Supply chain collaborative scheduling management method and system

Abstract

The invention discloses a supply chain collaborative scheduling management method and system, and relates to the technical field of data management. The method comprises the following steps: in response to a received supply chain scheduling demand, constructing a dynamic virtual resource pool based on schedulable resources of each supply chain node, obtaining a digital mirror image and an intelligent cooperation carrier of each supply chain node on a supply chain, constructing a joint intelligence wedge, carrying out collaborative decision game training on the joint intelligence wedge through a collaborative gain co-molding device, and obtaining a joint intelligence model; the method comprises the following steps: establishing a co-evolutionary co-frequency joint intelligence wedge group, searching a candidate co-frequency joint intelligence wedge group with matched resource granularity from the co-frequency joint intelligence wedge group based on a dynamic virtual resource pool, establishing a temporary co-scheduling chain, dynamically executing a co-scheduling task, and obtaining an execution result of each co-frequency joint intelligence wedge. And reconstructing the same-frequency joint intelligence wedge group to execute the scheduling task. According to the method, the overall operation efficiency, the anti-risk capability and the sustainable development capability of the supply chain can be comprehensively improved, and powerful support is provided for efficient, agile and stable operation of the supply chain.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data management, and in particular to a supply chain collaborative scheduling management method and system. BACKGROUND

[0002] With the globalization and fine division of supply chain, the collaborative dependence between various participants in the supply chain (such as suppliers, manufacturers, logistics companies, and retailers) has significantly increased. However, existing collaborative scheduling solutions still face many core difficulties and are difficult to adapt to complex and dynamic supply chain environments. First, the conflict of interest collaboration is prominent. Each participant mainly aims to maximize their own interests. For example, suppliers prefer to prioritize high-margin order production, and logistics companies tend to choose short-distance transportation tasks with low empty rate, leading to a conflict between local optimization and overall supply chain efficiency improvement. Although traditional centralized scheduling systems attempt to coordinate the overall situation, they fail to form a long-term stable collaborative relationship due to the neglect of the autonomous decision-making willingness and interest bottom line of each participant. Second, the collaborative structure is not rigid enough. Existing collaborative models are mostly based on fixed protocols (such as long-term VMI inventory protocols and fixed logistics cooperation relationships). When facing sudden disturbances (such as a sharp increase in order quantity, supplier disruption, or regional logistics disruption), they cannot quickly reorganize resources and collaboration relationships, resulting in significant response lag and causing order delays, inventory accumulation, or resource waste. Third, the cost of integrating new participants into the supply chain is high. New participants (such as new regional suppliers and logistics companies) need to adapt to existing collaborative rules from scratch, and there is a lack of standardized rapid deployment and training mechanisms, resulting in a long integration period and affecting the overall supply chain efficiency due to insufficient collaboration capabilities. In addition, there is a lack of dynamic optimization mechanisms throughout the cycle. Existing solutions mainly focus on the execution of single scheduling tasks and lack a collaborative effect monitoring and participant credit evaluation system based on real-time data, making it difficult to optimize subsequent collaborative strategies based on historical collaboration performance. This leads to a gradual decline in collaborative efficiency and a weakening of risk response capabilities in the long-term operation of the supply chain. These problems collectively result in the existing supply chain collaborative scheduling having shortcomings in efficiency, flexibility, and sustainability, which cannot meet the current demand for efficient, agile, and resilient supply chains. SUMMARY

[0003] The present application provides a supply chain collaborative scheduling management method, comprising: Step S1, in response to receiving a supply chain scheduling demand, constructing a dynamic virtual resource pool composed of schedulable resource particles based on the schedulable resources of each supply chain node; Step S2, obtaining the digital mirror and intelligent collaboration carrier of each supply chain node on the supply chain, and constructing a smart contract; Step S3, training the constructed smart contract through a collaborative gain co-shaper for collaborative decision-making game, and constructing a group of smart contracts with the same frequency of collaborative evolution; Step S4, find resource granularity matching candidate same frequency associated intelligence contract groups from the same frequency associated intelligence contract groups based on the dynamic virtual resource pool, construct a temporary collaborative scheduling chain, and dynamically execute collaborative scheduling tasks by each candidate same frequency associated intelligence contract group; Step S5, obtain the execution results of each same frequency associated intelligence contract group, and monitor the real-time updated associated intelligence contract group. According to the execution results and the updated associated intelligence contract group, return to step S2 to re-construct the same frequency associated intelligence contract group to execute the scheduling task.

[0004] The supply chain collaborative scheduling management method as described above, wherein the constructed associated intelligence contract group is trained by the collaborative gain co-molder for collaborative decision game, and the collaborative evolution same frequency associated intelligence contract group is constructed, including the following sub-steps: Step S31, set initial cooperative benefit credit for each associated intelligence contract group, and define initial self-operation benefit guide function; Step S32, based on the initial cooperative benefit credit and the initial self-operation benefit guide function, organize all associated intelligence contract groups to perform multiple rounds of collaborative game training through the collaborative gain co-molder, and construct the collaborative evolution same frequency associated intelligence contract group.

[0005] The supply chain collaborative scheduling management method as described above, wherein the constructed associated intelligence contract group is trained by the collaborative gain co-molder for collaborative decision game, and the collaborative evolution same frequency associated intelligence contract group is constructed, including the following sub-steps: Step S321, generate a supply chain simulation scene based on the full-link historical scheduling data of the supply chain and the preset disturbance through the collaborative gain co-molder; Step S322, based on the state of each associated intelligence contract group and the initial self-operation benefit guide function, each associated intelligence contract group makes autonomous decision in the supply chain simulation scene, and simulates scheduling execution through the collaborative gain co-molder to generate scheduling execution completion results; Step S323, based on the scheduling execution completion results, quantize the contribution share of each participating associated intelligence contract group to the global gain in this round of training through the collaborative gain co-molder algorithm, and upgrade the self-operation benefit guide function of each associated intelligence contract group to the cooperative benefit guide function, and update the cooperative benefit credit; Step S324, obtain the cooperation acceptance rate and global optimization target achievement rate of each associated intelligence contract group in each round of training, terminate the training based on the preset threshold, and generate the collaborative evolution same frequency associated intelligence contract group.

