Emergency material reserve distribution method under centralized purchase and storage medium

Abstract

The invention relates to an emergency material reserve distribution method under centralized purchase and a storage medium. The method comprises the following steps: acquiring information of a material supply point and a material demand point and option contract parameters; according to the constructed multi-coverage function and the government-enterprise joint reserve model, establishing a multi-objective optimization model which aims at maximizing the minimum coverage level of the demand point and minimizing the total pre-stored cost; chromosome coding is carried out on the reserve library construction level and the cooperation relation, and based on a multi-objective optimization model, material supply point information, material demand point information and option contract parameters of centralized purchase, an optimal solution in a cost single objective is used as an initial solution; performing iterative optimization on the initial solution through a comprehensive optimization algorithm; and repeating the iterative optimization step until a termination condition is met, and finally outputting a distribution scheme. Compared with a government independent storage mode, the multi-level cross-domain government-enterprise collaborative pre-storage mode has the advantages that the cost is obviously reduced, and the higher coverage level of the reserve library is ensured.

Description

Technical Field

[0001] The present invention relates to the field of emergency material pre-storage, and in particular to an emergency material reserve distribution method and storage medium under centralized procurement. Background Art

[0002] In recent years, the frequent occurrence of natural disasters, public health incidents, and other sudden disasters has placed higher demands on the efficiency and reliability of emergency material reserves and distribution. Traditional emergency material reserve allocation methods are mostly based on independent decision-making by a single entity (such as the government or enterprise), lacking a government-enterprise coordination mechanism, resulting in decentralized resource allocation and delayed response speed. In addition, existing methods have significant shortcomings in the following aspects:

[0003] The traditional model lacks effective coordination mechanisms between government and businesses, preventing them from optimizing resource allocation through joint stockpiling. Enterprise storage capabilities are not fully integrated, resulting in high emergency material stockpiling costs and limited flexibility, making it difficult to respond to sudden demand fluctuations.

[0004] Existing storage facilities are often limited to a single level or localized area, lacking a cross-domain collaborative network of national, provincial, and municipal-level storage facilities. When disasters strike, material dispatch efficiency is low, making it impossible to quickly meet the multi-level needs of affected areas.

[0005] In the existing technology, although some studies have attempted to introduce multi-objective optimization or game models, there are still problems such as high model complexity, low solution efficiency, and failure to combine the optimization problem with the government-enterprise game model for solution. Summary of the Invention

[0006] The present invention provides an emergency material reserve allocation method and storage medium under centralized procurement, which aims to solve the following technical problems in existing emergency material reserve allocation methods: insufficient consideration of the combination of government-enterprise collaboration and optimization; insufficient consideration of multi-level and cross-regional collaboration between reserve warehouses; difficulty in controlling the waste of resources in the face of the uncertainty of disaster occurrence, and difficulty in achieving a rapid response to material shortages in the early stages of a disaster.

[0007] To achieve the above objectives, the present invention provides a method for allocating emergency supplies reserves under centralized procurement, comprising the following steps:

[0008] Obtaining material supply point information, material demand point information, and centralized procurement option contract parameters;

[0009] Based on the pre-built multiple coverage functions and the government-enterprise joint reserve model, a multi-objective optimization model is established with the goals of maximizing the minimum coverage level of demand points and minimizing the total pre-storage cost;

[0010] Chromosome encoding is performed on the construction level and coordination relationship of the reserve warehouse, and based on the multi-objective optimization model, the material supply point information, the material demand point information and the option contract parameters of the centralized procurement, the optimal solution in the cost single objective is used as the initial solution;

[0011] The initial solution is iteratively optimized by a comprehensive optimization algorithm, including:

[0012] a) Convert economic objectives into constraints and retain fairness as the optimization goal;

[0013] b) generating a neighborhood solution using destruction and repair operations based on the constraints;

[0014] c) Allocate the amount of materials stored in the collaborative reserve warehouse through precise algorithms;

[0015] Repeat the above steps of iterative optimization until the termination conditions are met, and finally output the material reserve warehouse construction level, coordination relationship and reserve quantity allocation plan.

[0016] Furthermore, the material supply point information includes a set of available storage warehouses, the volume, capacity and construction cost of storage warehouses of different levels, and a set of enterprise storage points;

[0017] The demand point information includes the demand for material packages at the disaster-stricken point and the travel time between the supply point and the demand point;

[0018] The game parameters of the government-enterprise option contract include: the actual demand for emergency supplies, the probability of sudden disasters, the unit wholesale price of emergency supply packages, production costs, inventory costs, residual value income, unit shortage costs of supply packages, the unit option price of supply packages given by the government to enterprises, the option exercise price, and the proportion of regular supply packages purchased by the government from enterprises.

[0019] Furthermore, the expression of the constructed multiple coverage function is:

[0020]

[0021] Among them, C represents the comprehensive coverage rate of emergency material reserves, SU c represents the set of coverage levels, ω c represents the coverage weight corresponding to the coverage level c, i∈I represents the node set, i represents the specific node, l∈L represents the government reserve level set, l represents the specific level, P cl represents the coverage ratio of coverage level c to government emergency material reserve level l, Z il It is 1 when building a level l reserve i, otherwise it is 0, r∈R represents the set of enterprise storage points, r represents the specific storage point, P crepresents the coverage level c and the coverage rate of the enterprise emergency material reserve; T ir It is 1 when the reserve pool i and enterprise r have joint reserves, otherwise it is 0;

[0022] The expression of the constructed government-enterprise joint reserve model is:

[0023]

[0024] Among them, F is the government-enterprise joint reserve model function, F(Q i ,S r ,o) represents the government reserve and shortage cost function considering the game between government and enterprises, Q i represents the amount of emergency supplies stored in the government storage depot at storage point i; S r represents the storage volume of material packages at the storage point r, o represents the shortage of regional emergency material packages, ω represents the unit daily maintenance cost of government reserve materials, h represents the unit inventory cost of government reserve materials, o1 and o2 represent the unit material package option price paid by the government to enterprise 1 and enterprise 2, S r1 and S r2 represents the storage volume of enterprise material packages at storage points r1 and r2, θ represents the probability of sudden disasters, v is the unit residual value of remaining materials, x is the integral variable, which represents the actual demand for emergency materials, f(x) represents the probability density function of the demand for emergency materials, d(x) is the differential symbol in calculus, e1 and e2 represent the unit execution prices of material packages purchased by the government from enterprises 1 and 2, U is the maximum demand for emergency materials after the disaster, and G is the unit shortage cost of material packages.

