Resource allocation method, device and system based on multi-index grouping collaborative search

Abstract

The invention discloses a resource allocation method, device and system based on multi-index grouping collaborative search. The method comprises the following steps: carrying out multi-index calculation on a resource object; collaborative grouping is carried out based on multiple indexes; differential search is carried out on different groups, and a resource allocation scheme meeting constraint conditions is generated through a coordination control mechanism. By the adoption of the technical scheme, under the condition of limited computing resources, structured optimization of the constrained resource allocation problem is achieved, and the search efficiency, the solution stability and the implementation of the scheme are improved.

Description

Technical Field

[0001] This invention belongs to the field of resource allocation technology, specifically relating to a resource allocation method, apparatus, and system based on multi-index cluster collaborative search. Background Technology

[0002] With the development of information and intelligent technologies, resource allocation problems are widely modeled as computer-processable optimization problems and solved using heuristic or swarm intelligence algorithms. In constrained resource allocation scenarios, existing technologies typically employ rule-based decision-making methods or general swarm intelligence optimization methods to obtain feasible solutions under complex constraints.

[0003] Greedy algorithms, as a typical heuristic method, progressively construct solutions by selecting the best local decision at each step. While computationally efficient and easy to implement, these methods often fall into local optima due to a lack of global search capabilities. They struggle to obtain high-quality solutions in resource allocation problems with multiple constraints and objectives, especially when the objects are highly heterogeneous or the constraints are complex, where their optimization performance is significantly insufficient.

[0004] To overcome the limitations of greedy methods, existing technologies widely employ general swarm intelligence optimization methods such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). These methods perform global searches in the solution space through swarm evolution or swarm cooperation mechanisms, possessing a certain ability to escape local optima, and have been applied to various resource allocation and combinatorial optimization problems. However, existing general swarm intelligence algorithms typically treat decision variables as homogeneous objects, employing a uniform search strategy, failing to fully consider the differences between different resource objects in practical applications and their varying contributions to the overall objective.

[0005] Furthermore, in real-world resource allocation scenarios, there are often constraints such as limited computational budgets, limited adjustment ranges, and high requirements for solution stability. Existing swarm intelligence algorithms mostly focus on searching for the optimal solution in the solution space, while rarely designing specifically for solution stability, asymptotic adjustment characteristics, and policy feasibility. This leads to insufficient search efficiency under limited computational resources, or the generated solution schemes are difficult to implement smoothly in practical applications, thus reducing the practical value of the algorithm.

[0006] Therefore, existing technologies generally suffer from the following shortcomings in constrained resource allocation problems: (1) The global optimization capability of regular or greedy methods is insufficient, making it difficult to obtain high-quality solutions; (2) General-purpose swarm intelligence algorithms lack structured utilization of the differences in resource objects, resulting in limited search efficiency; (3) Failed to balance the quality, stability and feasibility of the solution within a limited computational budget.

[0007] For the reasons mentioned above, it is necessary to propose a resource allocation method and system that can group resource objects by combining multiple indicators under limited computing resources and improve optimization efficiency and solution stability through a collaborative search mechanism, so as to overcome the shortcomings of existing technologies. Summary of the Invention

[0008] To address the problems existing in the prior art, this invention provides a resource allocation method, apparatus, and system based on multi-index cluster collaborative search, which can improve the solution efficiency, solution stability, and practical feasibility of resource allocation problems under limited computing resources.

[0009] To achieve the above objectives, the present invention provides the following solution: A resource allocation method based on multi-index cluster collaborative search includes: Step S1: Obtain the set of resource objects to be configured; Step S2: For each resource object, calculate multiple evaluation indicators to describe its current matching status and potential contribution capability; Step S3: Based on real-time matching indicators and value contribution indicators, resource objects are collaboratively grouped according to preset thresholds or grouping rules; Step S4: Perform differentiated search operations for different subgroups, and adjust the search results uniformly through a coordination and control mechanism; Step S5: When the preset termination condition is met, terminate the search process and output the final resource configuration scheme as the optimization result.

[0010] Preferably, in step S2, a Real-Time Matching Index (RMI) is defined to measure the degree of matching between the current configuration status of the resource object and the target requirement. Its calculation method is shown in Equation (1): ; in, Indicates the first The current number of resource objects matched is [number]. This indicates the corresponding target demand. Meanwhile, the Value Contribution Index (VCI) is defined to measure the potential contribution capability of resource objects in the overall configuration, and its calculation method is shown in Equation (2): ; in, Indicates the first The overall contribution value of each resource object Indicates the total number of resource objects.

