Task-oriented equipment support resource region configuration model establishment method

Abstract

The invention discloses a task-oriented equipment support resource regional configuration model establishment method, and aims to solve the problem that a support resource configuration model is insufficient in generality in the aspects of an open environment, a combined target and hierarchical constraints in the task execution process of regional equipment. In order to solve the problems of the prior art, such as the lack of a support strategy and a configuration scheme on system analysis of an influence mechanism level, on the basis of the definition of a node set of a support object and a support mechanism, the number matching of node resources is converted into probability matching, an equipment support resource region configuration model is established based on an optimal transmission theory, and a basic algorithm for model solution is given. Case analysis shows that the configuration model provided by the invention is better in performance in intermittent demand, and shows a local zigzag structure for different resource types.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment support resource allocation, and particularly to a method for establishing a regional allocation model of equipment support resources oriented to tasks. Background Technique

[0002] Equipment support resource allocation is the intermediate link between support demand prediction and support resource scheduling, and also the material starting point of equipment support. Resource allocation is a process of pre-distributing human resources, spare parts, equipment and tools, etc. to relevant support agencies to meet the equipment support resource requirements during the equipment's mission execution. The research on resource allocation mainly focuses on three aspects: allocation methods, model construction, and support strategies.

[0003] Regarding the aspect of allocation methods: Jing Jun divided the optimization of equipment maintenance support resources into strategic and operational levels, and pointed out that different methods should be adopted at different levels from the perspective of top-level design. Qi Xiaogang studied the resource allocation problem of a multi-center equipment support system, and focused on analyzing the parameters and theoretical methods of the optimization of maintenance personnel and maintenance spare parts resources. Su Yong studied the complex equipment system with arbitrary distributions of component life and repair time according to the characteristics of two-level support, and proposed an optimization allocation method of support resources based on genetic algorithm and Monte Carlo simulation. Chen Yunxiang and Chen Xingxing proposed a maintenance support resource allocation method based on hierarchical timed colored Petri nets and an optimal allocation method of various support resources based on task stage division for the problem of aircraft maintenance support resource allocation. Wang Zhengyuan proposed a method for allocating wartime spare parts resources considering maintenance capabilities.

[0004] Regarding the aspect of model construction: Yao Lu analyzed the R-T law of project activities and three implementation modes from the perspective of project management, and established a project resource allocation model based on multi-mode selection; Yu Fengzhu analyzed the stage task division, resource element composition and their relationships of ship equipment, and constructed an optimization allocation model of support resources based on task priority; Wu Yingzhen proposed an equipment maintenance support resource allocation model oriented to tasks based on clear task requirements and maintenance support resource conditions; Gao Long started from the support requirements of combat tasks, and proposed a resource allocation network model under the collaborative support mode for the characteristics of dispersed, small-scale and unconnected support resources in the equipment support system. Regarding the aspect of support strategies: Liu Shengyu proposed an idea of replenishing goods at any time under the accompanying maintenance strategy, combined the regional danger coefficient with the current cost of the maintenance team, and studied the multi-objective joint optimization problem of resource allocation and task scheduling. Li Peng analyzed the influence of regular maintenance strategy, program preventive strategy, sudden failure maintenance strategy, wartime maintenance strategy, and phased strategy update on equipment mission execution through simulation.

[0005] However, there are two problems in existing research: (1) Although various configuration methods have been proposed from different perspectives and various model construction ideas have been put forward for configuration models starting from task requirements, these methods and models are applicable to specific situations and are not general for open environments, combined objectives, hierarchical constraints, etc.; (2) Regarding how the guarantee strategy affects the configuration plan, there is only research on the correspondence between the strategy and the result, lacking a systematic analysis at the level of the influence mechanism. At the same time, with the normalization of equipment tasks facing regions, existing technologies mainly focus on a single aspect of single equipment or single task. How to allocate resources for regional tasks has become a hot issue for current equipment support personnel to study. Summary of the Invention

[0006] In view of the above-mentioned shortcomings of the existing technology, the present invention proposes a method for establishing a regional configuration model of equipment support resources for tasks. Based on the equipment operation tasks in the region, it defines the node sets of the support objects and support agencies, and based on the optimal transport theory, converts the quantity matching of node resources into probability matching of node resources, thereby establishing a regional configuration model of equipment support resources.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A method for establishing a regional configuration model of equipment support resources for tasks, comprising the following steps:

[0009] Step 1: Construct a basic model for regional configuration of equipment support resources according to the support agencies and support objects in the region;

[0010] Step 2: Define the basic principles of the basic model for regional configuration of equipment support resources;

[0011] Step 3: Solve the basic model for regional configuration of equipment support resources to obtain the amount of resources to be configured at each supply point;

[0012] Step 4: Obtain a support resource configuration plan based on each supply point in the region according to the amount of resources to be configured at each supply point.