[0025] Furthermore, a method for establishing a multi-objective optimization model with the objectives of maximizing the minimum coverage level of demand points and minimizing the total pre-storage cost includes: establishing a first objective function with the objective of maximizing the minimum coverage level of demand points and establishing a second objective function with the objective of minimizing the total pre-storage cost; wherein:

[0026] Methods for establishing the first objective function with the goal of maximizing the minimum coverage level of demand points include:

[0027] Calculate the coverage level of each demand point based on the coverage level weight of the reserve and the corresponding coverage rate;

[0028] Taking the maximization of the lowest coverage level among all demand points as the optimization goal, the objective function expression is:

[0029]

[0030] Among them, ∏1 is the value of the first objective function, which represents the maximization result of the minimum coverage level among the demand points, a∈A represents the set of disaster-affected points, and a is the specific disaster-affected point;

[0031] Methods for establishing the second objective function with the goal of minimizing the total pre-storage cost include:

[0032] The sum of the fixed construction cost, collaborative storage cost and material storage cost of the storage depot is taken as the optimization target, and the objective function expression is:

[0033]

[0034] Among them, ∏2 is the second objective function value, which represents the minimization result of the total pre-storage cost, H il Represents the construction and daily operating expenses of level l reserve i, UC ij U represents the cost required for the coordinated reserve between reserve i and reserve j, ij It is 1 when reserve i and reserve j cooperate in reserves, otherwise it is 0, i≠j, j represents a specific reserve.

[0035] Furthermore, the constraints of the multi-objective optimization model are:

[0036] Each node can build at most one level of government reserve, and the constraint expression is:

[0037]

[0038] Only when the node builds a government reserve, joint reserve with enterprises is allowed. The constraint expression is:

[0039]

[0040] Each government reserve pool can jointly reserve with at most two enterprise storage points. The constraint expression is:

[0041]

[0042] When node i has a government reserve warehouse, emergency supplies are stored. The constraint expression is:

[0043]

[0044] Among them, S il Indicates the storage quantity of material packages in level l storage depot i; M is a real number;

[0045] When the node i enterprise conducts joint storage, the enterprise stores emergency supplies on behalf of the storage point, and the constraint expression is:

[0046]

[0047] Among them, S r Indicates the storage quantity of material packages in the storage point r;

[0048] When both nodes have government reserves, multi-level cross-domain collaborative reserves are allowed, and the constraint expression is:

[0049]

[0050] The quantity of emergency supplies in the reserve warehouse cannot exceed the capacity of the reserve warehouse. The constraint expression is:

[0051]

[0052] Among them, S il Cap represents the storage quantity of material packages in level l storage depot i, V represents the unit volume of material packages; il represents the capacity of level l reserve i;

[0053] The total demand for the region is equal to the total supply of the government reserves and the joint reserves of enterprises in the region plus the shortage. The constraint expression is:

[0054]

[0055] in, It represents the estimated demand for material packages at disaster site a;

[0056] All decision variables satisfy positive real number constraints and 0-1 variable constraints:

[0057] Positive real number constraint:

[0058]

[0059] 0-1 variable constraints:

[0060] Z il ,T ir ,U ij ∈{0,1}.

[0061] Furthermore, a method for performing chromosome encoding on the reserve warehouse construction level and the coordination relationship, and using the optimal solution in the cost single objective as the initial solution based on the multi-objective optimization model, the material supply point information, the material demand point information, and the option contract parameters of the centralized procurement, includes:

[0062] The chromosome code is designed as a list of length N+N×N, where the first segment contains N genes, which are used to encode the reserve construction level of N nodes. The gene value range is 0 to 4, and the table represents no construction, construction of county-level, city-level, provincial-level, and national-level reserves;

[0063] The second segment contains N × N genes, which are used to encode the collaborative reserve relationship between nodes. The gene is 0 or 1. If and only if both nodes build a reserve, collaborative reserve is allowed. 1 indicates collaborative reserve, and 0 indicates no collaboration.

[0064] Based on the multi-objective optimization model, the information of material supply points, the information of material demand points, and the option contract parameters of centralized procurement, with minimizing the total pre-storage cost as the single objective, an optimization algorithm is called to solve and obtain the optimal cost solution, and the optimal solution is used as the initial solution of the chromosome encoding;

[0065] During the initial solution generation process, the constraints include: each node can build at most one reserve, the collaborative reserve relationship only exists between nodes that have built reserve libraries, and the reserve amount does not exceed the capacity limit.

[0066] Further methods of converting economic objectives into constraints and retaining fairness objectives as optimization objectives include:

[0067] The economic target is converted into a constraint condition by using the ε-constraint method, and an upper limit value of the economic target is set;

[0068] Initializing the upper limit value to the minimum value of the economic target, and gradually increasing the upper limit value based on a preset step size until the objective function value cannot be updated;

[0069] In each iteration, the fairness goal is used as the optimization goal to solve the single-objective optimization problem that meets the current constraints;

[0070] The Pareto frontier solution set of the multi-objective optimization problem is generated through multiple iterations, wherein the Pareto frontier solution set includes a trade-off between the demand point coverage level and the total pre-storage cost under different cost constraints.

[0071] Furthermore, based on the constraints, a method for generating a neighborhood solution using destruction and repair operations includes:

[0072] Designing a destruction operator and a repair operator, wherein the destruction operator includes a random destruction operator and a greedy destruction operator, and the repair operator includes a random repair operator and a greedy repair operator;

[0073] Adaptively update the weight of each operator based on historical performance, and dynamically select the combination of destruction and repair operators in the current iteration through a roulette wheel selection mechanism;

[0074] Perform a destruction operation on the initial solution, remove some reserve construction levels and coordination relationship codes, and generate a destroyed intermediate solution;

[0075] Performing a repair operation on the intermediate solution, reinserting the reserve construction level and the collaborative relationship code to generate a new solution;

[0076] Based on the Metropolis criterion of the simulated annealing algorithm, determine whether to accept the new solution, and update the current solution and operator weights according to the acceptance result;

[0077] Among them, the random destruction operator removes the reserve construction level and associated collaborative relationship by randomly selecting city nodes, and the random repair operator rebuilds the reserve construction level and collaborative relationship by randomly generating insertion positions; the greedy destruction operator gives priority to removing the reserve nodes that have the greatest impact on the total government cost, and the greedy repair operator gives priority to inserting the nodes that will minimize the increase in the total government cost.

