Enterprise resource scheduling optimization system based on nested tree structure

Abstract

The application belongs to the technical field of resource scheduling, and discloses an enterprise resource scheduling optimization system based on a nested tree structure; comprising: collecting resource demand data of each enterprise task, and constructing a competitive relationship matrix; mapping each enterprise task into a tree node, identifying parent-child nodes and sibling nodes between the tree nodes according to the resource demand data and the competitive relationship matrix, and constructing a competitive nested tree structure; establishing a multi-party game model between each group of sibling nodes, and identifying conflict nodes in an equilibrium state; sequentially determining the hierarchical position and dilemma type of each conflict node, designing differentiated incentive strategies for different conflict nodes; and performing global optimization on the competitive nested tree structure, and obtaining a resource scheduling scheme with the highest comprehensive optimization score as an equilibrium scheduling scheme; the application can guide each task subject to change from local interest maximization to global coordination, and significantly improve the overall resource utilization level and operation efficiency of the enterprise.

Description

Technical Field

[0001] This invention relates to the field of resource scheduling technology, and more specifically, to an enterprise resource scheduling optimization system based on a nested tree structure. Background Technology

[0002] In the modern enterprise operating environment, resource scheduling optimization is a key link to improve enterprise efficiency and competitiveness. As enterprises expand and business complexity increases, multiple tasks often need to compete for a limited pool of resources, such as computing resources, human resources, and equipment resources. Traditional resource scheduling methods mainly use linear programming, heuristic algorithms, or simple priority ranking. These methods usually simplify the resource allocation problem into a single-objective optimization problem, ignoring the complex competitive relationships and mutual influences between tasks.

[0003] In recent years, some improved resource scheduling techniques have begun to incorporate multi-objective optimization and simple competitive relationship modeling in an attempt to improve the rationality of resource allocation. However, these methods still lack in-depth analysis of the game behavior between tasks and cannot effectively identify and resolve conflicts in the resource allocation process. In particular, when multiple tasks compete for the same resource pool at the same time, each task subject often makes decisions based on maximizing its own interests, which can easily lead to game dilemmas such as the prisoner's dilemma, resulting in a decrease in the overall efficiency of resource allocation. In addition, existing methods usually adopt a uniform scheduling strategy and fail to design differentiated incentive mechanisms according to the characteristics and competition intensity of different tasks, making it difficult to guide the task subjects to cooperate towards the global optimum.

[0004] In view of this, the present invention proposes an enterprise resource scheduling optimization system based on a nested tree structure to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: an enterprise resource scheduling optimization system based on a nested tree structure, comprising:

[0006] The resource acquisition module is used to collect resource requirement data for tasks from various enterprises.

[0007] The competition quantification module is used to select the corresponding set of competitive tasks for each pre-built resource pool based on resource demand data, and calculate the competitive intensity between every two enterprise tasks in each set of competitive tasks to build a competitive relationship matrix.

[0008] The nested building module is used to map each enterprise task to a tree node. Based on resource demand data and competition relationship matrix, it identifies the parent-child nodes and sibling nodes between tree nodes, and builds a competitive nested tree structure based on the parent-child nodes and sibling nodes.

[0009] The conflict identification module is used to establish a multi-party game model between each group of sibling nodes, and to iteratively solve the Nash equilibrium based on the pre-built utility function to identify conflicting nodes in the equilibrium state.

[0010] The incentive design module is used to determine the hierarchical position and dilemma type of each conflict node in sequence, and to design differentiated incentive strategies for different conflict nodes based on the hierarchical position and dilemma type.

[0011] The equilibrium optimization module uses the incentive parameters in the differentiated incentive strategy as decision variables to perform global optimization on the competitive nested tree structure, obtain the Pareto optimal solution set, and select the resource scheduling scheme with the highest comprehensive optimization score from the Pareto optimal solution set as the equilibrium scheduling scheme.

[0012] Furthermore, resource demand data includes demand type, demand quantity, demand period, acceptable delay duration, and task priority;

[0013] The content of the corresponding set of competitive tasks selected for each resource pool includes:

[0014] For each resource pool, enterprise tasks with the same corresponding demand type are integrated to obtain a set of enterprise tasks for each resource pool; task time periods are constructed according to the demand time periods corresponding to all enterprise tasks; and task time periods are evenly divided according to the preset time period span to obtain multiple sub-time periods.

[0015] Based on the demand period corresponding to each enterprise task, determine the sub-period corresponding to each enterprise task; divide the enterprise tasks in the enterprise task set corresponding to each resource pool according to the corresponding sub-period, and obtain multiple sets of time period tasks corresponding to each resource pool.

[0016] Based on the number of enterprise tasks with the same demand type in each time period task set corresponding to each resource pool, calculate the total demand of each resource pool in each sub-time period; obtain the resource capacity of each resource pool and compare it with the total demand in each corresponding sub-time period; if the resource capacity is less than or equal to the total demand, it is determined that there is resource competition, and the time period task set corresponding to the corresponding resource pool is taken as the competing task set.

[0017] Furthermore, the content of constructing the competition relationship matrix includes:

[0018] Each pair of enterprise tasks in each set of competing tasks is integrated sequentially to obtain multiple sets of computing tasks. The time overlap length of each set of computing tasks is calculated based on the demand period corresponding to each enterprise task in each set of computing tasks. The resource bottleneck degree of each set of computing tasks is calculated based on the time overlap length of each set of computing tasks and the resource capacity of the corresponding resource pool. The total amount of competition for each set of computing tasks is calculated, and the competition intensity of each set of computing tasks is calculated in combination with the resource bottleneck degree.

[0019] Traverse all computation sets and integrate those containing two enterprises with the same tasks to obtain multiple sets of integrated computation sets. Calculate the overall competitive intensity of each set of integrated computation sets based on the competitive intensity of each set. Construct an initial matrix by adding the overall competitive intensity of each set of integrated computation sets to the initial matrix according to the enterprise tasks they correspond to, thus constructing a competition relationship matrix.

[0020] Furthermore, identifying the parent-child and sibling nodes among tree nodes includes:

[0021] Each pair of tree nodes is integrated to obtain multiple sets of nodes; each set of nodes is then identified for inclusion relationships, and the sets of nodes with identified inclusion relationships are marked as parent-child sets, with the tree nodes in each set of parent-child sets being the parent and child nodes; for each set of parent-child sets, the tree node that covers the other tree node in both the demand time period and demand type is taken as the parent node, and the tree nodes that are not taken as parent nodes are taken as child nodes.

[0022] All node sets not marked as parent-child sets are marked as analysis sets. The comprehensive competitive intensity corresponding to each analysis set is obtained sequentially from the competition relationship matrix. If the comprehensive competitive intensity is not 0, the corresponding analysis set is marked as a sibling set, and the tree nodes in each sibling set are sibling nodes.