[0006] The supply chain collaborative scheduling management method as described above, wherein the constructed associated intelligence contract group is trained by the collaborative gain co-molder for collaborative decision game, and the collaborative evolution same frequency associated intelligence contract group is constructed, including the following sub-steps: Step S41, real-time acquisition of supply chain scheduling requirements, and parsing into demand combination characteristics of specific matching resource particles; Step S42, based on the demand combination characteristics, matching and binding the candidate intelligent contract bodies with the corresponding matching resource particles and the optimal sequence guide level in the same frequency intelligent contract body group; Step S43, combining the candidate intelligent contract bodies and calculating the collaborative scheduling values of each combination, selecting the scheduling execution intelligent contract body to form a temporary collaborative scheduling chain, and locking the corresponding matching resource particles in the dynamic virtual resource pool to dynamically execute the collaborative scheduling task.

[0007] The supply chain collaborative scheduling management method described above, wherein the execution results of each same frequency intelligent contract body are acquired, and the real-time updated intelligent contract bodies are monitored, and according to the execution results and the updated intelligent contract bodies, the candidate same frequency intelligent contract body group is returned to execute the scheduling task, including the following sub-steps: Step S51, in the collaborative scheduling task execution process, the execution results of each same frequency intelligent contract body are acquired in real time, the intelligent contract holographic double-layer monitoring board is built, and the collaborative execution performance archives are established and updated for each intelligent contract body in combination with the cooperation benefit credibility; Step S52, based on the collaborative execution performance archives of each intelligent contract body, the key optimization indicators are extracted, the parameters of the gain co-molding algorithm are updated, and the same frequency intelligent contract body group is reconstructed.

[0008] The application also provides a supply chain collaborative scheduling management system, comprising: A dynamic virtual resource pool construction module, which, in response to receiving a supply chain scheduling requirement, constructs a dynamic virtual resource pool composed of matching resource particles based on the schedulable resources of each supply chain node; An intelligent contract body construction module, which acquires the digital images and intelligent collaboration carriers of each supply chain node on the supply chain, and constructs intelligent contract bodies; A same frequency intelligent contract body group construction module, which trains the constructed intelligent contract bodies through a collaborative gain co-molder to construct a collaborative evolution same frequency intelligent contract body group; A collaborative scheduling task execution module, which finds a candidate same frequency intelligent contract body group with resource granularity matching from the same frequency intelligent contract body group based on the dynamic virtual resource pool, constructs a temporary collaborative scheduling chain, and dynamically executes a collaborative scheduling task by each candidate same frequency intelligent contract body; A same frequency intelligent contract body group reconstruction module, which acquires the execution results of each same frequency intelligent contract body, and monitors the real-time updated intelligent contract bodies, and according to the execution results and the updated intelligent contract bodies, returns the execution intelligent contract body construction module to reconstruct the same frequency intelligent contract body group to execute the scheduling task.

[0009] The supply chain collaborative scheduling management system described above, wherein the same frequency intelligent contract body group construction module specifically comprises: The association contract initialization setting submodule sets initial mutual benefit credibility for each association contract and defines an initial self-operated benefit guide function; The co-evolution synchronous association contract group construction submodule organizes all association contracts through a co-gain co-molder for multiple rounds of co-game training based on the initial mutual benefit credibility and the initial self-operated benefit guide function, to construct a co-evolution synchronous association contract group.

[0010] The supply chain collaborative scheduling management system as described above, wherein the co-evolution synchronous association contract group construction submodule specifically comprises: The supply chain simulation scenario generation submodule generates a supply chain simulation scenario based on the full-link historical scheduling data of the supply chain and a preset disturbance through the co-gain co-molder; The scheduling execution completion result generation submodule generates a scheduling execution completion result based on the self state of each association contract and the initial self-operated benefit guide function, in which each association contract makes autonomous decisions in the supply chain simulation scenario and simulates scheduling execution through the co-gain co-molder; The association contract parameter updating submodule quantifies the contribution share of each participating association contract to the global gain in the current round of training based on the scheduling execution completion result through the co-gain co-molder, remolds the self-operated benefit guide function of each association contract into a mutual benefit co-guide function, and updates the mutual benefit credibility; The training termination submodule obtains the cooperation acceptance rate of each association contract and the global optimization target achievement rate in each round of training, terminates the training based on a preset threshold, and generates a co-evolution synchronous association contract group.

[0011] The supply chain collaborative scheduling management system as described above, wherein the collaborative scheduling task execution module specifically comprises: The demand combination characteristic acquisition submodule acquires supply chain scheduling demands in real time and parses them into demand combination characteristics of specific matching resource particles; The candidate association contract selection submodule matches and binds the candidate association contract with the corresponding matching resource particle and the optimal benefit sequence guide level in the synchronous association contract group based on the demand combination characteristics; The scheduling task execution submodule combines the candidate association contract, calculates the collaborative scheduling value of each combination, selects a scheduling execution association contract for series connection, forms a temporary collaborative scheduling chain, locks the corresponding matching resource particle in the dynamic virtual resource pool, and dynamically executes the collaborative scheduling task.

[0012] The supply chain collaborative scheduling management system as described above, wherein the synchronous association contract group reconstruction module specifically comprises: The execution effect monitoring submodule acquires the execution result of each synchronous association contract in real time during the collaborative scheduling task execution process, builds an association contract holographic double-layer monitoring board, and establishes and updates the collaborative execution performance archives of each association contract in combination with the mutual benefit credibility. The Synchronous Intelligence Co-operation Group Reconstruction Submodule extracts key optimization indicators based on the collaborative execution performance profiles of each Synchronous Intelligence Co-operation Group, updates the parameters of the gain co-shaping algorithm, and reconstructs the Synchronous Intelligence Co-operation Group.