[0011] Preferably, in step S4... The overall optimization objective function can be expressed as equation (3): ; in, Represents resource objects Utility functions under the current configuration These are the weighting coefficients; The above optimization process must simultaneously satisfy resource constraints and adjustment constraints, as shown in equation (4): ; in, This indicates the upper limit of the total resource budget. This indicates the configuration value of the resource object in the previous iteration. Indicates the maximum allowable adjustment range; By coordinating and controlling the search results of different clusters, the final resource allocation scheme can optimize the objective function value while satisfying the constraints described in equation (4). .

[0012] The present invention also provides a resource allocation device based on multi-index cluster collaborative search, comprising: The first processing module is used to obtain a set of resource objects to be configured; The second processing module is used to calculate multiple evaluation indicators for each resource object to describe its current matching status and potential contribution capability. The third processing module is used to collaboratively group resource objects based on real-time matching indicators and value contribution indicators, according to preset thresholds or grouping rules. The fourth processing module is used to perform differentiated search operations for different subgroups and to uniformly adjust the search results through a coordination and control mechanism. The fifth processing module is used to terminate the search process when the preset termination conditions are met, and output the final resource configuration scheme as the optimization result.

[0013] Preferably, the second processing module is used to define the Real-Time Matching Index (RMI) to measure the degree of matching between the current configuration state of the resource object and the target requirement. Its calculation method is shown in Equation (1): ; in, Indicates the first The current number of resource objects matched is [number]. This indicates the corresponding target demand. The second processing module is used to define the Value Contribution Index (VCI) to measure the potential contribution of resource objects in the overall configuration. Its calculation method is shown in Equation (2): ; in, Indicates the first The overall contribution value of each resource object Indicates the total number of resource objects.

[0014] As a preferred option, the fourth processing module used to optimize the objective function can be expressed as equation (3): ; in, Represents resource objects Utility functions under the current configuration These are the weighting coefficients; The above optimization process must simultaneously satisfy resource constraints and adjustment constraints, as shown in equation (4): ; in, This indicates the upper limit of the total resource budget. This indicates the configuration value of the resource object in the previous iteration. Indicates the maximum allowable adjustment range; By coordinating and controlling the search results of different clusters, the final resource allocation scheme can optimize the objective function value while satisfying the constraints described in equation (4). .

[0015] The present invention also provides a resource allocation system based on multi-index cluster collaborative search, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a resource allocation method based on multi-index cluster collaborative search when executed by the processor.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. To address the issues of insufficient global search capability and susceptibility to local optima in greedy search methods, a swarm intelligence search mechanism and structured coordination strategy are introduced to improve the overall optimization capability in multi-constraint resource allocation problems.

[0017] 2. To address the issue that general-purpose swarm intelligence algorithms fail to differentiate between resource objects and have limited search efficiency, a resource grouping mechanism based on multiple indicators is constructed. This allows different resource objects to assume differentiated search roles during the search process, thereby improving the targeting and efficiency of the search process.

[0018] 3. To address the problem that existing technologies struggle to balance solution quality and stability under limited computational budgets, a coordination and constraint control mechanism is introduced during the search process to obtain a stable and feasible resource allocation scheme under limited evaluation resources.

[0019] 4. To address the issue that solutions generated by existing optimization methods require excessive adjustments and are difficult to implement in practical applications, this paper proposes to constrain the adjustment range and optimization path to achieve incremental optimization of resource allocation schemes, thereby improving the feasibility of solutions in real-world application scenarios.

[0020] Through the above-mentioned technical means, the present invention aims to achieve efficient optimization of constrained resource allocation problems without significantly increasing computational complexity, and to provide a stable and practical technical solution for computer-aided decision support systems. Attached Figure Description

[0021] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of a resource allocation method based on multi-index cluster collaborative search according to an embodiment of the present invention; Figure 2 The average fitness value varies with problem size; Figure 3 For stability during multiple independent runs. Detailed Implementation

[0023] 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.

[0024] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] Example 1 like Figure 1As shown, this invention provides a resource allocation method based on multi-index clustering and collaborative search. By quantitatively evaluating resource objects using multiple indicators and implementing clustering and collaborative search based on the evaluation results, it achieves efficient optimization and stable output of resource allocation schemes under limited computing resources and multiple constraints. The method includes: Step S1: Resource object modeling and initialization Get the collection of resource objects to be configured Furthermore, a resource allocation decision model was established, and initial settings were made for total resource constraints, adjustment range constraints, and calculation budget constraints.

