belongs to the technical field of system platform construction, and specifically relates to a method for constructing a smart logistics platform based on a multi-layer network cascading failure model. Background Art

[0002] Smart logistics platforms are gradually becoming the main driving force for the development of the logistics industry. They realize automation, visualization, controllability, intelligence and networking in the logistics process by applying technologies such as the Internet of Things, big data and artificial intelligence. The efficient operation of the smart logistics platform depends on the careful combination of multiple subsystem modules within it, which is one of the core tasks of project managers. Taking Huawei's Songshan Lake Logistics Center as an example, the project team carefully selected submodules such as radio frequency, electronic tag picking, and rotary conveyor belts in order to achieve efficient operation of multi-functional integration such as material reception, storage, selection, set-up, and distribution. They worked with third-party technology suppliers to ensure that the system and software can be seamlessly integrated with the platform, and worked with other third-party hardware providers to obtain the necessary sensors and scanning equipment. However, if the technology supplier fails to provide software updates in a timely manner, or the equipment delivered by the hardware provider does not meet the specifications, it may cause submodules such as electronic tag picking to fail to operate normally, thereby affecting the performance of the entire platform and even causing the risk of system collapse. This phenomenon is called inter-project cascading failure. Therefore, project cascading failure is a factor that cannot be ignored when building a smart logistics platform.

[0003] Although existing research has made some progress in module design, process optimization and implementation strategies, it often ignores the multi-layer network coupling relationship formed between functions, projects and third parties during the platform construction process, especially the risk of project cascading failure. At present, the research on project cascading failure is mostly focused on the field of single project management and enterprise management, while its application in the construction of smart logistics platforms is relatively lacking. Among them, project-level network construction is the basis of project cascading failure analysis, but the existing related technologies mainly rely on expert evaluation and cannot solve the problem of large-scale project network construction and analysis in the process of smart logistics platform construction. At the same time, it ignores the third party as one of the important stakeholders, which is not conducive to the construction of smart logistics platforms. Summary of the invention

[0004] The purpose of the present invention is to provide a method for constructing a smart logistics platform based on a multi-layer network cascading failure model, so as to meet the core functional requirements under given resource and technical conditions, while reducing the systemic risks caused by cascading failures of factor module construction projects.

[0005] In order to achieve the above tasks, the present invention adopts the following technical solutions:

[0006] A method for constructing a smart logistics platform based on a multi-layer network cascading failure model, comprising:

[0007] For each submodule construction project combination in a plurality of submodule construction project combinations, a "function-project-third party" multi-layer network is constructed; the "function-project-third party" multi-layer network includes a function layer network, a project layer network and a third party layer network, the function layer network includes all functions that need to be implemented by the smart logistics platform, the project layer network uses the project ontology model to describe each submodule construction project in the submodule construction project combination; the third party layer network includes a third party; the "function-project-third party" multi-layer network is formed by using the support relationship between the third party and the submodule construction project, and the contribution relationship between the submodule construction project and the function;

[0008] A failure analysis is performed on the "function-project-third party" multi-layer network corresponding to each sub-module construction project combination; during the failure analysis, the initial load and capacity of each node in the project layer network and each node in the third-party layer network are first determined, and the load distribution strategy when the node fails is set; based on the initial load, capacity and the load distribution strategy, a failure analysis is performed on the nodes in the project layer network or the nodes in the third-party network layer, and the capacity value of the project layer network smart logistics platform function realization and the capacity value of the third-party network smart logistics platform function realization are iteratively calculated, and finally the optimal sub-module construction project combination is selected through the resilience measurement index of the sub-module construction project combination.

[0009] In one embodiment of the present invention, the nodes in the functional layer network are each function. If there is a collaborative relationship between the functions, there is an edge between the functions; the collaborative relationship is the function ES k Impact function ES k′ , or function ES k′ Impact function ES k ; The edges between functions have directions, when the function ES k Impact function ES k′ When the direction of the edge is ES k to ES k′ ; Set functional weights for nodes in the functional layer network.

[0010] In one embodiment of the present invention, the project layer network uses a project ontology model to describe each submodule construction project in the submodule construction project combination, including:

[0011] The IPO-based project ontology model describes each submodule construction project in the submodule construction project portfolio; the ontology model includes project input, project processing, project output, project start and end time, and the functions supported by the project; the dependency relationship between submodule construction projects is clarified, and a dependency matrix between submodule construction projects is constructed. The dependency matrix is ​​used to clarify the relationship between each submodule construction project in the project layer network and the degree of closeness of the relationship. Each node in the project layer network represents a submodule construction project in the smart logistics platform; wherein:

[0012] The number of dependency types between submodule construction projects is used to characterize the dependency strength between submodule construction projects. The dependency types include technology dependency, resource dependency, strategic dependency, and benefit dependency. If the submodule construction project P j′ P j If there is at least one of the four dependencies, then P j′ Depends on P j ; When the dependency intensity between submodule construction projects reaches 1 or above, there is an edge between submodule construction projects, and the weight of the edge represents the dependency intensity between submodule construction projects; the edge has a direction, when P j′ Depends on P j When the direction of the edge between the two is from node P j′ To P j .

[0013] In one embodiment of the present invention, each third party in the third party layer network acts as a node; when two third party PCs i 、PC i′ When participating in the same submodule construction project, there is an edge between the two third parties, the direction of the edge is bidirectional and the weight is 1, otherwise, there is no edge between the third party nodes.

