Offshore wind plant 5G base station deployment decision-making method, system and equipment and storage medium

Abstract

The invention provides an offshore wind plant 5G base station deployment decision-making method, system and device, and a medium. The method comprises the following steps: S1, constructing and generating a required set and matrix; s2, establishing a bipartite graph model according to the matrix; s3, finding an initial solution by using a greedy algorithm; s4, disturbing the initial solution to generate a primary population, and calculating the deployment cost of individuals in the population; s5, performing crossover operation and mutation operation on the population by using a genetic algorithm to obtain a new individual deployment cost; s6, more superior individuals are selected to form a new population according to the deployment cost and the coverage condition; and S7, verifying whether the algorithm reaches a termination condition. A precise coverage relation model is established based on a bipartite graph, and network effectiveness and deployment economy are synchronously optimized through a differentiated base station type selection strategy on the premise of ensuring full fan coverage; a two-stage intelligent optimization algorithm is designed, a feasible solution is rapidly generated through a greedy algorithm, redundant configuration is eliminated through heuristic iteration, and finally a high-cost-performance deployment scheme is obtained.

Description

Technical Field

[0001] The present invention belongs to the field of communication technologies, and particularly relates to a method, a system, a device, and a storage medium for deploying and making decisions on 5G base stations in an offshore wind farm. Background Art

[0002] With the transformation of the global energy structure towards clean and renewable energy, offshore wind farms have gradually become new energy projects that are key for various countries to develop. Offshore wind farms not only have advantages such as rich resources, high power generation efficiency, and no occupation of land space. However, the construction and operation and maintenance of offshore wind farms face many challenges, among which an efficient and stable communication network is a key factor for ensuring the normal operation, remote monitoring, and intelligent management of the wind farm.

[0003] Currently, offshore communication mainly relies on satellite communication, microwave communication, and 4G / 5G mobile communication technologies, and each of these technologies has its own characteristics: Satellite communication: Although it has the advantage of wide coverage, it has deficiencies such as large communication latency, high cost, and limited bandwidth, and it is difficult to meet the needs of high-frequency data transmission; Microwave communication, although having the characteristics of low latency and high bandwidth and being suitable for land-sea data transmission, is easily affected by weather conditions and has complex deployment, making it difficult to cover the communication needs between each wind turbine within the wind farm; For 4G / 5G communication technologies, especially 5G communication technology, with its advantages of large bandwidth, low latency, and high reliability, it provides a new communication means for offshore wind farms and can support intelligent applications such as high-definition video backhaul, drone inspection, and sensor data collection.

[0004] However, the application of 5G communication technology in offshore wind farms still faces many problems. First, due to the limited coverage range of 5G base stations and the relatively scattered distribution of wind turbines in offshore wind farms, deploying multiple 5G base stations is often required to achieve full coverage; at the same time, when deploying multiple 5G base stations, because the types of base stations are different, their coverage ranges will also change, further increasing the complexity of the deployment scheme design; in addition, there are significant differences in the construction costs and maintenance costs of deploying different types of base stations. Therefore, how to reduce the deployment cost while meeting the coverage requirements is still a difficult problem that urgently needs to be solved. Summary of the Invention

[0005] The first object of the present invention is to provide a method for making decisions on the deployment of 5G base stations in an offshore wind farm in view of the above-mentioned problems.

[0006] To this end, the above object of the present invention is achieved by the following technical solutions:

[0007] A method for making decisions on the deployment of 5G base stations in an offshore wind farm includes the following steps:

[0008] S1. Construct a set of wind turbines \(M\), a set of candidate deployment points \(N\), and a set of base stations \(K\), and generate a coverage relationship matrix \(R\), a deployment cost matrix \(c\), and a unit cost coverage rate matrix \(W\);

[0009] S2. Based on the coverage relationship matrix \(R\) generated in step S1, establish a bipartite graph model \(G\), and model the deployment problem of base stations as a bipartite graph covering problem;

[0010] S3. For the bipartite graph model \(G\), find an initial solution that can meet the requirement of covering all wind turbines based on the greedy algorithm;

[0011] S4. Generate a diverse initial population by perturbing the initial solution, and calculate the deployment cost of individuals in the population using the deployment cost matrix \(C\), and use the initial population as the input of the genetic algorithm;

[0012] S5. Use the genetic algorithm to perform crossover operations and mutation operations on the population to obtain new individuals, and calculate the deployment cost of the new individuals using the deployment cost matrix \(C\);

[0013] S6. Comprehensively consider the deployment cost and coverage of individuals, and select more superior individuals to form a new population;

[0014] S7. Verify whether the algorithm reaches the termination condition. If the termination condition is met, end the algorithm and output the deployment decision. Otherwise, go to step S5.

