Logistics coordination scheduling and management method and system for food supply chain traceability

Abstract

The present application relates to a logistics collaborative scheduling and management method and system for food supply chain traceability, belonging to the technical field of logistics scheduling and supply chain management. Among them, the method comprises: obtaining link structure and logistics order information and performing associated coding calculation, performing associated domain self-organizing aggregation on the logistics objects to be scheduled to form a logistics collaborative association domain; based on the disturbance propagation gradient between the association mapping calculation domains to construct the scheduling disturbance conduction chain, real-time update the dynamic scheduling priority; through the priority identification of the link to be adjusted, the collaborative link reconstruction deduction is performed and the link reconstruction benefit is calculated, the scheduling reconstruction sequence is generated, and the logistics collaborative scheduling scheme is generated combined with the collaborative scheduling time limit joint evaluation model; the scheduling evolution track is formed by tracking the logistics node state to dynamically correct the association domain and the conduction chain, and the batch flow information and the scheduling record are associated and packaged to generate the traceability data set. The disturbance conduction driven active collaborative scheduling and full link traceable scheduling management closed loop are realized.

Description

Technical Field

[0001] This invention belongs to the field of logistics scheduling and supply chain management technology, specifically relating to a logistics collaborative scheduling and management method and system for food supply chain traceability. Background Technology

[0002] With the rapid development of fresh food e-commerce, cold chain logistics, and the food industry, the collaborative scheduling and end-to-end traceability management of the food supply chain have received increasing attention. The food supply chain involves multiple levels of logistics nodes, from production, processing, warehousing, and transportation to final sales. These nodes have complex spatiotemporal dependencies and resource competition relationships. Multiple logistics orders are interconnected and influence each other while sharing transportation resources. A scheduling deviation at any node can propagate along the logistics chain to upstream and downstream, triggering a chain reaction.

[0003] Currently, most solutions widely used in the food supply chain logistics scheduling field are based on static path planning and passive anomaly response mechanisms. These solutions generate fixed scheduling plans before transportation based on preset delivery times and capacity. During transportation, they monitor the execution status of each node through real-time location tracking or status reporting. When anomalies occur, scheduling adjustments are triggered manually or through a rule engine. However, these solutions have the following shortcomings: First, the scheduling plan is generated at the granularity of individual orders, failing to fully consider the spatiotemporal correlation and capacity synergy potential between multiple orders, and lacking a mechanism for aggregating and coordinating logistics objects across related domains. Second, the anomaly response mechanism is passively triggered, unable to predict the propagation path and impact range of the deviation along the logistics link in the early stages of scheduling deviations, and lacks a mechanism for quantifying the propagation gradient and constructing the transmission chain of scheduling disturbances across different logistics collaborative domains. Adjustment measures often lag behind the speed of deviation spread.

[0004] In terms of multi-batch coordinated scheduling, existing solutions typically treat each transport batch as an independent scheduling unit, ignoring resource competition and mutual influence between different batches within the same capacity unit. When a batch experiences a timeliness anomaly, existing solutions cannot quantify the degree of disturbance to other batches, nor can they dynamically adjust the scheduling scheme to absorb the impact of the disturbance through link substitution margin assessment and coordinated link reconstruction simulation.

[0005] In terms of traceability management, existing solutions mostly record batch information and logistics node transfer records for food, forming a static traceability chain. Such solutions only record transfer nodes and cannot reflect the dynamic adjustment trajectory and scheduling evolution process during scheduling. When timeliness disputes occur, it is impossible to trace the causal relationship between scheduling decisions and execution results, and it is also difficult to associate and map batch transfer information with scheduling records at each stage to form a complete scheduling traceability dataset.

[0006] In summary, how to construct a logistics collaborative association domain and a scheduling disturbance transmission chain, quantify the transmission impact of scheduling deviations through disturbance propagation gradient, and achieve proactive scheduling decisions by combining collaborative link reconstruction simulation and collaborative scheduling timeliness joint evaluation model, while forming a traceable scheduling evolution trajectory and source dataset, are the technical problems that need to be solved in this field. Summary of the Invention

[0007] To address the aforementioned problems in the existing technology, this invention provides a logistics collaborative scheduling and management method and system for food supply chain traceability. The objective of this invention can be achieved through the following technical solutions: S1: Obtain the link structure information and logistics order information based on the food category supply chain, perform association coding calculation on the link structure information and logistics order information, identify the logistics objects to be scheduled, and perform association domain self-organization aggregation on the logistics objects to be scheduled to form logistics collaborative association domains, while establishing association mapping relationships between each logistics collaborative association domain; S2: Calculate the disturbance propagation gradient between each logistics collaboration domain based on the aforementioned association mapping relationship, form a scheduling disturbance transmission chain, and update the scheduling priority of each logistics collaboration domain in real time; S3: Identify the links to be adjusted within the logistics collaboration domain through the scheduling priority; perform collaborative link reconstruction deduction on candidate links that meet the substitution conditions and calculate the link reconstruction benefits to generate multiple sets of scheduling reconstruction sequences; combine the collaborative scheduling timeliness joint evaluation model to calculate the overall scheduling efficiency value of each scheduling reconstruction sequence and generate a logistics collaborative scheduling scheme. S4: Obtain execution feedback of the logistics collaborative scheduling scheme, track the task completion status and resource change status of each logistics node, form a scheduling evolution trajectory and correct the logistics collaborative correlation domain and scheduling disturbance transmission chain; associate and store the logistics batch flow information of the traceable food with the scheduling records of each stage, and generate supply chain logistics collaborative scheduling and traceability results.

[0008] Specifically, the association encoding calculation includes: Extract the topological adjacency matrix of the logistics nodes in the link structure information, and construct a node connection graph based on the topological adjacency matrix; extract the order timeliness window and capacity constraint vector in the logistics order information, and write the order timeliness window and capacity constraint vector as node attributes into each node in the node connection graph to form a graph topological association coding space; In the graph topology association coding space, graph topology association coding vectors for each logistics object to be scheduled are generated based on the node attributes and connection relationships of each node. A spatiotemporal association graph between the logistics objects to be scheduled is constructed based on the similarity between the graph topology association coding vectors, and the spatiotemporal association relationship is determined.