[0014] Optionally, the supporting relationship between the third party and the submodule construction project, and the contribution relationship between the submodule construction project and the function are used to form the "function-project-third party" multi-layer network, including:

[0015] The support relationship between the third party and the submodule construction project represents the degree of support provided by the third party to the implementation of the submodule construction project, denoted as e ij , represents the i-th third-party PC i For the jth submodule construction project P j When a third party supports the submodule construction project, there is an edge between the two, and the weight of the edge is the support degree e ij ; The contribution relationship between the submodule construction project and the function represents the contribution of the submodule construction project to the realization of the function, denoted by e jk , indicating the submodule construction project Pj To realize the kth function ES k When the submodule construction project has a contribution relationship with the function, there is an edge between the two, and the weight of the edge is the contribution degree e jk ;e ij and e jk Given by the project manager;

[0016] By utilizing the support relationship and contribution degree, connections are established between the third-party layer network and the project layer network, and between the project layer network and the function layer network, thereby forming a "function-project-third party" multi-layer network.

[0017] Preferably, the initial load and capacity calculation method of the nodes in the project layer network is as follows:

[0018] L j =(1+α)(ω j ) β

[0019]

[0020]

[0021] Among them, L j Represents the jth submodule construction project P in the project layer network j The initial load, C j Indicates capacity; ω j Represents the submodule construction project P j The weight of α and β are the parameters for controlling the initial load strength of the node and α≥0 and β>0. Represents the submodule construction project P j The out-degree strength and in-degree strength of F j P j investment amount; η3 and η4 represent the importance that the project manager attaches to the dependencies between sub-module construction projects and the investment amount of sub-module construction projects, and satisfy 0≤η3+η4≤1, where η3, η4∈[0,1], λ and γ are capacity adjustment parameters, λ>0, γ>0.

[0022] Furthermore, the setting of the load distribution strategy when a node fails includes:

[0023] For the project-level network, when one or some submodule construction projects fail, the load carried by the failed submodule construction project needs to be distributed to other submodule construction projects that depend on the submodule construction project:

[0024]

[0025] Among them, △L jj′(t) represents the jth submodule construction project P at time t j Passed to the j′th submodule construction project P j′ Load, L j (t) represents the submodule construction project P j The load at time t, Indicates P j The amount of tasks carried is passed to P j′ Ratio:

[0026]

[0027] Among them, L j′ (t), C j′ Respectively represent the submodule construction project P j′ Load and capacity at time t; L h (t), C h They represent the load and capacity of the h-th submodule construction project at time t; Γ j Is dependent on the submodule that failed to build the project node P j A collection of project nodes.

[0028] Furthermore, the jth submodule construction project P is set j The support obtained is The i-th third-party PC i P j The support degree is e ij , I represents the number of third parties; set the threshold when When P j is in normal condition and is not affected by failure of third parties; When P j Sub-module construction projects that fail due to third-party failures will subsequently affect other sub-module construction projects through a cascading effect.

[0029] In one embodiment, the failure analysis of the nodes in the project layer network or the nodes in the third-party network layer, and iterative calculation of the capability value of the project layer network smart logistics platform function realization and the capability value of the smart logistics platform function realization of the third-party network include:

[0030] The failure analysis of the project layer network is as follows: the "function-project-third party" multi-layer network with determined initial load and capacity is used as the initialization network; for the initialization network, one or part of the nodes in the project layer network are randomly selected as failed nodes, and these failed nodes and the edges associated with them are deleted; then the load of the failed node is redistributed to the nodes that depend on the node and have not failed, and the new load of the non-failed nodes in each project layer network is calculated, and the state of each node in the project layer network at this time is judged. If the load of the node exceeds the capacity, the node is marked as a failed node. After deleting the failed node and the edges associated with it, the load of the failed node is redistributed to the nodes that depend on the node and have not failed; until the load of all non-failed nodes in the network does not exceed the capacity, an iteration is completed, and after the iteration is completed, the calculation By setting different numbers of failed nodes for the initialized network and going through multiple iterations, until Less than the preset capacity value;

[0031] The failure analysis of the third-party network is the same as the failure analysis of the project-level network. After each iteration, it is necessary to calculate After multiple iterations, Less than the preset capacity value;

[0032] in, and They represent the functional realization values ​​of the θth iteration of the sub-module construction project portfolio b when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2.

[0033] Furthermore, the capability value-based method selects the optimal submodule construction project combination through the resilience measurement index of the submodule construction project combination. The specific formula is:

[0034]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040] Among them, CC b is the expected capability value, g represents the g-th iteration; represents the final functional realization value of the submodule construction project combination b when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2; K is the number of functions, J represents the number of submodule construction projects in the submodule construction project combination b, ψ k represents the kth function ES k The node weights, where e kk′ represents the kth and k′th functions ES k ,ES k′ The synergistic relationship between Indicates function ES k A collection of other functions affected, OT means ES k The sum of the weights of the edges that affect other functions; η1,η2∈[0,1]; represents the ES determined by the experts k The functional weight, Θ k Indicates the function ES k The set of functions that have an impact; jk 、e jk′ Represents the jth submodule construction project P j Function ES k ,ES k′ the degree of contribution achieved; and They represent the number of nodes in the project layer network when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2, respectively. is the preset value, Ω1 represents the maximum number of failed submodule construction projects when the functional realization value is less than the preset realization value in all candidate submodule project combinations; Ω2 represents the maximum number of failed third parties when the functional realization value is less than the preset realization value in all candidate submodule project combinations; ST b It represents the resilience measurement index of submodule construction project portfolio b;

[0041] The sub-module construction project combination with the largest resilience measurement index is selected as the optimal sub-module construction project combination, and the sub-modules are constructed according to the sub-module construction projects in the sub-module construction project combination to realize the functions of the smart logistics platform.

