A device rotation scheduling multi-objective decision-making method based on conflict analysis graph model

Abstract

The present application relates to the technical field of equipment scheduling management, and particularly relates to a device rotation scheduling multi-objective decision-making method based on a conflict analysis graph model; the method comprises the following steps: constructing a scheduling conflict model, defining decision makers and their strategy options, and generating a feasible state set; establishing a target system for each decision maker, and quantitatively evaluating the consequences of each target of each feasible state; based on the multi-attribute utility theory, the comprehensive utility value of each party to each feasible state is calculated through standardization and weighted aggregation, and then individual preference ranking is generated; the obtained preference ranking is imported into the conflict analysis graph model, the stability of each feasible state to each decision maker is analyzed, and finally a globally balanced scheduling scheme set that is stable to all decision makers is solved; the present application realizes accurate characterization of multi-party game behavior, scientific and quantitative trade-off of multi-target conflict, and can automatically generate a stable, interpretable and consensus scheduling scheme, thereby providing systematic and transparent intelligent decision support for complex device rotation scheduling.

Description

Technical Field

[0001] This invention relates to the field of equipment scheduling and management technology, and specifically to a multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model. Background Technology

[0002] In the context of modern enterprise and organizational management, high-performance, high-value specialized equipment is a key element supporting the overall operational system. Such equipment is typically expensive, limited in quantity, and difficult to fully deploy. Therefore, a scientific equipment rotation and scheduling mechanism must be established to maximize the overall utilization efficiency of equipment and system availability. The rotation and scheduling process involves multiple stakeholders, including downstream business departments that submit usage requests, midstream technical support teams responsible for maintenance and support, and upstream management departments that coordinate the overall process. These stakeholders have different and often conflicting goals in scheduling: downstream departments pursue higher equipment availability, longer continuous usage time, and lower failure rates; midstream teams focus on the rational allocation of maintenance resources and the balance of workload; and upstream departments need to coordinate overall system efficiency, ensure fair scheduling, and guarantee responsiveness to high-priority tasks.

[0003] Traditional equipment scheduling methods are mostly based on static rules or simple optimization models, such as linear programming and heuristic algorithms. These methods have the following limitations: First, they usually treat each participant as a passive executor, ignoring their strategic interaction as rational decision-makers, thus making it difficult to accurately characterize and predict the game behavior that each party may take during the scheduling process; Second, when multiple conflicting objectives are involved, they often rely on subjective weighting or simple aggregation, making it difficult to give a complete, consistent, and interpretable priority ranking for all feasible solutions; Third, they have poor interpretability, failing to clearly explain the underlying reasons for choosing a certain solution, which is not conducive to identifying the root cause of the problem and finding negotiation space.

[0004] To overcome the aforementioned limitations, particularly the inadequacy in modeling multi-party strategic interactions, an analytical framework capable of characterizing the rational game-theoretic behavior of decision-makers is needed. Among these, conflict analysis graph models are rigorous mathematical tools derived from game theory, used to analyze and resolve strategic conflicts. By defining decision-makers, strategy options, feasible states, and state transition relationships, they can intuitively describe the structure of multi-party games. However, when traditional conflict analysis graph models are applied to equipment rotation scheduling problems, they typically require decision-makers to directly provide a complete preference ranking for all possible states. This requirement is difficult to achieve in complex scheduling scenarios involving multiple objectives and multiple participants. On the one hand, decision-makers struggle to directly form a global, consistent preference ranking in an environment with conflicting multiple objectives, resulting in strong subjectivity and significant practical difficulties. On the other hand, the models fail to effectively integrate with mature multi-objective decision-making theories in resource scheduling, and cannot systematically handle the quantitative trade-offs and comprehensive evaluation needs involved in the scheduling process.