[0013] Further, the specific steps of Step 1 include:

[0014] Step 1.1: Denote the support agency node set as {A1, A2... A m}, and the i-th node of the support agency can be expressed as A i = {α i , p i , x i , u i}, where m is the number of support agencies, α i is the node importance, p i is the node spatial position, p i = {pi1 , p i2}, p i1 , p i2 represent the abscissa and ordinate in Euclidean space respectively, x i is the relationship structure between nodes, represented by an adjacency matrix, u i is the node resource feature, represented by a vector of resource allocation quantities, u i = {u i1 , u i2 … u ik}}, u ik is the quantity of the k-th resource of the i-th node, where k is the type of resource;

[0015] Step 1.2: Denote the set of guaranteed object nodes as {B1, B2… B n}}, and the j-th node of the guaranteed object can be expressed as B j = {β j , q j , y j , v j}, where n is the number of guaranteed objects, β j is the node importance, q j is the node spatial position, q j = {q j1 , q j2}, q j1 , q j2 represent the abscissa and ordinate in Euclidean space respectively, y j is the relationship structure between nodes, represented by an adjacency matrix, v j is the node resource feature, represented by a vector of resource demand quantities, v j = {v j1 , v j2 … v jk}}, v jk is the quantity of the k-th resource of the j-th node, where k is the type of resource;

[0016] Step 1.3: Define the cost of transporting resources from A i to B j as C ij , and the probability of transporting resources from A i to B j as PT ij ;

[0017] Step 1.4: Based on the above definitions, establish the basic model for regional equipment support resource allocation:

[0018] 1) The first-layer goal is to optimize the resource satisfaction rate and utilization rate

[0019]

[0020] Among them, represents the satisfaction rate of the k-th type of resource; represents the utilization rate of the k-th type of resource;

[0021] 2) The second-layer goal is to minimize the resource transportation cost:

[0022]

[0023] In the formula, the first term is the resource feature matching term, and the second term is the importance feature resolution term. The third and fourth terms are the matching terms of importance α i and β j respectively. The fifth term |C x (i,k)-C y (j,l)| ε *PT ij PT kl is the positional relationship structure matching term. The sixth term |C p (i,k)-C q (j,l)| ε *PT ij PT kl is the functional relationship (system level) structure matching term. PT kl represents the probability that the resource is transported from A k to B l . A and B represent the set of support agency nodes and the set of support object nodes respectively; ε represents the norm type, σ represents the resolution parameter, γ represents the importance balance parameter, τ represents the positional relationship matching weight, represents the functional relationship structure matching weight, σ, γ, τ, are all non-negative numbers.

[0024] Furthermore, the constraint conditions of the basic model for regional equipment support resource allocation established in step 1.4 include:

[0025] ∑ i α i = 1 (8)

[0026] ∑ j β j = 1 (9)

[0027]

[0028] Among them, Equation (8) represents the importance of each node of the guarantee institution, Equation (9) represents the importance of each node of the guarantee object, Equation (10) represents the transportation probability and constraint from each guarantee institution node to all guarantee object nodes, Equation (11) represents the transportation probability and constraint from all guarantee institution nodes to each guarantee object node, Equation (12) represents that for each type of resource, the sum of the resources configured by the guarantee institution nodes is greater than the resources of this type for each guarantee demand node, and Equation (13) represents that for each type of resource, the sum of the resources demanded by the guarantee object nodes is less than the resources of this type for each guarantee institution node;

[0029] v jk represents the guarantee demand, and:

[0030] The amount of the kth resource that should be configured for the ith guarantee institution node is:

[0031]

[0032] In the formula, T represents the threat level, a jk represents the basic value of the guarantee demand when the threat is 0, and b jk represents the growth rate of the guarantee demand as the threat level increases.

[0033] Furthermore, the specific steps of Step 3 include:

[0034] Step 3.1: Input the parameters σ, γ, τ, ω, ε, θ ij , ρ ij and the variables α i , p i , x i , β j , q j , y j , v j ;

[0035] Step 3.2: Initialize the supply resource vector u i ;

[0036] Step 3.3: Calculate C ij , C x (i, k), C y (j, l), C p (i, k) and C q (j, l) respectively;

[0037] Step 3.4: Use the fmincon optimizer to solve the second-layer objective that satisfies the above constraints to obtain the resource allocation probability scheme PT i,j :

[0038]

[0039] Step 3.5: Substitute PT ij into the first-layer objective function to check whether PT i,j meets the requirements of the satisfaction rate and utilization rate. If it meets the requirements, go to Step 3.6; otherwise, return to Step 3.2.

[0040] Step 3.6: Calculate the resource quantities to be allocated at each supply point

[0041] Therefore, the present invention adopts the above-mentioned method for establishing a task-oriented regional allocation model of equipment support resources for the problem of support resource allocation during the task execution of equipment within a region, and has the following beneficial effects:

[0042] First, the present invention proposes a definition of the node set of equipment support supply and demand. Traditional equipment support resource allocation models show general deficiencies in aspects such as open environments, combined objectives, and hierarchical constraints. The analysis of the influence mechanism of support strategies on allocation schemes only focuses on the point-to-point mapping relationship, and the link mechanism and multi-mechanism analysis of how support strategies affect allocation schemes are insufficient. Therefore, the present invention abstracts the equipment support problem into a hierarchical and classified comprehensive matching of support agency nodes, support object nodes, and node sets, and uses probabilistic thinking to model the probabilistic characteristic attributes of resources. The model established thereby is more adaptable and general.

[0043] Second, the present invention establishes a regional allocation model of equipment support resources. Based on the definition of the node set, the quantity matching of node resources is converted into probabilistic matching. Based on the optimal transport theory, a regional allocation model of equipment support resources is established, and five basic principles for supply-demand matching are given by explaining the physical connotation of the model.

[0044] Third, the results of case analysis show that the resource allocation model proposed by the present invention has superiority for intermittent demands, and thus has more application value, and shows a local sawtooth structure for different resource types.

[0045] In summary, based on the definition of the supply-demand node set and the perspective of probabilistic matching, the present invention uses the optimal transport idea to establish a regional allocation model of equipment support resources, effectively establishing the correlation between support demands and resource allocation schemes.

[0046] Next, through the accompanying drawings and embodiments, the technical solutions of the present invention will be further described in detail. Description of the Drawings

[0047] Figure 1 It is an example diagram for explaining the basic model of regional allocation of equipment support resources constructed by the present invention.