[0078] Furthermore, the method of allocating the material reserves of the collaborative reserve warehouse through a precise algorithm includes:

[0079] Determine the total reserve of government emergency supplies after multi-level collaborative reserve based on the government-enterprise game model;

[0080] Constructing a precise allocation model with the goal of minimizing stockouts within a region, where the decision variables of the precise allocation model are the specific stockpiles of material packages in each government reserve;

[0081] The total reserve of government emergency supplies is used as a constraint condition, so that the sum of the reserve of each government reserve is equal to the total reserve;

[0082] The Gurobi solver is called to solve the exact allocation model to obtain the optimal allocation plan for the material packages of each government reserve warehouse to minimize the shortage.

[0083] To achieve the above-mentioned purpose, the third aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, the steps of the emergency material reserve allocation method under centralized procurement are executed.

[0084] Beneficial effects of the present invention:

[0085] Compared with the prior art, the present invention provides a method and storage medium for allocating emergency material reserves under centralized procurement. The present invention forms a multi-objective model by integrating multiple coverage models and the government-enterprise joint reserve model under the centralized procurement mechanism, and solves the problems of insufficient government-enterprise collaboration, lack of multi-level cross-domain collaboration, waste of resources and delayed response in the traditional allocation of emergency material reserves through a comprehensive optimization algorithm. Specifically, first, a government-enterprise game framework is constructed to deeply integrate the enterprise storage capacity with the government reserve resources, optimize the government-enterprise reserves and benefit distribution through contract constraints, reduce the construction cost of redundant reserve warehouses and the cost of material reserves, and improve the flexibility of material scheduling. Secondly, by designing multiple coverage functions for emergency material reserves, the coverage capacity of different levels of reserve warehouses (national, provincial, municipal, and county levels) and government-enterprise collaboration on disaster-stricken areas is quantified, and combined with the three-level time response standard, it is ensured that the rapid response in the early stage of the disaster matches the multi-level needs. On this basis, a multi-level cross-domain collaborative optimization model with the dual objectives of "maximizing the minimum coverage level" and "minimizing the total cost of pre-storage" was established. Through chromosome encoding, the reserve construction level and collaborative relationship were mapped into the solution space of the adaptive large neighborhood search algorithm, and the optimal solution of the single cost objective was used as the initial solution to ensure the feasibility and efficiency of the model solution.

[0086] To address the multi-objective optimization challenge, this paper employs a hierarchical integrated optimization algorithm. The outer layer transforms economic objectives into dynamic constraints using the ε-constraint method, retaining fairness as the primary optimization direction, and generating a Pareto frontier solution set covering different cost constraints. The middle layer incorporates an adaptive large neighborhood search algorithm and designs a randomized and greedy destruction-repair operator to dynamically adjust the relationship between reserve construction and coordination. Using a simulated annealing criterion, it accepts some inferior solutions to avoid local optimality, enabling global exploration of the solution space. The inner layer uses the Gurobi solver to accurately calculate the allocation of materials to government reserves, ensuring that shortages are minimized and capacity constraints are strictly met under the government-enterprise game. This three-layer collaborative mechanism significantly improves the model's solution efficiency and robustness. Ultimately, through iterative optimization, the output of the reserve construction level, cross-domain collaborative network, and material allocation plan achieves rapid multi-level response in the early stages of a disaster while controlling resource waste, providing scientific decision-making support for emergency management that balances economy and fairness. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments.

[0088] Figure 1 This is a flow chart of a method for allocating emergency supplies reserves under centralized procurement disclosed in an embodiment of the present invention.

[0089] Figure 2This is a schematic diagram of multiple coverage of emergency material reserves disclosed in an embodiment of the present invention.

[0090] Figure 3 This is a comprehensive algorithm framework diagram disclosed in an embodiment of the present invention.

[0091] Figure 4 This is a schematic diagram of a comprehensive optimization algorithm solution process disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0092] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0093] According to an embodiment of the present invention, it should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the following production method, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0094] The present invention designs a multi-level cross-domain government-enterprise collaborative reserve multi-objective optimization model and a comprehensive optimization algorithm for emergency material reserves under centralized procurement; by establishing multiple coverage functions for emergency material reserves and constructing a government-enterprise joint reserve model based on option contracts, the multiple coverage site selection function is combined with the reserve model of government-enterprise game to construct a multi-level cross-domain government-enterprise collaborative reserve multi-objective optimization model for emergency reserve warehouses; the comprehensive optimization algorithm involved in the present invention is a comprehensive optimization algorithm of an outer layer-constraint, an intermediate layer adaptive large neighborhood search algorithm and an inner layer Gurobi, the outer layer-constraint algorithm transforms the model, and retains the fairness goal as the optimization goal through the -constraint method, while converting the economic goal into a constraint; the intermediate layer uses cost as a constraint, and seeks a neighborhood solution with lower cost through the destruction and repair of the adaptive large neighborhood search algorithm; the inner layer uses the Gurobi solver to allocate the emergency material reserves of each reserve warehouse and enterprise.

[0095] like Figure 1 As shown, the following is a detailed description of a method for allocating emergency supplies reserves under centralized procurement provided by an embodiment of the present invention, which includes the following steps:

[0096] S1: Obtain material supply point information, material demand point information, and option contract parameters under centralized procurement.

[0097] S101: Material supply point information includes: the set of available storage depots (i∈I), the volume, capacity, and construction cost of emergency material storage depots at different levels (national, provincial, municipal, and county); and the set of enterprise storage points (r∈R).