[0013] The beneficial effects achieved by this invention are as follows: This invention enables precise, granular management and dynamic integration of resources, effectively improving resource utilization efficiency. Simultaneously, it promotes a shift from isolated decision-making to collaborative decision-making among participating entities, reducing scheduling conflicts, enhancing the scientific rigor and efficiency of overall decision-making, and ensuring the synchronization and stability of supply chain collaboration. For real-time supply chain scheduling needs, it can construct temporary collaborative scheduling chains, enabling dynamic adaptation and rapid execution of scheduling tasks, flexibly responding to changes in supply chain demand. Furthermore, it allows for real-time monitoring of the scheduling process across all dimensions, regulating entity behavior through credit constraints, and feeding back into the continuous optimization of the synchronized intelligent supply chain, forming a closed-loop management system of "monitoring-evaluation-optimization," thus improving scheduling reliability. At the same time, it lowers the entry threshold and cycle time for new entities, enhances the scalability and adaptability of the supply chain system, ensures that the supply chain dynamically adjusts with changes in participating entities, and maintains consistently high-efficiency collaborative scheduling capabilities. Ultimately, it comprehensively improves the overall operational efficiency, risk resistance, and sustainable development capabilities of the supply chain, providing strong support for the efficient, agile, and stable operation of the supply chain. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0015] Figure 1 This is a flowchart of a supply chain collaborative scheduling management method provided in Embodiment 1 of this application; Figure 2 This is a schematic diagram of a supply chain collaborative scheduling management system provided in Embodiment 2 of this application. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Example 1 like Figure 1 As shown in Embodiment 1 of this application, a supply chain collaborative scheduling management method is provided, which includes the following steps: Step S1, in response to receiving the supply chain scheduling demand, constructing a dynamic virtual resource pool composed of matching resource particles based on the schedulable resources of each supply chain node; Specifically, based on the received supply chain scheduling demand, each supply chain node outputs its schedulable resources according to its role in the supply chain and classifies them by resource type. Based on the business scenarios and scheduling accuracy requirements of each type of resource, it defines matching resource particles that include standardized core attribute formats. The core attributes of matching resource particles include identification attributes (particle UID, intelligent contract ID), resource nature attributes, space-time attributes, economic and state attributes, constraint and preference attributes, etc. The disassembled matching resource particles are associated with the corresponding supply chain nodes, and the state information (available / locked / revoked) of the matching resource particles is synchronized. Each node manages the full life cycle of the particles under its jurisdiction. After the association is completed, the pool registration is automatically triggered, the matching resource particles are sent to the dynamic virtual resource pool for integrity and uniqueness verification, and after the verification is passed, they are registered and stored. If it is not passed, it will be returned for modification. The dynamic virtual resource pool uses a NoSQL database to record the existing particle registration information and data, and builds a composite index engine based on particle attributes and states for subsequent fast retrieval. Each node regularly reports particle status to the resource pool, and if no status information is received for a long time, the corresponding particle status will be modified to unknown.

[0018] Step S2, obtaining the digital image and intelligent collaboration carrier of each supply chain node on the supply chain, and constructing an intelligent contract body; Specifically, holographic ontology cognitive data of each supply chain node is collected. This data includes core feature data across all dimensions, such as basic attribute data, operational capability data, decision-making goal data, interest demand data, collaborative interaction data, resource attribute data, risk tolerance data, and privacy protection requirements. The holographic ontology cognitive data is cleaned and standardized, unifying data formats, indicator definitions, and units of measurement, and removing outliers and redundant information. Based on the relationships between data, an ontology cognitive knowledge graph for each node is constructed. Graph neural network technology is used to extract precise holographic feature vectors representing the core features of each node from the knowledge graph. This serves as the core data foundation for analyzing the physical state, business capabilities, and collaborative interactions of supply chain nodes. Core characteristics such as preferences are digitally replicated one-to-one to construct digital mirrors of each node. Based on these digital mirrors and combined with the overall supply chain collaboration needs, core functional modules for each node are built. These include an intelligent decision-making module responsible for autonomous decision-making, a collaborative interaction module responsible for interacting with other nodes, a resource binding module responsible for resource management, and an anomaly handling module responsible for risk response. Each module is assembled according to preset logic to construct an intelligent collaboration carrier for each node. A real-time data synchronization channel is established through IoT sensors and system interfaces to dynamically receive updates on the physical status of nodes and changes in the supply chain environment, forming a smart alliance for each node. The resource particles associated with each node are bound to the corresponding smart alliance to achieve associated management of resources and alliances.

[0019] Step S3: Train the constructed intelligent entity through collaborative decision-making game using a collaborative gain co-shaper to build a co-evolving intelligent entity group with the same frequency. Furthermore, the constructed intelligent alliance is trained through collaborative decision-making game theory using a collaborative gain co-shaper to build a co-evolving, synchronous intelligent alliance group, which includes the following sub-steps: Step S31: Set the initial synergistic benefit product confidence for each intelligence alliance and define the initial self-operated benefit derivative function; Specifically, based on the historical collaboration data of the corresponding nodes of each intelligence alliance, the initial synergy product confidence quantification formula is used. Calculate the initial product confidence level, where, For the first Initial synergy product confidence of each intelligent entity The range of values ​​is , For the number of Lianzhiqi entities, This is the benchmark value for the benefit of the cooperative. For the first The number of synergistic benefit assessment indicators for each intelligent cooperation entity. The range of values ​​is , For the first The first of the joint intelligence community a confidence coefficient of the benefit accumulation credit evaluation index, is the first is the first is the first is the first is the first is the industry correlation correction index of the first is the data integrity adjustment coefficient of the first

[0020] Based on the core benefit target of each alliance contract body, the compliance index of each benefit index is defined, and the initial self-operated benefit guide function of each alliance contract body is constructed , wherein is the initial self-operated benefit guide value of the first is the number of benefit index parameters of the first The value range of is , is the benefit sensitivity coefficient of the first is the first is the first is the first is the first is the number of risk and constraint penalty parameters of the first The value range of is , is the tolerance value of the first is the first is the first is the benefit penalty value of the first is the first

[0021] Step S32, based on the initial benefit accumulation credit and the initial self-operated benefit guide function, the multi-round cooperative game training of all alliance contract bodies is organized by the cooperative gain co-molder, and the co-evolution of the same frequency alliance contract body group is constructed; Specifically, through the cooperative gain co-molder, all alliance contract bodies are repeatedly trained in the simulation scene, and the alliance contract bodies are guided to evolve from each other to win-win cooperation. The cooperative gain co-molder includes a simulation scene generation module, a disturbance control module, a global state tracking module, a game interaction module, an incremental quantization module, a rule remodeling module, a credit record module, and a training termination determination module. Each module communicates and calls data through a data interface.