[0009] Specifically, the self-organizing aggregation of related domains includes: Using the graph topology association encoding vector as the clustering input feature, and combining the delivery time constraint boundary and the logistics function coordination boundary extracted from logistics order information, a multi-dimensional constraint boundary is constructed. Within the multidimensional constraint boundary, collaborative aggregation is performed on the logistics objects to be scheduled. When the aggregation degree reaches the preset aggregation degree threshold, the aggregation stops and the logistics objects to be scheduled that meet the preset aggregation degree threshold are aggregated into a logistics collaborative association domain. During the iteration process, the aggregation degree threshold is adaptively adjusted according to the aggregation change trend until the aggregation result meets the preset stability condition.

[0010] Specifically, the formation of the scheduling disturbance propagation chain includes: When the logistics collaboration association domain generates a scheduling deviation, the affected downstream association domain is determined according to the association mapping relationship; the propagation gradient of the scheduling deviation is calculated according to the weight of the association edges between the logistics collaboration association domains, and the propagation gradient is iteratively calculated by combining the attenuation rate of the scheduling deviation and the dynamic amplification coefficient. Obtain the propagation gradient feedback, and correct the weights of the association edges between logistics collaboration association domains in the association mapping relationship based on the propagation gradient feedback; The iteration terminates when the change in the propagation gradient is less than the preset convergence threshold. The logistics collaboration correlation domain that generates the scheduling deviation is taken as the disturbance source, and the scheduling disturbance propagation chain is formed along the directed connection relationship of the propagation gradient.

[0011] Specifically, the process of updating the scheduling priority of each logistics collaboration domain includes: Along the forward propagation path of the scheduling disturbance propagation chain, the cumulative disturbance intensity value of each logistics collaboration domain is calculated level by level; The cumulative disturbance intensity value is compared with the preset tolerance threshold of each logistics collaboration domain. The priority increase increment is determined based on the comparison result. The original scheduling priority is superimposed with the priority increase increment. For logistics collaboration domains whose scheduling status has returned to normal, the scheduling priority is gradually reduced according to the preset disturbance release rate until it returns to the original preset level.

[0012] Specifically, the identification process of the link to be adjusted includes: Target logistics collaborative association domains with scheduling priorities higher than a preset priority threshold are selected; the current scheduling links are traversed within the target logistics collaborative association domains, and the link substitution margin of each scheduling link is calculated. The link substitution margin is obtained by fusing the number of alternative paths, node switching cost, and subsequent logistics collaborative association domain disturbance propagation increment, and is used to characterize the feasibility of the current link being replaced by alternative links in the logistics network and the disturbance impact on downstream association domains after replacement. The scheduled links whose link replacement margin is lower than a preset margin threshold are marked as the links to be adjusted.

[0013] Specifically, the collaborative link reconstruction simulation includes: Based on the starting point and ending point of the link to be adjusted, candidate links that meet the requirements of capacity level and path connectivity are retrieved from the global backup capacity pool and backup route library. The candidate links are subjected to a benefit-driven iterative deduction using a graph topology collaborative reconstruction deduction algorithm. In each round of deduction, the selection probability of each candidate link is updated according to the link reconstruction benefit. Candidate links with benefits higher than a preset retention threshold are retained, and candidate links with benefits lower than the preset retention threshold are recombined. When the improvement in iteration benefits is less than a preset convergence threshold, the deduction is terminated, and the currently retained candidate links are taken as the scheduling and reconstruction sequence; the improvement in link continuity and node load balancing of each scheduling and reconstruction sequence are calculated as the link reconstruction benefits.

[0014] Specifically, the collaborative scheduling timeliness joint evaluation model adopts a hierarchical progressive game evaluation architecture, including: The basic evaluation layer calculates link continuity score and node load balancing score, and outputs the calculation results to the constraint correction layer. The constraint correction layer receives the output of the basic evaluation layer and performs compliance verification on the expected arrival time of each logistics node through the path time window constraint matrix to obtain the expected time window deviation value. The expected time window deviation value is used as a correction factor to correct the output of the basic evaluation layer. The collaborative game layer receives the output of the constraint correction layer, performs multi-objective collaborative game convergence calculation on the expected time window deviation value, the corrected link continuity score, the corrected node load balance score and the link reconstruction benefit, and outputs the overall scheduling efficiency value. The scheduling reconstruction sequence with the optimal overall scheduling efficiency value is taken as the logistics collaborative scheduling scheme.

[0015] Specifically, the formation of the scheduling evolution trajectory includes: The task receipt timestamps, remaining transport capacity of transport vehicles, and available loading and unloading capacity of each logistics node are collected in real time as the task completion status and resource change status. Using absolute time as the axis, the task completion status and resource change status of each logistics node at each timestamp are aligned and mapped with the corresponding logistics collaboration association domain in a time-series dimension to generate a scheduling evolution trajectory that represents the dynamic execution process of the scheduling scheme.

[0016] Specifically, the associated storage and generation of tracing results includes: Using the logistics batch transfer information of the food to be traced as the primary key, a mapping index is established between the logistics batch transfer information and the scheduling records of each stage. The scheduling evolution trajectory is used as process data and associated with the mapping index to form a traceable batch flow and scheduling decision association index, generating supply chain logistics collaborative scheduling traceability data results.

[0017] Specifically, a logistics collaborative scheduling and management system for food supply chain traceability includes: The collaborative association domain construction module acquires and parses the link structure information and logistics order information based on the food category supply chain, performs association coding calculation to identify the logistics objects to be scheduled, performs association domain self-organization aggregation on the logistics objects to be scheduled, and generates logistics collaborative association domains and association mapping relationships. The disturbance propagation control module calculates the disturbance propagation gradient based on the associated mapping relationship, generates a scheduling disturbance propagation chain based on the directed connection relationship of the disturbance propagation gradient, and dynamically updates the scheduling priority of each logistics collaboration associated domain. The link reconfiguration decision module identifies links to be adjusted based on the scheduling priority, performs iterative deduction driven by benefits on candidate links and calculates the benefits of link reconfiguration, generates a scheduling reconfiguration sequence, calculates the overall scheduling efficiency value of each scheduling reconfiguration sequence through a collaborative scheduling efficiency joint evaluation model, and selects the scheduling reconfiguration sequence with the optimal overall scheduling efficiency value as the logistics collaborative scheduling scheme. The scheduling evolution traceability module obtains execution feedback and generates scheduling evolution trajectory, corrects the logistics collaboration association domain and scheduling disturbance transmission chain, associates and encapsulates the logistics batch flow information of the food to be traced with the scheduling records of each stage, and generates supply chain logistics collaborative scheduling traceability data results.