[0023] Furthermore, the content of establishing a multi-party game model includes:

[0024] The tree nodes in the sibling set points are regarded as participants. Each participant makes choices in the strategy space and is subject to resource constraints. The strategy space includes quantity space, time space, and priority space. The resource constraints include quantity constraints, time constraints, priority constraints, and allocation rules. Through the definition of participants, the setting of the strategy space, and the limitation of resource constraints, a multi-party game model is formed.

[0025] The content of iteratively solving for Nash equilibrium based on pre-constructed utility functions includes:

[0026] Each participant randomly selects a scheduling policy from the policy space. The scheduling policy includes the amount of resources, execution time, and execution priority. A convergence threshold is preset. The scheduling policy of each participant is iteratively updated based on the utility function. At the end of each round of iteration, the policy change is calculated and compared with the convergence threshold. If the policy change is less than the convergence threshold, the policy combination formed by the scheduling policies of all participants has reached Nash equilibrium, and the iteration ends. If the policy change is greater than or equal to the convergence threshold, the policy combination formed by the scheduling policies of all participants has not reached Nash equilibrium, and the iteration continues.

[0027] Furthermore, the content of identifying conflicting nodes in an equilibrium state includes:

[0028] The total amount of resources is obtained by summing the resources in the scheduling strategy corresponding to each participant when Nash equilibrium is reached, and then comparing the total amount of resources with the resource capacity of the corresponding resource pool.

[0029] Based on the utility function, calculate the ideal utility and equilibrium utility for each participant; calculate the difference between the ideal utility and equilibrium utility for each participant to obtain the utility difference; compare each utility difference with a utility threshold.

[0030] If the total amount of resources exceeds the resource capacity, or there is a utility difference greater than or equal to the utility threshold, or the number of iterations exceeds the preset number of iterations threshold, then all tree nodes in the corresponding sibling set will be marked as conflicting nodes.

[0031] If the total resource amount is less than or equal to the resource capacity, and all utility differences are less than the utility threshold, and Nash equilibrium is reached, then the tree nodes in the corresponding sibling set will not be marked as conflict nodes.

[0032] Furthermore, determining the hierarchical position of each conflict node sequentially includes:

[0033] Mark the topmost tree node in the nested competitive tree structure as the root node. Starting from the level corresponding to the root node in the nested competitive tree structure, assign incrementally increasing numerical labels to each level from top to bottom and mark them as level labels. Sequentially obtain the level labels of each conflicting node from the nested competitive tree structure. Calculate the size of the child nodes and the degree of competition for each conflicting node.

[0034] Different influence weights are set for hierarchical labels, child node size, and competitive correlation, and the influence range index of each conflict node is obtained by weighted summation of hierarchical labels, child node size, and competitive correlation based on the influence weights.

[0035] The influence range index of each conflict node is compared with a preset classification threshold, which includes a core threshold and an edge threshold. If the influence range index is greater than the core threshold, the corresponding conflict node is determined to be at the core level. If the influence range index is less than or equal to the core threshold but greater than the edge threshold, the corresponding conflict node is determined to be at the middle level. If the influence range index is less than or equal to the edge threshold, the corresponding conflict node is determined to be at the edge level.

[0036] Furthermore, the content of determining the dilemma type for each conflict node in sequence includes:

[0037] Each conflict node's sibling set is marked as a conflict set. The resource overload, task duration, and priority dispersion of each conflict set are calculated sequentially and compared with preset dilemma thresholds, including overload threshold, time threshold, and dispersion threshold. If the resource overload is greater than the overload threshold, an overload instruction is generated for the corresponding conflict set. If the time conflict is greater than the time threshold, a time instruction is generated for the corresponding conflict set. If the priority dispersion is less than the dispersion threshold, a dispersion instruction is generated for the corresponding conflict set.

[0038] If only overload instructions are generated, the dilemma type of each conflict node in the corresponding conflict set is determined to be a resource overload dilemma; if only time instructions are generated, the dilemma type of each conflict node in the corresponding conflict set is determined to be a time conflict dilemma; if only distribution instructions are generated, the dilemma type of each conflict node in the corresponding conflict set is determined to be a priority disorder dilemma; if at least two of the overload, time, and distribution instructions are generated, the dilemma type of each conflict node in the corresponding conflict set is determined to be a combined conflict dilemma.

[0039] Furthermore, the design of differentiated incentive strategies for different conflict nodes includes:

[0040] For the conflict nodes corresponding to the resource overload dilemma, design resource reduction incentive strategies;

[0041] For the conflict nodes corresponding to time-conflict dilemmas, design time adjustment incentive strategies;

[0042] For conflict nodes corresponding to priority confusion dilemmas, design priority negotiation incentive strategies;

[0043] For the conflict nodes corresponding to comprehensive conflict-type dilemmas, design combined incentive strategies;

[0044] Different level adjustment coefficients are set for different level positions. The level adjustment coefficient of each conflict node is multiplied by the corresponding incentive strategy to obtain the differentiated incentive strategy for each conflict node.

[0045] Furthermore, obtaining the Pareto optimal solution set includes:

[0046] Incentive parameters include resource reduction, time adjustment, and priority adjustment; construct a multi-objective optimization function, which includes a utility difference minimization objective function, a utility variance minimization objective function, and an incentive cost minimization objective function;

[0047] Based on multi-objective optimization functions, decision variables, and resource constraints, a multi-objective evolutionary algorithm is used to perform global optimization on a competitive nested tree structure to obtain a Pareto optimal solution set. The Pareto optimal solution set includes multiple resource scheduling schemes, and each resource scheduling scheme includes a scheduling scheme corresponding to each enterprise task.

[0048] The technical effects and advantages of the enterprise resource scheduling optimization system based on nested tree structure of the present invention are as follows:

[0049] By constructing a competitive nested tree structure, the system clearly expresses the hierarchical and competitive relationships of resource demands among enterprise tasks, overcoming the shortcomings of traditional methods in expressing complex relationships between tasks. It introduces a game theory-based multi-party game model and Nash equilibrium solution mechanism, accurately identifying conflict nodes and their utility losses during resource allocation. This approach is more precise and aligned with real-world competitive scenarios compared to traditional methods that rely solely on resource capacity. The system proposes a method to design differentiated incentive strategies based on the hierarchical position and dilemma type of conflict nodes. For conflicts with different characteristics, it employs resource reduction incentives, time adjustment incentives, priority negotiation incentives, or combined incentive strategies, avoiding a "one-size-fits-all" scheduling approach. A multi-objective optimization framework encompassing minimizing utility difference, utility variance, and incentive cost is adopted. A multi-objective evolutionary algorithm is used for global optimization on the competitive nested tree structure, achieving a balance between resource utilization efficiency, scheduling fairness, and incentive cost. This systematically solves the problem of reduced resource allocation efficiency caused by the game-playing behavior of task stakeholders in enterprise resource scheduling, guiding task stakeholders to shift from maximizing local interests to global collaboration, significantly improving the overall resource utilization level and operational efficiency of the enterprise. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of an enterprise resource scheduling optimization system based on a nested tree structure, according to Embodiment 1 of the present invention. Detailed Implementation

[0051] 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 embodiments of the present invention, and not all embodiments. 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.