[0022] ​​​​Further, based on the initial mutual benefit credit and the initial self-operated benefit guide function, all associated intelligence contracts are organized by the synergistic gain co-molder to conduct multiple rounds of synergistic game training, and a synergistically evolved same-frequency associated intelligence contract group is constructed, including the following sub-steps: Step S321, based on the full-link historical scheduling data of the supply chain and the preset disturbance, a supply chain simulation scenario is generated by the synergistic gain co-molder; Specifically, the full-link historical scheduling data of the supply chain is layered according to the complexity of the scene, the core elements of each scene are extracted from each layer of data, and multiple basic simulation scenarios without disturbance are generated by a simulation scenario generation module; based on the preset disturbance, a disturbance control module injects random disturbance events into the basic simulation scenarios according to the set disturbance probability and specific training targets, and generates multiple customized simulation scenarios with labeled disturbance information.

[0023] Step S322, based on the state of each associated intelligence contract and the initial self-operated benefit guide function, each associated intelligence contract makes autonomous decisions in the supply chain simulation scenario, and the scheduling execution is simulated and executed by the synergistic gain co-molder to generate a scheduling execution completion result; Specifically, the public information of the supply chain simulation scenario to be trained is broadcast to all related associated intelligence contracts by the global state tracking module of the synergistic gain co-molder, each associated intelligence contract makes autonomous decisions based on its own state and the initial self-operated benefit guide function to maximize its own interests, outputs decision actions, and collects the decision actions of all associated intelligence contracts. When there is a resource conflict, a preset conflict elimination rule is used for elimination; the game interaction module simulates the execution of the scheduling task in the supply chain simulation scenario based on all the final determined decision actions, updates the state of the supply chain simulation scenario, and generates a scheduling execution completion result. The scheduling execution completion result includes global results and individual results, and the global results and individual results include negative data for decision-making to maximize self-interest.

[0024] Step S323, based on the scheduling execution completion result, the contribution share of each participating associated intelligence contract to the global gain is quantified by the synergistic gain co-molding algorithm, the self-operated benefit guide function of each associated intelligence contract is remolded into a mutual benefit guide function, and the mutual benefit credit is updated; Specifically, the incremental quantification module calculates the global net value created by the decision of the current training through the decision global net price formula

[0025] The global net value created by the decision of the current training, is the global net value, is the income conversion coefficient, is the total income of the supply chain actually realized by the decision of the current training, is the expected income without synergistic decision, is the number of resource operation cost parameters, the value range of , For the first The time value factor of resource operation costs For the first The uncoordinated benchmark operating cost of similar resource operating costs This is the first training decision made after this round. The actual operating cost of resource-related operations. For the number of service value parameters, The range of values ​​is , For the first The value coefficient of the service value parameter For the first Service increment of a service value parameter For the number of Value at Risk parameters, The range of values ​​is , For the first The value coefficient of the risk value parameter For the first Risk increment of a risk value parameter For the number of sustainability value parameters, The range of values ​​is , For the first The value coefficient of a sustainability value parameter For the first Sustainability increment of a kind of sustainability value parameter To coordinate the number of cost parameters, The range of values ​​is , For the first The weighting coefficients of each coordination cost parameter. For the first The cost value of each coordination cost parameter. For the number of uncertain scenarios, The range of values ​​is , For the first The probability of an uncertain scenario occurring. For the first Net price loss index for uncertain scenarios.

[0026] Based on global net value, a collaborative gain co-shaping algorithm is used.

[0027] Quantify the contribution share of each participating intelligence consortium to the global gain in this round of training decision-making, among which, For the first The contribution share of each interconnected intelligence entity to the global gain. For the first The impact coefficient of the marginal contribution of each intellectual community The collaborative value when all stakeholders participate. To exclude the first The collaborative value of a unified intelligence platform. For the number of Lianzhiqi entities, The range of values ​​is , The range of values ​​is , For the first Resource input impact coefficient of each intellectual consortium The number of dimensions contributing to resource input. The range of values ​​is , For the first Impact coefficient of resource input contribution dimension For the first The first of the joint intelligence community Resource input contribution value in each dimension For the first Risk-taking impact coefficient of each intelligent alliance The number of risk-taking dimensions, The range of values ​​is , For the first The impact coefficient of each risk-taking dimension For the first The first of the joint intelligence community Risk tolerance values ​​in each dimension To determine the number of strategic influence dimensions, The range of values ​​is The impact coefficient of each strategic influence dimension For the first The first of the joint intelligence community Strategic impact values ​​in each dimension For overall net value, To adaptively adjust the number of factors, The range of values ​​is , For the first The adjustment coefficient of each adaptive adjustment factor For the first The first of the joint intelligence community An adaptive adjustment factor.

[0028] The rule reshaping module reshapes the self-operated benefit derivative function of each intelligence alliance based on the contribution share of each alliance, upgrading it to a collaborative benefit derivative function. ,in, For the first The first of the joint intelligence community The co-derivative value of the training round, The co-benefit coefficients are the co-benefit coefficients from the previous round of training of the co-benefit common derivative function. For the first The first of the joint intelligence community The co-derivative value of the training round, when hour, This is the initial self-operated benefit function value. The coefficient of synergistic contribution. For the first The first of the joint intelligence community Contribution share of training rounds, The co-leading conversion index for contribution share, For the first The coefficient of synergy of the synergy product confidence of a consortium of intellectual entities. For the first The first of the joint intelligence community The credit product reliability of the training round, when hour, The initial product confidence level, The common derivative conversion index of the product of benefits confidence. For the first The environmental adaptation modulation index of an individual intelligent entity.