[0018] The beneficial effects of this invention are as follows: (1) Based on graph topology encoding and self-organizing aggregation, a logistics collaborative association domain is constructed. Combined with the disturbance gradient iteration mechanism, a scheduling disturbance transmission chain is built to realize the upgrade from passive anomaly handling to active prediction of disturbances across the entire domain. This invention no longer uses a single order as the scheduling unit, but aggregates logistics objects with capacity collaboration potential based on spatiotemporal similarity, fully releasing the space for multi-order collaborative scheduling. At the same time, the transmission intensity of scheduling deviation to downstream association domains can be quantified, the inter-domain disturbance association weights can be dynamically updated, and the scheduling priority of each region can be adjusted according to the cumulative disturbance intensity. Priority is given to tilting scheduling resources to high-disturbance links, thus curbing the spread of local delays into network-wide scheduling imbalance from the source.

[0019] (2) By introducing a revenue-driven graph topology reconstruction deduction and a three-layer progressive game evaluation model, automated scheduling optimization under disturbance scenarios is achieved. At the same time, a time-series scheduling evolution trajectory and batch traceability dataset are constructed to form a closed-loop management of scheduling, execution, and review. This solution can automatically generate multiple sets of alternative link combinations, quantify the comprehensive scheduling efficiency from multiple dimensions such as link connectivity, node load, and time constraints, and select the globally optimal scheduling scheme. The time-series operation data of the entire process can form a complete scheduling evolution trajectory, and bind food batch information to generate an integrated traceability dataset, providing complete data basis for anomaly review and responsibility definition. Attached Figure Description

[0020] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0021] Figure 1 This is a flowchart illustrating a logistics collaborative scheduling and management method for food supply chain traceability according to the present invention. Figure 2 This is an architecture diagram of a logistics collaborative scheduling and management system for food supply chain traceability, as described in this invention. Detailed Implementation

[0022] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following detailed description of the specific implementation methods, structures, features and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0023] Please see Figures 1-2 A logistics collaborative scheduling and management method for food supply chain traceability includes the following steps: S1: Obtain the link structure information and logistics order information based on the food category supply chain, perform association coding calculation on the link structure information and logistics order information, identify the logistics objects to be scheduled, and perform association domain self-organization aggregation on the logistics objects to be scheduled to form logistics collaborative association domains, while establishing association mapping relationships between each logistics collaborative association domain; S2: Calculate the disturbance propagation gradient between each logistics collaboration domain based on the aforementioned association mapping relationship, form a scheduling disturbance transmission chain, and update the scheduling priority of each logistics collaboration domain in real time; S3: Identify the links to be adjusted within the logistics collaboration domain through the scheduling priority; perform collaborative link reconstruction deduction on candidate links that meet the substitution conditions and calculate the link reconstruction benefits to generate multiple sets of scheduling reconstruction sequences; combine the collaborative scheduling timeliness joint evaluation model to calculate the overall scheduling efficiency value of each scheduling reconstruction sequence and generate a logistics collaborative scheduling scheme. S4: Obtain execution feedback of the logistics collaborative scheduling scheme, track the task completion status and resource change status of each logistics node, form a scheduling evolution trajectory and correct the logistics collaborative correlation domain and scheduling disturbance transmission chain; associate and store the logistics batch flow information of the traceable food with the scheduling records of each stage, and generate supply chain logistics collaborative scheduling and traceability results.

[0024] Specifically, the association encoding calculation includes: Extract the topological adjacency matrix of the logistics nodes in the link structure information, and construct a node connection graph based on the topological adjacency matrix; extract the order timeliness window and capacity constraint vector in the logistics order information, and write the order timeliness window and capacity constraint vector as node attributes into each node in the node connection graph to form a graph topological association coding space; In the graph topology association coding space, graph topology association coding vectors for each logistics object to be scheduled are generated based on the node attributes and connection relationships of each node. A spatiotemporal association graph between the logistics objects to be scheduled is constructed based on the similarity between the graph topology association coding vectors, and the spatiotemporal association relationship is determined.

[0025] Specifically, the self-organizing aggregation of the logistics collaborative association domain includes: Using the graph topology association encoding vector as the clustering input feature, and combining the delivery time constraint boundary and the logistics function coordination boundary extracted from logistics order information, a multi-dimensional constraint boundary is constructed. Within the multidimensional constraint boundary, collaborative aggregation is performed on the logistics objects to be scheduled. When the aggregation degree reaches the preset aggregation degree threshold, the aggregation stops and the logistics objects to be scheduled that meet the preset aggregation degree threshold are aggregated into a logistics collaborative association domain. During the iteration process, the aggregation degree threshold is adaptively adjusted according to the aggregation change trend until the aggregation result meets the preset stability condition.

[0026] Specifically, the formation of the scheduling disturbance propagation chain includes: When the logistics collaboration association domain generates a scheduling deviation, the affected downstream association domain is determined according to the association mapping relationship; the propagation gradient of the scheduling deviation is calculated according to the weight of the association edges between the logistics collaboration association domains, and the propagation gradient is iteratively calculated by combining the attenuation rate of the scheduling deviation and the dynamic amplification coefficient. Obtain the propagation gradient feedback, and correct the weights of the association edges between logistics collaboration association domains in the association mapping relationship based on the propagation gradient feedback; The iteration terminates when the change in the propagation gradient is less than the preset convergence threshold. The logistics collaboration correlation domain that generates the scheduling deviation is taken as the disturbance source, and the scheduling disturbance propagation chain is formed along the directed connection relationship of the propagation gradient.

[0027] Specifically, the process of updating the scheduling priority of each logistics collaboration domain includes: Along the forward propagation path of the scheduling disturbance propagation chain, the cumulative disturbance intensity value of each logistics collaboration domain is calculated level by level; The cumulative disturbance intensity value is compared with the preset tolerance threshold of each logistics collaboration domain. The priority increase increment is determined based on the comparison result. The original scheduling priority is superimposed with the priority increase increment. For logistics collaboration domains whose scheduling status has returned to normal, the scheduling priority is gradually reduced according to the preset disturbance release rate until it returns to the original preset level.