[0052] Example 1

[0053] Please see Figure 1 As shown in the figure, the enterprise resource scheduling optimization system based on a nested tree structure described in this embodiment includes a resource acquisition module, a competition quantification module, a nested construction module, a conflict identification module, an incentive design module, and a balance optimization module; each module is connected via wired and / or wireless means to realize data transmission between modules.

[0054] The resource acquisition module is used to collect resource requirement data for tasks from various enterprises.

[0055] Enterprise tasks refer to the activities or project units that an enterprise needs to perform in order to achieve specific goals or complete specific work; resource requirement data includes, but is not limited to, requirement type, requirement quantity, requirement period, acceptable delay duration, task priority, etc., which are obtained through internal enterprise management systems such as PMS (Project Management System), ERP (Enterprise Resource Planning), and MES (Manufacturing Execution System);

[0056] Among them, demand type refers to the type of resources required by the enterprise task, such as computing resources (e.g., CPU, GPU, network bandwidth, etc.), human resources (e.g., engineers, programmers, quality inspectors, etc.), and equipment resources (e.g., CNC machine tools, oscilloscopes, stamping machines, etc.); demand quantity refers to the specific quantitative demand of the enterprise task for a particular resource type, such as the number of equipment units, personnel, and computing units; demand period refers to the time range for the enterprise task to use a particular resource type, expressed as start and end times; acceptable latency refers to the acceptable latency for the enterprise task when resources are insufficient, used to flexibly adjust resource allocation; task priority refers to the urgency or importance of the enterprise task in resource allocation, used to determine the scheduling order in case of conflict.

[0057] The competition quantification module is used to select the corresponding set of competitive tasks for each pre-built resource pool based on resource demand data, and calculate the competition intensity between every two enterprise tasks in each set of competitive tasks to build a competition relationship matrix.

[0058] The content of the corresponding set of competitive tasks selected for each resource pool includes:

[0059] A resource pool refers to a collection of resources of the same type within an enterprise that can be shared or competed for by multiple enterprise tasks. Different resource pools correspond to different types of resources. Resource pools are pre-built by those skilled in the art based on the enterprise's schedulable resources (including computing resources, human resources, equipment resources, etc.) for centralized management and unified scheduling of resources of the same type.

[0060] For each resource pool, enterprise tasks with the same corresponding demand type are integrated to obtain a set of enterprise tasks for each resource pool; task time periods are constructed according to the demand time periods corresponding to all enterprise tasks; wherein, the start time of the task time period is the earliest start time among all demand time periods corresponding to all enterprise tasks, and the end time of the task time period is the latest end time among all demand time periods corresponding to all enterprise tasks; according to the preset time period span, the task time period is evenly divided to obtain multiple sub-time periods, and the time period span is preset by those skilled in the art according to the actual situation;

[0061] Based on the demand period corresponding to each enterprise task, the sub-period corresponding to each enterprise task is determined; the enterprise tasks in the enterprise task set corresponding to each resource pool are divided according to the corresponding sub-period to obtain the sub-task set corresponding to each resource pool; among them, the sub-task set includes multiple time period task sets, and each time period task set includes one or more enterprise tasks. The time period task set corresponds one-to-one with the sub-period, that is, one time period task set corresponds to one sub-period.

[0062] For each resource pool, the number of enterprise tasks with the same demand type in each time period task set is added together to obtain the total demand for each resource pool in each sub-time period. The resource capacity of each resource pool is obtained and compared with the total demand for each corresponding sub-time period. If the resource capacity is less than or equal to the total demand, resource competition is determined, and the time period task set corresponding to the resource pool is taken as the competing task set. If the resource capacity is greater than the total demand, resource competition is determined. Here, resource capacity refers to the maximum amount of available resources that a resource pool can schedule or use in a sub-time period, which is obtained through enterprise internal management systems such as PMS, ERP, and MES.

[0063] The content of constructing the competition relationship matrix includes:

[0064] Each set of competing tasks is sequentially integrated with every two enterprise tasks to obtain multiple computation sets. The time overlap length of each computation set is calculated based on the demand period corresponding to each enterprise task within that set. The ratio of the time overlap length to the resource capacity of the corresponding resource pool is calculated to obtain the resource bottleneck degree of each computation set. The sum of the demand quantities corresponding to the two enterprise tasks in each computation set is calculated to obtain the total competition amount of each computation set. The product of the resource bottleneck degree and the total competition amount for each computation set is calculated to obtain the competition intensity of each computation set. The calculation process for the time overlap length is as follows: the start and end times of the two enterprise tasks in the computation set are compared, the latest start time is marked as the overlap start time, and the earliest end time is marked as the overlap end time. The difference between the overlap end time and the overlap start time is calculated to obtain the time overlap length.

[0065] Iterate through all computation sets, integrating those containing two identical enterprise tasks to obtain multiple composite computation sets. Each composite computation set includes multiple sets containing two identical enterprise tasks. Sum the competition intensity of each composite computation set to obtain the composite competition intensity of each set, representing the degree of resource conflict between the two enterprise tasks corresponding to each set under multiple resource demands and multiple time segments. Construct an initial matrix with size [size missing]. , The number of enterprise tasks; based on the enterprise tasks corresponding to each set of comprehensive calculations, the comprehensive competitive intensity corresponding to each set of comprehensive calculations is added to the initial matrix in turn to construct a competitive relationship matrix, which is used to represent the comprehensive competitive intensity between all enterprise tasks.

[0066] The nested building module is used to map each enterprise task to a tree node. Based on resource demand data and a competition relationship matrix, it identifies the parent-child and sibling nodes between tree nodes and builds a competing nested tree structure based on the parent-child and sibling nodes.

[0067] Identifying parent-child and sibling nodes in a tree includes:

[0068] Each pair of tree nodes is integrated to obtain multiple sets of nodes. Each set of nodes contains two different tree nodes, and each set of nodes is unique. The inclusion relationship of each set of nodes is identified, and the set of nodes with the inclusion relationship is marked as a parent-child set, that is, the tree nodes in each set of parent-child sets are parent and child nodes. For each set of parent-child sets, the tree node that covers the other tree node in both the demand time period and demand type is taken as the parent node, and the tree node that is not a parent node is taken as the child node.