[0005] Therefore, there is an urgent need in this field to develop a systematic method that can simultaneously and accurately characterize game theory, quantify multi-objective trade-offs, and automatically generate stable solutions. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention proposes a multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model. This method integrates multi-attribute decision theory with the conflict analysis graph model, aiming to scientifically deduce decision-makers' preferences and systematically analyze conflict stability, thereby providing transparent, interpretable, and balanced decision support for complex scheduling problems.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] This invention proposes a multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model, comprising the following steps:

[0009] S1. Determine the set of decision-makers in the equipment rotation scheduling decision; define one or more independent strategy options for each decision-maker in the set of decision-makers, wherein the strategy options are binary strategy options; combine all strategy options of all decision-makers to generate an initial state set; filter the initial state set according to preset logical constraints, remove infeasible states, and obtain a feasible state set consisting of all feasible states.

[0010] S2. Establish a target system for each decision-maker. For each feasible state obtained in S1, quantitatively predict the achievement of each decision-maker's target under that feasible state, and obtain the quantitative consequence vector of each decision-maker-state pair.

[0011] S3. Standardize each quantified consequence vector obtained in S2 to obtain the standardized consequence value of each decision-maker-state pair on each objective; determine the weight vector corresponding to the objective system of each decision-maker; based on the standardized consequence value and weight vector, calculate the comprehensive utility value of each decision-maker for each feasible state, and generate the preference ranking of each decision-maker for all feasible states.

[0012] S4. Import all preference rankings obtained in S3 into the conflict analysis graph model, define the unilateral movement rules for each decision-maker between feasible states, and construct a global state transition graph. Based on the global state transition graph, examine the stability of each feasible state for each decision-maker according to the definitions of Nash stability, general meta-rational stability, symmetric meta-rational stability, and sequential stability. If a feasible state satisfies at least one definition of stability for a decision-maker, then the feasible state is determined to be stable for that decision-maker. If a feasible state is stable for all decision-makers, then the feasible state is determined to be a global equilibrium scheduling scheme. All global equilibrium scheduling schemes constitute a global equilibrium scheduling scheme set.

[0013] Furthermore, in S1, the decision-making set includes at least the downstream business department that proposes the usage requirements, the midstream technical team responsible for maintenance and support, and the upstream management department that conducts overall coordination.

[0014] Furthermore, in S2, the target system of the downstream business department includes: business support capability, task success rate, and equipment wear and tear; the target system of the midstream technical team includes: workload balance, resource consumption cost, and preventive maintenance level; and the target system of the upstream management department includes: overall system efficiency, scheduling fairness, and emergency task guarantee capability.

[0015] Furthermore, in S3, the standardization process specifically involves: positive standardization for benefit-oriented objectives and negative standardization for cost-oriented objectives.

[0016] Furthermore, in S1, the preset logical constraints include at least one of mutual exclusion rules, dependency rules, and decision consistency rules.

[0017] Furthermore, in S2, the quantitative prediction is implemented by: joint prediction based on historical operating data and simulation models, or by expert scoring based on preset rules.

[0018] Furthermore, in S3, the weight vector is determined by the analytic hierarchy process, the entropy weight method, or the direct assignment method.

[0019] Furthermore, in S3, the comprehensive utility value is calculated using a linear weighted sum model.

[0020] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0021] (1) This invention introduces a conflict analysis graph model, defining each participant, traditionally regarded as a passive executor, as a decision-maker with autonomous choice. Through systematic modeling, a set of feasible states covering all possible strategy combinations is generated, and unilateral movement rules between feasible states are defined according to strategy options, thereby constructing a global state transition graph that intuitively presents the dynamic evolution process of multi-party strategy interaction and conflict. This fundamentally overcomes the inherent defect of traditional static optimization models in failing to reflect the strategic game behavior of decision-makers, and provides a foundation for predicting the conflict evolution in complex scheduling scenarios.