[0029] The credit record module uses a contribution-based credit formula based on the contribution share of each intellectual property entity. Update the confidence level of the product of mutual benefit, whereby, For the updated product confidence level, For the first The first joint intelligence alliance Confidence of the product of synergies after training rounds For the first The first of the joint intelligence community Contribution share of training rounds, This is a credit increment identification index. For the first The synergistic benefit accumulation time decay factor of the joint intelligence contract, For the first Each intelligent entity has a synergistic incentive factor.

[0030] Step S324: Obtain the collaboration acceptance rate and global optimization goal achievement rate of each intelligence alliance during each training round, terminate training based on a preset threshold, and generate a co-evolving intelligence alliance group; Specifically, at the end of each training round, the training termination determination module uses a collaborative acceptance formula. The cooperation acceptance rate of each alliance is calculated, wherein, The cooperation acceptance rate of the first The cooperation weight of the regular cooperative task of the first The cooperation acceptance rate of the first The cooperation weight of the emergency task of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation weight of the emergency task of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first The cooperation acceptance rate of the first

[0031] The global optimization target achievement formula is obtained by global optimization The global optimization target achievement rate is calculated, wherein, The global optimization target achievement rate is calculated, wherein, The global optimization target achievement rate is calculated, wherein, The global optimization target achievement rate is calculated, wherein, The global optimization target achievement rate is calculated, wherein, The optimization key coefficient of the first The actual achievement value of the first The actual achievement value of the first The actual achievement value of the first The actual achievement value of the first The actual achievement value of the first

[0032] Based on the cooperation acceptance rate of each alliance, the global optimization target achievement rate and the preset threshold, convergence judgment is performed, and when the two indexes are stable and exceed the threshold, the training is terminated, at this time, the behavior mode of all alliances is highly cooperative, a synergistic evolution of the same frequency alliance group is generated, and the last training of the cooperation benefit common guide function and the cooperation benefit cumulative confidence is saved.

[0033] In step S4, based on the dynamic virtual resource pool, a candidate same frequency alliance group with resource granularity matching is found from the same frequency alliance group, and a temporary cooperative scheduling chain is constructed, and each candidate same frequency alliance dynamically executes a cooperative scheduling task; Further, based on the dynamic virtual resource pool, a candidate same frequency alliance group with resource granularity matching is found from the same frequency alliance group, and a temporary cooperative scheduling chain is constructed, and each candidate same frequency alliance dynamically executes a cooperative scheduling task, including the following sub-steps: Step S41, real-time acquisition of supply chain scheduling demand, and parsing into demand combination characteristics of specific matching resource particles; Specifically, the collaborative scheduling tasks of the supply chain scheduling demand are acquired in real time through the scheduling demand interface of each node, the urgency of the corresponding collaborative scheduling tasks is identified according to demand initiators, time requirements, influence ranges and other elements, the scheduling execution priorities of each collaborative scheduling task are marked, the demand characteristics of each collaborative scheduling task are extracted and the task is decomposed in order of scheduling execution priority, and the dependency relationship between the subtasks is parsed into multiple subtasks including "resource type dimension + attribute constraint dimension + time and space constraint dimension" three-dimensional demand characteristics. The subtasks and their three-dimensional demand characteristics are mapped into demand combination characteristics corresponding to the matching resource particle data according to the particle query template preset by the dynamic virtual resource pool composite index engine.

[0034] Step S42, based on the demand combination characteristics, matching and binding the corresponding matching resource particles and the candidate intelligent contract bodies with the optimal order level in the same frequency intelligent contract body group; Specifically, based on the demand combination characteristics, the corresponding matching resource particles with available search state in the dynamic virtual resource pool are searched through the composite index engine, and the intelligent contract body ID bound thereto is acquired. Based on the multi-dimensional order elements of the intelligent contract body and the collaborative execution performance archive, the order level of each intelligent contract body is calculated through the order level formula The order level of each intelligent contract body is calculated, wherein, the order level of the first intelligent contract body, the demand matching factor of the first intelligent contract body, the number of order credit elements, the value range of , the weight coefficient of the first order credit element of the first intelligent contract body, the order credit score of the first order credit element of the first intelligent contract body.

[0035] The intelligent contract bodies are classified according to the types of the executed scheduling tasks, the intelligent contract bodies of various types are sorted according to the order level, and the intelligent contract bodies of various types with the top preset number of order levels are selected as candidate intelligent contract bodies.

[0036] Step S43, combining the candidate intelligent contract bodies and calculating the collaborative scheduling values of each combination, selecting the scheduling execution intelligent contract bodies for series connection to form a temporary collaborative scheduling chain, and locking the corresponding matching resource particles in the dynamic virtual resource pool to dynamically execute the collaborative scheduling tasks; Specifically, each candidate joint intelligence contract is combined according to the type, and the task scheduling coordination degree of each combination is identified, and the candidate joint intelligence contract corresponding to the best combination of the task scheduling coordination degree is selected as the scheduling execution joint intelligence contract. According to the dependency relationship and execution process of each subtask in the collaborative scheduling task, the scheduling execution joint intelligence contract is connected in series to form a temporary collaborative scheduling chain. Each node in the temporary collaborative scheduling chain represents a scheduling subtask responsible for each scheduling execution joint intelligence contract, and the edge between the nodes represents the dependency relationship and execution process between each subtask. When constructing the temporary collaborative scheduling chain, the state of the corresponding matching resource particle is modified to be locked in the dynamic virtual resource pool, and the task execution parameters are configured for each node based on the matching resource particle. After the configuration is completed, the collaborative scheduling task is automatically executed. When the task execution is completed, the temporary collaborative scheduling chain is automatically dissolved. Based on the actual state of the resource corresponding to each matching resource particle after the completion of the collaborative scheduling task, the state of the matching resource particle is modified to be available or cancelled in the dynamic virtual resource pool, and the collaborative scheduling task is finally completed.

[0037] Step S5, the execution results of each same-frequency joint intelligence contract are obtained, and the real-time updated joint intelligence contract is monitored. According to the execution results and the updated joint intelligence contract, the execution step S2 is returned to reselect the candidate same-frequency joint intelligence contract group to execute the scheduling task.