[0048] Figure 2 It is the regional support agency in the case analysis of the embodiment.

[0049] Figure 3 For the regional equipment grouping in the case analysis of the embodiment.

[0050] Figure 4 For the calculation results of the case data under Resource Allocation Plan 1 and Resource Allocation Plan 2.

[0051] Figure 5 For the calculation results of the intermittent demand data under Resource Allocation Plan 3 and Resource Allocation Plan 4.

[0052] Figure 6 For the influence of parameter τ at supply point 1 under the tree structure.

[0053] Figure 7 For the calculation results of the case data under Resource Allocation Plan 5 and Resource Allocation Plan 6.

[0054] Figure 8 For the influence of parameter τ at supply point 1 under Relationship 1.

[0055] Figure 9 For the influence of parameter τ at supply point 1 under Relationship 2.

[0056] Figure 10 For the influence of parameter τ at supply point 1 under Structure 1.

[0057] Figure 11 For the influence of parameter τ at supply point 1 under Structure 2.

[0058] Figure 12 For the calculation results of the case data under three strategies.

[0059] Figure 13 For the calculation results of the intermittent demand data under three strategies.

[0060] Figure 14 For parameter The analysis results of the influence of the resource allocation plan on the case data.

[0061] Figure 15 For parameter The analysis results of the influence of the resource allocation plan on the intermittent data.

[0062] Figure 16 For the resource allocation plans obtained under the three situations of no threat, less threat, and greater threat in the embodiment.

[0063] Figure 17 For the line chart of Resource Allocation Plan 10 obtained by the method proposed in the present invention and Resource Allocation Plan 11 obtained by the traditional method in the embodiment. Detailed implementation manners

[0064] In the description of the present invention, it should also be noted that unless otherwise clearly specified and defined, these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims of this application.

[0065] A method for establishing a task-oriented regional allocation model of equipment support resources proposed by the present invention includes the following steps:

[0066] Step 1: Construct a basic model for regional allocation of equipment support resources according to the support agencies and support objects within the region;

[0067] Step 2: Define the basic principles of the basic model for regional allocation of equipment support resources;

[0068] Step 3: Solve the basic model for regional allocation of equipment support resources that meets the basic principles to obtain the resource quantities to be allocated at each supply point;

[0069] Step 4: Obtain a support resource allocation plan based on each supply point within the region according to the resource quantities to be allocated at each supply point.

[0070] The following is a detailed introduction to the task-oriented regional allocation model of equipment support resources:

[0071] 1. Basic model for regional allocation of equipment support resources

[0072] Assume that there are m support agencies and n support objects within the region. Denote the set of support agency nodes as {A1, A2... A m}, and the i-th node of the support agency can be expressed as A i = {α i , p i , x i , u i}, where:

[0073] α i is the node importance;

[0074] p i is the node spatial position, represented by Cartesian coordinates, p i = {p i1 , p i2}, p i1 , p i2 respectively represent the abscissa and ordinate in the Euclidean space;

[0075] x i is the relationship structure (or functional structure) between the node and other nodes, represented by an adjacency matrix;

[0076] u i is the node resource feature, represented by a vector of resource allocation quantities. u i ={u i1 , u i2 … u ik}, where u ik is the quantity of the k-th resource of the i-th node, and k is the resource type;

[0077] Next, denote the set of guaranteed object nodes as {B1, B2… B n}, and the j-th node of the guaranteed object is represented as B j ={β j , q j , y j , v j}, where:

[0078] β j is the node importance;

[0079] q j is the node spatial position, represented by Cartesian coordinates. q j ={q j1 , q j2}, where q j1 , q j2 represent the abscissa and ordinate in the Euclidean space respectively;

[0080] y j is the relationship structure (or functional structure) between the node and other nodes, represented by an adjacency matrix;

[0081] v j is the node resource feature, represented by a vector of resource demand quantities. v j ={v j1 , v j2 … v jk}, where v jk is the quantity of the k-th resource of the j-th node, and k is the resource type.

[0082] Define the cost of transporting resources from A i to B j as C ij , and define the probability of transporting resources from A i to B j as PT ij , where C ij is the comprehensive cost considering various factors, which will be elaborated in detail later. PT ij is the decision variable to be solved by the model, and is obtained through the model solving algorithm;

[0083] Equipment support resource allocation is to reasonably distribute limited resources to support agency nodes to meet the task needs of support object nodes. Essentially, it is to find the best match between supply and demand resources. In task-oriented equipment support activities, due to the short task window time and the weak role of maintenance, the equipment support mainly adopts the mode of front replacement and back repair. Assuming that the replacement time is not considered, resource allocation is to solve the problem of matching support resources to support agencies, and then the support agencies deliver them to the corresponding support objects according to a certain support mechanism. Based on this, the matching between support agencies and support objects can be decomposed into two parts. The first part is the matching of {α i ,p i ,u i} and {β j ,q j ,v j}, that is, the matching of supply and demand nodes of resource characteristics, which belongs to the typical point cloud matching problem. The second part is the matching of {α i ,p i ,u i} and x i with {β j ,q j ,v j} and y j , that is, the matching of supply and demand nodes of "feature + structure", which belongs to the typical FGW problem. Therefore, based on the support resource scheduling model constructed based on the optimal transport theory in the previous literature "Research on Regional Support Scheduling Model from the Perspective of Resource Distribution, Systems Engineering and Electronics, 2024, 46(09): 3093-3102" and "Research on Regional Support Scheduling Model Based on Structured Data. Systems Engineering and Electronics", according to the point cloud matching theory and considering the double-layer objective, the basic model of regional equipment support resource allocation is established as follows:

[0084] The first-layer objective is to optimize the resource satisfaction rate and utilization rate:

[0085]

[0086] The second-layer objective is to minimize the resource delivery cost:

[0087]

[0088]

[0089] In the formula, the first item is the resource feature matching item, the second item is the importance feature resolution item, the third and fourth items are the matching items of importance α i and β j respectively, and the fifth item |C x (i,k)-C y (j,l)| ε*PT ij PT kl is the matching item of the position relationship structure, the sixth item |C p (i,k)-Cqj,lε*PTijPTkl is the matching item of the functional relationship (system level) structure, and PTkl represents the probability that resources are transported from Ak to B l ; A and B respectively represent the set of support agency nodes and the set of support object nodes;

[0090] Among them, the similarity between pairwise distances is defined as the structural similarity, which is represented by a four-dimensional tensor, that is:

[0091] L i,j,k,l (C x ,C y ) = |C x (i,k)-C y (j,l)| ε (4)

[0092] L i,j,k,l (C p ,C q ) = |C p (i,k)-C q (j,l)| ε (5)

[0093] The GW distance of the structure is:

[0094]

[0095] GW(C p ,C q ) = ∑ i,j,k,l (L(C p ,C q ) * PT ij PT kl ) (7)

[0096] Therefore, Equation (3) is the FGW distance of the supply and demand resource distribution, and it is used as the objective function. In the equation, ε represents the norm type, generally taking 1 and 2, σ represents the resolution parameter, γ represents the importance balance parameter, τ represents the position relationship matching weight, represents the functional relationship (system level) structure matching weight, σ, γ, τ, are all non-negative numbers.

[0097] The constraint conditions are:

[0098] ∑ i α i = 1 (8)

[0099] ∑ j β j= 1 (9)

[0100]

[0101] Among them, formula (8) represents the importance of each node in the support organization, reflecting the contribution degree of the node in the support activity. Formula (9) represents the importance of each node of the support object, reflecting the contribution degree of the node in the combat activity. Formula (10) represents the delivery probability and constraint from each support organization node to all support object nodes. Formula (11) represents the delivery probability and constraint from all support organization nodes to each support object node. Formula (12) means that for each type of resource, the sum of the resources configured by the support organization nodes should be greater than the resources of this type for each support demand node. Formula (13) means that for each type of resource, the sum of the resources required by the support object nodes should be less than the resources of this type for each support organization node.

[0102] In the model, C p (i, k) = d A (p i , p k ), C q (j, l) = d B (q j , q l ), C x (i, k) = d A (x i , x k ), C y (j, l) = d B (y j , y l ), respectively represent the positional distance between nodes in the support organization system, the positional distance between nodes in the support object system, the structural distance between nodes in the support organization system, and the structural distance between nodes in the support object system; p i , p k , q j , q l are respectively the structural information of the node position relationship in the two sets, and x i , x k , y j , y l are the structural information of the node functional relationship (system level) in the two sets. The functional relationship is considered at low resolution of the system structure (coarse-grained level), and the system level is considered at high resolution (fine-grained level).

[0103] In the first-level objective, the satisfaction rate of the k-th type of resource is:

[0104]

[0105] The utilization rate of the k-th type of resource is:

[0106]

[0107] The i-th security agency node should be configured with the k-th type of resource

[0108]

[0109] During the combat mission process, there may also be threats from the enemy. Different threat levels result in different quantities of support resources required. Using the classical exponential function model to simulate the relationship between threat level and support requirements, the quantity of the k-th type of resource required for the j-th node among the supported objects is:

[0110]

[0111] Among them, v jk represents the support requirement, T represents the threat level, with a value range of [0, 1], where 0 means no threat and 1 means a great threat, a jk represents the basic value of the support requirement when the threat is 0. This basic value is predicted based on equipment reliability parameters and equipment working time (or mission execution time), and has been implemented in previous research (the literature "Research on Equipment Support Requirement Prediction Model Based on Grouping and Stratification [J]. Systems Theory and Practice"), b jk represents the growth rate of the support requirement as the threat level increases.

[0112] In summary, equations (8) - (13) give the constraint conditions of the basic model for regional equipment support resource allocation, and equations (14) - (18) give the relevant calculation formulas.

[0113] 2. Explanation of the Regional Allocation Model of Equipment Support Resources

[0114] Based on the above basic model for regional equipment support resource allocation, combined with the optimal transport theory, the physical connotations of each item in the model are explained as follows:

[0115] (1) Cost

[0116] The cost C in the basic model for regional equipment support resource allocation ij considers factors such as resource characteristics and transportation distance:

[0117] C ij =|pu i -pv j | ω *θ ij *d ij *ρ ij (18)

[0118] Among them, ω is a positive integer. Generally, it takes 1 or 2. When taking 1, it corresponds to the Euclidean distance of the eigenvector. When taking 2, it corresponds to the square of the Euclidean distance of the eigenvector. This kind of metric can better reflect the non-linear growth and can measure the similarity or difference. Different values represent different measurement methods of cost. In actual use, whether to take 1 or 2 depends on the nature of the specific problem and the optimization goal. When taking 2, the influence of noise can be reduced. The square of the distance will amplify large differences and reduce the influence of small differences;

[0119] θ ij ={1, 0}, 1 means u i contains v j all resource types, 0 means u i does not contain v j all resource types, d ij represents A i and B j physical distance, ρ ij is the guarantee relationship, ρ ij ={1, 0}, 1 means there is a guarantee relationship between i and j, 0 means there is no guarantee relationship between i and j.