[0042] Compared with the prior art, the present invention has the following technical features:

[0043] 1. The present invention proposes a project dependency generation algorithm based on the project input-processing-output (IPO) model, which can effectively reduce the complexity and subjectivity of large-scale project dependency determination and realize the convenient generation of large-scale project layer networks. By recording and analyzing the interaction data of various projects in the logistics platform in the multi-layer network of functions, projects and third parties, the present invention breaks away from the traditional way of relying on expert evaluation and improves the accuracy and efficiency of dependency determination. This algorithm provides more reliable data support for project portfolio selection and management in the smart logistics platform and optimizes project collaboration in logistics operations.

[0044] 2. The present invention expands the traditional two-layer network and constructs a "function-project-third party" multi-layer network cascading failure model. The introduction of the third-party network layer can more accurately reflect the complex collaborative relationship in the smart logistics platform, enhance the systematicness and credibility of the analysis results of the cascading failure model, and solve the problem that the traditional model cannot meet the actual logistics needs due to ignoring the influence of third parties. The multi-layer network model helps to more comprehensively analyze and respond to the cascading failure phenomenon in the logistics platform, and improve the overall risk resistance and operational efficiency of the logistics platform.

[0045] 3. This invention brings the functional modules of the smart logistics platform into specific logistics distribution scenarios, selects the optimal solution through project combination, and provides strong technical support for the construction and operation of the logistics system. This application not only improves the efficiency and reliability of logistics distribution, but also reduces operating costs and enhances the market competitiveness of the logistics system by optimizing resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of the process of the present invention;

[0047] Figure 2 This is a schematic diagram of the multi-layer network of "function-project-third party";

[0048] Figure 3 It is a multi-layer network of “function-project-third party” built based on PP1 in the embodiment;

[0049] Figure 4 It is the performance of multi-layer network under project-level cascading failure;

[0050] Figure 5 This is the performance of a multi-layer network under cascading failure of the third layer. DETAILED DESCRIPTION

[0051] In the construction of smart logistics platform, the relationship between submodules and platform functions is complex and dynamic. A submodule may support multiple platform functions, and a platform function may also rely on the collaborative work of multiple submodules. For example, the "order management" function in the smart logistics platform may rely on the joint support of the "inventory management submodule", "distribution scheduling submodule" and "data analysis submodule"; at the same time, the "inventory management submodule" not only supports the "order management" function, but also may support the "inventory optimization" function and "supply chain management" function. Each submodule construction project is responsible for the development and construction of a specific submodule, and its success or failure directly affects the functional realization of the submodule, and then affects the stability and reliability of the overall platform function. Therefore, the relationship between submodules and platform functions determines the complexity of the selection of submodule construction project combinations, and the overall performance of the platform must be ensured through precise design and optimization.

[0052] The technical problem that the present invention aims to solve is how to effectively identify and quantify the risk of project cascading failure in the multi-layer network of "function-project-third party" during the construction of the smart logistics platform, and select the optimal sub-module construction project combination. This method does not simply eliminate certain sub-modules, but optimizes the combination of sub-modules to ensure that the platform can meet the core functional requirements under given resource and technical conditions, and reduce the systemic risks caused by cascading failures of factor module construction projects.

[0053] The present invention innovatively proposes a method for constructing a smart logistics platform based on a multi-layer network cascading failure model, wherein the smart logistics platform needs to implement multiple functions, each function needs to be supported by a sub-module, and one sub-module can support multiple functions at the same time; the sub-module construction project used to construct the sub-module requires the participation of a third party, and a third party may participate in multiple sub-module construction projects at the same time.

[0054] Since a function may be realized through multiple sub-modules, there may be multiple sub-module combinations for the realization of all functions that the smart logistics platform needs to realize, that is, corresponding to multiple sub-module construction project combinations. This plan pre-sets multiple sub-module construction project combinations, and the sub-module combination constructed by each sub-module construction project combination can meet the realization of all required functions of the smart logistics platform; the purpose of this plan is to select the optimal sub-module construction project combination from multiple alternative sub-module construction project combinations, and use this combination to realize the construction of the corresponding sub-module combination.

[0055] See attached Figure 1 The present invention provides a method for constructing a smart logistics platform based on a multi-layer network cascading failure model, comprising the following steps:

[0056] Step 1, for each sub-module construction project combination in multiple sub-module construction project combinations, construct a "function-project-third party" multi-layer network; the "function-project-third party" multi-layer network includes a function layer network, a project layer network and a third-party layer network, the function layer network includes all functions that need to be implemented by the smart logistics platform, the project layer network uses the project ontology model to describe each sub-module construction project in the sub-module construction project combination; the third-party layer network includes a third party; the support relationship between the third party and the sub-module construction project, and the contribution relationship between the sub-module construction project and the function are used to form the "function-project-third party" multi-layer network.

[0057] Step 1.1, the functional layer network uses ES = {ES k |k=1,2,...,K} represents, where ES k is the kth function that the smart logistics platform needs to realize, and K is the number of functions that need to be realized. The nodes in the functional layer network are each function. If there is a collaborative relationship between functions, there are edges (connections) between functions; the collaborative relationship is the function ES k Impact function ES k′ , or function ES k′ Impact function ES k ; The edges between functions have directions, when the function ES k Impact function ES k′ When the direction of the edge is ES k to ES k′ .

[0058] Since the number of functions that need to be implemented in the smart logistics platform is limited, the functional weight of each node in the functional layer network and the collaborative relationship of the functions can be set through expert judgment, and the collaborative relationship of the functions can be used as the weight of the edge; among which, the functional weight is used to characterize the importance of the functional node.