[0028] Specifically, the identification process of the link to be adjusted includes: Target logistics collaborative association domains with scheduling priorities higher than a preset priority threshold are selected; the current scheduling links are traversed within the target logistics collaborative association domains, and the link substitution margin of each scheduling link is calculated. The link substitution margin is obtained by fusing the number of alternative paths, node switching cost, and subsequent logistics collaborative association domain disturbance propagation increment, and is used to characterize the feasibility of the current link being replaced by alternative links in the logistics network and the disturbance impact on downstream association domains after replacement. The scheduled links whose link replacement margin is lower than a preset margin threshold are marked as the links to be adjusted.

[0029] Specifically, the collaborative link reconstruction simulation includes: Based on the starting point and ending point of the link to be adjusted, candidate links that meet the requirements of capacity level and path connectivity are retrieved from the global backup capacity pool and backup route library. The candidate links are subjected to a benefit-driven iterative deduction using a graph topology collaborative reconstruction deduction algorithm. In each round of deduction, the selection probability of each candidate link is updated according to the link reconstruction benefit. Candidate links with benefits higher than a preset retention threshold are retained, and candidate links with benefits lower than the preset retention threshold are recombined. When the improvement in iteration benefits is less than a preset convergence threshold, the deduction is terminated, and the currently retained candidate links are taken as the scheduling and reconstruction sequence; the improvement in link continuity and node load balancing of each scheduling and reconstruction sequence are calculated as the link reconstruction benefits.

[0030] Specifically, the collaborative scheduling timeliness joint evaluation model adopts a hierarchical progressive game evaluation architecture, including: The basic evaluation layer calculates link continuity score and node load balancing score, and outputs the calculation results to the constraint correction layer. The constraint correction layer receives the output of the basic evaluation layer and performs compliance verification on the expected arrival time of each logistics node through the path time window constraint matrix to obtain the expected time window deviation value. The expected time window deviation value is used as a correction factor to correct the output of the basic evaluation layer. The collaborative game layer receives the output of the constraint correction layer, performs multi-objective collaborative game convergence calculation on the expected time window deviation value, the corrected link continuity score, the corrected node load balancing score and the link reconstruction benefit, and outputs the overall scheduling efficiency value.

[0031] Specifically, the formation of the scheduling evolution trajectory includes: The task receipt timestamps, remaining transport capacity of transport vehicles, and available loading and unloading capacity of each logistics node are collected in real time as the task completion status and resource change status. Using absolute time as the axis, the task completion status and resource change status of each logistics node at each timestamp are aligned and mapped with the corresponding logistics collaboration association domain in a time-series dimension to generate a scheduling evolution trajectory that represents the dynamic execution process of the scheduling scheme.

[0032] Specifically, the associated storage and generation of tracing results includes: Using the logistics batch transfer information of the food to be traced as the primary key, a mapping index is established between the logistics batch transfer information and the scheduling records of each stage. The scheduling evolution trajectory is used as process data and associated with the mapping index to form a traceable batch flow and scheduling decision association index, generating supply chain logistics collaborative scheduling traceability data results.

[0033] Specifically, a logistics collaborative scheduling and management system for food supply chain traceability includes: The collaborative association domain construction module acquires and parses the link structure information and logistics order information based on the food category supply chain, performs association coding calculation to identify the logistics objects to be scheduled, performs association domain self-organization aggregation on the logistics objects to be scheduled, and generates logistics collaborative association domains and association mapping relationships. The disturbance propagation control module calculates the disturbance propagation gradient based on the associated mapping relationship, generates a scheduling disturbance propagation chain based on the directed connection relationship of the disturbance propagation gradient, and dynamically updates the scheduling priority of each logistics collaboration associated domain. The link reconfiguration decision module identifies links to be adjusted based on the scheduling priority, performs iterative deduction driven by benefits on candidate links and calculates the benefits of link reconfiguration, generates a scheduling reconfiguration sequence, calculates the overall scheduling efficiency value of each scheduling reconfiguration sequence through a collaborative scheduling efficiency joint evaluation model, and selects the scheduling reconfiguration sequence with the optimal overall scheduling efficiency value as the logistics collaborative scheduling scheme. The scheduling evolution traceability module obtains execution feedback and generates scheduling evolution trajectory, corrects the logistics collaboration association domain and scheduling disturbance transmission chain, associates and encapsulates the logistics batch flow information of the food to be traced with the scheduling records of each stage, and generates supply chain logistics collaborative scheduling traceability data results.

[0034] Example 1: Spatiotemporal topological association coding and association domain self-organizing aggregation mechanism In this embodiment, within the context of a multi-level food supply chain, a graph topology method is used to reconstruct multi-source discrete logistics orders and transportation resources, forming associated domain entities with collaborative characteristics: 1. Construction and Vectorization of Graph Topological Association Encoding Space The collaborative association domain construction module first retrieves the link structure information and logistics order information based on the food category supply chain; the module then extracts the topological adjacency matrix M of the distribution nodes at each level and the logistics transit hubs in the link structure information. adj If there is a physical transport route between node i and node j, then the corresponding element M in the adjacency matrix... adj =1, otherwise 0. Based on this matrix, a topology baseline input graph is used to construct the engine, forming a multi-level network connection graph G. topo =(V node E link ), where V node E represents the set of vertices of a multi-level logistics node. link The table contains the set of directed edges connecting the nodes.

[0035] Synchronously extract the order feature data corresponding to each order from the logistics order information; and set the latest delivery time window parameter T for each order. win As a time-constraint factor, the rated load capacity parameter C of the special vehicle is used. cap As a capacity constraint factor, it is explicitly mapped and written into the network connection graph G in the form of attribute feature vectors. topo The corresponding vertices V in node In this process, a high-dimensional feature-interwoven graph topological association coding space Ω is generated within the system. code .

[0036] In the graph topology association coding space Ω code In this module, a graph structure encoding operator is used to generate a dense graph topology association encoding vector for each logistics object to be scheduled, based on the neighborhood connection topology of each node and the fused node attribute features. The preset feature vector of the logistics object to be scheduled, i, is v. i The feature vector of the logistics object j to be scheduled is v j The system uses the following cosine similarity equation to calculate the spatiotemporal similarity S between the two. sim (i,j): , S sim (i,j): Represents the similarity value of the spatiotemporal relationship between logistics object i and logistics object j to be scheduled; v i ·v j : Represents the inner product of two eigenvectors; ||v i ·v j ||: These represent the magnitudes of the corresponding vectors.

[0037] The system calculates the similarity S. sim Construct a spatiotemporal relationship graph between the logistics objects to be scheduled (i,j); iterate through all pairs of logistics objects to be scheduled if and only if the similarity S sim (i,j) is greater than the preset spatiotemporal correlation threshold constant S threshold In this way, connecting edges are established between corresponding nodes in the graph topology, thereby accurately determining the objective spatiotemporal relationships.