[0015] While adopting the above technical solution, the present invention can also adopt or combine the following technical solutions:

[0016] As a preferred technical solution of the present invention: In step S1,

[0017] The coverage relationship matrix \(R\) is used to describe the coverage relationship between different base stations deployed at each candidate deployment point and wind turbines. The coverage relationship matrix \(R\) is a three-dimensional matrix, and the three dimensions respectively represent wind turbines, candidate deployment points, and base station types, with a size of \(|M|\times|N|\times|K|\). Its element \(R\) m,n,k represents the coverage relationship of the \(m\)-th wind turbine after the \(k\)-th base station is deployed at the \(n\)-th candidate deployment point. 0 means unable to cover, and 1 means can cover;

[0018] The deployment cost matrix \(C\) is used to describe the cost of deploying different base stations at each candidate deployment point. The deployment cost matrix \(C\) is a two-dimensional matrix, and the two dimensions respectively represent candidate deployment points and base station types, with a size of \(|N|\times|K|\). Its element \(C\) n,k represents the cost of deploying the \(k\)-th base station at the \(n\)-th candidate deployment point;

[0019] The unit - cost coverage matrix \(W\) is used to describe the number of wind turbines that can be covered per unit cost. The unit - cost coverage matrix \(W\) is a two - dimensional matrix. The two dimensions respectively represent the candidate deployment points and the base - station types, with a size of \(|N|\times|K|\). Its element \(W\) n,k

[0020] represents the number of wind turbines that can be covered per unit cost when deploying the \(k\) - th type of base - station at the \(n\) - th candidate deployment point. The formula is as follows:

[0021]

[0022] As a preferred technical solution of the present invention: In step S2, the bipartite - graph model \(G\) is constructed according to the coverage - relation matrix \(R\), where \(G=(V, E)\). Its vertex set \(V\) can be divided into two non - overlapping subsets \(V1\) and \(V2\), and \(V1\) and \(V2\) satisfy and \(V1\cup V2 = V\);

[0023] Among them, the vertices in subset \(V1\) correspond one - to - one with the wind - turbine dimension in the coverage - relation matrix \(R\). For subset \(V2\), if the element in the coverage - relation matrix \(R\) satisfies then add a vertex to subset \(V2\), indicating the deployment of the \(k\) - th type of base - station at the \(n\) - th candidate deployment point;

[0024] If the element \(R\) in the coverage - relation matrix \(R\) m,n,k \( = 1\), then correspondingly add an edge to the edge set \(E\) of the bipartite - graph model \(G\), indicating that the \(m\) - th wind turbine can be covered when deploying the \(k\) - th type of base - station at the \(n\) - th candidate deployment point.

[0025] As a preferred technical solution of the present invention: In step S2, the bipartite - graph covering problem is to find an edge set such that all vertices \(v\in V1\) in subset \(V1\) belong to at least one edge in set \(X\).

[0026] As a preferred technical solution of the present invention: Step S3 further includes the following sub - steps:

[0027] S31. Initialization, construct empty sets \(S\), Covered, \(E'\) and \(E''\). Among them, set \(S\) is used as the solution set, set Covered is used to record the coverage situation of wind turbines, and sets \(E'\) and \(E''\) are used to record the edges that change during the process;

[0028] S32. Find all vertices \(v1\) with degree 1 in subset \(V1\), find their adjacent vertices \(u1\in V2\), add all adjacent nodes \(N1(v)=\{v\in V1|(u1, v)\in E\}\) of vertex \(u1\) to set Covered, and add vertex \(u1\) to the solution set \(S\);

[0029] S33. Delete the edge in the bipartite - graph \(G\) That is, delete the associated edges between vertex u1 and its adjacent nodes N1(v);

[0030] S34. In the bipartite graph, delete all the associated edges of the vertices N2(u) that are the same as the candidate deployment points represented by vertex u1 Meanwhile, let