[0059] Step 1.2, describe each sub-module construction project in the sub-module construction project portfolio based on the IPO (Input-Process-Output) project ontology model; the ontology model includes project input, project processing, project output, project start and end time, and the functions supported by the project; clarify the dependencies between sub-module construction projects, construct a dependency matrix between sub-module construction projects, and use the dependency matrix to clarify the relationship between each sub-module construction project in the project layer network and the closeness of the relationship.

[0060] The submodule construction project is a systematic process. Within a specified time, based on input resources, it integrates relevant technologies and knowledge for processing and outputs corresponding products or services to support the realization of corporate strategy. First, the ontology model is introduced to systematically describe the input, processing, output, start and end time, and support functions of a submodule construction project, so as to clarify the dependencies between submodule construction projects and ensure the accurate construction of the project layer network.

[0061] (1) Project ontology model based on IPO.

[0062] The IPO-based project ontology model proposed by the present invention for the sub-module construction project combination consists of five parts, namely project input, project processing, project output, project start and end time, and the functions supported by the project.

[0063] Project input refers to the resources required for the submodule construction project. The jth submodule construction project P j The required resource set is denoted as in Indicates P j The first resource required, L j Indicates P j The number of resource types required; Project processing refers to the knowledge or technology used in the implementation of the submodule construction project; j The set of knowledge or technology used is denoted as in Indicates P j The nth type of knowledge or technology used, N j Indicates P j The number of types of knowledge or technology used; j The project output can be divided into two parts: one is the results, which can be recorded as in Indicates P j The mth result of is the number of achievements; the second is the reusable resources, recorded as in Indicates P j The m′th reusable resource, is the number of reusable resource types; the project start and end time refers to the submodule construction project P j Start time ST j and end time ET j ; The functions supported by the project refer to the smart logistics platform consisting of submodule construction projects P j The set of platform functions that are successfully implemented is denoted as in Represents the submodule construction project P jThe cth platform function that was successfully realized, C j P j The number of functions supported.

[0064] (2) Dependency types between sub-module construction projects.

[0065] This plan sets up four types of inter-project dependencies: technology dependency, resource dependency, strategic dependency, and revenue dependency, which are used to determine the dependencies between sub-module construction projects based on the IPO relationships between the sub-module construction projects.

[0066] If the j′th submodule construction project P j′ With P j If the formula (1) is satisfied, it means that the submodule construction project P j′ Technology depends on P j ; If formula (2) is satisfied, it means P j′ Resources depend on P j ; If formula (3) is satisfied, it means P j′ and strategy depends on P j ; If formula (4) is satisfied, it means P j′ The return depends on P j .

[0067]

[0068]

[0069] Jaccard(S j , S j′ )>0 (3)

[0070]

[0071] Among them, T j′ , R j′ , S j′ , R j′ 、ST j′ Respectively represent the submodule construction project P j′ The set of knowledge or technologies used, the set of resources required, the set of platform functions achieved through the successful construction of the sub-module project, the set of resources required, and the start time. OR means or, and and means and.

[0072] Indicates the similarity between two sets X1 and X2, and the two sets X1 and X2 do not belong to the same submodule construction project; for example, Jaccard (T j , T j′ ) represents the set T j and the set T j′The similarity of and j≠j′, the meanings of the parameters in other formulas are explained in the same way and will not be repeated here.

[0073] (3) Dependency matrix between sub-module construction projects.

[0074] This solution uses the number of dependency types between submodule construction projects to characterize the dependency strength between submodule construction projects, namely 0, 1, 2, 3 and 4. For example, according to the above formulas (1) to (4), it is determined that there are two types of technical dependency and resource dependency between two submodule construction projects, and the dependency strength between them is 2.

[0075] The dependency strength between each submodule construction project is calculated respectively, and the dependency matrix between submodule construction projects is constructed. Each element in the dependency matrix represents the dependency strength between the corresponding two submodule construction projects, as shown in Table 1: jj′ Indicates P j Depends on P j′ The strength of , Q is the number of items, j′, j∈[1, 2, ...Q].

[0076] Table 1 Example of inter-project dependency matrix

[0077] <![CDATA[P1]]> <![CDATA[P2]]> ... <![CDATA[P Q ]]> <![CDATA[P1]]> 0 <![CDATA[Y 12 ]]> ... <![CDATA[Y 1Q ]]> <![CDATA[P2]]> <![CDATA[Y 21 ]]> 0 ... <![CDATA[Y 2Q ]]> ... ... ... ... ... <![CDATA[P Q ]]> <![CDATA[Y Q1 ]]> <![CDATA[Y Q2 ]]> ... 0

[0078] (4) Project level network.

[0079] Construct a project layer network; each node in the project layer network represents a submodule construction project in the smart logistics platform. If the submodule construction project P j′ P j If there is at least one of the four dependencies, then P j′ Depends on P j .

[0080] When the dependency intensity between submodule construction projects reaches 1 or above, there is an edge between submodule construction projects. The weight of the edge indicates the closeness of the relationship between submodule construction projects, that is, the dependency intensity. The edge has a direction. When P j′ Depends on P j When the direction of the edge between the two nodes is from node P j′ To P j .

[0081] Step 1.3: When establishing the third-party layer network, the third-party set of the smart logistics platform is recorded as PC = {PC i |i=1,2,...,I},PC i represents the i-th third party, I is the number of third parties; PC i The collection of submodule construction projects involved is recorded as Indicates a third-party PC i The jth submodule construction project participated in, J i For PC i The number of submodule construction projects participated in.

[0082] In the third-party layer network, each third party is regarded as a node; when two third-party PCs i 、PC i′ When participating in the same submodule construction project, there is an edge between the two third parties, the direction of the edge is bidirectional and the weight is 1, that is, e ii′ =e i′i =1; otherwise, there is no edge between the third nodes, that is, e ii′ =e i′i =0; where e ii′ Indicates a third-party PC i To a third-party PC i′ The weight of the edge.