[0038] Further, the execution results of each same-frequency joint intelligence contract are obtained, and the real-time updated joint intelligence contract is monitored. According to the execution results and the updated joint intelligence contract, the execution step S2 is returned to reselect the candidate same-frequency joint intelligence contract group to execute the scheduling task, including the following substeps: Step S51, during the execution of the collaborative scheduling task, the execution results of each same-frequency joint intelligence contract are obtained in real time, and a joint intelligence holographic double-layer monitoring board is built. The collaborative execution performance archives are established and updated for each joint intelligence contract in combination with the cooperation benefit credibility; Specifically, the execution effect monitoring data in the collaborative scheduling task execution process is collected in real time through the Internet of Things sensor, and each data is associated with task ID, joint intelligence contract ID, collection timestamp and other traceable information. Global data is extracted from the collected data, and the state information of all temporary collaborative scheduling chains being executed is displayed in real time in the form of a supply chain network diagram, and the holographic supply chain global collaborative efficiency is also displayed. Local data of each joint intelligence contract is extracted from the collected data, and the state information of each joint intelligence contract during task execution is displayed, and the efficiency of the holographic joint intelligence contract is also displayed, so as to build a joint intelligence holographic double-layer monitoring board.

[0039] Based on the historical scheduling performance information of the corresponding nodes of each association contract body, combined with the initial value of the cooperative benefit trust, the initialization cooperative performance file of each association contract body is constructed, which contains three core data: basic information, historical performance data, and initial cooperative benefit trust. After completing each cooperative scheduling task, based on the whole-process data monitored by the holographic double-layer monitoring board of the association contract, the execution efficiency, cooperative contribution, abnormal handling performance, and resource utilization level of the association contract body in this task are extracted, and the cooperative performance file is dynamically updated in combination with the real-time changes of the task difficulty coefficient and the cooperative benefit trust, to supplement the complete execution record and performance score of this task, and accurately reflect the latest cooperative ability and performance level of the association contract body.

[0040] Step S52, based on the cooperative performance file of each association contract body, the key optimization indicators are extracted, the parameters of the gain co-shaping algorithm are updated, and the same-frequency association contract body group is reconstructed; Specifically, the cooperative performance files of all association contract bodies are synchronized to the cooperative gain co-shaping device, the file data is automatically analyzed, the training requirements are disassembled, the targeted simulation scenarios are generated based on the disassembled training requirements, and the same-frequency association contract body group is organized to carry out multiple rounds of cooperative game retraining. In the training process, the historical data in the cooperative performance file is taken as a reference to set differentiated training targets, and the retraining performance of each contract body is collected in real time and compared with the historical data in the file. After the retraining is completed, the same-frequency association contract body group is reconstructed according to the comprehensive evaluation results of the training performance of each association contract body and the cooperative performance file. At the same time, the core parameters of the cooperative gain co-shaping algorithm are updated based on the retraining data.

[0041] Embodiment Two As shown in Figure 2 , the embodiment two of the present application provides a supply chain cooperative scheduling management system, which comprises: A dynamic virtual resource pool construction module 21, which constructs a dynamic virtual resource pool composed of schedulable resource particles based on the schedulable resources of each supply chain node in response to receiving a supply chain scheduling demand; An association contract body construction module 22, which acquires the digital images and intelligent collaboration carriers of each supply chain node on the supply chain to construct an association contract body; A same-frequency association contract body group construction module 23, which trains the constructed association contract bodies through cooperative decision game training by a cooperative gain co-shaping device to construct a cooperative evolution same-frequency association contract body group; Further, the same-frequency association contract body group construction module 23 comprises the following submodules: An association contract body initialization setting submodule, which sets an initial cooperative benefit trust for each association contract body and defines an initial self-operated benefit guide function; The co-evolution same-frequency joint intelligence contract group construction submodule constructs a co-evolution same-frequency joint intelligence contract group through a plurality of rounds of co-evolution game training of all joint intelligence contracts based on initial mutual benefit credibility and initial self-operation benefit guide function through a co-evolution gain co-molder; Further, the co-evolution same-frequency joint intelligence contract group construction submodule comprises the following submodules: The supply chain simulation scenario generation submodule generates a supply chain simulation scenario based on historical scheduling data of the supply chain and a preset disturbance through a co-evolution gain co-molder; The scheduling execution completion result generation submodule generates a scheduling execution completion result through a co-evolution gain co-molder based on the self state of each joint intelligence contract and the initial self-operation benefit guide function, and each joint intelligence contract makes autonomous decisions in the supply chain simulation scenario and simulates scheduling execution. The joint intelligence contract parameter updating submodule quantifies the contribution share of each participating joint intelligence contract to the global gain through a co-evolution gain co-molding algorithm based on the scheduling execution completion result, remolds the self-operation benefit guide function of each joint intelligence contract into a mutual benefit co-guide function, and updates the mutual benefit credibility. The training termination submodule obtains the cooperation acceptance rate of each joint intelligence contract and the global optimization target achievement rate in each round of training, terminates the training based on a preset threshold, and generates a co-evolution same-frequency joint intelligence contract group. The co-scheduling task execution module 24 finds a candidate co-scheduling task execution module 24 from the co-scheduling task execution module 24 based on a dynamic virtual resource pool, constructs a temporary co-scheduling chain, and dynamically executes a co-scheduling task by each candidate co-scheduling task execution module 24. Further, the co-scheduling task execution module 24 comprises the following submodules: The demand combination characteristic acquisition submodule acquires supply chain scheduling demand in real time and parses it into demand combination characteristics of specific matching resource particles; The candidate joint intelligence contract selection submodule matches and binds the candidate joint intelligence contract with the corresponding matching resource particle and the optimal benefit sequence guide level in the co-scheduling task execution module 24 based on the demand combination characteristics. The scheduling task execution submodule combines the candidate joint intelligence contract and calculates the co-scheduling value of each combination, selects a scheduling execution joint intelligence contract, forms a temporary co-scheduling chain, locks the corresponding matching resource particle in the dynamic virtual resource pool, and dynamically executes a co-scheduling task. The co-scheduling task execution module 24 finds a candidate co-scheduling task execution module 24 from the co-scheduling task execution module 24 based on a dynamic virtual resource pool, constructs a temporary co-scheduling chain, and dynamically executes a co-scheduling task by each candidate co-scheduling task execution module 24.