[0120] Based on the definition of cost, the following basic principle 1 is obtained:

[0121] Basic principle 1: On the premise that the resource feature θ ij = 1 is satisfied, the closer the resource vectors are, the more preferentially they are matched.

[0122] (2) Importance consistency

[0123] Importance consistency E ij considers the KL divergence of the importance between PT ij and the guarantee institution and the guarantee object.

[0124]

[0125] According to the importance consistency, the following basic principle 2 is obtained:

[0126] Basic principle 2: On the premise that the resource feature θ ij = 1 is satisfied, the closer the importances are, the more preferentially they are matched.

[0127] (3) Positional relationship

[0128] According to the literature "Gao Yingjie, Cao Jiping, Lei Ning, Wang Qingxiao. Research review on the optimization of equipment maintenance support resource allocation [J]. Journal of Gun Launch & Control, 2020, 41(03): 102 - 106", considering the spatial position, which is manifested as the positional relationship structure, it is obtained that:

[0129]

[0130] This positional relationship structure corresponds to the sub-item τ in the above basic model. 2 ∑ i,j,k,l |C x (i,k)-C y (j,l)| ε *PTi,jPTk,l. When considering the spatial position, the corresponding sub-item in the original basic model will be replaced by Equation (20).

[0131] Based on the above formula, v j and v l are related, that is, demand-related. Then the correlation coefficient The supply link refers to u i and u k are related. Then the correlation coefficient

[0132] According to the definition of the positional relationship structure, the following basic principle 3 is obtained:

[0133] Basic principle 3: On the premise of satisfying the guarantee relationship, demand correlation and supply link, the closer the distance is, the more preferentially it is matched.

[0134] (4) Functional relationship (system level)

[0135] According to the above literature, considering the network structure, we get:

[0136] ∑ i,j,k,l |C p (i,k)-C q (j,l)| ε *PT ij PT kl (21)

[0137] In the case of low resolution, the system structure shows a functional relationship structure. C p (i,k) and C q (j,l) represent the functional relationship between the guarantee institution and the guarantee object in the mission profile.

[0138] Basic principle 4: On the premise of satisfying the resource characteristics, guarantee relationship and task correlation, the more concentrated the functions are, the more preferentially they are matched.

[0139] In the case of high resolution, considering the system hierarchical structure, C u (i,k) and C v (j,l) represent the structural relationship between the guarantee institution and the guarantee object in the system hierarchy. The present invention only considers the case of high resolution.

[0140] Basic principle 5: On the premise of satisfying the resource characteristics θ ij = 1, guarantee relationship and structural correlation (xik = 1, y jl On the premise that (= 1), the more concentrated the structure is, the higher the priority of matching.

[0141] Furthermore, the matching between the m guarantee institutions and the n guarantee objects in the present invention has three meanings: one is resource matching. The guarantee institutions provide resources and the guarantee objects need resources, and the best matching of resources is carried out between the two according to the task requirements; the second is task matching. The resources provided by the guarantee institutions are carried by the guarantee forces, and the guarantee forces are guided by tasks. The resources needed by the guarantee objects often need to be transformed into the form of guarantee tasks to facilitate guarantee command. Therefore, there is also the best matching of guarantee tasks between the guarantee institutions and the guarantee objects, which is the intermediate medium between resource matching and force matching; the third is force matching. On the basis of the meaning of task matching, the resources needed by the guarantee objects can also be transformed into the needed forces, because the manifestation form of the bottom-up aggregation of underlying resources is force. Therefore, what is manifested between the guarantee institutions and the guarantee objects at the macro level is the best matching of forces, which is the basic logic of command. On this basis, the corresponding transformation function can be designed to realize model solution.

[0142] 3. Explanation example

[0143] Such as Figure 1 shown, taking 3 demand points and 3 supply points as examples for explanation, focusing on explaining the second-layer target position relationship and system-level breakdown. The supply-demand matching is the hierarchical and classified comprehensive matching of multiple indicators of the two. The types and quantities of resource allocations of the supply points obtained after matching are the resource allocation plans. When B1 is matched to A1, it is found that A3 and A1 are very close, and the matching plan tends to match B1 to A3 as well. When B3 and B1 are very close, it also tends to match the combination of B3 and B1 with the combination of A1 and A3. The distance here includes both the physical position relationship and the hierarchical structure relationship. It can be analyzed from this example that:

[0144] (1) Open environment embodiment. In the model, σ, γ, τ, ω, ε are environmental parameters, σ is the importance resolution parameter, γ is the importance matching parameter, which should match both α i and β j match, τ is the position relationship matching parameter, is the structure relationship matching parameter, ω is the supply-demand matching parameter of the resource probability in the cost item, and ε is the supply-demand matching parameter in the position relationship and functional relationship. Different parameters represent different environments. For example, when the value of τ is relatively large, it represents the position priority strategy, when the value is relatively large, it represents the structure priority strategy, when the value of σ is relatively large, it represents a lower importance resolution, and when the value of γ is relatively large, it represents a lower importance matching degree.

[0145] (2) Combined target expression. There are two layers of targets in the model. The first layer is the resource satisfaction rate and utilization rate, and the second layer is the weighted combination of cost items, importance matching items, location relationship matching items, and functional relationship matching items. The combination of inter-layer targets is reflected in the inter-layer based on the probability matrix PT (such as resources from A i Transport to B j The probability of PT ij )’s inter-layer target values ​​interact, and the intra-layer target combination is reflected as the nonlinear weighting of each matching item, and the weight parameter is related to the environment.