[0083] Step 1.4, after the functional layer network, project layer network and third-party layer network are constructed, by clarifying the support relationship between the third party and the submodule construction project, and the contribution relationship between the submodule construction project and the function, the connection between the third-party layer network and the project layer network, and the project layer network and the functional layer network is established, thus forming a multi-layer network of "function-project-third party" that is interconnected and coordinated. In this process, it is first necessary to clarify the relationship between the third party and the project, and between the project and the function.

[0084] The support relationship between the third party and the submodule construction project represents the degree of support provided by the third party to the implementation of the submodule construction project, denoted as e ij , i.e. PC i P j When a third party supports the submodule construction project, there is an edge between the two, and the weight of the edge is the support degree e ij ; The contribution relationship between the submodule construction project and the function represents the contribution of the submodule construction project to the realization of the function, denoted by e jk , that is, P j To achieve the function ES k When the submodule construction project has a contribution relationship with the function, there is an edge between the two, and the weight of the edge is the contribution degree e jk ;e ij and e jk can be given by the project manager. The larger the value, the greater the corresponding degree. ij The value of e is [0,1], jk The value of is [0.5,1.5]; the final constructed "function-project-third party" multi-layer network is as follows Figure 2 shown.

[0085] Step 2, perform failure analysis on the "function-project-third party" multi-layer network corresponding to each sub-module construction project combination; during the failure analysis, first determine the initial load and capacity of each node in the project layer network and each node in the third-party layer network, and set the load distribution strategy when the node fails; based on the initial load, capacity and the load distribution strategy, randomly attack the nodes in the project layer network or the nodes in the third-party network layer, iteratively calculate the capacity value of the project layer network smart logistics platform function realization and the capacity value of the third-party network smart logistics platform function realization, and finally select the optimal sub-module construction project combination through the resilience measurement index of the sub-module construction project combination.

[0086] The nodes or edges of a complex network can bear a certain load, but their capacity is limited. When one or more nodes in the network fail, the load will be transferred to its adjacent nodes. If the adjacent nodes cannot bear the additional load, the node will be affected, causing a chain reaction and leading to the collapse of the entire network. This is the cascading failure phenomenon.

[0087] Among them, in the project layer network, when a sub-module construction project fails, the node corresponding to the sub-module construction project will fail; in the third-party layer network, when the third party cannot provide the support level preset by the project manager, the node corresponding to the third party will fail; the failure of nodes in the project layer network and the node in the third-party layer network will be transmitted through the network.

[0088] Step 2.1, determine the initial load and capacity of each node in the "function-project-third party" multi-layer network; use the "function-project-third party" multi-layer network with the determined initial load and capacity as the initialization network.

[0089] The functional layer network represents the functions of the smart logistics platform that need to be realized by the sub-module construction project combination, and does not involve the concepts of initial load and capacity.

[0090] The initial load of the node in the project layer network represents the amount of tasks carried by the submodule construction project to realize the functions of the smart logistics platform, as shown in formula (5); j Represents the submodule construction project P j The importance of can be calculated by formula (6); the node capacity C j It can be determined by formula (7).

[0091] L j =(1+α)(ω j ) β (5)

[0092]

[0093] Among them, ωj Represents the submodule construction project P j The weight of α and β are the parameters for controlling the initial load strength of the node and α≥0 and β>0. Represents the submodule construction project P j The out-degree strength (from node P j The sum of all weights of the edges pointing to other nodes) and the in-degree strength (from other nodes to node P j The sum of all weights of the edges of j P j investment amount; η3 and η4 represent the importance the project manager attaches to the dependency relationship between sub-module construction projects and the investment amount of sub-module construction projects, and satisfy 0≤η3+η4≤1, where η3,η4∈[0,1].

[0094]

[0095] Among them, λ (λ>0) and γ (γ>0) are capacity adjustment parameters.

[0096] The initial load of the nodes in the third-party layer network represents the resources that the third party needs to provide for the success of the submodule construction project; i The initial load L i , Capacity C i The calculation is also based on formula (5) and formula (7). The difference is that the data used in the calculation is the relevant data in the third-party layer network, and the third-party P in the third-party layer network i The weight of the node ω i (corresponding to formula 6) is calculated using the following formula:

[0097]

[0098] Among them, M i Refers to the set of third parties that have edges with third party i, R i (R i ∈[0, 1]) represents the intimacy between the company where the smart logistics platform is located and the third party. The higher the intimacy, the more conducive it is to the smooth implementation of the project. Its specific value is given by the project manager.

[0099] Step 2.2, determine the load distribution strategy when a node fails.

[0100] This scheme assumes that after a node crashes, the cascading effect is first transmitted within the layer where it is located. When it stabilizes, it is then transmitted to other layers through the inter-layer relationship, and the cycle continues until the cascading failure ends.

[0101] For the project layer network, when one or some submodule construction projects fail, the load carried by the failed submodule construction project should be distributed to other submodule construction projects that depend on the submodule construction project; as shown in formula (8).

[0102]

[0103] Among them, △L jj′ (t) represents the jth submodule construction project P at time t j Passed to the j′th submodule construction project P j′ The load, Π jj′ Indicates P j The amount of tasks carried is passed to P j′ The sharing ratio strategy can be determined according to the remaining capacity of the load-accepting node, that is, the more remaining capacity, the more load accepted, as shown in formula (9).

[0104]

[0105] Among them, Γ j Is dependent on the submodule that failed to build the project node P j A collection of project nodes.

[0106] For the third-party layer, when a failed node appears, its load distribution process is the same as the load distribution method of the failed node in the project layer network (Formula (8)); when one (some) third parties fail, the project they participated in may fail due to insufficient investment.

