A phased decision method for offshore emergency resource allocation

Abstract

This invention discloses a phased allocation decision-making method for maritime emergency resources, belonging to the technical field of maritime emergency search and rescue resource scheduling and multi-agent collaborative decision-making. The method first acquires information such as the set of accident points; secondly, it constructs a three-objective resource allocation model with the objectives of minimizing response time, minimizing scheduling costs, and optimizing service capacity balance; it uses the NSGA-III algorithm for integer encoding, feasible population initialization, non-dominated sorting, and reference point association selection to obtain a Pareto optimal candidate scheme set; then, for each resource scheduling scheme in the Pareto candidate scheme set, it constructs a cooperative alliance of responsible entities and a characteristic function, calculates the marginal contribution and responsibility share of each rescue base according to the Shapley value formula, and finally constructs a comprehensive evaluation function based on the normalized objective function and fairness index to perform a second screening of the Pareto candidate schemes, outputting a final scheduling scheme that balances response efficiency, scheduling cost, service balance, and responsibility fairness.

Description

Technical Field

[0001] This invention relates to the fields of maritime emergency search and rescue resource scheduling, multi-objective optimization, and cooperative game decision-making technology, and in particular to a phased allocation decision-making method for maritime emergency resources. Background Technology

[0002] Currently, maritime emergency search and rescue missions typically require rapid matching between the accident location, rescue base, rescue vessels, helicopters, and other emergency resources. Due to the suddenness, complexity, and time-sensitive nature of maritime accidents, dispatch plans not only need to meet resource requirements but also need to comprehensively consider response time, dispatch costs, base load, and fairness in multi-stakeholder collaboration.

[0003] Currently, common practices in existing technologies involve determining resource scheduling schemes through single-objective optimization, manual experience ranking, or simple weighted methods. While these methods can improve response speed to some extent, they still have the following drawbacks: First, single-objective or fixed-weight methods struggle to simultaneously balance response time, scheduling costs, and service capacity. Second, after obtaining the Pareto solution set using general multi-objective optimization algorithms, it is still necessary to select the final scheme from multiple non-dominated solutions; existing methods typically rely on subjective human judgment and lack clear criteria for secondary selection. Third, different rescue bases, as different responsible entities, have varying resource inputs, task assignments, and marginal contributions; existing methods struggle to quantify the contributions and responsibility shares of each entity. Fourth, directly mixing fairness with efficiency and cost in the same optimization stage may lead to the coupling of fairness and efficiency objectives, making it difficult to clearly explain the trade-off between efficiency and fairness in the final scheme.

[0004] Therefore, there is a need for a phased decision-making method for the allocation of maritime emergency resources that can first use NSGA-III to obtain uniformly distributed Pareto candidate schemes, and then use Shapley values ​​to evaluate the fairness of the candidate schemes and perform secondary screening. Summary of the Invention

[0005] To address the problems existing in current technologies, this invention discloses a phased allocation decision-making method for maritime emergency resources. The method includes: constructing a set of accident sites, rescue bases, and resource types; establishing a response time expression and resource scheduling decision variables; constructing a three-objective optimization model balancing response time, scheduling cost, and service capacity; generating a Pareto optimal candidate solution set using the NSGA-III algorithm; constructing a cooperative alliance of responsible entities and a characteristic function for each candidate solution; calculating the marginal contribution of each rescue base using the Shapley value formula; constructing a fairness evaluation function using the dispersion of the Shapley value; and finally, performing a secondary screening of the Pareto candidate solution set through a comprehensive evaluation function to output the final resource scheduling scheme and responsibility allocation results.

[0006] The technical solution of this method specifically adopts the following steps: S1. Obtain basic data on the allocation of maritime emergency resources, including a set of accident sites. Rescue base collection Resource type set , accident point For resource types demand ,base Corresponding resource type Available quantity ,base and the accident site Distance between Resource types At the base average speed Unit scheduling cost and the base's comprehensive service capacity coefficient ; S2. Construct a response time expression based on the aforementioned basic data:

[0007] And construct resource scheduling decision variables:

[0008] in, Indicated by the base To the accident site Allocation Number of resource types; S3. Based on the resource scheduling decision variables, construct a three-objective optimization model for maritime emergency resource allocation. The three-objective optimization model includes the objectives of minimizing response time, minimizing scheduling cost, and balancing service capacity, expressed as:

[0009] in:

[0010]

[0011]

[0012]

[0013] S4. Set demand constraints, base resource capacity constraints, resource service radius constraints, and non-negative integer constraints for the three-objective optimization model, as expressed in the following expression:

[0014]

[0015]

[0016]

[0017] S5. Solve the three-objective optimization model using the NSGA-III algorithm to obtain the Pareto optimal candidate solution set. ; S6. For the Pareto optimal candidate solution set any candidate solution in Construct a set of responsible entities and define the characteristic function. Represents the main subset The system benefits achievable under the candidate solution; S7. Based on the characteristic function Calculate the first Shapley values ​​for each responsible entity:

[0018] in, Representing the subject The marginal contribution; S8. Construct a fairness evaluation function based on the Shapley values ​​of each responsible party:

[0019] in:

[0020] S9. Obtaining the Pareto optimal candidate solution set Then, a comprehensive evaluation function is constructed based on the normalized objective function and the fairness evaluation function:

[0021] in, This represents the normalized objective function. These are the weighting coefficients; S10. Based on the comprehensive evaluation function, evaluate the Pareto optimal candidate solution set. Perform a second sorting and output the final maritime emergency resource dispatch plan, Shapley values ​​of each rescue base, responsibility share, and task-rescue base responsibility allocation matrix.

[0022] Furthermore, in step S5, the NSGA-Ⅲ algorithm uses integer encoding to encode all... Unfold into one-dimensional entities in a fixed order:

[0023] Let the population size be The initial population was:

[0024] Furthermore, a constraint repair strategy is employed to ensure that individuals within the population meet demand constraints and resource capacity constraints.

[0025] Furthermore, in step S5, the individual The fitness function is:

[0026] in:

[0027]

[0028]

[0029] Furthermore, in step S5, the non-dominated ranking judgment rule is that the individual Dominant Individual If and only if:

[0030] And combined with the set of reference points Reference point association selection is performed to maintain the uniformity of the Pareto candidate scheme set in the target space.

[0031] Furthermore, in step S5, the genetic operation includes a crossover operator and a mutation operator. The crossover operator is:

[0032] in, The mutation operator is:

[0033] in, It is a small-range random integer perturbation.

[0034] Furthermore, in step S5, when the maximum number of iterations is reached... If the convergence condition is met, output the set of non-dominated solutions in the final population:

[0035] The non-dominated solution set constitutes the Pareto front, which is used as input for candidate solutions in the Shapley value fairness evaluation.

[0036] Furthermore, the characteristic function in step S6 Constructed from the normalized values ​​of the objective function of the candidate solutions, used to characterize the main subset. The system benefits achieved under a certain resource scheduling scheme.

[0037] Furthermore, the responsibility share output in step S10 is obtained by normalizing the Shapley value and is used to explain the contribution ratio and responsibility sharing ratio of each rescue base in the final dispatch plan.

[0038] By employing the above-mentioned technical solutions, the phased allocation decision-making method for maritime emergency resources provided by this invention has the following advantages: It places multi-objective optimization and fairness evaluation in different decision-making stages, preserving various non-dominated trade-offs between response time, scheduling costs, and service equilibrium before conducting responsibility fairness evaluation and selection. It uses a unified response time function, three-objective function, constraints, fitness function, Shapley value function, fairness evaluation function, and comprehensive evaluation function to ensure clear and complete model expression. The reference point mechanism of NSGA-III improves the problem of uneven distribution of candidate solutions in the three-objective optimization scenario. The Shapley value quantifies the average marginal contribution of each rescue base in the candidate solutions, making the responsibility share interpretable. The fairness evaluation function and comprehensive evaluation function perform a secondary sorting of the Pareto solution set, reducing the dependence of the final solution selection on subjective human judgment. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0040] Figure 1 This is a diagram illustrating the overall framework for the phased allocation of maritime emergency resources in this invention. Figure 2 This is a flowchart of the NSGA-Ⅲ process for generating Pareto candidate schemes in this invention. Figure 3 This is a flowchart of the Shapley fairness evaluation process in this invention; Figure 4 This is an example diagram of the Pareto candidate schemes and fairness evaluation results in this invention; Figure 5 This is an example diagram of the task-rescue base responsibility share matrix in this invention. Detailed Implementation

[0041] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. 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 should fall within the scope of protection of the present invention.