[0146] (3) Hierarchical constraint embodiment. Considering the resolution, the guarantee relationship and functional relationship have the same-layer and inter-layer relationship constraints in high and low resolution situations. Since this paper only considers the high-resolution situation, it only reflects the importance and constraints, matching probability and constraints, and supply exceeding demand constraints in the high-resolution situation.

[0147] 4. Model solving algorithm

[0148] The present invention also provides an algorithm for solving the above-constructed equipment support resource regional configuration model, which includes the following steps:

[0149] (1) Input parameters σ, γ, τ, ω,ε,θ ij ,ρ ij and variable α i ,p i ,x i ,β j ,q j ,y j ,v j ;

[0150] (2) Initialize the supply resource vector u i ;

[0151] (3) Normalized u i and v j , get pu i and pv j ;

[0152] (4) Calculated by the following formula:

[0153] d ij =|p i -q j |

[0154] C ij =|pu i -pv j | ω *θ ij *d ij *ρ ij

[0155] (5) Calculate the location distances of the nodes within the support institution system, the location distances of the nodes within the support object system, the structural distances of the nodes within the support institution system, and the structural distances of the nodes within the support object system respectively according to the above formulas;

[0156] (6) Use the fmincon optimizer to solve the second-layer objective that satisfies the above constraints to obtain the resource allocation probability scheme PT ij :

[0157]

[0158] (7) Substitute PT ij into the first-layer objective function (Equation 1) to check whether PT ij meets the requirements of the satisfaction rate and utilization rate. If it meets the requirements, go to (8); if it does not meet the requirements, return to (2);

[0159] (8) Calculate the resource quantities that should be allocated at each supply point

[0160] Through this algorithm, a support resource allocation scheme based on each supply point within the region can be obtained.

[0161] In the present invention, ε is taken as 1, and θ ij has been described. Other parameters σ, γ, τ, are non-negative numbers. The larger the resolution parameter σ during macro-layer matching of the model, the smaller the resolution parameter σ during micro-layer matching. The supply resource vector u i The consideration for initialization is that since the initial value will affect the calculation result, in this paper, the average value of the support object demand aggregated at the support institution is used as the initial value. In practice, the initial value can be given according to experience.

[0162] Embodiment

[0163] The present invention proposes a method for establishing a task-oriented regional allocation model of equipment support resources. In order to verify this method, the following experiments are carried out.

[0164] 1. Experimental case

[0165] Before a certain mission, according to the plan, there are a total of 4 groups and 20 types of equipment units for the combat mission, and a total of 5 support institutions and 12 support units participate in the support mission. The equipment grouping, support institution structure, and importance are as Figure 2 and Figure 3 shown. The information such as the units, locations, resource supply and demand quantities of the equipment grouping and support institutions is shown in Table 1 and Table 2, and the support relationship is shown in Table 3. A support scheduling plan before the mission is formulated. The data in square brackets in Table 1 and Table 2 are the location coordinates for the first two, and the supply and demand quantities of resource A and resource B are the last two respectively.

[0166] Table 1 Relevant Information of Support Units

[0167]

[0168] Table 2 Relevant Information of Equipment Units

[0169]

[0170] Table 3 Support Relationships between Support Agencies and Equipment Groupings

[0171]

[0172] The further support relationships are obtained from Table 3. The support relationships between all support units and equipment units within each support agency and equipment grouping are the same, and are the support relationship values of this agency and grouping. For example, the support relationships between X1 and Y1 are both 1e-3.

[0173] 2. Calculation Results

[0174] In this case, m = 12, n = 20, k = 4, take σ = 1, γ = 1, τ = 1, w = 2, ∈ = 0.5, and the calculation results are as Figure 4 shown. The rows in the figure represent support units, and the columns represent four types of resources.

[0175] Based on the requirements that the resource satisfaction rate is not less than 0.8 and the utilization rate is not less than 0.6, Resource Allocation Schemes 1 and 2 are the calculation results of the tree structure and the simple network structure under the free competition support relationship respectively. Among them, according to the case data, the support relationship is defined as the fixed combination support relationship, abbreviated as Relationship 2. When all matrix elements are adjusted to 1, the support relationship is defined as the free competition support relationship, abbreviated as Relationship 1. The tree structure is abbreviated as Structure 1, and the simple network structure is abbreviated as Structure 2.

[0176] Considering that in practice, there are a large number of zero values in the support requirement matrix during the equipment's mission execution, showing intermittent requirements. The intermittent requirement distribution function SimID(20, 4, 2.5, 2.5, 3) is edited in the matlab software, and the intermittent requirement matrix is obtained as [0, 0, 1, 1; 0, 0, 1, 0; 0, 1, 14, 0; 0, 0, 1, 0; 0, 0, 0, 1; 0, 0, 0, 1; 0, 0, 1, 0; 0, 1, 0, 11; 6, 1, 0, 3; 0, 1, 0, 0; 4, 0, 0, 2; 1, 0, 0, 0; 0, 0, 3, 0; 0, 0, 1, 1; 0, 0, 0, 0; 0, 0, 1, 1; 0, 0, 0, 0; 1, 0, 0, 0; 0, 0, 0, 0; 0, 0, 0, 1]. At this time, the calculation results are as Figure 5As shown, the rows represent the support units and the columns represent the four types of resources. Resource allocation plans 3 and 4 are the calculation results of the tree structure and the simple network structure for relationship 1 respectively.