[0042] Further, the co-scheduling task execution module 24 comprises the following submodules: An execution effect monitoring submodule acquires the execution results of each same-frequency associated intelligence body in real time in the collaborative scheduling task execution process, builds an associated intelligence holographic double-layer monitoring board, and establishes and updates a collaborative execution performance archive for each associated intelligence body in combination with the collaborative benefit credibility; A same-frequency associated intelligence body group reconstruction submodule extracts key optimization indexes based on the collaborative execution performance archives of the associated intelligence bodies, updates the parameters of the gain co-molding algorithm, and reconstructs the same-frequency associated intelligence body group; Corresponding to the above-mentioned embodiments, the embodiments of the present application provide a computer storage medium, comprising at least one memory and at least one processor. The memory is used for storing one or more program instructions. The processor is used for running one or more program instructions to execute a supply chain collaborative scheduling management method.

[0043] Corresponding to the above-mentioned embodiments, the embodiments of the present application provide a computer readable storage medium, and the computer storage medium contains one or more program instructions, and the one or more program instructions are used for a processor to execute a supply chain collaborative scheduling management method.

[0044] The embodiments disclosed in the present application provide a computer readable storage medium, and the computer readable storage medium stores computer program instructions, and when the computer program instructions run on a computer, the computer executes the above-mentioned supply chain collaborative scheduling management method.

[0045] In the embodiments of the present application, the processor can be an integrated circuit chip with signal processing capability. The processor can be a general processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.

[0046] The disclosed methods, steps and logic block diagrams in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can be any conventional processor. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as hardware code processor execution or executed by a combination of hardware and software modules in the code processor. The software module can be located in a random memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register or other mature storage medium in the art. The processor reads the information in the storage medium and combines the hardware to complete the steps of the above-mentioned method.

[0047] The storage medium can be a memory, for example, can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories.

[0048] Among them, the non-volatile memory can be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically EPROM (EEPROM), or a flash memory.

[0049] The volatile memory can be a Random Access Memory (RAM) used as an external cache. By way of example, and not limitation, many forms of RAM can be used, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous Dynamic RAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced SDRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DRRAM).

[0050] The storage medium described in the embodiments of the present application is intended to include, but not limited to, these and any other suitable types of memory.

[0051] Those skilled in the art should be aware that in one or more of the above examples, the functions described in the present application can be implemented in combination of hardware and software. When the software is applied, the corresponding functions can be stored in a computer readable medium or transmitted as one or more instructions or codes on a computer readable medium. The computer readable medium includes a computer storage medium and a communication medium, wherein the communication medium includes any medium that facilitates the transfer of computer programs from one place to another. The storage medium can be any available medium that can be accessed by a general or special purpose computer.

[0052] The above detailed description of the specific embodiments of the present application has been given to illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. Any modification, equivalent replacement, improvement, etc. made on the basis of the technical solutions of the present application shall be included in the protection scope of the present application.

Claims

1. A supply chain collaborative scheduling management method, characterized in that, include: Step S1: In response to receiving a supply chain scheduling request, construct a dynamic virtual resource pool composed of allocable resource granules based on the schedulable resources of each supply chain node; Step S2: Obtain digital images and intelligent collaboration carriers of each supply chain node to build a smart alliance; Step S3: Train the constructed intelligent entity through collaborative decision-making game using a collaborative gain co-shaper to build a co-evolving intelligent entity group with the same frequency. Step S4: Based on the dynamic virtual resource pool, find candidate co-frequency intelligent alliance groups with matching resource granularity from the co-frequency intelligent alliance group, construct a temporary collaborative scheduling chain, and have each candidate co-frequency intelligent alliance group dynamically execute collaborative scheduling tasks. Step S5: Obtain the execution results of each synchronized intelligence alliance and monitor the real-time updated intelligence alliances. Based on the execution results and the updated intelligence alliances, return to step S2 and reselect the candidate synchronized intelligence alliance group to execute the scheduling task.

2. The supply chain collaborative scheduling management method as described in claim 1, characterized in that, The construction of a co-evolutionary, synchronous intelligent entity group is achieved by training the constructed intelligent entity through collaborative decision-making game using a collaborative gain co-shaper, and includes the following sub-steps: Step S31: Set the initial synergistic benefit product confidence for each intelligence alliance and define the initial self-operated benefit derivative function; Step S32: Based on the initial synergistic benefit product confidence and the initial self-managed benefit derivative function, organize all intelligent entities through a synergistic gain co-shaper to conduct multi-round synergistic game training, and construct a co-evolving synchronous intelligent entity group.

3. The supply chain collaborative scheduling management method as described in claim 1, characterized in that, Based on the initial synergistic benefit product confidence and the initial self-managed benefit derivative function, all intelligent cooperative entities are organized through a synergistic gain co-shaper to conduct multi-round synergistic game training, constructing a co-evolutionary group of intelligent cooperative entities with the same frequency. The process includes the following sub-steps: Step S321: Based on the historical scheduling data of the entire supply chain and preset disturbances, a supply chain simulation scenario is generated through a collaborative gain co-shaper. Step S322: Based on the self-state and initial self-operated benefit function of each intelligent entity, each intelligent entity makes autonomous decisions in the supply chain simulation scenario, and simulates scheduling execution through the collaborative gain co-shaper to generate scheduling execution completion results; Step S323: Based on the scheduling execution completion result, quantify the contribution share of each participating intelligence alliance to the global gain in this round of training through the collaborative gain co-shaping algorithm, reshape the self-managed benefit derivative function of each intelligence alliance to upgrade it to the collaborative benefit co-derived function, and update the collaborative benefit product confidence. Step S324: Obtain the collaboration acceptance rate and global optimization goal achievement rate of each intelligence partner during each training round, terminate training based on a preset threshold, and generate a co-evolving, frequency-matching intelligence partner group.