[0038] 2. Self-organized aggregation of association domains under adaptive boundary constraints After obtaining the graph topological association encoding vector, the module uses it as the feature input vector F of the clustering feature space. cluster In order to introduce external rigid operational management boundaries, the system retrieves the specific delivery time constraint boundaries B for each food category. time Boundary of Collaboration with Logistics Functions B coop Full-scale cascaded cross-compilation generates the multidimensional constraint boundary matrix M. bound .

[0039] In the multidimensional constraint boundary matrix M bound Within the defined solution space, the algorithm performs collaborative agglomerative clustering iterations on the logistics objects to be scheduled. The specific clustering aggregation and boundary dynamic update logic is as follows: First, the system labels the current degree of collaborative aggregation as AG. n It is defined as the pairwise similarity S among all the logistics objects to be scheduled within the same local cluster in the nth round of clustering iteration. sim The mathematical expectation value. The algorithm sets the initial clustering aggregation degree threshold to θ. n During the iterative process, the system continuously monitors the current AG. n ; When the iteration reaches the nth round and the synergistic aggregation degree index satisfies AG n ≥θ n When this occurs, the control flow state machine is interrupted and clustering aggregation is stopped; the system will then satisfy the adaptive aggregation degree threshold θ. n A series of logistics objects to be scheduled are aggregated into an independent logistics collaboration domain D. coop .

[0040] To prevent clustering results from being too dense or too sparse due to a fixed threshold, the controller dynamically adjusts the aggregation degree threshold adaptively during iterative execution based on the centripetal variance convergence trend of the current partitioning results. This adaptive threshold adjustment process follows the following recursive equation: θ n+1 =θ n +k adj ·(AG n -AG n-1 ), where θ n θ represents the current adaptive aggregation degree threshold of the system instantiation during the nth round of clustering iteration; n+1 : Represents the updated adaptive aggregation degree threshold boundary used for the next round of clustering decisions; k adj : Represents the preset adaptive fine-tuning scaling coefficient, used to adjust the response sensitivity of the threshold; AG n -AG n-1 : Represents the rate of change of the degree of cohesion within a local agglomeration cluster between two adjacent iterations.

[0041] Until the difference in the aggregation degree index between two adjacent iterations is less than the preset small convergence limit value. ,Right now: When the preset stability condition is met, the system stops iterating and outputs the latched multiple logistics collaborative association domains D. coop At the same time, establish various logistics collaboration domains D coop The topological association mapping relationship between them.

[0042] Example 2: Dynamic Scheduling Perturbation Propagation Chain Construction and Priority Self-Update Algorithm In this embodiment, when a sudden disturbance occurs at a node in the supply chain network, the management decision logic involves network impedance propagation and dynamic priority adjustment: 1. Generation of scheduling perturbation propagation chain under gradient iterative diffusion The disturbance propagation control module continuously monitors the overall network status. When a specific logistics collaboration domain D... coop Due to sudden anomalies, scheduling deviations occur (the spatiotemporal residual variability is denoted as D). error When the module performs a reverse topology tracing based on the association mapping relationship established in the previous stage, it locks all downstream association domains affected by it.

[0043] The system is based on the logistics collaboration domain D coop Dynamic weight W of inter-edge connections ijCalculating the propagation gradient of the scheduling deviation; in order to accurately simulate the absorption and cascading amplification effect of disturbances in a complex multi-level network, the module introduces the attenuation rate parameter d of the scheduling deviation (with a rigid constraint in the range of 0<d<1) and the dynamic amplification factor α caused by the timeliness competition of food flow, and performs the cascading iterative solution of the propagation gradient as follows: , where : represents the scalar value of the disturbance propagation gradient updated after the current (k+1)-th iteration, which is used to quantify the marginal diffusion intensity of the unit disturbance; : represents the scalar value of the disturbance propagation gradient from the upstream associated domain i to the downstream associated domain j during the k-th network iteration; W ij : represents the dynamic topological weight value of the connecting edge between the entities of the upstream associated domain i and the downstream associated domain j in the current association mapping relationship; d: represents the inherent natural disturbance attenuation coefficient of the network, which is used to characterize the absorption overhead of the buffer zone on the disturbance; α: represents the cascading dynamic amplification factor, which is used to characterize the capacity seizing amplification effect caused by the deterioration of food transportation timeliness.

[0044] After receiving the propagation gradient feedback output in the current round, the module reversely fine-tunes the weight W of the association edge between the logistics collaborative associated domains ij ; when the gradient variation of each association edge in the entire network in two adjacent iterations satisfies the partial norm convergence criterion, that is, the absolute value of the maximum gradient deviation is less than the preset convergence threshold constant | - (where is the preset convergence threshold constant), the iteration is terminated; at this time, the system takes the logistics collaborative associated domain D that generates the scheduling deviation coop as the disturbance source, and generates a global scheduling disturbance conduction chain L along the directed connection topological path of the propagation gradient dist .

[0045] 2. Dynamic priority update based on accumulated disturbance Along the forward conduction path of the generated scheduling disturbance conduction chain L dist , the module performs sequential integral accumulation step by step on each logistics collaborative associated domain D on the chain coop , and solves the accumulated disturbance intensity value I borne by each associated domain at the current moment cum : , where G m : represents the disturbance propagation gradient of the m-th affected associated node on the scheduling disturbance conduction chain; ∆t: Represents the time step window for the continuous accumulation of disturbance. The disturbance propagation control module will calculate the cumulative disturbance intensity value I. cum The inherent tolerance threshold I configured within the corresponding associated domain th Perform a comparison; if I is satisfied cum >I th The condition for determining the opposite indicates that the local delay has crossed the system's tolerable safety boundary, triggering an incremental adjustment of the system's execution priority, and determining the corresponding priority increase increment ∆P=β·(I) based on the excess ratio. cum -I th (where β is a preset dynamic adjustment sensitivity coefficient), the original scheduling priority P of the associated domain is... old Corrected and adjusted to P new =P old +∆P.

[0046] Simultaneously, for the logistics collaboration domain D whose scheduling status has returned to normal... coop The system initiates a priority smooth rollback mechanism, gradually reducing the scheduling priority according to the preset disturbance release rate until it seamlessly rolls back to the original preset level, ensuring the self-healing allocation of network resource capacity.