[0177] During the solution process of this model, it is found that based on the case data, the matrix PT and the constraint error are 15.31%, and based on the intermittent demand data, the matrix PT and the constraint error are 3.37%. The normal distribution, Weibull distribution, uniform distribution data and intermittent distribution data are compared and verified, and it is obtained that: under the intermittent demand condition, the model shows superiority in meeting the matrix PT and the constraints. The influence analysis of parameters σ, γ, τ, is carried out, and it is obtained that: in the case of intermittent demand, the resource allocation plan is robust. Due to space limitations, only the influence of parameter τ at supply point 1 under the tree structure is given here as Figure 6 shown. Since the support demand shows intermittent characteristics in actual situations, this model has good practical application value.

[0178] 3. Performance Analysis

[0179] (1) Algorithm Performance

[0180] Five scales are defined, and the running times are shown in Table 4.

[0181] Table 4 Algorithm Running Times under Five Scales

[0182]

[0183] It can be seen from Table 4 that as the scale increases, the time for generating the support plan increases non-linearly, and the support benefit drops sharply. Therefore, this algorithm is mainly applicable to small-scale conditions.

[0184] (2) Influence of Different Support Relationships

[0185] In this case, resource allocation plans 5 and 6 are obtained for relationship 1 and relationship 2 respectively, as Figure 7 shown, and the mean square deviations of the four types of resources in the two plans are 11.18, 11.61, 10.90, and 11.04 respectively.

[0186] Taking supply point 1 as an example, based on the case data and the simple network, the influence of parameter τ on resource allocation in the two cases of relationship 1 and relationship 2 is considered, as Figure 8 and Figure 9As shown in the figure. In the case of Relationship 1, the influence of parameter τ on resource allocation shows a generally gentle trend, and there is a trough at τ = 6. In the case of Relationship 2, the influence of parameter τ on resource allocation shows a phased gentle trend, and there are two troughs at τ = 4 and 9. Based on the calculated data, it is found that the same rule applies to other supply points. At the same time, for supply points 2, 3, and 6, there are two troughs at τ = 6 and 8 in the case of Relationship 1, and the trough at τ = 8 is not significant.

[0187] (3) Influence of Different Hierarchical Structures

[0188] Taking supply point 1 as an example, based on the analysis of intermittent demand data, the changes in the quantity of resource allocation for parameter τ in the cases of Structure 1 and Structure 2 are as Figure 10 and Figure 11 shown. Generally speaking, the changes in the quantity of resource allocation under intermittent demand are relatively stable, without obvious peaks and valleys. Structure 1 shows a local zigzag structure, and Structure 2 shows a local trapezoidal structure.

[0189] (4) Mechanism Analysis of Configuration Strategies

[0190] Based on the case data, the following parameter values are taken. When adopting the cost - priority strategy, σ = 1, γ = 1, τ = 0.01, when adopting the location - priority strategy, σ = 1, γ = 1, τ = 100, when adopting the structure - priority strategy, σ = 1, γ = 1, τ = 1, The calculation results are as Figure 12 shown. It can be seen that compared with supply points 1 - 6, supply points 7 - 12 need to allocate more resources. When adopting the cost - priority strategy and the location - priority strategy, the difference is more obvious. When adopting the structure - priority strategy, the difference is smaller, indicating that the guarantee is more balanced under the structure - priority strategy.

[0191] Based on the previous intermittent data and taking the above - mentioned parameter values, the calculation results for the three strategies are as Figure 13 shown. It can be seen that under the cost - priority strategy and the location - priority strategy, the law of resource allocation at supply points is the same as the calculation results of the case data. And under the structure - priority strategy, the allocation of Resource 1, Resource 2, and Resource 3 at each supply point is exactly the same, and there is a 0.1 difference in the allocation of Resource 4 at supply points 7, 9, 11, and 12.

[0192] According to the analysis of the above two situations, in this case, the influence of the cost - priority, location - priority, and structure - priority strategies on the resource allocation results has the characteristics of insensitivity to resource types and local aggregation at supply points.

[0193] (5) Influence Analysis of Environmental Parameters

[0194] Both the demand type and environmental parameters affect the resource allocation plan. Taking the parameter For example, distinguish case data and intermittent data to obtain all supply points The analysis results with values ranging from 1 to 10 are as Figure 14 and Figure 15 shown

[0195] According to Figure 14 and Figure 15 Analysis shows that: ① The model shows relatively large fluctuations under case data conditions. Generally speaking, the resource allocation rules among supply points can be summarized into two situations. The resource allocation of supply points 1 - 6 shows a downward trend, and that of supply points 7 - 12 shows an upward trend. However, the resource allocation rules for different resource types are consistent; ② The model shows a smoother performance under intermittent data conditions. The resource allocation rules among supply points are more consistent. The resource allocation rules for different resource types show overall consistency and exhibit a local sawtooth structure for different resource types

[0196] (6) Influence analysis of threat level

[0197] Under the condition of mission confrontation, considering the threat from the enemy, the resource quantity given in the case is the basic value a for ensuring demand jk , if b jk takes 1, considering T = 0, 0.2, 0.6 corresponding to three situations of no threat, relatively small threat, and relatively large threat respectively, the obtained resource allocation schemes are as Figure 16 shown

[0198] Obviously, when the threat level increases, the resource allocation quantity increases. In this case, with the increase of the threat level, the increase in resource allocation is positively correlated with the basic value of ensuring demand

[0199] (7) Comparative analysis of modeling methods

[0200] For the resource allocation problem, the analysis perspective of the present invention is based on two points: one is not considering the part replacement time, and the other is normalizing both supply and demand resources. Thus, the spatial distribution of resources is obtained, and on this basis, the optimal transport theory is used for modeling. Based on this model, a comparative analysis is carried out for two situations: the supply - demand resource distribution perspective (this method) and the quantity perspective (traditional method). On the basis of the previous intermittent demand, the demand quantity is magnified by 10 times to magnify the difference in the allocation schemes of the two methods. The line charts of resource allocation schemes 10 and 11 are as Figure 17 shown