[0069] The process of identifying the inclusion relationship of a node set is as follows: The requirement types corresponding to two tree nodes are compared to determine whether the requirement type of one tree node completely covers the requirement type of the other tree node; that is, each requirement type of the child node can be found in the requirement type corresponding to the parent node. If they do not completely cover each other, then the corresponding node sets do not have an inclusion relationship. If they completely cover each other, the requirement time periods corresponding to the two tree nodes are compared to determine whether the requirement time period of one tree node completely covers the requirement time period of the other tree node; that is, the start time of the child node is later than or equal to the start time of the parent node, and the end time is earlier than or equal to the end time of the parent node. If they do not completely cover each other, then the corresponding node sets do not have an inclusion relationship; if they completely cover each other, then the corresponding node sets have an inclusion relationship.

[0070] All node sets not marked as parent-child sets are marked as analysis sets. The comprehensive competitive intensity corresponding to each analysis set is obtained sequentially from the competition relationship matrix. If the comprehensive competitive intensity is not 0, the corresponding analysis set is marked as a sibling set, that is, the tree nodes in each sibling set are sibling nodes. If the comprehensive competitive intensity is 0, the corresponding analysis set is not marked. It should be noted that if multiple sibling sets correspond to the same parent node, and the enterprise tasks corresponding to all tree nodes in multiple sibling sets are in the same enterprise task set, then the multiple sibling sets are merged to form a sibling set containing multiple tree nodes, which is used to represent all competing enterprise tasks under the corresponding parent node.

[0071] The content of constructing a competing nested tree structure includes:

[0072] A vertical hierarchical structure of nested trees is established based on parent-child nodes, where the parent node is at the upper level and the child nodes are nested at the lower level of the corresponding parent node, forming the trunk and branches of the nested tree. Sibling nodes are placed at the same level in the nested tree and marked with horizontal associations to indicate that there is a resource competition relationship between sibling nodes. Finally, a competitive nested tree structure is formed that reflects both the hierarchical structure of tasks (parent-child nesting) and the competitive relationship between tasks at the same level (sibling competition). The connection relationship between tree nodes clearly shows the resource demand coverage relationship and the overall competition intensity between different tasks.

[0073] The conflict identification module is used to establish a multi-party game model between each group of sibling nodes, and to iteratively solve the Nash equilibrium based on the pre-built utility function to identify conflicting nodes in the equilibrium state.

[0074] The content of establishing a multi-party game model includes:

[0075] The tree nodes in the sibling aggregation points are regarded as independent game participants. Each participant makes choices in the strategy space and is subject to resource constraints, simulating the resource competition relationship between tasks of enterprises at the same level. The strategy space includes a quantity space (the range of resources that participants can choose), a time space (the range of execution time that participants can choose), and a priority space (the range of execution priorities that participants can choose). The resource constraints include quantity constraints, time constraints, priority constraints, and allocation rules. That is, each participant must meet the resource constraints when making choices in the strategy space.

[0076] The quantity and time constraints differ for different participants. The quantity constraints are specifically as follows: , The specific time constraint is the number of participants required. , This refers to the start time of the corresponding time period required by the participants. This is the sum of the participant's desired end time for their designated time period and their acceptable delay duration; the priority constraint is specifically... Allocation rules refer to the logic or mechanism used to allocate limited resources during the game process, specifically allocation according to priority; a multi-party game model is formed by defining participants, setting the strategy space, and limiting resource constraints.

[0077] The content of iteratively solving for Nash equilibrium based on pre-constructed utility functions includes:

[0078] Each participant randomly selects a scheduling strategy from the strategy space. Each scheduling strategy includes a resource quantity, execution time, and execution priority, and satisfies resource constraints. A convergence threshold is preset, which is pre-set by those skilled in the art based on actual conditions. The scheduling strategy of each participant is iteratively updated based on the utility function. At the end of each iteration, the strategy change is calculated and compared with the convergence threshold. If the strategy change is less than the convergence threshold, the strategy combination formed by the scheduling strategies of all participants reaches Nash equilibrium, and the iteration ends. If the strategy change is greater than or equal to the convergence threshold, the strategy combination formed by the scheduling strategies of all participants has not reached Nash equilibrium, and the iteration continues.

[0079] The expression for the utility function is: In the formula, Represents the utility function. Indicates resource satisfaction. Indicates the cost of resource competition. , All are utility weights;

[0080] The resource satisfaction calculation process is as follows: calculate the ratio of the resource quantity in the scheduling strategy corresponding to the participant to the demand quantity in the corresponding resource demand data to obtain the resource satisfaction. The resource competition cost calculation process is as follows: calculate the difference between the start time in the scheduling strategy corresponding to the participant and the start time in the corresponding resource demand data, and multiply it by a preset waiting cost coefficient to obtain the waiting cost; obtain the competition quantity corresponding to the participant, which is the number of tree nodes in the participant's corresponding sibling set; calculate the product of the competition quantity and the preset coordination cost coefficient to obtain the coordination cost; calculate the sum of the waiting cost and the coordination cost to obtain the resource competition cost. It should be noted that the utility weight, waiting cost coefficient, and coordination cost coefficient are all preset by those skilled in the art according to the actual situation.

[0081] The expression for iterative updates is: ;

[0082] In the formula, Indicates in In the first iteration Scheduling strategies for each participant Indicates the first Individual participants in strategy combinations The value of the utility function. Indicates the first Scheduling strategies available to each participant Indicates except the first In addition to the individual participants, all participants are in The strategy combination constituted by the scheduling strategies in round iteration, This represents the scheduling strategy that maximizes the utility function.

[0083] The policy change is defined as the Euclidean distance between the policy combination formed by the scheduling policies of all participants after the current iteration update and the policy combination formed by the scheduling policies of all participants before the current iteration update.

[0084] The content that identifies conflicting nodes in an equilibrium state includes:

[0085] The total amount of resources is obtained by summing the resources in the scheduling strategy corresponding to each participant when Nash equilibrium is reached, and then comparing the total amount of resources with the resource capacity of the corresponding resource pool.