[0047] Example 3: Cross-regional link reconstruction simulation and multi-objective game timeliness evaluation model In this specific embodiment, an active reconfiguration scheme is used to reorganize the network of high-risk links to be adjusted, and a multi-objective model is used to output the optimal scheduling decision.

[0048] 1. Calculation of link replacement margin and precise location of links to be adjusted The link reconfiguration decision module first performs a full network scan to filter and select the current scheduling priority P. new Higher than the preset priority threshold P limit Target logistics collaborative domain D coop Within this associated domain, the module traverses and scans all currently executing core scheduling links, calculating the link substitution margin M for each scheduling link based on the following multivariate algebraic equation. sub : , M sub This represents the feasibility and resilience margin (i.e., link replacement margin) of replacing the current scheduled link with an alternative link in the logistics network. N path : Represents the absolute number of available physical parallel alternative paths connecting the same start and end nodes in the current spatiotemporal topology graph; C switch : Represents the time and space organization costs and scheduling expenses required to perform emergency vehicle allocation, transit replacement or delivery hub switching; ∆Gdown This represents the incremental disturbance propagation in downstream adjacent domains caused by the competition for transportation resources after the link is switched. w1, w2, w3: Represents the system's preset multivariate scaling weight coefficients used to balance the relationship between the number of paths, switching costs, and downstream sweep surface.

[0049] When the link replacement margin M of a certain scheduling link sub Below the preset margin threshold M th If the link has lost its resilience against disturbances, it is marked as a link to be adjusted.

[0050] 2. Generation of scheduling and reconfiguration sequences under revenue-driven iteration Based on the identified start and end points of the links to be adjusted, the module accesses the global backup capacity pool and backup route database to retrieve and extract a set of all candidate links that meet the requirements of capacity matching level (such as having cold chain temperature control qualifications) and path connectivity. Subsequently, the system activates the collaborative link reconstruction simulation algorithm.

[0051] This simulation employs a reward-driven iterative mechanism, and its specific algorithmic control flow and probabilistic evolution path are implemented as follows: First, the module initializes the candidate link set and assigns the same initial selection probability to each candidate link in the set in the weight distribution. In subsequent iterations, the algorithm uses the link continuity improvement parameter ∆L (characterizing the reduction of path inflection points and breakpoints) and the node load balancing improvement parameter ∆B (characterizing the reduction of hub queuing) generated by the reconstructed network topology to calculate the link reconstruction benefit R of the current round by weighted summation. gain : , Where α and β are preset weighting coefficients for the comprehensive score allocation, and the system assigns weights based on this return R. gain Dynamically update the weight distribution and selection probability distribution P of each candidate link in the overall candidate link weight distribution. r Specifically, the link reconstruction benefit R calculated in the current round is... gain In the feedback adjustment factor input probability update module, the selection probability of candidate links with increased returns is increased proportionally, while the selection probability of candidate links with decreased returns is decreased simultaneously.

[0052] After the probability distribution is updated, the system performs a reconstruction state determination: the link reconstruction benefit R calculated for each candidate link in the current nth round is used. gain Compared with the system's preset retention threshold R s Perform real-time comparison; When the link reconstruction benefit R is satisfied gainHigher than or equal to the preset retention threshold R s When the condition is met, the execution state of the candidate link is preserved and locked in the current carrier path topology; When the link reconstruction benefit R is satisfied gain Below the preset retention threshold R s When the condition is met, a reorganization and replacement operation is triggered, removing the low-yield link from the current path and adjusting it according to the updated selection probability distribution P. r Random sampling and matching are performed again from the backup route library to introduce new links and complete the recombination.

[0053] After completing the current nth round of reorganization, the system calculates the total network reconstruction benefit R for the current nth round. total (n) The total network reconstruction revenue R of the adjacent n-1 rounds prior. total The absolute value of the difference between (n-1) is taken as the improvement in iterative profit ∆R. total .

[0054] The system continuously stream monitors the improvement in iterative returns, ∆R. total The convergence trajectory, when the iterative gain ∆R increases after multiple consecutive iterations. total Less than the preset convergence threshold When the system determines that the overall network reconstruction benefit has reached a saturated and stable state, the reconstruction simulation terminates. At this point, the state machine sequentially encapsulates the currently retained candidate links, generating multiple sets of candidate scheduling and reconstruction sequences S. seq .

[0055] 3. Establishment and Solution of a Hierarchical, Layered, Progressive Game Evaluation Model To select the plan with the highest overall network efficiency from multiple scheduling and reconfiguration sequences, the system establishes a hierarchical progressive game evaluation model that includes a basic evaluation layer, a constraint correction layer, and a collaborative game layer. Its arithmetic pipeline hierarchical architecture is as follows: Basic evaluation layer: Input a specific scheduling and reconfiguration sequence, and directly calculate the underlying link continuity basic score S. L Basic score for node load balancing S B The two-dimensional scoring matrix results are used as state factors input to the constraint correction layer.

[0056] Constraint Correction Layer: Receives the output from the basic evaluation layer and performs compliance checks on the estimated arrival times of each logistics node using the path time window constraint matrix to obtain the expected time window deviation value D. tw The expected time window deviation value D tw As a nonlinear correction factor, a penalty correction is applied to the output of the basic evaluation layer, and the corrected high-confidence score set S is calculated. L 'and SB Its modified model is as follows: S L '=S L ·exp(-λ1·D tw ), S B =max(S) B -λ2·D tw ,0), In the formula, exp represents the exponential function operator with the natural logarithm as the base. Here, it is used to construct a nonlinear acceleration penalty mechanism, so that the link score collapses exponentially with the increase of delay time, in order to simulate the physical property characteristic that "once there is a serious timeout, the continuous relay value of the entire supply chain becomes completely invalid".

[0057] λ1 (first time-sensitivity penalty constant): represents the time-sensitivity adjustment coefficient for link continuity, used to control the exponential decay rate of the expected time window deviation on the link continuity score; λ2 (Second Time-Sensitivity Penalty Constant): Represents the time-sensitivity adjustment coefficient for node load balancing. It is used to determine the linear deduction step size of the node load balancing score per unit time window deviation, in order to simulate the characteristic that "the queuing pressure of transit nodes increases uniformly for each unit delay time". The maximum value function `max` represents the zero-value lower bound truncation operator, used to apply rigid boundary constraints to the linear subtraction result, ensuring the corrected node load balancing score `S`. B 'It remains non-negative in scenarios with large delays, ensuring the mathematical stability of convergence calculation in subsequent multi-objective collaborative games.'