[0026] Step S2: Multi-indicator calculation and quantitative evaluation For each resource object Calculate multiple evaluation metrics to describe its current matching status and potential contribution capability.

[0027] In one embodiment, a Real-Time Matching Index (RMI) is defined to measure the degree of matching between the current configuration state of a resource object and the target requirement, and its calculation method is shown in Equation (1): ; in, Indicates the first The current number of resource objects matched is [number]. This indicates the corresponding target demand.

[0028] Meanwhile, the Value Contribution Index (VCI) is defined to measure the potential contribution capability of resource objects in the overall configuration, and its calculation method is shown in Equation (2): ; in, Indicates the first The overall contribution value of each resource object Indicates the total number of resource objects.

[0029] Step S3: Collaborative clustering based on multiple indicators Based on the real-time matching index calculated in step S2 With value contribution metrics Resource objects are collaboratively grouped according to preset thresholds or grouping rules.

[0030] In step S3, the specific process of collaborative grouping is as follows: A real-time matching threshold τ_R and a value contribution threshold τ_V are pre-set. For each resource object, its calculated RMI_i and VCI_i are compared with these dual thresholds. Based on the comparison results, all resource objects are precisely divided into four groups with different optimization strategies: The first group consists of objects with RMI_i ≥ τ_R and VCI_i ≥ τ_V, considered high-quality resources, and a fine-tuning strategy is adopted; the second group consists of objects with RMI_i ≥ τ_R but VCI_i < τ_V, considered high-matching but low-contribution resources, with the strategy of maintaining matching and improving their value; the third group consists of objects with RMI_i < τ_R but VCI_i ≥ τ_V, considered high-potential resources, with the strategy of focusing on strengthening their matching with actual needs; the fourth group consists of objects with RMI_i < τ_R and VCI_i < τ_V, considered resources to be optimized, and an exploratory strategy or restart mechanism is adopted. This grouping process is automatically completed based on quantitative indicators and explicit rules, providing a structured foundation for subsequent differentiated collaborative search.

[0031] Different clusters correspond to different search priorities and adjustment strategies to reflect the differentiated roles of resource objects in the optimization process.

[0032] Step S4: Collaborative Search and Coordination Control Different search operations are performed for different subgroups, and the search results are uniformly adjusted through a coordination and control mechanism.

[0033] In one embodiment, the overall optimization objective function can be expressed as equation (3): ; in, Represents resource objects Utility functions under the current configuration This is a weighting coefficient used to reflect the importance of different resource objects.

[0034] The above optimization process must simultaneously satisfy resource constraints and adjustment constraints, as shown in equation (4): ; in, This indicates the upper limit of the total resource budget. This indicates the configuration value of the resource object in the previous iteration. Indicates the maximum allowable adjustment range.

[0035] By coordinating and controlling the search results of different clusters, the final resource allocation scheme can optimize the objective function value while satisfying the constraints described in equation (4). .

[0036] Step S5: Termination determination and result output When the preset termination conditions are met (including reaching the maximum number of evaluations or the convergence condition of the objective function), the search process terminates and the final resource allocation scheme is output as the optimization result.

[0037] In step S5, after each round of collaborative search iteration in step S4, a dual judgment is performed. First, it checks whether the cumulative function evaluation count has reached a preset upper limit, which was initialized as a computational budget constraint in step S1. Second, it monitors the objective function value F of the global optimum. If its improvement over K consecutive iterations is less than the convergence threshold ε, it is considered converged. Once either condition is triggered, the coordination control mechanism immediately terminates the search operators for all clusters. Subsequently, all feasible solutions that fully satisfy the total resource constraint B and the adjustment magnitude constraint Δ_i are selected from the entire optimization process, and the configuration scheme {x_i} with the highest objective function value F is selected as the final output. This process ensures that the search effectively stops within limited computational resources and transforms the iterative dynamic results into deterministic optimization decisions.

[0038] In summary, the technical solution of this invention establishes a resource decision-making model and constraints in step S1; in step S2, RMI and VCI are directly calculated based on model variables (such as demand D_i), providing structured input for subsequent searches; the grouping rules in step S3 need to refer to the adjustment range constraint Δ_i to set a threshold to ensure reasonable classification; in step S4, the search process directly optimizes the model objective function F; the total resource constraint B is guaranteed by a solution repair mechanism, and if the solution exceeds the budget, the allocation is reduced according to preset rules; the adjustment range constraint Δ_i is naturally satisfied by designing differentiated search operators—limiting the search step size for groups that need fine-tuning, and adjusting groups that need reconstruction within the range of Δ_i; the computational budget constraint is directly reflected in the upper limit of the number of iterations. The coordination control mechanism arbitrates the suggestions of each group in real time to ensure that the final solution satisfies all constraints; the termination condition in step S5 is the execution of the computational budget constraint. This invention can improve the solution efficiency, solution stability, and practical feasibility of resource allocation problems under limited computational resource conditions.