[0086] Based on the utility function, the ideal utility and equilibrium utility for each participant are calculated. Ideal utility refers to the value of the participant's utility function in a state of exclusive resource ownership (i.e., without competition from other participants for resources). Equilibrium utility is the value of the utility function of a participant when Nash equilibrium is reached; the difference between the ideal utility and the equilibrium utility for each participant is calculated to obtain the utility difference; each utility difference is compared with a utility threshold, which is preset by a person skilled in the art based on the actual situation;

[0087] If the total amount of resources exceeds the resource capacity, or if there is a utility difference greater than or equal to the utility threshold, or if the number of iterations exceeds the preset number of iterations threshold, then all tree nodes in the corresponding sibling set will be marked as conflicting nodes; the number of iterations threshold will be preset by those skilled in the art based on the actual situation.

[0088] If the total resource amount is less than or equal to the resource capacity, and all utility differences are less than the utility threshold, and Nash equilibrium is reached, then the tree nodes in the corresponding sibling set will not be marked as conflict nodes.

[0089] The incentive design module is used to determine the hierarchical position and dilemma type of each conflict node in sequence, and to design differentiated incentive strategies for different conflict nodes based on the hierarchical position and dilemma type.

[0090] The content that determines the hierarchical position of each conflict node in sequence includes:

[0091] Mark the topmost tree node in the competing nested tree structure as the root node. Starting from the level corresponding to the root node in the competing nested tree structure, assign an incrementing numerical label to each level from top to bottom and mark it as the level label. The level label of the level corresponding to the root node is 0, and the level label of the level corresponding to the tree node directly connected to the root node is 1. Obtain the level label of each conflicting node from the competing nested tree structure in sequence.

[0092] Count the number of child nodes corresponding to each conflict node to obtain the size of the child nodes of each conflict node; where the number of child nodes is equal to the sum of the number of direct child nodes and the number of indirect child nodes. Direct child nodes refer to the next level tree nodes directly connected to the conflict node, and indirect child nodes refer to the next level tree nodes connected to the conflict node through one or more parent-child relationships.

[0093] For example, tree node A is connected to tree node B and tree node C respectively, and tree node B is connected to tree node B1 and tree node B2; therefore, the direct child nodes of tree node A are tree node B and tree node C, and the indirect child nodes are tree node B1 and tree node B2.

[0094] Extract the overall competitive intensity between each conflicting node and all other tree nodes from the competitive relationship matrix, and count the number of tree nodes with a non-zero overall competitive intensity corresponding to each conflicting node to obtain the competitive correlation degree of each conflicting node;

[0095] Different influence weights are set for hierarchical labels, child node size, and competitive correlation, and the influence range index of each conflict node is obtained by weighted summation of hierarchical labels, child node size, and competitive correlation based on the influence weights; wherein, the influence weights are preset by those skilled in the art according to the actual situation.

[0096] The influence range index of each conflict node is compared with the preset classification threshold in turn. The classification threshold includes the core threshold and the edge threshold. The classification threshold is preset by those skilled in the art according to the actual situation.

[0097] If the scope of influence index is greater than the core threshold, the corresponding conflict node is determined to be at the core level, indicating that the corresponding conflict node is located at a key position in the competing nested tree structure and has a wide scope of influence.

[0098] If the influence range index is less than or equal to the core threshold and greater than the edge threshold, then the corresponding conflict node is determined to be at the middle level, indicating that the corresponding conflict node has a certain influence but is not the core.

[0099] If the influence range index is less than or equal to the edge threshold, the corresponding conflict node is determined to be at the edge level, indicating that the corresponding node is located at the edge of the competing nested tree structure and has a limited influence range.

[0100] The content for determining the dilemma type of each conflict node in sequence includes:

[0101] Mark the sibling set of each conflict node as a conflict set, obtain the total resource amount corresponding to each conflict set; calculate the ratio between the total resource amount of each conflict set and the total resource amount of the corresponding resource pool to obtain the resource overload of each conflict set.

[0102] Based on the execution time of each conflict node in the scheduling strategy when reaching Nash equilibrium, the time overlap length between every two conflict nodes in each conflict set is calculated sequentially. The time overlap lengths corresponding to the same conflict set are added together to obtain the total time overlap of each conflict set. Based on the execution time of each conflict node in each conflict set, the task time length of each conflict set is calculated. The ratio between the total time overlap of each conflict set and the corresponding task time length is calculated to obtain the time conflict degree of each conflict set. The calculation process for the task time length is as follows: from all execution times corresponding to the conflict set, the earliest time (earliest start time) and the latest time (latest end time) are obtained. The difference between the latest time and the earliest time is calculated to obtain the task time length.

[0103] Based on the execution priority of each conflict node in the scheduling policy when reaching Nash equilibrium, the standard deviation of all execution priorities corresponding to each conflict set is calculated sequentially to obtain the priority dispersion of each conflict set.

[0104] The resource overload, task duration, and priority dispersion of each conflict set are compared with preset dilemma thresholds, including overload threshold, time threshold, and dispersion threshold. The dilemma thresholds are all preset by those skilled in the art based on the actual situation.

[0105] If the resource overload is greater than the overload threshold, an overload instruction is generated for the corresponding conflict set; if the time conflict is greater than the time threshold, a time instruction is generated for the corresponding conflict set; if the priority dispersion is less than the dispersion threshold, a dispersion instruction is generated for the corresponding conflict set.

[0106] If only overload instructions are generated, then the dilemma type of each conflict node in the corresponding conflict set is determined to be a resource overload dilemma, indicating that the main contradiction is insufficient resources;

[0107] If only time instructions are generated, then the dilemma type of each conflict node in the corresponding conflict set is determined to be a time conflict dilemma, indicating that the main contradiction is excessive time overlap;

[0108] If only distributed instructions are generated, the dilemma type of each conflict node in the corresponding conflict set is determined to be a priority disorder dilemma, indicating that the main contradiction is that the priority setting is unreasonable;

[0109] If at least two of the following instructions are generated: overload instruction, time instruction, and distributed instruction, then the dilemma type of each conflict node in the corresponding conflict set is determined to be a comprehensive conflict dilemma, indicating the existence of multiple contradictions.

[0110] The content of designing differentiated incentive strategies for different conflict nodes includes:

[0111] For conflict nodes corresponding to resource overload dilemmas, design resource reduction incentive strategies to encourage enterprise tasks corresponding to conflict nodes to reduce the amount of resources while ensuring minimum resource requirements.

[0112] For time-conflict type dilemmas, design time adjustment incentive strategies to encourage the enterprise tasks corresponding to the conflict nodes to shift their execution time within an acceptable delay period.

[0113] For conflict nodes corresponding to priority confusion dilemmas, a priority negotiation incentive strategy is designed to encourage the enterprise tasks corresponding to the conflict nodes to renegotiate their execution priorities based on their actual importance, thereby widening the priority gap.

[0114] For the conflict nodes corresponding to comprehensive conflict-type dilemmas, design combined incentive strategies;

[0115] Based on the hierarchical position of each conflict node, the corresponding incentive strategy is adjusted to obtain a differentiated incentive strategy for each conflict node.