[0058] Specifically, the constraint correction layer employs differentiated correction strategies based on the different business characteristics of the two types of basic indicators: Link continuity is a highly sensitive indicator: if the time window deviation exceeds the critical value, it will directly lead to link disconnection. Therefore, an exponential decay nonlinear penalty is adopted. Node load balancing is a weakly sensitive indicator: timeouts will only gradually increase the backlog of pressure on nodes and will not directly cause the balance to collapse. Therefore, a gradual penalty with linear deduction is adopted.

[0059] Collaborative game layer: Receives the entire score set corrected by the constraint correction layer; sets the expected time window deviation value D... tw Corrected link continuity score S L ', Corrected node load balancing score S B 'Benefits R from link reconstruction derived from the previous stage' gain The unified input multi-objective cooperative game convergence equation is used to perform multi-objective cooperative convergence calculation, and finally outputs a unique global integrated scheduling efficiency value E. total : E total =μ1·SL +μ2·S B '-μ3·D tw +μ4·R gain , μ1, μ2, μ3, μ4: These represent the weighted game coefficients corresponding to each core element. The system ultimately selects E. total The best-performing set of scheduling reconstruction sequences will be used as the final logistics collaborative scheduling scheme to be issued and executed.

[0060] Example 4: Tracking the Evolutionary Trajectory of In-Transit Scheduling and Collaborative Tracing of Multi-Source Data Chains In this embodiment, the streaming tracking feedback after the scheme is executed, and the non-encrypted transparent management process that rigidly intertwines macro-scheduling events with micro-level food batch flow, are as follows: 1. Construction of scheduling evolution trajectory under bidirectional spatiotemporal alignment mapping The scheduling evolution tracing module maintains streaming communication with each execution terminal. During the implementation of the logistics collaborative scheduling scheme, the task receipt timestamps t of each key logistics node in the supply chain are collected in real time in a distributed manner. sign The remaining transport capacity C of vehicles en route rem And the available loading and unloading capacity A of the node physical forward warehouse load As a core basic indicator representing the status of task completion and resource changes.

[0061] The system introduces a globally unified timeline T abs ; with this absolute time axis T abs As a rigid reference, the multidimensional discrete state output by each logistics node at each time stamp is correlated with the corresponding logistics collaborative association domain D. coop The topology changes are aligned with the time-series dimension alignment mapping, thereby connecting and generating the scheduling evolution trajectory that represents the dynamic execution of the scheduling scheme throughout its entire lifecycle.

[0062] 2. Closed-loop feedback correction mechanism Simultaneously with generating the scheduling evolution trajectory, the module performs online deviation analysis on the trajectory to obtain the correspondence between disturbances and adjustments; once the spatiotemporal residual D between the actual execution trajectory and the theoretical collaborative scheduling scheme is captured... error If the error exceeds the stability boundary, it indicates a change in the network topology; the system immediately uses this residual signal as the error input source for feedback correction, and fine-tunes the edge weights w of the adjacency matrix in the aforementioned embodiments. ij Directly correct and reset the logistics collaboration association domain D coop Coverage boundary and refresh scheduling disturbance propagation chain L dist This achieves a complete closed loop between control flow and state flow.

[0063] 3. Association and encapsulation of supply chain logistics collaborative scheduling and traceability data results In the end-to-end storage and traceability generation process, the system abandons blockchain and cryptographic encryption logic, focusing entirely on the data flow binding of core decision-making data. The system assigns a unique batch code ID to each food product to be traced. batch As a globally unique primary key, and using this primary key as the core index anchor point, a mapping index is established in the database between logistics batch flow information and scheduling records of various historical stages.

[0064] Subsequently, the module performs multi-source data fusion and encapsulation: the dynamic scheduling evolution trajectory, rich in spatiotemporal topological change information and used as process record data, is checked and associated with the mapping index and encapsulated. For each stage of generated composite data, the system sequentially embeds the data integrity check value V from the previous historical stage. last The final data integrity check value V at the current stage check The calculation is performed following the data integrity verification function equation: V check =f check (Data current +V last +ID batch ), Where V check : Represents the latest data integrity verification value generated in the current stage, used to ensure the logical credibility and integrity of the entire link scheduling traceability chain; Data current : Represents the current stage's core input scheduling evolution trajectory and mapping index data stream; V last : Represents the check value passed in the previous historical stage; f check : Represents the standard data integrity verification function. This verification function is only used for internal integrity verification of the source dataset and does not participate in the core calculation of scheduling optimization.

[0065] Through the above calculations, the system ultimately associates and encapsulates the scheduling evolution trajectory as process data with the mapping index, forming a traceable batch flow and scheduling decision association index, and organizes and generates a continuous scheduling traceability chain with strong verification relationships, thereby directly outputting the supply chain logistics collaborative scheduling traceability data results.

[0066] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A logistics collaborative scheduling and management method for food supply chain traceability, characterized in that, Includes the following steps: S1: Obtain the link structure information and logistics order information based on the food category supply chain, perform association coding calculation on the link structure information and logistics order information, identify the logistics objects to be scheduled, and perform association domain self-organization aggregation on the logistics objects to be scheduled to form logistics collaborative association domains, while establishing association mapping relationships between each logistics collaborative association domain; S2: Calculate the disturbance propagation gradient between each logistics collaboration domain based on the aforementioned association mapping relationship, form a scheduling disturbance transmission chain, and update the scheduling priority of each logistics collaboration domain in real time; S3: Identify the links to be adjusted within the logistics collaboration domain through the scheduling priority, perform collaborative link reconstruction deduction on candidate links that meet the substitution conditions and calculate the link reconstruction benefits, and generate multiple sets of scheduling reconstruction sequences; By combining the collaborative scheduling timeliness joint evaluation model, the overall scheduling efficiency value of each scheduling reconstruction sequence is calculated, and a logistics collaborative scheduling scheme is generated. S4: Obtain execution feedback of the logistics collaborative scheduling scheme, track the task completion status and resource change status of each logistics node, form a scheduling evolution trajectory and correct the logistics collaborative correlation domain and scheduling disturbance transmission chain; associate and store the logistics batch flow information of the traceable food with the scheduling records of each stage, and generate supply chain logistics collaborative scheduling and traceability results.