[0039] like Figure 2 As shown, under the same resource constraints and computational budget, the method described in this application can achieve high objective function values ​​across different problem sizes. Compared with greedy methods and general swarm intelligence optimization methods, this application effectively improves search efficiency through multi-index clustering and coordinated search mechanisms, and maintains good solution quality as the problem size increases, demonstrating good scalability.

[0040] In addition, such as Figure 3As shown, under multiple independent simulation runs, the optimization results obtained by the method described in this application are more concentrated and have a smaller fluctuation range, indicating that the method has good stability and repeatability in constrained resource allocation problems. This technical effect is particularly important for resource allocation problems in practical applications that require smooth adjustments and stable decisions.

[0041] In summary, this application, by introducing a multi-index collaborative clustering and coordinated search mechanism, achieves efficient optimization of constrained resource allocation problems under limited computing resources, significantly improving search efficiency, solution stability, and the practical feasibility of the solution.

[0042] In constrained resource allocation problems, a multi-index quantitative evaluation system is constructed to collaboratively group resource objects, and coordinated search and constraint control are implemented based on the grouping results, thereby achieving an efficient, stable and feasible resource allocation optimization scheme under limited computing resources.

[0043] Specifically, this application has the following innovative features: (1) Technical means of multi-indicator quantitative evaluation This application quantifies resource objects by introducing at least two evaluation indicators to characterize their status and potential contribution, rather than relying solely on a single indicator or rule. This multi-indicator evaluation provides structured input for subsequent clustering and searching, serving as a fundamental technical means to achieve differentiated search and coordinated control.

[0044] (2) Collaborative clustering mechanism based on multiple indicators Based on the multi-index evaluation results, resource objects are collaboratively clustered, enabling different resource objects to assume different search roles and adjustment strategies during the optimization process. This clustering mechanism breaks through the existing approach of treating resource objects as homogeneous variables and adopting a unified search strategy, structurally improving search efficiency and targeting.

[0045] (3) Cluster-driven coordinated search strategy Different search methods are adopted for different subgroups, and a coordination control mechanism is used to jointly optimize the search results of each subgroup. This generates a consistent resource allocation scheme while satisfying resource constraints and adjustment constraints. This coordinated search strategy avoids the conflicts and instability caused by independent optimization of each subgroup.

[0046] (4) Optimization control mechanism under constraints During the optimization process, total resource constraints, adjustment range constraints, and computational budget constraints are explicitly introduced. Through a coordination and control mechanism, the search process is ensured to be carried out under limited assessment resources, thereby balancing the quality, stability, and practical feasibility of the solution.

[0047] By organically combining the above-mentioned key technologies, this application can achieve structured optimization of constrained resource allocation problems without relying on specific optimization algorithm forms, overcoming the shortcomings of existing technologies in terms of search efficiency, solution stability, and practical application feasibility.

[0048] Example 2 The present invention also provides a resource allocation device based on multi-index cluster collaborative search, comprising: The first processing module is used to obtain a set of resource objects to be configured; The second processing module is used to calculate multiple evaluation indicators for each resource object to describe its current matching status and potential contribution capability. The third processing module is used to collaboratively group resource objects based on real-time matching indicators and value contribution indicators, according to preset thresholds or grouping rules. The fourth processing module is used to perform differentiated search operations for different subgroups and to uniformly adjust the search results through a coordination and control mechanism. The fifth processing module is used to terminate the search process when the preset termination conditions are met, and output the final resource configuration scheme as the optimization result.

[0049] As one embodiment of the present invention, the second processing module is used to define a real-time matching index (RMI) to measure the degree of matching between the current configuration state of the resource object and the target requirement, and its calculation method is shown in equation (1): ; in, Indicates the first The current number of resource objects matched is [number]. This indicates the corresponding target demand. The second processing module is used to define the Value Contribution Index (VCI) to measure the potential contribution of resource objects in the overall configuration. Its calculation method is shown in Equation (2): ; in, Indicates the first The overall contribution value of each resource object Indicates the total number of resource objects.