[0201] It can be seen from the figure that under the parameter selection of this case, this method can better reflect the demand distribution differences from the overall perspective, and thus this difference is reflected in the resource allocation scheme. The traditional method is based on the quantity perspective and does not show obvious differences in this case

[0202] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions of the present invention or make equivalent replacements, and these modifications or equivalent replacements do not enable the modified technical solutions to depart from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for establishing a task-oriented regional allocation model of equipment support resources, characterized in that It includes the following steps: Step 1: Construct a basic model for the regional allocation of equipment support resources according to the support institutions and support objects within the region; Step 2: Define the basic principles of the basic model for the regional allocation of equipment support resources; Step 3: Solve the basic model for the regional allocation of equipment support resources to obtain the amount of resources that should be allocated at each supply point; Step 4: Obtain a support resource allocation plan based on each supply point within the region according to the amount of resources that should be allocated at each supply point.

2. The method for establishing a task-oriented equipment support resource regional allocation model according to claim 1, characterized in that The specific steps of Step 1 include: Step 1.1: Denote the set of guarantee institution nodes as {A1, A2…A m}, and the i-th node of the guarantee institution can be expressed as A i = {α i , p i , x i , u i}, where m is the number of guarantee institutions, α i is the node importance, p i is the node spatial position, p i = {p i1 , p i2}, p i1 , p i2 represent the abscissa and ordinate in the Euclidean space respectively, x i is the relationship structure between the node and other nodes, represented by an adjacency matrix, u i is the node resource feature, represented by a vector of resource allocation quantities, u i = {u i1 , u i2 …u ik}, u ik is the quantity of the k-th resource of the i-th node, and k is the type of resource; Step 1.2: Denote the set of protected object nodes as {B1, B2…B n}, and the j-th node of the protected object can be represented as B j = {β j , q j , y j , v j}, where n is the number of protected objects, β j is the node importance, q j is the node spatial position, q j = {q j1 , q j2}, q j1 , q j2 represent the abscissa and ordinate in the Euclidean space respectively, y j is the relationship structure between the node and other nodes, represented by an adjacency matrix, v j is the node resource feature, represented by a vector of resource demand quantities, v j = {v j1 , v j2 …v jk}, v jk is the quantity of the k-th resource of the j-th node, and k is the type of resource; Step 1.3: Transfer the resources from A i to B j The cost is defined as C ij The resources are transferred from A i to B j The probability is defined as PT ij ; Step 1.4: Based on the above definitions, establish a basic model for the regional allocation of equipment support resources: 1) The first-layer objective is to optimize the resource satisfaction rate and utilization rate Among them, represents the satisfaction rate of the k-th resource; represents the utilization rate of the k-th resource; 2) The second-layer objective is to minimize the resource transportation cost: In the formula, the first term is the resource feature matching term, and the second term is the importance feature resolution term. The third and fourth terms are the matching terms for importance α i and β j respectively. The fifth term |C x (i,k) - C y (j,l) | ε * PT ij PT kl is the positional relationship structure matching term. The sixth term |C p (i,k) - C q (j,l)| ε * PT ij PT kl is the functional relationship (system level) structure matching term. PT kl represents the probability that the resource is transported from A k to B l . A and B represent the set of support agency nodes and the set of support object nodes respectively; ε represents the norm type, σ represents the resolution parameter, γ represents the importance balance parameter, τ represents the positional relationship matching weight, represents the functional relationship structure matching weight, and σ, γ, τ, are all non - negative numbers.

3. The method for establishing a task-oriented equipment support resource regional allocation model according to claim 2, wherein The constraint conditions of the basic model for the regional allocation of equipment support resources established in Step 1.4 include: ∑ i α i ＝1 (8) ∑ j β j ＝1 (9) Among them, Equation (8) represents the importance of each node of the support institution, Equation (9) represents the importance of each node of the support object, Equation (10) represents the transportation probability and constraint from each support institution node to all support object nodes, Equation (11) represents the transportation probability and constraint from all support institution nodes to each support object node, Equation (12) represents that for each type of resource, the sum of the resources of this type allocated at the support institution node should be greater than the resources of this type at each support demand node, and Equation (13) represents that for each type of resource, the sum of the resources demanded by the support object node should be less than the resources of this type at each support institution node; v jk represents a safeguard requirement, and: The amount of the kth type of resource that should be allocated at the ith support institution node is: where T represents the threat level, a jk represents the basic value of the security requirement when the threat is 0, and b jk represents the growth rate of the security requirement as the threat level increases.

4. The method for establishing a task-oriented equipment support resource regional allocation model according to claim 3, characterized in that, The specific steps of Step 3 include: Step 3.1: Input parameters σ, γ, τ, ω, ε, θ ij , ρ ij and variable α i , p i , x i , β j , q j , y j , v j ; Step 3.2: Initialize the supply resource vector u i ; Step 3.3: Calculate C ij , C x (i, k), C y (j, l), C p (i, k) and C q (j, l); Step 3.4: Use the fmincon optimizer to solve the second-layer objective that satisfies the above constraints, and obtain the resource allocation probability scheme PT i,j : Step 3.5: Substitute PT ij into the first-layer objective function to check whether PT i,j meets the requirements of the satisfaction rate and utilization rate. If it meets the requirements, go to Step 3.6; if it does not meet the requirements, go back to Step 3.2; Step 3.6: Calculate the resource quantity to be configured at each supply point

Images