[0042] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0043] like Figure 1 The method shown is a phased allocation decision-making method for maritime emergency resources, which specifically includes the following steps: Step 1: Establish a basic dataset for maritime emergency resource allocation. Define the accident point set as follows:

[0044] The rescue bases are as follows:

[0045] The resource type set is as follows:

[0046] accident point For resource types The demand is denoted as ,base Corresponding resource type The available quantity is denoted as ,base and the accident site The distance between them is Resource type At the base The average speed is The corresponding response time is:

[0047] This response time is used to measure the time cost of resources traveling from the base to the incident site.

[0048] Step 2: Establish resource scheduling decision variables. Definition:

[0049] in, Indicated by the base To the accident site Allocation The number of class resources. When At that time, it was stated that the base did not participate in this type of resource rescue mission at the accident site.

[0050] Step 3: Construct a three-objective optimization model for the allocation of maritime emergency resources.

[0051] First, regarding response efficiency, a minimization objective function based on weighted response time is established:

[0052] Secondly, in terms of economics, considering the operating costs incurred during resource scheduling, a cost minimization objective function is constructed:

[0053] Furthermore, to avoid the problem of individual bases being overburdened or underutilized during resource allocation, a base definition is established. The load level is:

[0054] Taking into account the differences in base station capabilities, the load is normalized, and a service capacity balancing target is constructed:

[0055] In summary, the following multi-objective optimization problem is formed:

[0056] Step 4: Set model constraints. To meet the resource requirements of the accident site, the following conditions should be met:

[0057] Considering the limited resource supply capacity of each base, the following should be met:

[0058] Because different resource types have service radius limitations, when the distance between the base and the incident point exceeds the maximum reachability range, that resource cannot participate in scheduling. Therefore:

[0059] in, Indicates the first The maximum service distance of the resource class. In addition, the variable should satisfy the non-negative integer constraint:

[0060] Step 5: Solve the three-objective optimization model using the NSGA-III algorithm. For example... Figure 2 As shown, for decision variables Individuals are represented using integer encoding, and all variables are expanded into a one-dimensional vector in a fixed order:

[0061] Each individual represents a complete resource scheduling scheme. Let the population size be... The initial population was:

[0062] Based on the three-objective model, individual The fitness function is:

[0063] in:

[0064]

[0065]

[0066] In each generation of evolution, a non-dominated ranking is performed on the population to define individuals. Dominate If and only if:

[0067] Based on this relationship, the population is divided into multiple non-dominant levels. In the environment selection phase, a set of reference points is introduced:

[0068] By mapping individuals to a reference point, the distribution of the solution set is guided. The crossover operator can be represented as:

[0069] in, The generated individuals are then rounded and constraint-corrected. The mutation operator can be expressed as:

[0070] in, The system uses small-range random integer perturbations and employs a repair mechanism to ensure that the constraints are met. When the maximum algebraic value is reached... If the convergence condition is met, output the set of non-dominated solutions in the final population:

[0071] Step six: Evaluate fairness based on the Shapley score. For example... Figure 3 As shown, let the set of responsible parties involved in the search and rescue be:

[0072] Define characteristic function Represents the main subset The system benefits achievable under a given scheduling scheme. Then the main body... The Shapley value is defined as:

[0073] in, Representing the subject The marginal contribution of each entity can be calculated. By calculating the Shapley value of each entity, the distribution of their contribution to the overall search and rescue mission can be obtained.

[0074] Step 7: Design fairness evaluation metrics. To measure the fairness of the scheduling scheme, the dispersion of the Shapley value distribution is introduced as an evaluation metric:

[0075] in:

[0076] This indicator reflects the degree of balance in the distribution of contributions among various entities; the smaller the value, the higher the fairness.

[0077] Step 8: Construct a phased comprehensive decision-making mechanism. This is done after obtaining the Pareto solution set. Then, for each candidate solution Calculate the corresponding Shapley value distribution. And obtain fairness indicators Based on this, a comprehensive evaluation function is constructed:

[0078] in, This represents the normalized objective function. where represents the weighting coefficient. By performing a secondary sorting of the Pareto solution set, the optimal scheduling scheme that balances efficiency, cost, and fairness is selected.