[0022] (2) This invention addresses the diverse and conflicting decision-making objectives in equipment rotation scheduling by constructing a target system directly related to the core interests of each decision-maker. Based on this target system, a multi-dimensional consequence quantification assessment is performed on each feasible state. Furthermore, based on multi-attribute utility theory, standardization eliminates dimensional differences, and weighted aggregation calculations are used to objectively derive the preference ranking of each decision-maker for all feasible states. This method transforms the traditional fuzzy trade-offs relying on subjective experience into data-driven, logically clear scientific calculations, greatly improving the objectivity of decision-making and the transparency and traceability of the decision-making process.

[0023] (3) This invention imports preference ranking into a conflict analysis graph model and comprehensively applies Nash stability, general meta-rational stability, symmetric meta-rational stability, and sequential stability for systematic analysis. It can automatically identify and solve a set of globally balanced scheduling schemes that are stable for all decision-makers. The schemes in the set of globally balanced scheduling schemes represent the compromise results that are most likely to be accepted and stably executed by all decision-makers, thereby effectively breaking the decision-making deadlock caused by goal conflict. It not only provides managers with a clear set of preferred schemes, but also deeply reveals the stable logic and compromise space behind the schemes, realizing the improvement from experience-based decision-making to intelligent and consensus-based decision-making. Attached Figure Description

[0024] Figure 1 This is an overall flowchart of the multi-objective decision-making method for equipment rotation scheduling proposed in this invention;

[0025] Figure 2 This is the global state transition diagram in an embodiment of the present invention;

[0026] Figure 3 This is a unilateral promotion state transition diagram in an embodiment of the present invention. Detailed Implementation

[0027] 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.

[0028] Example

[0029] This embodiment uses a simplified equipment rotation scheduling scenario as an example to further illustrate the method proposed in this invention. Assume an organization has two sets of equipment (equipment 1 and equipment 2), two downstream business departments (department A and department B), a midstream technical team (department M), and an upstream management department (department C). Equipment 1 and equipment 2 need to be rotated between department A and department B according to a plan. Department M is responsible for the maintenance and upkeep of both equipment, while department C is responsible for the scheduling approval and coordination decisions for both equipment. Currently, equipment 1's usage period in department A is about to end, and it needs to be rotated to department B according to the plan. Simultaneously, equipment 2's usage period in department B is about to end, and it needs to be rotated to department A according to the plan. During this process, conflicting requests such as extended usage or early maintenance may arise from various parties, making it difficult to execute according to the plan.

[0030] To solve the aforementioned multi-objective conflict decision-making problem, refer to Figure 1 This embodiment proposes a multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model, which is implemented through the following steps:

[0031] S1. Define a binary strategy option that can be independently selected for each decision-maker in the equipment rotation scheduling decision. Obtain a set of feasible states based on all binary strategy options of all decision-makers. This includes the following sub-steps:

[0032] S101. Determine the set of decision-makers D in the equipment rotation scheduling decision. In this embodiment, D = {Department A, Department B, Department M, Department C}.

[0033] S102. Define a binary strategy option that can be chosen independently for each decision-maker in the decision-maker set D. In this embodiment, the strategy options for each decision-maker are shown in Table 1.

[0034] Table 1. Decision Makers and Their Binary Strategy Options

[0035]

[0036] S103. Combine all 7 policy options from S102 to generate an initial state set. The initial state set contains... There are 128 possible states. However, some of these states are logically infeasible in reality. Therefore, it is necessary to filter the initial set of states through pre-defined logical constraints, eliminating infeasible states. The remaining states are all feasible states, and all feasible states constitute the feasible state set. The pre-defined logical constraints include the following three categories:

[0037] Mutual exclusion rules: Strategy option 3 and strategy option 4 cannot be selected at the same time, that is, department M cannot select recall 1 and recall 2 at the same time; strategy option 5 and strategy option 6 cannot be selected at the same time, that is, department C cannot select approval 1 and approval 2 at the same time.

[0038] Dependency rules: Policy option 5 depends on policy option 1, i.e., approval 1 depends on deferred A; policy option 6 depends on policy option 2; policy option 7 depends on policy option 3 or policy option 4.

[0039] Decision consistency rule: If department C chooses to approve a department's extension application for a certain equipment, it will not approve the recall of that equipment, and vice versa.