[0107] To this end, set P j The support obtained is PC i P j The support degree is e ij , I represents the number of third parties; managers can set the threshold This study sets f j * (e ij )=1; when f j (e ij )≥f j * (e ij ), P j is in a normal state and is not affected by the failure of a third party; when f j (e ij )<f j * (e ij ), P j Sub-module construction projects that fail due to third-party failures will subsequently affect other sub-module construction projects through a cascading effect.

[0108] Step 2.3, conduct cascading failure analysis on the "function-project-third party" network established based on a sub-module construction project portfolio; randomly attack nodes in the project layer network or nodes in the third-party network layer. The specific strategy is:

[0109] The failure analysis of the project-level network is as follows: for the initialized network, randomly select one or part of the nodes in the project-level network as failed nodes, delete these failed nodes and the edges associated with them; then redistribute the load of the failed nodes to the nodes that depend on the node and have not failed, calculate the new load of the non-failed nodes in each project-level network, and judge the status of each node in the project-level network at this time. If the load of the node exceeds the capacity, mark the node as a failed node, delete the failed node and the edges associated with it, and redistribute the load of the failed node to the nodes that depend on the node and have not failed; until the load of all non-failed nodes in the network does not exceed the capacity, an iteration is completed, and after the iteration is completed, the capacity value of the project-level network smart logistics platform function is calculated By setting different numbers of failed nodes for the initialized network and going through multiple iterations, until Less than the preset capacity value 10 -3 .

[0110] The failure analysis of the third-party network is also carried out on the basis of the initialization network. The method is the same as the failure analysis of the project layer network. After each iteration, the capability value of the smart logistics platform function of the third-party network needs to be calculated. After multiple iterations, Less than the preset capacity value 10 -3 .

[0111] Step 2.4, select the optimal sub-module construction project portfolio through the sub-module construction project portfolio resilience measurement index.

[0112] After the cascading failure process is over, the function realization value and project portfolio resilience of all alternative sub-module construction project portfolio networks are calculated and compared, and the project portfolio with high function realization value and strong resilience is selected as the optimal sub-module construction project portfolio.

[0113] The functional realization value is measured by formulas (10) to (14), and the resilience measurement index of project portfolio b is determined by formula (15):

[0114]

[0115]

[0116]

[0117]

[0118]

[0119] in, and They represent the functional realization value of the θth iteration of the submodule construction project portfolio b when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2, CC b is the expected capability value, represents the final functional realization value of the submodule construction project combination b when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2; K is the number of functions, J represents the number of submodule construction projects in the submodule construction project combination b, ψ k Indicates the function node ES k The node weights, where e kk′ Indicates function ES k ,ES k′ The synergistic relationship between Indicates function ES k A collection of other functions affected, OT means ES k The sum of the weights of the edges that affect other features; represents the ES determined by the experts k The functional weight, Θ k Indicates the node ES k The set of nodes that have an impact; and They represent the number of nodes in the project layer network when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2, respectively. The default value is 10. -3 , g represents the g-th iteration.

[0120]

[0121] Among them, Ω1 means that among all the candidate sub-module project combinations, the function realization value is less than the preset value of 10 -2 Ω2 indicates that among all the candidate submodule project combinations, the function realization value is less than the preset value of 10 -2 The maximum number of failed third-party nodes when

[0122] The resilience measurement index ST of the submodule construction project portfolio b bThe larger the value, the stronger the resilience of the sub-module construction project combination b. After all sub-module construction project combinations are calculated according to the above method, the sub-module construction project combination with the largest resilience measurement index is selected as the optimal sub-module construction project combination, and the sub-modules are constructed according to the sub-module construction projects in the sub-module construction project combination to realize the functions of the smart logistics platform.

[0123] The smart logistics platform built based on the multi-layer network cascading failure model can not only efficiently coordinate the operation of various functional modules, but also has the ability to deal with complex failure scenarios, thus providing a solid guarantee for the intelligent management and efficient operation of the supply chain.

[0124] Example:

[0125] The implementation example provided by the present invention originates from a global logistics company, which plans to build a smart logistics platform based on a multi-layer network cascading failure model, aiming to achieve efficient collaboration and intelligent management of the supply chain. The platform consists of multiple interdependent sub-modules, including warehouse management, order management, transportation management, distribution management, etc. The goal of the platform is to select 60 optimal sub-modules from 100 alternative sub-modules to achieve the established functional goals while meeting the constraints of the capital budget; three sub-module construction project combinations that can simultaneously meet all the functional realizations of the platform are determined: PP1, PP2 and PP3; in order to select the combination with the highest resilience measurement index in the optional sub-module construction project combination, the smart logistics platform construction method based on the multi-layer network cascading failure model is adopted in the embodiment.

[0126] 1. Build a multi-layer network of "function-project-third party".

[0127] The functions of the smart logistics platform are subdivided into multiple functions, and experts are invited to evaluate the importance of each function and its synergy; the importance of the function is used as the functional weight of the functional layer node; in this embodiment, the platform function is divided into 20 sub-functions, and the importance of the sub-function is determined based on the weight.

[0128] Each project manager implements project-level network construction based on the IPO (Input-Process-Output) project dependency relationship.

[0129] A third-party layer network is generated based on whether the third party participates in the common sub-module construction project; the support of each third party to the sub-module construction project is evaluated by the project manager based on the capabilities of the third party.