2. The method according to claim 1, characterized in that, The association coding calculation specifically includes: Extract the topological adjacency matrix of the logistics nodes in the link structure information, and construct a node connection graph based on the topological adjacency matrix; extract the order timeliness window and capacity constraint vector in the logistics order information, and write the order timeliness window and capacity constraint vector as node attributes into each node in the node connection graph to form a graph topological association coding space; In the graph topology association coding space, graph topology association coding vectors for each logistics object to be scheduled are generated based on the node attributes and connection relationships of each node. A spatiotemporal association graph between the logistics objects to be scheduled is constructed based on the similarity between the graph topology association coding vectors, and the spatiotemporal association relationship is determined.

3. The method according to claim 1, characterized in that, The self-organizing aggregation of related domains specifically includes: Using the graph topology association encoding vector as the clustering input feature, and combining the delivery time constraint boundary and the logistics function coordination boundary extracted from logistics order information, a multi-dimensional constraint boundary is constructed. Within the multidimensional constraint boundary, collaborative aggregation is performed on the logistics objects to be scheduled. Aggregation stops when the aggregation degree reaches a preset aggregation degree threshold. The logistics objects to be scheduled that meet the preset aggregation degree threshold are aggregated into a logistics collaborative association domain. The aggregation degree threshold is adaptively adjusted according to the aggregation change trend until the aggregation result meets the preset stability condition.

4. The method according to claim 1, characterized in that, The formation of the scheduling disturbance propagation chain specifically includes: When the logistics collaboration association domain generates a scheduling deviation, the affected downstream association domain is determined according to the association mapping relationship; the propagation gradient of the scheduling deviation is calculated according to the weight of the association edges between the logistics collaboration association domains, and the propagation gradient is iteratively calculated by combining the attenuation rate of the scheduling deviation and the dynamic amplification coefficient. Obtain the propagation gradient feedback, and correct the weights of the association edges between logistics collaboration association domains in the association mapping relationship based on the propagation gradient feedback; The iteration terminates when the change in the propagation gradient is less than the preset convergence threshold. The logistics collaboration correlation domain that generates the scheduling deviation is taken as the disturbance source, and the scheduling disturbance propagation chain is formed along the directed connection relationship of the propagation gradient.

5. The method according to claim 1, characterized in that, The identification process of the link to be adjusted specifically includes: Select target logistics collaboration association domains whose scheduling priority is higher than a preset priority threshold; traverse the current scheduling links within the target logistics collaboration association domains and calculate the link substitution margin for each scheduling link; The scheduled links whose link replacement margin is lower than a preset margin threshold are marked as the links to be adjusted.

6. The method according to claim 1, characterized in that, The collaborative link reconstruction simulation specifically includes: Based on the starting point and ending point of the link to be adjusted, candidate links that meet the requirements of capacity level and path connectivity are retrieved from the global backup capacity pool and backup route library. The candidate links are subjected to a benefit-driven iterative deduction using a graph topology collaborative reconstruction deduction algorithm. In each round of deduction, the selection probability of each candidate link is updated according to the link reconstruction benefit. When the improvement in iteration benefits is less than a preset convergence threshold, the deduction is terminated, and the currently retained candidate links are taken as the scheduling and reconstruction sequence; the improvement in link continuity and node load balancing of each scheduling and reconstruction sequence are calculated as the link reconstruction benefits.

7. The method according to claim 1, characterized in that, The collaborative scheduling timeliness joint evaluation model adopts a hierarchical progressive game evaluation architecture, including: The basic evaluation layer calculates link continuity score and node load balancing score, and outputs the calculation results to the constraint correction layer. The constraint correction layer receives the output of the basic evaluation layer and performs compliance verification on the expected arrival time of each logistics node through the path time window constraint matrix to obtain the expected time window deviation value. The expected time window deviation value is used as a correction factor to correct the output of the basic evaluation layer. The collaborative game layer receives the output of the constraint correction layer, performs multi-objective collaborative game convergence calculation on the expected time window deviation value, the corrected link continuity score, the corrected node load balancing score and the link reconstruction benefit, and outputs the overall scheduling efficiency value.

8. The method according to claim 1, characterized in that, The formation of the scheduling evolution trajectory specifically includes: The task receipt timestamps, remaining transport capacity of transport vehicles, and available loading and unloading capacity of each logistics node are collected in real time as the task completion status and resource change status. Using absolute time as the axis, the task completion status and resource change status of each logistics node at each timestamp are aligned and mapped with the corresponding logistics collaboration association domain in a time-series dimension to generate a scheduling evolution trajectory that represents the dynamic execution process of the scheduling scheme.

9. The method according to claim 1, characterized in that, The associated storage and generation of tracing results specifically include: Using the logistics batch transfer information of the food to be traced as the primary key, a mapping index is established between the logistics batch transfer information and the scheduling records of each stage. The scheduling evolution trajectory is used as process data and associated with the mapping index to form a traceable batch flow and scheduling decision association index, generating supply chain logistics collaborative scheduling traceability data results.

10. A logistics collaborative scheduling and management system for food supply chain traceability, characterized in that: A method for performing the method of claims 1-9, comprising: The collaborative association domain construction module acquires and parses the link structure information and logistics order information based on the food category supply chain, performs association coding calculation to identify the logistics objects to be scheduled, performs association domain self-organization aggregation on the logistics objects to be scheduled, and generates logistics collaborative association domains and association mapping relationships. The disturbance propagation control module calculates the disturbance propagation gradient based on the associated mapping relationship, generates a scheduling disturbance propagation chain based on the directed connection relationship of the disturbance propagation gradient, and dynamically updates the scheduling priority of each logistics collaboration associated domain. The link reconfiguration decision module identifies links to be adjusted based on the scheduling priority, performs iterative deduction driven by benefits on candidate links and calculates the benefits of link reconfiguration, generates a scheduling reconfiguration sequence, calculates the overall scheduling efficiency value of each scheduling reconfiguration sequence through a collaborative scheduling efficiency joint evaluation model, and selects the scheduling reconfiguration sequence with the optimal overall scheduling efficiency value as the logistics collaborative scheduling scheme. The scheduling evolution traceability module obtains execution feedback and generates scheduling evolution trajectory, corrects the logistics collaboration association domain and scheduling disturbance transmission chain, associates and encapsulates the logistics batch flow information of the food to be traced with the scheduling records of each stage, and generates supply chain logistics collaborative scheduling traceability data results.

Images