[0040] In this embodiment, after filtering the 128 states in the initial state set according to preset logical constraints, 15 feasible states remain. These 15 feasible states constitute the feasible state set, denoted as […]. Each feasible state ( The strategy is uniquely determined by the values ​​of 7 policy options, with the following rules: Y represents selecting or adopting the policy, and N represents not selecting or not adopting the policy. The policy option combinations for each feasible state are shown in Table 2.

[0041] Table 2. Strategy option combinations for each feasible state

[0042]

[0043] S2. Analyze the core interests and needs of each decision-maker, and establish a target system for each decision-maker:

[0044] Department A: Primarily concerned with equipment utilization efficiency, its target system includes business support capability (B1, efficiency-oriented), task success rate (B2, efficiency-oriented), and equipment wear and tear level (Q1, cost-oriented).

[0045] Department B's goal system is the same as Department A's;

[0046] Department M: Primarily concerned with the economy and sustainability of maintenance work, its target system includes: workload balance (B1, efficiency-oriented), resource consumption cost (Q1, cost-oriented), and preventive maintenance level (B2, efficiency-oriented).

[0047] Department C: It needs to weigh the overall situation, and its target system includes: overall system effectiveness (B1, benefit-oriented), scheduling fairness (B2, benefit-oriented), and emergency task support capability (B3, benefit-oriented).

[0048] For each feasible state ( By combining historical operational data with equipment performance simulation models, the achievement of each decision-maker's objectives under this feasible state is quantified. The evaluation results are expressed as percentage scores, and the quantitative consequence vector of each decision-maker-state pair is obtained. The quantitative consequence vector of each decision-maker-state pair is shown in Table 3.

[0049] Table 3. Decision Maker-State Quantification Consequence Vector Table

[0050]

[0051] In Table 3, the column headings follow the naming rule of "decision-objective". For example, A-B1 represents the business support capability score of department A, M-Q1 represents the resource consumption cost score of department M, and so on.

[0052] S3. Based on the quantified consequence vector obtained in S2, through standardization, weighted aggregation, and other processing, calculate the comprehensive utility value of each decision-maker for each feasible state, and then generate the preference ranking of each decision-maker for all feasible states. This includes the following sub-steps:

[0053] S301. To eliminate differences in the dimensions and magnitudes of different objectives, each quantified consequence vector obtained in S2 is standardized to obtain the standardized consequence value of each decision-maker-state pair for each objective. The standardization process is as follows:

[0054] For benefit-oriented objectives, positive standardization is adopted:

[0055] ;

[0056] For cost-related objectives, negative standardization is employed:

[0057] ;

[0058] In the formula, For the first The first decision-maker The objective is feasible. The standardized consequence value, For the first The first decision-maker The objective is feasible. The original score below, For the first The first decision-maker The original minimum score for each objective across all feasible states; For the first The first decision-maker The highest original score for each objective across all feasible states.

[0059] S302 uses the analytic hierarchy process (AHP), entropy weighting method, or direct assignment method to determine the relative importance each decision-maker places on each objective, generating a weight vector corresponding to each decision-maker's objective system. This embodiment uses the direct assignment method, obtaining the following weight vector:

[0060] Weight vector corresponding to the target system of department A = (0.5, 0.3, 0.2), which means that the weight assigned to business support capability is 0.5, the weight assigned to task success rate is 0.3, and the weight assigned to equipment wear level is 0.2.

[0061] Weight vector corresponding to the target system of department B and same;

[0062] Weight vector corresponding to the target system of department M = (0.4, 0.4, 0.2), which means that the weight allocated to workload balance is 0.4, the weight allocated to resource consumption cost is 0.4, and the weight allocated to preventive maintenance level is 0.2;

[0063] Weight vector corresponding to the target system of department C = (0.4, 0.3, 0.3), which means that the weight allocated to the overall system effectiveness is 0.4, the weight allocated to the scheduling fairness is 0.3, and the weight allocated to the emergency task support capability is 0.3.