[0130] First, determine the relationship between submodule construction projects and functions. The contribution of each submodule construction project to each functional goal can be given by each single project management team / project manager, with a value of [0,1]. The contribution degree gradually increases from 0 to 1, and the contribution degree of each submodule construction project to each functional goal in the three project portfolios is obtained; secondly, determine the support degree of the third party to the submodule construction project. The enterprise and the third party work together to complete the established project, and each third party contributes to the completion of the project according to the division of labor; but the third party itself may have more capabilities to complete more work, but it is limited to the division of labor requirements. Therefore, each single project manager can determine the support degree of the third party to the submodule construction project based on the third party's capabilities, and take the value in the range of [0.5,1.5]. The larger the value, the stronger the support degree.

[0131] Through the above steps, the "function-project-third party" multi-layer network model under different sub-module construction project combinations is completed. Taking PP1 as an example, its "function-project-third party" multi-layer network model is as follows Figure 3 As shown, ES1~ES20 represent functions, P1~P98 represent sub-module construction projects, and PC1~PC41 represent third parties.

[0132] 2. Cascading failure analysis.

[0133] First, determine the initial load and capacity of each node in the project layer network and the third-party layer network. Then, judge the status of the nodes in the project layer network and the third-party layer network. According to the node load distribution process and strategy, determine the resilience measurement index.

[0134] According to the above steps, the multi-layer network cascading failure results under the participation of three different sub-module construction project combinations of PP1, PP2 and PP3 are calculated. When the failure nodes are 1, 2, 3... until the function realization value is close to 0, the multi-layer network cascading failure results under different numbers of failure nodes in the project layer network and the third-party layer network are obtained respectively, and the results are visualized, see Figure 4 , Figure 5 ; And it was calculated that the resilience measurement indicators of PP1, PP2 and PP3 were -0.1236, -0.3616, and -0.0407 respectively. The multi-layer network formed by PP1 had the strongest resilience and was finally selected.

[0135] The implementation of the present invention not only demonstrates its application potential in the construction of a specific logistics platform, but also proves its practical utility in improving the intelligence and resilience of supply chain management. Through the present invention, logistics companies can more effectively select and manage sub-module projects to ensure that they remain competitive in a complex and changing market environment.

[0136] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for constructing a smart logistics platform based on a multi-layer network cascading failure model, characterized in that: include: For each submodule construction project combination in the multiple submodule construction project combinations, a "function-project-third party" multi-layer network is constructed; the "function-project-third party" multi-layer network includes a function layer network, a project layer network and a third party layer network. The function layer network includes all functions that need to be implemented by the smart logistics platform. The project layer network uses the project ontology model to describe each submodule construction project in the submodule construction project combination; the third party layer network includes a third party; the "function-project-third party" multi-layer network is formed by using the support relationship between the third party and the submodule construction project and the contribution relationship between the submodule construction project and the function; Perform failure analysis on the "function-project-third party" multi-layer network corresponding to each sub-module construction project combination; during failure analysis, first determine the initial load and capacity of each node in the project layer network and each node in the third-party layer network, and set the load distribution strategy when the node fails; Based on the initial load, capacity and the load distribution strategy, failure analysis is performed on the nodes in the project layer network or the nodes in the third-party network layer, the capability value of the smart logistics platform function realized by the project layer network and the capability value of the smart logistics platform function realized by the third-party network are calculated, and the optimal sub-module construction project combination is selected based on the capability value through the resilience measurement index of the sub-module construction project combination; The optimal submodule construction project combination is selected based on the capability value through the resilience measurement index of the submodule construction project combination. The specific formula is: Among them, CC b is the expected capability value, and They represent the functional realization values ​​of the submodule construction project combination b at the θth iteration when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2, respectively, and g represents the gth iteration; represents the final functional realization value of the submodule construction project combination b when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2; K is the number of functions to be realized, J represents the number of submodule construction projects in the submodule construction project combination b, ψ k represents the kth function ES k The node weights, where e kk′ represents the kth and k′th functions ES k ,ES k′ The synergistic relationship between Indicates function ES k A collection of other functions affected, OT means ES k The sum of the weights of the edges that affect other features; η1,η2∈[0,1]; represents the ES determined by the experts k The functional weight, Θ k Indicates the function ES k The set of functions that have an impact; jk 、e jk′ Represents the jth submodule construction project P j For function ES k ,ES k′ the degree of contribution achieved; and They represent the number of nodes in the project layer network when the number of failed nodes in the project layer network is χ1 and the number of failed nodes in the third-party layer network is χ2, respectively. is the preset value, Ω1 represents the maximum number of failed submodule construction projects when the functional realization value is less than the preset realization value in all candidate submodule project combinations; Ω2 represents the maximum number of failed third parties when the functional realization value is less than the preset realization value in all candidate submodule project combinations; ST b It represents the resilience measurement index of submodule construction project portfolio b; The sub-module construction project combination with the largest resilience measurement index is selected as the optimal sub-module construction project combination, and the sub-modules are constructed according to the sub-module construction projects in the sub-module construction project combination to realize the functions of the smart logistics platform.

2. The method for constructing a smart logistics platform based on a multi-layer network cascading failure model according to claim 1 is characterized in that: Each node in the functional layer network is a function. If there is a collaborative relationship between functions, there is an edge between the functions. The collaborative relationship is the function ES k Impact function ES k′ , or function ES k′ Impact function ES k ; The edges between functions have directions, when the function ES k Impact function ES k′ When the edge direction is ES k To ES k′ ; Set functional weights for nodes in the functional layer network.