[0116] The specific expression for the resource reduction incentive strategy is:

[0117] ;

[0118] In the formula, Indicates the intensity of incentives for resource reduction. Indicates the incentive coefficient for resource reduction. Indicates the amount of resource reduction. This indicates the adjusted resource quantity. Indicates the quantity required;

[0119] The specific expression for the time-adjusted incentive strategy is:

[0120] ;

[0121] In the formula, Indicates that the incentive intensity is adjusted over time. Indicates the time-adjusted incentive coefficient. Indicates the amount of time adjustment. This indicates the adjusted delay duration (i.e., the difference between the adjusted end time and the original end time). Indicates the acceptable delay duration;

[0122] The specific expression for the priority negotiation incentive strategy is:

[0123] ;

[0124] In the formula, Indicates the strength of priority negotiation incentives. This represents the priority negotiation incentive coefficient. Indicates the priority adjustment amount. This indicates the adjusted priority dispersion. This indicates the priority dispersion before adjustment;

[0125] It should be noted that the incentive coefficients for resource reduction, time adjustment, and priority negotiation are preset by those skilled in the art based on factors such as corporate strategic goals and industry experience.

[0126] The combined incentive strategy is as follows: different incentive weights are set for the incentive intensity of resource reduction, the incentive intensity of time adjustment, and the incentive intensity of priority negotiation. The incentive weights are preset by those skilled in the art based on the actual situation. Based on the incentive weights, the incentive intensity of resource reduction, the incentive intensity of time adjustment, and the incentive intensity of priority negotiation are weighted and summed to obtain the combined incentive intensity.

[0127] Different level adjustment coefficients are set for different level positions. The level adjustment coefficients are preset by those skilled in the art according to the actual situation. Among them, the level adjustment coefficient corresponding to the core level is greater than 1, the level adjustment coefficient corresponding to the middle level is equal to 1, and the level adjustment coefficient corresponding to the edge level is less than 1. The level adjustment coefficient corresponding to the level position of each conflict node is multiplied by the corresponding incentive strategy to obtain the differentiated incentive strategy for each conflict node.

[0128] The equilibrium optimization module uses the incentive parameters in the differentiated incentive strategy as decision variables to perform global optimization on the competitive nested tree structure, obtain the Pareto optimal solution set, and select the resource scheduling scheme with the highest comprehensive optimization score from the Pareto optimal solution set as the equilibrium scheduling scheme.

[0129] Obtaining the Pareto optimal solution set includes:

[0130] The incentive parameters in the differentiated incentive strategy are used as decision variables. The incentive parameters include resource reduction amount, time adjustment amount, and priority adjustment amount.

[0131] For each enterprise task, the resource reduction amount is calculated as follows: calculate the difference between the corresponding demand quantity and the resource quantity in the scheduling plan to obtain the resource reduction amount; the time adjustment amount is calculated as follows: calculate the average of the start and end times corresponding to the corresponding demand period to obtain the intermediate time; calculate the average of the start and end times corresponding to the execution time in the scheduling plan to obtain the execution time; calculate the difference between the execution time and the intermediate time to obtain the time adjustment amount; the priority adjustment amount is calculated as follows: calculate the difference between the corresponding task priority and the execution priority in the scheduling plan to obtain the priority adjustment amount.

[0132] The scheduling scheme is the same as the scheduling strategy, that is, the scheduling scheme also includes the number of resources, execution time and execution priority.

[0133] Construct a multi-objective optimization function, which includes an objective function to minimize the utility difference, an objective function to minimize the utility variance, and an objective function to minimize the incentive cost;

[0134] The utility difference minimization objective function aims to minimize the sum of utility differences for all enterprise tasks when resources are allocated according to the corresponding scheduling scheme, thereby reducing the overall gap between the actual utility and ideal utility of each enterprise task. The calculation process of the utility difference is as follows: based on the utility function, calculate the ideal utility and actual utility corresponding to the enterprise task, where the actual utility is the value of the utility function when the enterprise task is allocated resources according to the corresponding scheduling scheme; calculate the difference between the ideal utility and the actual utility to obtain the utility difference.

[0135] The utility variance minimization objective function is used to minimize the variance of the utility difference between enterprise tasks when resources are scheduled according to the corresponding scheduling scheme, so as to reduce the utility difference between enterprise tasks and make resource allocation more fair and consistent.

[0136] The incentive cost minimization objective function aims to minimize the sum of incentive costs acquired by all enterprise tasks when resource scheduling is performed according to the corresponding scheduling scheme, thereby reducing the total incentive cost required to implement the scheduling scheme. The calculation process of incentive cost is as follows: different cost coefficients are set for different incentive strategies, and the cost coefficients are preset by those skilled in the art based on the actual situation; the incentive intensity of each enterprise task is calculated based on the differentiated incentive strategy of each conflict node; the incentive intensity of each enterprise task is multiplied by the corresponding cost coefficient to obtain the incentive cost of each enterprise task.

[0137] Based on multi-objective optimization functions, decision variables, and resource constraints, multi-objective evolutionary algorithms (such as NSGA-II, NSGA-III, and MOEA / D algorithms) are used to perform global optimization on a competitive nested tree structure to obtain a Pareto optimal solution set. The Pareto optimal solution set includes multiple resource scheduling schemes, each of which includes a scheduling scheme corresponding to each enterprise task. It should be noted that multi-objective evolutionary algorithms are existing technologies, and the specific process will not be elaborated upon here.

[0138] The calculation of the comprehensive optimization score includes:

[0139] Obtain the objective function value corresponding to each resource scheduling scheme in the Pareto optimal solution set; the objective function value includes the objective function value for minimizing the utility difference, the objective function value for minimizing the utility variance, and the objective function value for minimizing the incentive cost; normalize the objective function value of each resource scheduling scheme to obtain the standard objective value; set different objective weights for different standard objective values, and calculate the weighted sum of the standard objective values ​​of each resource scheduling scheme based on the objective weights to obtain the comprehensive optimization score of each resource scheduling scheme.

[0140] It should be noted that during the normalization process, the direction of each objective function value is adjusted sequentially, that is, all objective function values ​​are uniformly converted into the form of "the larger the value, the better", so as to facilitate the subsequent weighted summation calculation of the comprehensive optimization score; the normalization process is an existing technology, and the specific process will not be described in detail here; the objective weights are preset by those skilled in the art according to the actual situation.