[0050] As one embodiment of the present invention, the fourth processing module is used to optimize the objective function, which can be expressed as equation (3): ; in, Represents resource objects Utility functions under the current configuration These are the weighting coefficients; The above optimization process must simultaneously satisfy resource constraints and adjustment constraints, as shown in equation (4): ; in, This indicates the upper limit of the total resource budget. This indicates the configuration value of the resource object in the previous iteration. Indicates the maximum allowable adjustment range; By coordinating and controlling the search results of different clusters, the final resource allocation scheme can optimize the objective function value while satisfying the constraints described in equation (4). .

[0051] Example 3 The present invention also provides a resource allocation system based on multi-index cluster collaborative search, comprising: a memory and a processor, wherein the memory stores a computer program executed by the processor, and the computer program executes a resource allocation method based on multi-index cluster collaborative search when executed by the processor.

[0052] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A resource allocation method based on multi-index cluster collaborative search, characterized in that, include: Step S1: Obtain the set of resource objects to be configured; Step S2: For each resource object, calculate multiple evaluation indicators to describe its current matching status and potential contribution capability; Step S3: Based on real-time matching indicators and value contribution indicators, resource objects are collaboratively grouped according to preset thresholds or grouping rules; Step S4: Perform differentiated search operations for different subgroups, and adjust the search results uniformly through a coordination and control mechanism; Step S5: When the preset termination condition is met, terminate the search process and output the final resource configuration scheme as the optimization result.

2. The resource allocation method based on multi-index cluster collaborative search as described in claim 1, characterized in that, In step S2, a Real-Time Matching Index (RMI) is defined to measure the degree of matching between the current configuration state of a resource object and the target requirement. Its calculation method is shown in Equation (1): ； in, Indicates the first The current number of resource objects matched is [number]. This indicates the corresponding target demand. Meanwhile, the Value Contribution Index (VCI) is defined to measure the potential contribution capability of resource objects in the overall configuration, and its calculation method is shown in Equation (2): ； in, Indicates the first The overall contribution value of each resource object Indicates the total number of resource objects.

3. The resource allocation method based on multi-index cluster collaborative search as described in claim 2, characterized in that, In step S4, The overall optimization objective function can be expressed as equation (3): ； in, Represents resource objects Utility functions under the current configuration These are the weighting coefficients; The above optimization process must simultaneously satisfy resource constraints and adjustment constraints, as shown in equation (4): ； in, This indicates the upper limit of the total resource budget. This indicates the configuration value of the resource object in the previous iteration. Indicates the maximum allowable adjustment range; By coordinating and controlling the search results of different clusters, the final resource allocation scheme can optimize the objective function value while satisfying the constraints described in equation (4). .

4. A resource allocation device based on multi-index cluster collaborative search, characterized in that, include: The first processing module is used to obtain a set of resource objects to be configured; The second processing module is used to calculate multiple evaluation indicators for each resource object to describe its current matching status and potential contribution capability. The third processing module is used to collaboratively group resource objects based on real-time matching indicators and value contribution indicators, according to preset thresholds or grouping rules. The fourth processing module is used to perform differentiated search operations for different subgroups and to uniformly adjust the search results through a coordination and control mechanism. The fifth processing module is used to terminate the search process when the preset termination conditions are met, and output the final resource configuration scheme as the optimization result.

5. The resource allocation device based on multi-index cluster collaborative search as described in claim 4, characterized in that, The second processing module is used to define the Real-Time Matching Index (RMI) to measure the degree of matching between the current configuration state of the resource object and the target requirements. Its calculation method is shown in Equation (1): ； in, Indicates the first The current number of resource objects matched is [number]. This indicates the corresponding target demand. The second processing module is used to define the Value Contribution Index (VCI) to measure the potential contribution of resource objects in the overall configuration. Its calculation method is shown in Equation (2): ； in, Indicates the first The overall contribution value of each resource object Indicates the total number of resource objects.

6. The resource allocation device based on multi-index cluster collaborative search as described in claim 5, characterized in that, The fourth processing module is used to optimize the objective function, which can be expressed as equation (3): ； in, Represents resource objects Utility functions under the current configuration These are the weighting coefficients; The above optimization process must simultaneously satisfy resource constraints and adjustment constraints, as shown in equation (4): ； in, This indicates the upper limit of the total resource budget. This indicates the configuration value of the resource object in the previous iteration. Indicates the maximum allowable adjustment range; By coordinating and controlling the search results of different clusters, the final resource allocation scheme can optimize the objective function value while satisfying the constraints described in equation (4). .

7. A resource allocation system based on multi-index cluster collaborative search, characterized in that, include: A memory and a processor, wherein the memory stores a computer program executed by the processor, the computer program, when executed by the processor, performs the resource allocation method based on multi-index clustering collaborative search as described in any one of claims 1-3.

Images