[0098] S102: Material demand point information includes: the demand for material packages at disaster site a Travel time between supply and demand points;

[0099] S103: The game parameters involved in the option contract under centralized procurement include:

[0100] x: The actual demand for emergency supplies, x follows a random distribution, U is the maximum demand for emergency supplies after a disaster, its probability density function is f(x), and its cumulative distribution function is F(x), where F(x) is an increasing function;

[0101] θ: probability of sudden disaster occurrence, 0≤θ≤1;

[0102] γ: unit wholesale price of emergency supply packages;

[0103] c: unit production cost of emergency supplies package, c<γ;

[0104] h: Unit inventory cost of emergency supplies package, h <c;

[0105] v: residual value of remaining materials per unit, v <c;

[0106] G: unit material package out-of-stock cost;

[0107] o k : The unit option price of the material package given by the government to enterprise k, k = 1, 2;

[0108] e k : The unit option exercise price for the government to purchase emergency supplies from enterprise k, k = 1, 2;

[0109] α k : The proportion of regular emergency material packages purchased by the government from enterprise k, α1>0, α2>0, α1+α2=1.

[0110] S2: Pre-built multiple coverage functions and government-enterprise joint reserve models are used to establish a multi-objective optimization model with the goals of maximizing the minimum coverage level of demand points and minimizing the total pre-storage cost.

[0111] S201: Taking the maximum minimum coverage level under multiple coverage as the objective function, in view of the differences in capacity and coverage capabilities of different levels of reserves, the definition of coverage level is expanded, and a multiple coverage level function for the site selection of multi-level emergency reserve warehouses under the joint efforts of government and enterprises is constructed. Emergency material reserve facilities (government reserve warehouses) are divided into four levels: central level, provincial (autonomous region, municipality) level, city, and county (district) level, represented by l = {1, 2, 3, 4}, l∈L. Government reserve warehouses of different levels are jointly reserved with two enterprise reserve warehouses of the same level. Reserve warehouses of different levels have different capacities and geographical coverage. Therefore, for each demand point, its coverage rate can be calculated by dividing it into sets of different levels according to the time from the emergency material reserve facilities to the demand point. Use SU c represents the set of emergency material storage depots that achieve C-level coverage for demand points, c = {1, 2, 3}, that is, each demand point is covered by core coverage, regional coverage, and edge coverage, and the time it takes for the three levels of coverage to reach the demand point is 2, 5, and 10 hours respectively. Figure 2 shown.

[0112] The coverage rates of government emergency material reserve warehouse levels and demand point coverage levels are shown in Table 1. The coverage rate of storage agent companies is 20% of the coverage rate of the corresponding level and coverage level of the cooperating government reserve warehouse.

[0113] Table 1 Coverage rates corresponding to government emergency material reserve levels and demand point coverage levels

[0114]

[0115] Therefore, the functional expression of multiple coverage of emergency material reserves is:

[0116]

[0117] Among them, C represents the comprehensive coverage rate of emergency material reserves, SU c represents the set of coverage levels, ω c represents the coverage weight corresponding to the coverage level c, where l is 1, 2, and 3, corresponding to the demand point being covered by core coverage, regional coverage, and edge coverage. The time for the three levels of coverage to the demand point is 2, 5, and 10 hours respectively. i∈I represents the node set, i represents the specific node, l∈L represents the government reserve level set, l represents the specific level, and P cl It represents the coverage ratio of the coverage level c to the government emergency material reserve level l, where l is 1, 2, 3, and 4, corresponding to the central, provincial, municipal, and county-level reserves, and Z il It is 1 when building a level l reserve i, otherwise it is 0, r∈R represents the set of enterprise storage points, r represents the specific storage point, P c represents the coverage level c and the coverage rate of the enterprise emergency material reserve; Tir It is 1 when reserve pool i and enterprise r have joint reserves, and 0 otherwise.

[0118] S202: The government-enterprise joint reserve model under centralized procurement is expressed as:

[0119]

[0120] Among them, F is the government-enterprise joint reserve model function, F(Q i ,S r ,o) represents the government reserve and shortage cost function considering the game between government and enterprises, Q i S represents the amount of emergency supplies stored in the government storage depot at storage point i; r represents the storage volume of material packages at the storage point r, o represents the shortage of regional emergency material packages, ω represents the unit daily maintenance cost of government reserve materials, h represents the unit inventory cost of government reserve materials, o1 and o2 represent the unit material package option price paid by the government to enterprise 1 and enterprise 2, S r1 and S r2 represents the storage volume of enterprise material packages at storage points r1 and r2, θ represents the probability of sudden disasters, v is the unit residual value of remaining materials, x is the integral variable, which represents the actual demand for emergency materials, f(x) represents the probability density function of the demand for emergency materials, d(x) is the differential symbol in calculus, e1 and e2 represent the unit execution prices of material packages purchased by the government from enterprises 1 and 2, U is the maximum demand for emergency materials after the disaster, and G is the unit shortage cost of material packages.

[0121] S203: Decision variable: U ij It is 1 when the reserve i and reserve j cooperate in storage, otherwise it is 0, i≠j; S il Indicates the storage quantity of material packages in level l storage depot i; Q i represents the reserve amount of the material package at point i after the coordination of the government reserve warehouse; S r Indicates the storage quantity of material packages in the storage point r;

[0122] S204: First objective function: Objective 1: The coverage level of the demand points corresponding to the emergency material reserve allocation plan should be as large as possible, that is, the minimum coverage level of each demand point should be maximized. The calculation formula is as follows:

[0123]

[0124] Among them, ∏1 is the value of the first objective function, which represents the maximization result of the minimum coverage level among the demand points, a∈A represents the set of disaster-affected points, and a is the specific disaster-affected point;

[0125] Second objective function: Objective 2: The reserve cost corresponding to the emergency material reserve allocation plan should be as small as possible, that is, the total pre-storage cost consisting of the fixed construction cost of the emergency material reserve, the coordination cost between government reserve warehouses, and the material reserve cost should be minimized. The calculation formula is as follows:

[0126]

[0127] Among them, ∏2 is the second objective function value, which represents the minimization result of the total pre-storage cost, H il Represents the construction and daily operating expenses of level l reserve i, UC ij U represents the cost required for the coordinated reserve between reserve i and reserve j, ij It is 1 when the reserve i and reserve j cooperate in storage, otherwise it is 0, i≠j, j represents a specific reserve; F(Q i ,S r ,o) is the government reserve and shortage cost function considering the game between government and enterprises;

[0128] S205: Constraint: Each node can establish at most one level of government reserve:

[0129]

[0130] Only when node i has built a government reserve pool can enterprises conduct joint reserves with the reserve pool:

[0131]

[0132] The government reserve of each node i can have joint reserves with at most two enterprises:

[0133]

[0134] Emergency supplies can only be stored when node i has a government reserve:

[0135]

[0136] Among them, S il Indicates the storage quantity of material packages in level l storage i; M represents a large real number;

[0137] When the node i enterprise conducts joint storage, the enterprise storage point will store emergency supplies on behalf of the enterprise:

[0138]

[0139] Among them, S r Indicates the storage quantity of material packages in the storage point r;

[0140] The amount of supplies jointly stored by government reserves and enterprises in the region plus the amount of shortages equals the regional demand for supplies:

[0141]

[0142] Among them, o represents the shortage of regional emergency supply packages; It represents the estimated demand for material packages at disaster site a;

[0143] Multi-level cross-domain collaborative reserves can only be carried out when both nodes i and j have government reserves:

[0144]

[0145] The quantity of emergency supplies in the reserve cannot exceed the capacity of the government reserve:

[0146]

[0147] Among them, V represents the unit volume of the material package; Cap il represents the capacity of level l reserve i;

[0148] Positive real number constraint:

[0149]

[0150] 0-1 variable constraints:

[0151] Z il ,T ir ,U ij ∈{0,1}

[0152] Step S3: Chromosome encoding is performed on the reserve warehouse construction level and the coordination relationship. Based on the multi-objective optimization model, the material supply point information, the material demand point information and the option contract parameters of the centralized procurement, the optimal solution in the cost single objective is used as the initial solution;

[0153] The levels of material reserve construction and their coordination relationships are encoded using the following encoding scheme: a chromosome is an N+N*N list consisting of two segments. The first segment encodes the level of reserve construction in n cities, and the second segment encodes the coordination relationship between the reserve warehouses in these n cities. The first segment encodes genes from 0 to 4, representing no construction, construction of county-level, city-level, provincial-level, and national-level reserves, respectively. The second segment encodes genes from 0 to 1, and whether or not a reserve warehouse is constructed in the first segment determines whether or not the reserve warehouses will coordinate with each other. Multi-level coordinated storage is only possible if both nodes have constructed a reserve warehouse. 0 represents no coordinated storage, and 1 represents coordinated storage. Maintaining the constraints unchanged, the objective function considers only the cost to construct an initial solution, using the optimal solution for the single cost objective as the initial solution.

[0154] Step S4: iteratively optimizing the initial solution using a comprehensive optimization algorithm, including:

[0155] a) Convert economic objectives into constraints and retain fairness as the optimization goal;

[0156] b) generating a neighborhood solution using destruction and repair operations based on the constraints;

[0157] c) Allocate the amount of materials stored in the collaborative reserve warehouse through precise algorithms;

[0158] For a), the specific methods include:

[0159] Step a1), converting the economic target into a constraint condition by using the ε-constraint method, and setting an upper limit value of the economic target;

[0160] Step a2), initializing the upper limit value to the minimum value of the economic target, and gradually increasing the upper limit value based on a preset step size until the objective function value cannot be updated;

[0161] Step a3) In each iteration, the fairness goal is used as the optimization goal to solve the single-objective optimization problem that satisfies the current constraints;

[0162] Step a4) Generate a Pareto frontier solution set for the multi-objective optimization problem through multiple iterations, wherein the Pareto frontier solution set includes a trade-off between the demand point coverage level and the total pre-storage cost under different cost constraints.

[0163] For b), the specific methods include:

[0164] Step b1) designing a destruction operator and a repair operator, wherein the destruction operator includes a random destruction operator and a greedy destruction operator, and the repair operator includes a random repair operator and a greedy repair operator;

[0165] Step b2) adaptively update the weights of each operator based on historical performance, and dynamically select the combination of destruction and repair operators in the current iteration through a roulette wheel selection mechanism;

[0166] Step b3) performing a destruction operation on the initial solution, removing some reserve construction levels and collaborative relationship codes, and generating a destroyed intermediate solution;

[0167] Step b4) performing a repair operation on the intermediate solution, reinserting the reserve construction level and collaborative relationship code to generate a new solution;

[0168] Step b5), based on the Metropolis criterion of the simulated annealing algorithm, determine whether to accept the new solution, and update the current solution and operator weights according to the acceptance result;

[0169] The greedy destruction operator preferentially removes the reserve nodes that have the greatest impact on the total government cost, and the greedy repair operator preferentially inserts the nodes that increase the total government cost the least.

[0170] For c), the specific methods include:

[0171] Step c1) determining the total reserve of government emergency supplies packages after multi-level collaborative reserve based on the government-enterprise game model;

[0172] Step c2) constructing a precise allocation model with the goal of minimizing the amount of stock-outs within the region, wherein the decision variable of the precise allocation model is the specific reserve quantity of the material packages in each government reserve warehouse;

[0173] Step c3) using the total reserve of the government emergency supplies package as a constraint condition, so that the sum of the reserve quantities of each government reserve depot is equal to the total reserve quantity;

[0174] Step c4) calling the Gurobi solver to solve the precise allocation model to obtain the optimal allocation plan for the material packages of each government reserve warehouse to minimize the shortage.

[0175] It is understandable that: Figure 3 As shown, the initial solution is iteratively calculated using the outer layer - constraint, the middle layer adaptive large neighborhood search algorithm, and the inner layer Gurobi comprehensive optimization algorithm; the outer layer - constraint algorithm transforms the model, retaining the fairness goal as the optimization goal through the - constraint method, while converting the economic goal into a constraint; the middle layer uses cost as a constraint, and seeks a neighborhood solution with lower cost through the destruction and repair of the adaptive large neighborhood search algorithm; the inner layer uses the Gurobi solver to allocate the emergency material reserves of the multi-level cross-domain collaborative reserve. Specifically:

[0176] S401: Outer ε-constraint algorithm:

[0177] The epsilon-constraint method retains the primary objective (equity) as the optimization goal, while converting the secondary objective (economy) into a constraint. By iteratively updating the value of ε, the range of economic objectives is continuously narrowed, thus transforming the dual-objective site selection and reserve model into a single-objective model.