[0141] This embodiment constructs a competitive nested tree structure to clearly express the hierarchical and competitive relationships of resource demands among enterprise tasks, overcoming the problem of traditional methods' inadequate expression of complex relationships between tasks. It introduces a game theory-based multi-party game model and Nash equilibrium solution mechanism, accurately identifying conflict nodes and their utility losses during resource allocation. This is more accurate and consistent with actual competitive scenarios compared to traditional methods that judge conflict solely based on resource capacity. It proposes a method to design differentiated incentive strategies based on the hierarchical position and dilemma type of conflict nodes, employing resource reduction incentives, time adjustment incentives, priority negotiation incentives, or combined incentive strategies for conflicts with different characteristics, avoiding a "one-size-fits-all" scheduling approach. A multi-objective optimization framework including minimizing utility difference, utility variance, and incentive cost is adopted, using a multi-objective evolutionary algorithm for global optimization on the competitive nested tree structure to achieve a balance between resource utilization efficiency, scheduling fairness, and incentive cost. This systematically solves the problem of reduced resource allocation efficiency caused by the game-playing behavior of task subjects in enterprise resource scheduling, guiding task subjects to shift from maximizing local interests to global collaboration, significantly improving the overall resource utilization level and operational efficiency of the enterprise.

[0142] Example 2

[0143] This application also provides an electronic device. The electronic device may include one or more processors and one or more memories. The memories store computer-readable code that, when executed by the one or more processors, can perform an enterprise resource scheduling optimization system based on a nested tree structure as described above.

[0144] The methods or systems according to embodiments of this application can also be implemented using the architecture of the electronic device shown in this application. The electronic device may include a bus, one or more CPUs, ROM, RAM, a communication port connected to a network, input / output, a hard disk, etc. The storage device in the electronic device, such as ROM or hard disk, may store an enterprise resource scheduling optimization system based on a nested tree structure provided in this application. Furthermore, the electronic device may also include a user interface. Of course, the architecture shown in this application is merely exemplary; when implementing different devices, one or more components in the electronic device shown in this application may be omitted according to actual needs.

[0145] Example 3

[0146] One embodiment of this application discloses a computer-readable storage medium. The computer-readable storage medium stores computer-readable instructions. When the computer-readable instructions are executed by a processor, an enterprise resource scheduling optimization system based on a nested tree structure according to an embodiment of this application, as described with reference to the above figures, can be executed. The storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0147] Furthermore, according to embodiments of this application, the processes described above with reference to the example figures can be implemented as computer software programs. For example, this application provides a non-transitory machine-readable storage medium storing machine-readable instructions that can be executed by a processor to perform instructions corresponding to the method steps provided in this application, such as an enterprise resource scheduling optimization system based on a nested tree structure. When this computer program is executed by a central processing unit (CPU), it performs the functions defined in the method of this application.

[0148] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0149] All formulas in this manual are dimensionless and calculated numerically. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0150] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A nested tree structure based enterprise resource scheduling optimization system, characterized in that, The method comprises the following steps: a resource collection module is used to collect resource demand data of each enterprise task; a competition quantification module is used to screen out a corresponding competition task set for each pre-constructed resource pool according to the resource demand data, and calculate the competition intensity between each two enterprise tasks in each competition task set in turn to construct a competition relationship matrix; a nested construction module is used to map each enterprise task to a tree node, identify the parent-child nodes and sibling nodes between the tree nodes according to the resource demand data and the competition relationship matrix, and construct a competition nested tree structure based on the parent-child nodes and the sibling nodes; a conflict identification module is used to establish a multi-party game model between each group of sibling nodes, and iteratively solve the Nash equilibrium based on a pre-constructed utility function to identify the conflict nodes in the equilibrium state; an incentive design module is used to determine the hierarchical position and dilemma type of each conflict node in turn, and design different incentive strategies for different conflict nodes based on the hierarchical position and the dilemma type; an equilibrium optimization module is used to take the incentive parameters in the different incentive strategies as decision variables, perform global optimization on the competition nested tree structure, obtain a Pareto optimal solution set, and obtain a resource scheduling scheme with the highest comprehensive optimization score from the Pareto optimal solution set as an equilibrium scheduling scheme.

2. The enterprise resource scheduling optimization system based on nested tree structure according to claim 1, characterized in that, The resource demand data includes demand type, demand quantity, demand period, acceptable delay duration, and task priority; The content of screening out the corresponding competition task set for each resource pool includes: For each resource pool, integrate the enterprise tasks with the same demand type to obtain the enterprise task set of each resource pool; construct the task period according to the demand period corresponding to all enterprise tasks; uniformly divide the task period according to a preset period span to obtain a plurality of subdivided periods; Determine the subdivided period corresponding to each enterprise task according to the demand period corresponding to each enterprise task; divide the enterprise tasks in the enterprise task set corresponding to each resource pool according to the corresponding subdivided period to obtain a plurality of period task sets corresponding to each resource pool; Calculate the total demand of each resource pool in each subdivided period according to the demand quantity corresponding to the enterprise tasks with the same demand type in each period task set corresponding to each resource pool; compare the resource capacity of each resource pool with the total demand in each subdivided period respectively; if the resource capacity is less than or equal to the total demand, it is determined that there is resource competition, and the period task set corresponding to the corresponding resource pool is taken as the competition task set.

3. The enterprise resource scheduling optimization system based on nested tree structure according to claim 2, characterized in that, The content of constructing the competition relationship matrix includes: Integrate each two enterprise tasks in each competition task set in turn to obtain a plurality of calculation sets; calculate the time overlap length of each calculation set according to the demand period corresponding to each enterprise task in each calculation set; calculate the resource bottleneck degree of each calculation set according to the time overlap length of each calculation set and the resource capacity of the corresponding resource pool; calculate the competition total of each calculation set, and calculate the competition intensity of each calculation set in combination with the resource bottleneck degree; Traverse all the calculation sets, integrate the calculation sets containing two enterprise tasks which are the same, and obtain multiple groups of comprehensive calculation sets; calculate the comprehensive competition strength of each group of comprehensive calculation sets according to the competition strength of each group of calculation sets corresponding to each group of comprehensive calculation sets; construct an initial matrix, and add the comprehensive competition strength corresponding to each group of comprehensive calculation sets into the initial matrix according to the enterprise task corresponding to each group of comprehensive calculation sets, to construct a competition relationship matrix.

4. The enterprise resource scheduling optimization system based on nested tree structure according to claim 3, characterized in that, The content of identifying the parent-child nodes and sibling nodes between the tree nodes includes: Integrate each two tree nodes respectively to obtain multiple groups of node sets; identify the containing relationship of each group of node sets in turn, and mark the node sets with the identified containing relationship as parent-child sets, the tree nodes in each parent-child set being parent-child nodes; for each parent-child set, take the tree node covering another tree node in the demand period and the demand type as a parent node, and take the tree node not as a parent node as a child node; Mark all the node sets not marked as parent-child sets as analysis sets, and obtain the comprehensive competition strength corresponding to each group of analysis sets from the competition relationship matrix in turn; if the comprehensive competition strength is not 0, mark the corresponding analysis set as a sibling set, and the tree nodes in each sibling set being sibling nodes.