3. The method for constructing a smart logistics platform based on a multi-layer network cascading failure model according to claim 1 is characterized in that: The project layer network uses the project ontology model to describe each sub-module construction project in the sub-module construction project portfolio, including: The IPO-based project ontology model describes each submodule construction project in the submodule construction project portfolio; the ontology model includes project input, project processing, project output, project start and end time, and the functions supported by the project; the dependency relationship between submodule construction projects is clarified, and a dependency matrix between submodule construction projects is constructed. The dependency matrix is ​​used to clarify the relationship between each submodule construction project in the project layer network and the degree of closeness of the relationship. Each node in the project layer network represents a submodule construction project in the smart logistics platform; wherein: The number of dependency types between submodule construction projects is used to characterize the dependency strength between submodule construction projects. The dependency types include technology dependency, resource dependency, strategic dependency, and benefit dependency. If the submodule construction project P j′ P j If there is at least one of the four dependencies, then P j′ Depends on P j ; When the dependency intensity between submodule construction projects reaches 1 or above, there is an edge between submodule construction projects, and the weight of the edge represents the dependency intensity between submodule construction projects; the edge has a direction, when P j′ Depends on P j When the direction of the edge between the two is from node P j′ To P j .

4. The method for constructing a smart logistics platform based on a multi-layer network cascading failure model according to claim 1 is characterized in that: Each third party in the third-party layer network acts as a node; when two third-party PCs i 、PC i′ When participating in the same submodule construction project, there is an edge between the two third parties, the direction of the edge is bidirectional and the weight is 1, otherwise, there is no edge between the third party nodes.

5. The method for constructing a smart logistics platform based on a multi-layer network cascading failure model according to claim 1 is characterized in that: By utilizing the support relationship between the third party and the submodule construction project, and the contribution relationship between the submodule construction project and the function, the "function-project-third party" multi-layer network is formed, including: The support relationship between the third party and the submodule construction project represents the degree of support provided by the third party to the implementation of the submodule construction project, denoted as e ij , represents the i-th third-party PC i For the jth submodule construction project P j When a third party supports the submodule construction project, there is an edge between the two, and the weight of the edge is the support degree e ij ; The contribution relationship between the submodule construction project and the function represents the contribution of the submodule construction project to the realization of the function, denoted by e jk , indicating the submodule construction project P j To realize the kth function ES k When the submodule construction project has a contribution relationship with the function, there is an edge between the two, and the weight of the edge is the contribution degree e jk ;e ij and e jk Given by the project manager; By utilizing the support relationship and contribution degree, connections are established between the third-party layer network and the project layer network, and between the project layer network and the function layer network, thereby forming a "function-project-third party" multi-layer network.

6. The method for constructing a smart logistics platform based on a multi-layer network cascading failure model according to claim 1 is characterized in that: The initial load and capacity calculation method of the nodes in the project layer network is as follows: Among them, L j Represents the jth submodule construction project P in the project layer network j The initial load, C j Indicates capacity; ω j Represents the submodule construction project P j The weight of α and β are the parameters for controlling the initial load strength of the node and α≥0 and β>0. Represents the submodule construction project P j The out-degree strength and in-degree strength of F j P j investment amount; η3 and η4 represent the importance that the project manager attaches to the dependencies between sub-module construction projects and the investment amount of sub-module construction projects, and satisfy 0≤η3+η4≤1, where η3, η4∈[0,1], λ and γ are capacity adjustment parameters, λ>0, γ>0.

7. The method for constructing a smart logistics platform based on a multi-layer network cascading failure model according to claim 1 is characterized in that: The setting of the load distribution strategy when a node fails includes: For the project-level network, when one or some submodule construction projects fail, the load carried by the failed submodule construction project needs to be distributed to other submodule construction projects that depend on the submodule construction project: Where, ΔL jj′ (t) represents the jth submodule construction project P at time t j Passed to the j′th submodule construction project P j′ Load, L j (t) represents the submodule construction project P j The load at time t, Indicates P j The amount of tasks carried is passed to P j′ Ratio: Among them, L j′ (t), C j′ Respectively represent the submodule construction project P j′ Load and capacity at time t; L h (t), C h They represent the load and capacity of the h-th submodule construction project at time t; Γ j Is dependent on the submodule that failed to build the project node P j A collection of project nodes.

8. The method for constructing a smart logistics platform based on a multi-layer network cascading failure model according to claim 1 is characterized in that: Set the jth submodule construction project P j The support obtained is The i-th third-party PC i P j The support degree is e ij , I represents the number of third parties; set the threshold when When P j is in normal condition and is not affected by failure of third parties; When P j Sub-module construction projects that fail due to third-party failures will subsequently affect other sub-module construction projects through a cascading effect.

9. The method for constructing a smart logistics platform based on a multi-layer network cascading failure model according to claim 1 is characterized in that: The failure analysis of the nodes in the project layer network or the nodes in the third-party network layer, and iterative calculation of the capability value of the project layer network smart logistics platform function realization and the capability value of the third-party network smart logistics platform function realization include: The failure analysis of the project layer network is as follows: the "function-project-third party" multi-layer network with determined initial load and capacity is used as the initialization network; for the initialization network, one or part of the nodes in the project layer network are randomly selected as failed nodes, and these failed nodes and the edges associated with them are deleted; then the load of the failed node is redistributed to the nodes that depend on the node and have not failed, and the new load of the non-failed nodes in each project layer network is calculated, and the state of each node in the project layer network at this time is judged. If the load of the node exceeds the capacity, the node is marked as a failed node, and after deleting the failed node and the edges associated with it, the load of the failed node is redistributed to the nodes that depend on the node and have not failed; until the load of all non-failed nodes in the network does not exceed the capacity, an iteration is completed, and after the iteration is completed, the calculation By setting different numbers of failed nodes for the initialized network and going through multiple iterations, until Less than the preset capacity value; The failure analysis of the third-party network is the same as the failure analysis of the project-level network. After each iteration, it is necessary to calculate After multiple iterations, Less than the preset capacity value.

Images