[0178] ∏1 is the first objective function value, i.e., the fairness objective function value, and ∏2 is the second objective function value, i.e., the economic objective function value. The maximum value of the economic goal is to establish sufficient emergency material storage facilities. It represents the optimal value when only cost objective is considered without considering coverage. Therefore, the economic objective satisfies the following constraints:

[0179]

[0180] make, The Pareto frontier is obtained by adjusting the value of ε with a constant step size a (a > 0), such that ε = ε + a and 0 ≤ ∏ 2 ≤ ε. The value of ε is continuously updated iteratively until the objective function value cannot be updated.

[0181] Therefore, the multi-objective optimization model can be rewritten as:

[0182] Objective function: max∏1;

[0183] Constraints: ∏2≤ε;

[0184] S402: Intermediate layer adaptive large neighborhood search algorithm:

[0185] Random Destruction Operator: Randomly select three city nodes for removal. First, change the reserve construction code of the selected city to 0, meaning no reserve is constructed. Since the removed city nodes do not have a reserve, no collaborative reserve is established. Consequently, the codes of the rows and columns containing these three cities in the collaborative matrix are all 0. Finally, the original reserve code and collaborative code of the removed city nodes are added to the removal list in sequence.

[0186] Greedy destruction operator: The greedy destruction operator refers to selecting the three node reserves that have the greatest impact on the total government cost from all city reserves for destruction and removal. First, a city node reserve is pre-selected for removal, its reserve construction code is changed to 0, and the codes of the rows and columns where the node is located in the collaboration matrix are changed to 0. The total government cost after removing the current node reserve is calculated one by one, and the first reserve with the greatest impact on the total government cost is determined, and then removed. Then, the second and third node reserves with the greatest impact on the total government cost are determined in the same destruction method and removed. Finally, the original reserve code and collaboration code of the city node to be removed are added to the removal list in sequence.

[0187] Random repair operator: Random destruction and repair are designed to ensure the diversity of solution searches and avoid falling into local optimality. Randomly repair the construction and coordination of the three city reserves. First, randomly generate an insertion index, and then insert the reserve code in the removal list into the randomly generated position. If a reserve has already been built at the insertion position, its original coordination sequence will not be changed; if a reserve has not been built at the insertion position, the corresponding coordination sequence in the removal list will be inserted into the corresponding position until the construction and coordinated repair of the three city reserves are completed. Whether a reserve has been built at each node after the destruction and repair has changed, so the coordination list after insertion needs to be repaired and adjusted. If a city node has not built a reserve, the coordination codes of the city's reserve will all be changed to 0.

[0188] Greedy Repair Operator: The greedy repair operator involves inserting a reservoir into the city node location that minimizes the increase in total government cost. The greedy repair of the construction and coordination of the three city reservoirs is implemented as follows: First, for the index position of the feasible solution where a reservoir has not been built, a random reservoir level is inserted, and the corresponding coordination sequence in the removal list is inserted. For the index position of the built reservoir, a random reservoir level is inserted without changing the coordination sequence. Then, the coordination list is repaired after insertion. If a city node does not have a reservoir, the coordination code of the city reservoir is all 0. Finally, the two city nodes with the smallest increase in total government cost are selected and inserted using the same repair method to complete the repair.

[0189] Adaptive design: The adaptive large neighborhood algorithm selects and adjusts the destruction operator and repair operator according to the operator weight and roulette wheel method during iteration.

[0190] Operator weights are updated based on past performance, specifically as follows:

[0191]

[0192] Among them, ω i is the weight of operator i; u i is the number of times operator i is successfully used; ρ is the weight update coefficient, which is used to control the speed of weight change; s i is the operator score, the initial value is 0, s i Each increase in η k units.

[0193] Initially, all operators have the same weight and score. During each iteration, scores are assigned in a step-by-step manner based on the operator's performance, with higher scores indicating better performance. During the search process, accepting only non-inferior solutions can easily lead to falling into a local optimum. To avoid this, the Metropolis criterion of the simulated annealing algorithm is often used in adaptive large-scale search. This accepts inferior solutions with a certain probability, which is:

[0194]

[0195] Among them, P T is the acceptance probability, x′ is the new solution, x is the current solution, and T is the temperature of the current simulated annealing algorithm.

[0196] Roulette wheel selection operator: During the algorithm iteration process, the system dynamically updates the operator weights and implements probabilistic selection of operators based on the roulette wheel selection mechanism, where the selection probability of each operator is proportional to its weight. The operator probability calculation formula is:

[0197]

[0198] S403: Exact algorithm:

[0199] Given the demand for each city node, an adaptive large neighborhood search algorithm is used to iteratively determine the construction level and location of the reserve. A government-enterprise game model is then used to determine the optimal reserve quantities for both the government and enterprises. However, the government's emergency supply reserve represents the total quantity of supplies after multi-level, cross-domain collaborative reserve preparation. The specific reserve quantity of each government emergency supply reserve is unknown and requires a precise algorithm to calculate and determine. The precise algorithm aims to allocate the reserve quantities of the coordinated reserve and determine the specific reserve quantities of emergency supplies for each government reserve to minimize regional shortages.

[0200] This precise algorithm model is solved using the Gurobi solver, and the decision variables are the reserve quantities of material packages in each government emergency material reserve;

[0201] The objective function of the exact algorithm is the fairness objective, which is expressed as minimizing the number of out-of-stock items in a region:

[0202] min o

[0203] Material reserve constraints:

[0204]

[0205] The government reserve at the supply point calculated by the government-enterprise game model is equal to the total amount of material packages after multi-level coordinated reserve:

[0206]

[0207] Among them, Q j represents the total amount of emergency supplies reserves at supply point j (e.g., a government reserve), which is calculated by accumulating the supply packages after multi-level cross-domain collaborative reserves. j is a specific reserve (e.g., national, provincial, etc.) in the supply point set I. ω c It represents the coverage weight corresponding to the coverage level c, reflecting the contribution weight of different levels of collaboration (such as core coverage, regional coverage, and edge coverage) to the reserve volume. il It represents the reserve capacity or maximum storage quantity of supply point i at the storage level l.