[0064] S303. Using a linear weighted sum model, calculate the comprehensive utility value for each decision-maker for each feasible state. Each decision-maker considers the feasible state. Overall utility value The calculation formula is:

[0065]

[0066] In the formula, For the first In the goal system of a decision-maker, the first The weights corresponding to each objective For the first The total number of objectives in a decision-maker's objective system.

[0067] The combined utility values ​​of each decision-maker for each feasible state obtained in this embodiment are shown in Table 4.

[0068] Table 4. Overall Utility Values ​​of Each Decision Maker

[0069]

[0070] S304. Sort each decision-maker's comprehensive utility value for all feasible states in descending order from high to low to generate a preference ranking for each decision-maker for all feasible states, as shown in Table 5.

[0071] Table 5 Ranking of Decision Makers' Preferences

[0072]

[0073] S4. Perform stability analysis on all preference rankings obtained in S3 to solve for the global equilibrium scheduling scheme set. The specific method is as follows:

[0074] S401. Import all preference rankings obtained in S3 into the conflict analysis graph model, which uses the feasible state set obtained in S1. Let be the set of nodes. Based on the strategy options of each decision-maker in Table 1, define the unilateral movement rule for each decision-maker between feasible states, that is, in any feasible state, the 1st... Each decision-maker can transition to another state in one step, and only by changing the policy options under their own control. The unilateral movement rules of all decision-makers collectively define all possible transition relationships between feasible states.

[0075] S402. Based on the unilateral movement rules defined in S401, construct a global state transition graph containing all possible transitions, such as... Figure 2 As shown. Figure 2 In this diagram, each node represents a feasible state, and directed edges represent single-step transitions from one feasible state to another, with the arrow direction controlled by the decision-maker. Subsequently, based on the imported preference ranking, one-sided boosting arcs are labeled. A one-sided boosting arc refers to a transition that allows the initiator to obtain a higher overall utility value. All one-sided boosting arcs constitute a network as shown below. Figure 3 The diagram shown depicts a unilateral improvement state transition, which fully illustrates the strategic changes that each decision-maker is willing to proactively initiate in order to improve their own situation.

[0076] S403. Based on the unilateral lift state transition diagram obtained in S402, the stability of each feasible state for each decision-maker is tested according to the following four stability definitions:

[0077] Nash stability: Each decision-maker in a feasible state If there are no unilateral improvements that can be implemented, then the condition is determined to be feasible. For the The decision-maker is Nash stable.

[0078] General Metarational Stability (GMR): The first Each decision-maker starts from the feasible state Any unilateral improvement issued may be subject to subsequent sanctions from other decision-makers, resulting in their situation not improving or worsening; in such cases, the situation is deemed feasible. For the The decision-makers are GMR stable.

[0079] Symmetric Meta-Rational Stability (SMR): Feasible states are For the first The first decision-maker is stable in GMR, and based on this, the second... Each decision-maker starts from the feasible state Any unilateral elevation issued after being subject to sanctions will be subject to the following: Even if a decision-maker initiates any unilateral improvement, it will not achieve the same result as... A better state is considered a feasible state. For the The decision-maker is SMR stable. This takes into account the scenario where the decision-maker anticipates a counterattack and attempts to mitigate the damage but is still unsuccessful.

[0080] Sequential stability (SEQ): the first Each decision-maker starts from the feasible state Any unilateral promotion issued is subject to at least one move initiated by another decision-maker, which has an impact on the third... If a decision-maker can impose sanctions that increase the overall utility of the sanctioning party, then the situation is considered feasible. For the The decision-maker is SEQ stable.

[0081] If feasible For the If each decision-maker satisfies at least one definition of stability, then a feasible state is determined. For the The decision-maker is stable.

[0082] S404. Traverse all feasible states to determine a globally balanced scheduling scheme. If a feasible state is stable for all decision-makers, then that feasible state is determined to be a globally balanced scheduling scheme. All globally balanced scheduling schemes constitute the globally balanced scheduling scheme set E.