4. The supply chain collaborative scheduling management method as described in claim 1, characterized in that, Based on a dynamic virtual resource pool, candidate intelligent alliance groups with matching resource granularity are identified from the pool of intelligent alliance groups. A temporary collaborative scheduling chain is constructed, and each candidate intelligent alliance group dynamically executes the collaborative scheduling task, including the following sub-steps: Step S41: Obtain supply chain scheduling requirements in real time and parse them into the demand combination characteristics of specific resource granules to be allocated; Step S42: Based on the demand combination characteristics, match and bind the corresponding resource particles and the candidate intelligent alliance group with the best benefit order in the same frequency intelligent alliance group; Step S43: Combine the candidate intelligent entities and calculate the collaborative scheduling value of each combination. Select the intelligent entities to be executed for scheduling and connect them in series to form a temporary collaborative scheduling chain. Lock the corresponding resource particles in the dynamic virtual resource pool and dynamically execute the collaborative scheduling task.

5. The supply chain collaborative scheduling management method as described in claim 1, characterized in that, Obtain the execution results of each synchronized intelligence alliance and monitor the real-time updates of the alliances. Based on the execution results and the updated alliances, return to reselect candidate synchronized intelligence alliance groups to execute the scheduling task, including the following sub-steps: Step S51: During the execution of the collaborative scheduling task, obtain the execution results of each co-frequency intelligent entity in real time, build a holographic dual-layer monitoring dashboard for the intelligent entity, and establish and update the collaborative execution performance file for each intelligent entity in combination with the collaborative benefit confidence. Step S52: Based on the collaborative performance profiles of each intelligence alliance, extract key optimization indicators, update the parameters of the gain co-shaping algorithm, and reconstruct the same-frequency intelligence alliance group.

6. A supply chain collaborative scheduling and management system, characterized in that, include: The dynamic virtual resource pool construction module, in response to receiving supply chain scheduling requirements, constructs a dynamic virtual resource pool composed of allocable resource granules based on the schedulable resources of each supply chain node. The Connectivity and Intelligence Building Module acquires digital images and intelligent collaboration carriers of each node in the supply chain to build a Connectivity and Intelligence Building system. The module for constructing a co-frequency intelligent entity group uses a collaborative gain co-shaper to train the constructed intelligent entities through collaborative decision-making game theory, thereby building a co-evolving co-frequency intelligent entity group. The collaborative scheduling task execution module searches for candidate co-frequency intelligent alliance groups with matching resource granularity from the co-frequency intelligent alliance group based on the dynamic virtual resource pool, constructs a temporary collaborative scheduling chain, and the collaborative scheduling tasks are dynamically executed by each candidate co-frequency intelligent alliance group. The Synchronous Intelligence Alliance Group Reconstruction Module obtains the execution results of each Synchronous Intelligence Alliance Group and monitors the real-time updates of the Synchronous Intelligence Alliance Group. Based on the execution results and the updated Synchronous Intelligence Alliance Group, it returns to the Synchronous Intelligence Alliance Group Construction Module.

7. A supply chain collaborative scheduling management system as described in claim 6, characterized in that, The module for building a community of shared intelligence and collaboration includes: The Lianzhiqiti initialization setting submodule sets the initial synergistic benefit product confidence for each Lianzhiqititi and defines the initial self-operated benefit derivative function; The co-evolutionary synchronous intelligence pool construction submodule organizes all intelligence pools through multiple rounds of collaborative game training based on the initial synergistic benefit product confidence and the initial self-managed benefit derivative function, thereby constructing a co-evolutionary synchronous intelligence pool.

8. A supply chain collaborative scheduling management system as described in claim 6, characterized in that, The sub-module for constructing a co-evolutionary, synchronized, and intelligent swarm includes: The supply chain simulation scenario generation submodule generates supply chain simulation scenarios based on the historical scheduling data of the entire supply chain and preset disturbances through a collaborative gain shaper. The scheduling execution completion result generation submodule, based on the self-state of each intelligent entity and the initial self-operated benefit function, allows each intelligent entity to make autonomous decisions in the supply chain simulation scenario, and simulates scheduling execution through the collaborative gain co-shaper to generate the scheduling execution completion result; The Joint Intelligence Entity Parameter Update Submodule, based on the scheduling execution completion result, quantifies the contribution share of each participating Joint Intelligence Entity to the global gain in this round of training through the collaborative gain co-shaping algorithm, reshapes the self-operated benefit derivative function of each Joint Intelligence Entity to upgrade it to the collaborative benefit co-derivative function, and updates the collaborative benefit product confidence. The training termination submodule obtains the collaboration acceptance rate and global optimization goal achievement rate of each intelligence partner during each training round, terminates training based on a preset threshold, and generates a co-evolving intelligence partner group.

9. A supply chain collaborative scheduling management system as described in claim 6, characterized in that, The collaborative scheduling task execution module specifically includes: The demand combination characteristic acquisition submodule acquires supply chain scheduling requirements in real time and parses them into demand combination characteristics for specific resource granules. The candidate intelligence alliance selection submodule matches and binds the corresponding resource granules and the best benefit sequence among the same frequency intelligence alliance groups based on the demand combination characteristics. The scheduling task execution submodule combines candidate intelligent entities and calculates the collaborative scheduling value of each combination. It selects the intelligent entities to be executed and connects them to form a temporary collaborative scheduling chain. It then locks the corresponding resource particles in the dynamic virtual resource pool and dynamically executes the collaborative scheduling task.

10. A supply chain collaborative scheduling management system as described in claim 6, characterized in that, The module for reconstructing a unified, intelligent, and collaborative ecosystem includes: The execution performance monitoring submodule acquires the execution results of each co-frequency intelligent alliance in real time during the execution of collaborative scheduling tasks, builds a holographic two-layer monitoring dashboard for the intelligent alliance, and establishes and updates the collaborative execution performance profile for each intelligent alliance based on the collaborative benefit confidence score. The Synchronous Intelligence Co-operation Group Reconstruction Submodule extracts key optimization indicators based on the collaborative execution performance profiles of each Synchronous Intelligence Co-operation Group, updates the parameters of the gain co-shaping algorithm, and reconstructs the Synchronous Intelligence Co-operation Group.

Images