[0208] The comprehensive optimization algorithm process and the comprehensive algorithm solution process are as follows Figure 4 The specific operations are as follows:

[0209] (1) Construct an initial solution, taking the optimal solution of the cost single objective as the initial solution;

[0210] (2) Determine whether the ε constraint condition is satisfied. If so, the process ends and the corresponding solution and Pareto frontier value are output; if not, proceed to step (3);

[0211] (3) Determine whether the maximum number of iterations is met. If it is met, output the optimal solution under different ε constraints. If it is not met, go to step (4);

[0212] (4) Determine whether the temperature update condition is met. If so, continue to determine whether the maximum number of iterations is met. If not, proceed to step (5);

[0213] (5) Update the annealing temperature, select the destruction operator and the repair operator, and call the precise algorithm for calculation;

[0214] (6) Get a new solution and decide whether to accept it. If the conditions are met, update the current solution to the new solution. If not, update the operator score.

[0215] (7) Determine whether the new solution is better than the optimal solution. If it is satisfied, update the new solution to the optimal solution. If it is not satisfied, update the operator score.

[0216] (8) After updating the operator score, the annealing temperature is updated again and the simulated annealing algorithm is calculated;

[0217] (9) Iterate the calculation until the final constraint conditions are met, output the corresponding solution and Pareto frontier value, and end.

[0218] Step S5: Repeat the above iterative optimization steps until the termination conditions are met, and finally output the material reserve warehouse construction level, coordination relationship and reserve quantity allocation plan.

[0219] Through the calculation of the comprehensive optimization algorithm, the final output is the material reserve warehouse construction level, coordination relationship and reserve quantity allocation plan, including: the coverage of demand points under different cost conditions and the material reserve warehouse construction level, coordination relationship and reserve quantity allocation plan.

[0220] According to another aspect of an embodiment of the present application, an electronic device is provided, including a processor and a memory, wherein the processor is configured to implement the steps of the method when executing a computer program stored in the memory.

[0221] In the above embodiments of the present invention, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0222] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only exemplary. For example, the division of the units can be a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0223] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0224] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk, etc. Various media that can store program codes.

[0225] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for allocating emergency supplies reserves under centralized procurement, characterized in that: The steps include: Obtaining material supply point information, material demand point information, and centralized procurement option contract parameters; Based on the pre-built multiple coverage functions and the government-enterprise joint reserve model, a multi-objective optimization model is established with the goals of maximizing the minimum coverage level of demand points and minimizing the total pre-storage cost; Chromosome encoding is performed on the construction level and coordination relationship of the reserve warehouse, and based on the multi-objective optimization model, the material supply point information, the material demand point information and the option contract parameters of the centralized procurement, the optimal solution in the cost single objective is used as the initial solution; The initial solution is iteratively optimized by a comprehensive optimization algorithm, including: a) Convert economic objectives into constraints and retain fairness as the optimization goal; b) generating a neighborhood solution using destruction and repair operations based on the constraints; c) Allocate the amount of materials stored in the collaborative reserve through precise algorithms; Repeat the above steps of iterative optimization until the termination conditions are met, and finally output the material reserve warehouse construction level, coordination relationship and reserve quantity allocation plan.

2. The method for allocating emergency supplies under centralized procurement as claimed in claim 1, characterized in that: The material supply point information includes the set of available storage warehouses, the volume, capacity and construction cost of storage warehouses of different levels, and the set of enterprise storage points; The demand point information includes the demand for material packages at the disaster-stricken point and the travel time between the supply point and the demand point; The game parameters of the government-enterprise option contract include: the actual demand for emergency supplies, the probability of sudden disasters, the unit wholesale price of emergency supply packages, production costs, inventory costs, residual value income, unit shortage costs of supply packages, the unit option price of supply packages given by the government to enterprises, the option exercise price, and the proportion of regular supply packages purchased by the government from enterprises.

3. The method for allocating emergency supplies under centralized procurement as claimed in claim 1, characterized in that: The expression of the constructed multiple coverage function is: Among them, C represents the comprehensive coverage rate of emergency material reserves, SU c represents the set of coverage levels, ω c represents the coverage weight corresponding to the coverage level c, i∈I represents the node set, i represents the specific node, l∈L represents the government reserve level set, l represents the specific level, P cl represents the coverage ratio of coverage level c to government emergency material reserve level l, Z il It is 1 when building a level l reserve i, otherwise it is 0, r∈R represents the set of enterprise storage points, r represents the specific storage point, P c represents the coverage level c and the coverage rate of the enterprise's emergency material reserve; T ir It is 1 when the reserve pool i and enterprise r have joint reserves, otherwise it is 0; The expression of the constructed government-enterprise joint reserve model is: Among them, F is the government-enterprise joint reserve model function, F(Q i ,S r ,o) represents the government reserve and shortage cost function considering the game between government and enterprises, Q i represents the amount of emergency supplies stored in the government storage depot at storage point i; S r represents the storage volume of material packages at the storage point r, o represents the shortage of regional emergency material packages, ω represents the unit daily maintenance cost of government reserve materials, h represents the unit inventory cost of government reserve materials, o1 and o2 represent the unit material package option price paid by the government to enterprise 1 and enterprise 2, S r1 and S r2 represents the storage volume of enterprise material packages at storage points r1 and r2, θ represents the probability of sudden disasters, v is the unit residual value of remaining materials, x is the integral variable, which represents the actual demand for emergency materials, f(x) represents the probability density function of the demand for emergency materials, d(x) is the differential symbol in calculus, e1 and e2 represent the unit execution prices of material packages purchased by the government from enterprises 1 and 2, U is the maximum demand for emergency materials after the disaster, and G is the unit shortage cost of material packages.

4. The method for allocating emergency supplies under centralized procurement as claimed in claim 3, characterized in that: The method for establishing a multi-objective optimization model with the objectives of maximizing the minimum coverage level of demand points and minimizing the total cost of pre-storage includes: establishing a first objective function with the objective of maximizing the minimum coverage level of demand points and establishing a second objective function with the objective of minimizing the total cost of pre-storage; wherein: Methods for establishing the first objective function with the goal of maximizing the minimum coverage level of demand points include: Calculate the coverage level of each demand point based on the coverage level weight of the reserve and the corresponding coverage rate; Taking the maximization of the lowest coverage level among all demand points as the optimization goal, the objective function expression is: Among them, ∏1 is the value of the first objective function, which represents the maximization result of the minimum coverage level among the demand points, a∈A represents the set of disaster-affected points, and a is the specific disaster-affected point; Methods for establishing the second objective function with the goal of minimizing the total pre-storage cost include: The sum of the fixed construction cost, collaborative storage cost and material storage cost of the storage depot is taken as the optimization target, and the objective function expression is: Among them, ∏2 is the second objective function value, which represents the minimization result of the total pre-storage cost, H il Represents the construction and daily operating expenses of level l reserve i, UC ij U represents the cost required for the coordinated reserve between reserve i and reserve j, ij It is 1 when reserve i and reserve j cooperate in reserves, otherwise it is 0, i≠j, j represents a specific reserve.