[0079] Example 1 Taking multiple rescue bases, multiple accident sites, and various types of search and rescue resources as an example, the system first obtains the accident site location, accident resource requirements, rescue base coordinates, available base resources, resource speed, scheduling cost, and base capability coefficient. Then, it constructs a three-objective optimization model according to steps one through four. Using the NSGA-III algorithm, multiple non-dominated resource scheduling schemes are obtained, which constitute a Pareto optimal candidate scheme set, such as... Figure 4 As shown. Subsequently, the Shapley value for each rescue base is calculated for each candidate solution, forming a task-rescue base responsibility share matrix, as shown. Figure 5 As shown. Finally, through a comprehensive evaluation function... The candidate schemes are ranked a second time, and the scheme with the best comprehensive evaluation value is selected as the final maritime emergency resource dispatch scheme.

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A phased allocation decision-making method for maritime emergency resources, characterized in that... include: S1. Obtain basic data on the allocation of maritime emergency resources, including a set of accident sites. Rescue base collection Resource type set , accident point For resource types demand ,base Corresponding resource type Available quantity ,base and the accident site Distance between Resource types At the base average speed Unit scheduling cost and the base's comprehensive service capacity coefficient ; S2. Construct a response time expression based on the aforementioned basic data: And construct resource scheduling decision variables: in, Indicated by the base To the accident site Allocation Number of resource types; S3. Based on the resource scheduling decision variables, construct a three-objective optimization model for maritime emergency resource allocation. The three-objective optimization model includes the objectives of minimizing response time, minimizing scheduling cost, and balancing service capacity, expressed as: in: S4. Set demand constraints, base resource capacity constraints, resource service radius constraints, and non-negative integer constraints for the three-objective optimization model, as expressed in the following expression: S5. Solve the three-objective optimization model using the NSGA-III algorithm to obtain the Pareto optimal candidate solution set. ; S6. For the Pareto optimal candidate solution set Any candidate solution in Construct a set of responsible entities and define the characteristic function. Represents the main subset The system benefits achievable under the candidate solution; S7. Based on the characteristic function Calculate the first Shapley values ​​for each responsible entity: in, Representing the subject The marginal contribution; S8. Construct a fairness evaluation function based on the Shapley values ​​of each responsible party: in: S9. Obtaining the Pareto optimal candidate solution set Then, a comprehensive evaluation function is constructed based on the normalized objective function and the fairness evaluation function: in, This represents the normalized objective function. These are the weighting coefficients; S10. Based on the comprehensive evaluation function, evaluate the Pareto optimal candidate solution set. Perform a second sorting and output the final maritime emergency resource dispatch plan, Shapley values ​​of each rescue base, responsibility share, and task-rescue base responsibility allocation matrix.

2. The method for phased allocation of maritime emergency resources according to claim 1, characterized in that: In step S5, the NSGA-Ⅲ algorithm uses integer encoding to encode all... Unfold into one-dimensional entities in a fixed order: Let the population size be The initial population was: Furthermore, a constraint repair strategy is employed to ensure that individuals within the population meet demand constraints and resource capacity constraints.

3. The method for phased allocation of maritime emergency resources according to claim 2, characterized in that: In step S5, the individual The fitness function is: in: 。 4. The method for phased allocation of maritime emergency resources according to claim 3, characterized in that: In step S5, the non-dominated ranking judgment rule is that the individual Dominant Individual If and only if: And combined with the set of reference points Reference point association selection is performed to maintain the uniformity of the Pareto candidate scheme set in the target space.

5. The method for phased allocation of maritime emergency resources according to claim 1, characterized in that: In step S5, the genetic operation includes a crossover operator and a mutation operator. The crossover operator is as follows: in, The mutation operator is: in, It is a small-range random integer perturbation.

6. The method for phased allocation of maritime emergency resources according to claim 1, characterized in that: In step S5, when the maximum number of iterations is reached... If the convergence condition is met, output the set of non-dominated solutions in the final population: The non-dominated solution set constitutes the Pareto front, which is used as input for candidate solutions in the Shapley value fairness evaluation.

7. The method for phased allocation of maritime emergency resources according to claim 1, characterized in that: The characteristic function in step S6 Constructed from the normalized values ​​of the objective function of the candidate solutions, used to characterize the main subset. The system benefits achieved under a certain resource scheduling scheme.

8. The method for phased allocation of maritime emergency resources according to claim 1, characterized in that: The responsibility share output in step S10 is obtained by normalizing the Shapley value and is used to explain the contribution and responsibility ratio of each rescue base in the final dispatch plan.

Images