5. The enterprise resource scheduling optimization system based on nested tree structure according to claim 1, characterized in that, The content of establishing a multi-party game model includes: regard the tree nodes in the sibling set as participants, each participant selecting in a strategy space while being limited by a resource constraint condition; the strategy space includes a quantity space, a time space and a priority space; the resource constraint condition includes a quantity constraint, a time constraint, a priority constraint and an allocation rule; through participant definition, strategy space setting and resource constraint condition limitation, form a multi-party game model; The content of iteratively solving the Nash equilibrium based on the pre-constructed utility function includes: each participant randomly selects a scheduling strategy in the strategy space, the scheduling strategy including resource quantity, execution time and execution priority; preset a convergence threshold, iteratively update the scheduling strategy of each participant based on the utility function, calculate the strategy variation amount when each round of iterative updating is completed, and compare the strategy variation amount with the convergence threshold; if the strategy variation amount is less than the convergence threshold, the strategy combination formed by the scheduling strategies corresponding to all participants reaches the Nash equilibrium, and the iterative updating ends; if the strategy variation amount is greater than or equal to the convergence threshold, the strategy combination formed by the scheduling strategies corresponding to all participants does not reach the Nash equilibrium, and the iterative updating continues.

6. The enterprise resource scheduling optimization system based on nested tree structure according to claim 5, wherein, The content of identifying the conflict nodes in the equilibrium state includes: add the resource quantity in the scheduling strategy corresponding to each participant when reaching the Nash equilibrium in turn to obtain the total resource quantity, and compare the total resource quantity with the resource capacity of the corresponding resource pool; calculate the ideal utility and the equilibrium utility corresponding to each participant based on the utility function; calculate the difference between the ideal utility and the equilibrium utility corresponding to each participant to obtain the utility difference; compare each utility difference with the utility threshold respectively; If the total resource amount is greater than the resource capacity, or there is a utility difference greater than or equal to the utility threshold, or the number of iteration updates is greater than the preset number of rounds threshold, the tree nodes in the corresponding sibling set are marked as conflict nodes; If the total resource amount is less than or equal to the resource capacity, and all utility differences are less than the utility threshold, and the Nash equilibrium is reached, the tree nodes in the corresponding sibling set are not marked as conflict nodes.

7. The enterprise resource scheduling optimization system based on nested tree structure according to claim 1, wherein, The content of sequentially determining the hierarchical position of each conflict node includes: Marking the tree node located at the uppermost layer in the competition nested tree structure as a root node, setting an incremental digital label for each layer from top to bottom, and marking it as a hierarchical label, taking the corresponding layer of the root node in the competition nested tree structure as the starting point; sequentially obtaining the hierarchical label of each conflict node from the competition nested tree structure; and counting the size of the child nodes corresponding to each conflict node and the competition correlation degree. Setting different influence weights for the hierarchical label, the size of the child nodes, and the competition correlation degree, and performing weighted summation calculation on the hierarchical label, the size of the child nodes, and the competition correlation degree of each conflict node based on the influence weights to obtain the influence range index of each conflict node. Sequentially comparing the influence range index of each conflict node with a preset classification threshold, the classification threshold including a core threshold and an edge threshold; if the influence range index is greater than the core threshold, the hierarchical position of the corresponding conflict node is determined as a core layer; if the influence range index is less than or equal to the core threshold and greater than the edge threshold, the hierarchical position of the corresponding conflict node is determined as an intermediate layer; if the influence range index is less than or equal to the edge threshold, the hierarchical position of the corresponding conflict node is determined as an edge layer.

8. The enterprise resource scheduling optimization system based on nested tree structure according to claim 1, wherein, The content of sequentially determining the dilemma type of each conflict node includes: Marking each sibling set in which each conflict node is located as a conflict set, sequentially calculating the resource overload degree, task time length, and priority dispersion degree of each conflict set, and comparing them with a preset dilemma threshold, the dilemma threshold including an overload threshold, a time threshold, and a dispersion threshold; if the resource overload degree is greater than the overload threshold, an overload instruction is generated for the corresponding conflict set; if the time conflict degree is greater than the time threshold, a time instruction is generated for the corresponding conflict set; if the priority dispersion degree is less than the dispersion threshold, a dispersion instruction is generated for the corresponding conflict set; If only the overload instruction is generated, the dilemma type of each conflict node in the corresponding conflict set is determined as a resource overload type dilemma; if only the time instruction is generated, the dilemma type of each conflict node in the corresponding conflict set is determined as a time conflict type dilemma; if only the dispersion instruction is generated, the dilemma type of each conflict node in the corresponding conflict set is determined as a priority confusion type dilemma; if at least two of the overload instruction, the time instruction, and the dispersion instruction are generated, the dilemma type of each conflict node in the corresponding conflict set is determined as a comprehensive conflict type dilemma.

9. The enterprise resource scheduling optimization system based on nested tree structure according to claim 8, wherein, The content of designing different incentive strategies for different conflict nodes includes: For the conflict nodes corresponding to the resource overload type dilemma, a resource reduction incentive strategy is designed; For the conflict nodes corresponding to the time conflict type dilemma, a time adjustment incentive strategy is designed; For the conflict nodes corresponding to the priority confusion type dilemma, a priority negotiation incentive strategy is designed; For the conflict nodes corresponding to the comprehensive conflict type dilemma, a combination incentive strategy is designed; Different hierarchical adjustment coefficients are set for different hierarchical positions, and the hierarchical adjustment coefficient of each conflict node corresponding to the hierarchical position is multiplied by the corresponding incentive strategy to obtain the differentiated incentive strategy of each conflict node.

10. The enterprise resource scheduling optimization system based on nested tree structure according to claim 1, wherein, The content of obtaining the pareto optimal solution set includes: The incentive parameters include resource reduction, time adjustment and priority adjustment; a multi-objective optimization function is constructed, and the multi-objective optimization function includes utility difference minimization objective function, utility variance minimization objective function and incentive cost minimization objective function; Based on the multi-objective optimization function, the decision variable and the resource constraint condition, a multi-objective evolutionary algorithm is used to perform global optimization on the competition nested tree structure to obtain the pareto optimal solution set; wherein the pareto optimal solution set includes multiple resource scheduling schemes, and each resource scheduling scheme includes the scheduling scheme corresponding to each enterprise task.

Images