5. The method for allocating emergency supplies under centralized procurement as claimed in claim 4, characterized in that: The constraints of the multi-objective optimization model are: Each node can build at most one level of government reserve, and the constraint expression is: Only when the node builds a government reserve, joint reserve with enterprises is allowed. The constraint expression is: Each government reserve pool can jointly reserve with at most two enterprise storage points. The constraint expression is: When node i has a government reserve warehouse, emergency supplies are stored. The constraint expression is: Among them, S il Indicates the storage quantity of material packages in level l storage depot i; M is a real number; When the node i enterprise conducts joint storage, the enterprise stores emergency supplies on behalf of the storage point, and the constraint expression is: Among them, S r Indicates the storage quantity of material packages in the storage point r; When both nodes have government reserves, multi-level cross-domain collaborative reserves are allowed, and the constraint expression is: The quantity of emergency supplies in the reserve warehouse cannot exceed the capacity of the reserve warehouse. The constraint expression is: Among them, S il Cap represents the storage quantity of material packages in level l storage depot i, V represents the unit volume of material packages; il represents the capacity of level l reserve i; The total demand for the region is equal to the total supply of the government reserves and the joint reserves of enterprises in the region plus the shortage. The constraint expression is: in, It represents the estimated demand for material packages at disaster site a; All decision variables satisfy positive real number constraints and 0-1 variable constraints: Positive real number constraint: 0-1 variable constraints: Z il ,T ir ,U ij ∈{0,1}。 6. The method for allocating emergency supplies under centralized procurement as claimed in claim 1, characterized in that: The method of performing chromosome encoding on the reserve warehouse construction level and the coordination relationship, and using the optimal solution in the cost single objective as the initial solution based on the multi-objective optimization model, the material supply point information, the material demand point information and the option contract parameters of the centralized procurement includes: The chromosome code is designed as a list of length N+N×N, where the first segment contains N genes, which are used to encode the reserve construction level of N nodes. The gene value range is 0 to 4, and the table represents no construction, construction of county-level, city-level, provincial-level, and national-level reserves; The second segment contains N × N genes, which are used to encode the collaborative reserve relationship between nodes. The gene is 0 or 1. If and only if both nodes build a reserve, collaborative reserve is allowed. 1 indicates collaborative reserve, and 0 indicates no collaboration. Based on the multi-objective optimization model, the information of material supply points, the information of material demand points, and the option contract parameters of centralized procurement, with minimizing the total pre-storage cost as the single objective, an optimization algorithm is called to solve and obtain the optimal cost solution, and the optimal solution is used as the initial solution of the chromosome encoding; During the initial solution generation process, the constraints include: each node can build at most one reserve, the collaborative reserve relationship only exists between nodes that have built reserve libraries, and the reserve amount does not exceed the capacity limit.

7. The method for allocating emergency supplies under centralized procurement as claimed in claim 1, characterized in that: Methods for converting economic objectives into constraints and retaining fairness objectives as optimization goals include: The economic target is converted into a constraint condition by using the ε-constraint method, and an upper limit value of the economic target is set; Initializing the upper limit value to the minimum value of the economic target, and gradually increasing the upper limit value based on a preset step size until the objective function value cannot be updated; In each iteration, the fairness goal is used as the optimization goal to solve the single-objective optimization problem that meets the current constraints; The Pareto frontier solution set of the multi-objective optimization problem is generated through multiple iterations, wherein the Pareto frontier solution set includes a trade-off between the demand point coverage level and the total pre-storage cost under different cost constraints.

8. The method for allocating emergency supplies under centralized procurement as claimed in claim 1, characterized in that: Based on the constraints, methods for generating neighborhood solutions using destruction and repair operations include: Designing a destruction operator and a repair operator, wherein the destruction operator includes a random destruction operator and a greedy destruction operator, and the repair operator includes a random repair operator and a greedy repair operator; Adaptively update the weight of each operator based on historical performance, and dynamically select the combination of destruction and repair operators in the current iteration through a roulette wheel selection mechanism; Perform a destruction operation on the initial solution, remove some reserve construction levels and coordination relationship codes, and generate a destroyed intermediate solution; Performing a repair operation on the intermediate solution, reinserting the reserve construction level and the collaborative relationship code to generate a new solution; Based on the Metropolis criterion of the simulated annealing algorithm, determine whether to accept the new solution, and update the current solution and operator weights according to the acceptance result; Among them, the random destruction operator removes the reserve construction level and associated collaborative relationship by randomly selecting city nodes, and the random repair operator rebuilds the reserve construction level and collaborative relationship by randomly generating insertion positions; the greedy destruction operator gives priority to removing the reserve nodes that have the greatest impact on the total government cost, and the greedy repair operator gives priority to inserting the nodes that will minimize the increase in the total government cost.

9. The method for allocating emergency supplies under centralized procurement as claimed in claim 1, characterized in that: Methods for allocating the amount of material reserves in collaborative reserves through precise algorithms include: Determine the total reserve of government emergency supplies after multi-level collaborative reserve based on the government-enterprise game model; Constructing a precise allocation model with the goal of minimizing stockouts within a region, where the decision variables of the precise allocation model are the specific stockpiles of material packages in each government reserve; The total reserve of government emergency supplies is used as a constraint condition, so that the sum of the reserve of each government reserve is equal to the total reserve; The Gurobi solver is called to solve the exact allocation model to obtain the optimal allocation plan for the material packages of each government reserve warehouse to minimize the shortage.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for allocating emergency supplies reserves under centralized procurement according to any one of claims 1 to 9 are executed.

Images