[0083] In this embodiment, the final globally balanced scheduling scheme set E= The stability results of each global balanced scheduling scheme are shown in Table 6. The "√" in Table 6 indicates that the feasible state shown in the row is stable under the corresponding stability definition.

[0084] Table 6. Stability results of the global balanced scheduling scheme

[0085]

[0086] The specific scheduling decisions corresponding to each globally balanced scheduling scheme in the global balanced scheduling scheme set are theoretically the schemes most likely to be accepted by all parties and implemented stably. Management can select the final implementation plan from this global balanced scheduling scheme set based on strategic priorities.

[0087] The specific embodiments of the present invention are provided to enable those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention.

[0088] It should be understood that the present invention is not limited to the content already described above, and various modifications and changes can be made without departing from its scope. The scope of the present invention is limited only by the appended claims.

Claims

1. A multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model, characterized in that, Includes the following steps: S1. Determine the set of decision-makers in the equipment rotation scheduling decision; define one or more independent strategy options for each decision-maker in the set of decision-makers, wherein the strategy options are binary strategy options; combine all strategy options of all decision-makers to generate an initial state set; filter the initial state set according to preset logical constraints, remove infeasible states, and obtain a feasible state set consisting of all feasible states. S2. Establish a target system for each decision-maker. For each feasible state obtained in S1, quantitatively predict the achievement of each decision-maker's target under that feasible state, and obtain the quantitative consequence vector of each decision-maker-state pair. S3. Standardize each quantified consequence vector obtained in S2 to obtain the standardized consequence value of each decision-maker-state pair on each objective. Determine the weight vector corresponding to the target system of each decision-maker; based on the standardized consequence value and weight vector, calculate the comprehensive utility value of each decision-maker for each feasible state, and generate the preference ranking of each decision-maker for all feasible states; S4. Import all the preference rankings obtained in S3 into the conflict analysis graph model, define the one-sided movement rules of each decision-maker between feasible states, and construct a global state transition graph; based on the global state transition graph, test the stability of each feasible state for each decision-maker according to the definitions of Nash stability, general meta-rational stability, symmetric meta-rational stability and sequential stability respectively. If a feasible state satisfies at least one definition of stability for a decision-maker, then the feasible state is determined to be stable for that decision-maker; if a feasible state is stable for all decision-makers, then the feasible state is determined to be a globally balanced scheduling scheme. All global balanced scheduling schemes constitute a global balanced scheduling scheme set.

2. The multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model according to claim 1, characterized in that, In S1, the decision-making set includes at least the downstream business departments that put forward usage requirements, the midstream technical team responsible for maintenance and support, and the upstream management department that conducts overall coordination.

3. The multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model according to claim 2, characterized in that, In S2, the target system of the downstream business department includes: business support capability, task success rate, and equipment wear and tear; the target system of the midstream technical team includes: workload balance, resource consumption cost, and preventive maintenance level; and the target system of the upstream management department includes: overall system efficiency, scheduling fairness, and emergency task guarantee capability.

4. The multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model according to claim 3, characterized in that, In S3, the standardization process specifically involves: positive standardization for benefit-oriented objectives and negative standardization for cost-oriented objectives.

5. The multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model according to claim 1, characterized in that, In S1, the preset logical constraints include at least one of mutual exclusion rules, dependency rules, and decision consistency rules.

6. The multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model according to claim 1, characterized in that, In S2, the quantitative prediction is implemented by either joint prediction based on historical operating data and simulation models, or by expert scoring based on preset rules.

7. The multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model according to claim 1, characterized in that, In S3, the weight vector is determined by the analytic hierarchy process, the entropy weight method, or the direct assignment method.

8. The multi-objective decision-making method for equipment rotation scheduling based on a conflict analysis graph model according to claim 1, characterized in that, In S3, the comprehensive utility value is calculated using a linear weighted sum model.

Images