[0031] S35. Find the vertex u2 with the largest degree in the vertex subset V2 at present, add all the adjacent nodes N3(v) = {v ∈ V1|(u2, v) ∈ E} of vertex u2 to the set Covered, and add vertex u2 to the solution set S;

[0032] S36. In the bipartite graph G, delete the edges

[0033] S37. In the bipartite graph, delete all the associated edges of the vertices N4(u) that are the same as the candidate deployment points represented by vertex u2 Meanwhile, let

[0034] S38. Judge whether the condition Covered = V1 is satisfied. If it is satisfied, go to step S39; otherwise, go to step S32;

[0035] S39. Construct a matrix Y1 of size |N|×|K|, whose elements represent whether to deploy the k-th type of base station at the n-th candidate deployment point. Set the elements corresponding to the vertices in the set S to 1, and set the values of the remaining elements to 0. Take the matrix Y1 as the initial solution of the deployment decision.

[0036] As a preferred technical solution of the present invention: In step S4, the deployment cost of an individual is calculated by taking the trace of the result matrix obtained by multiplying the deployment scheme matrix Y by the transpose of the deployment cost matrix C. The formula is as follows:

[0037] cost = Tr(Y·C T )

[0038] As a preferred technical solution of the present invention: Step S6 further includes the following sub-steps:

[0039] S61. Set the maximum population size |P max |, assign the coefficient α, divide the population into two parts and calculate the population sizes of the two parts respectively. The formula is as follows:

[0040] |P a | = α|P max |

[0041] |Pb | = (1 - α)|P max |

[0042] S62. Mix the parental population and the offspring population to obtain a mixed population P0, and divide the individuals in P0 that can meet the requirements of covering all wind turbines and those that cannot meet the requirements into populations P1 and P2;

[0043] S63. For the individuals in population P1, sort them in ascending order according to the deployment cost cost. If |P1| > |P a |, then select the first |P’1| = |P a | individuals and add them to the new population. If not, add all the individuals in P1 to the new population, and the insufficient part is supplemented by the individuals in P2;

[0044] S64. For the individuals in population P2, define the coverage rate as Then, according to the deployment cost index of each individual and the coverage rate index weighted sum to obtain its corresponding comprehensive index, and calculate its comprehensive index γ,

[0045] The formula for the deployment cost index is as follows:

[0046]

[0047] The formula for the coverage rate index φ is as follows:

[0048]

[0049] The formula for the comprehensive index γ is as follows:

[0050]

[0051] In the formula, cost i , ρ i respectively represent the deployment cost and the coverage rate of the i-th individual, represents the minimum deployment cost and the maximum coverage rate of each individual in population P2, and λ is a non-negative constant less than 1;

[0052] S65. Sort the individuals in population P2 in descending order according to the comprehensive index. If the condition |P1| > |P a | is satisfied, then select the first |P’2| = |P b | individuals and add them to the new population; if the condition is not satisfied, then select the first |P’2| = |P max | - |P1| individuals and add them to the new population.

[0053] The second object of the present invention is to provide a decision-making system for the deployment of 5G base stations in an offshore wind farm, including the following modules:

[0054] A data model construction module for constructing a set of wind turbines, a set of candidate deployment points, and a set of base stations, and generating a coverage relationship matrix, a deployment cost matrix, and a unit cost coverage rate matrix;

[0055] A bipartite graph modeling module for modeling the deployment problem of base stations as a bipartite graph coverage problem;

[0056] An initial solution generation module for finding an initial solution that can meet the requirement of covering all wind turbines based on the greedy algorithm;

[0057] A genetic algorithm optimization module for optimizing the genetic algorithm;

[0058] An output module for finding a coverage plan that meets the expected cost, or reaching the maximum number of iterations, and outputting a deployment matrix that covers all wind turbines and has the lowest cost.

[0059] The third object of the present invention is to provide an electronic device, which includes a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory complete mutual communication through the communication bus.

[0060] The memory is used to store a computer program;

[0061] The processor is used to execute the computer program stored on the memory to implement the steps of the method for making a decision on the deployment of 5G base stations in an offshore wind farm as described above.

[0062] Another object of the present invention is to provide a non-volatile storage medium, in which an executable program is stored. When the executable program is executed by a processor, the steps of the method for making a decision on the deployment of 5G base stations in an offshore wind farm as described above are implemented.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] 1) By constructing a multi-dimensional decision-making model of candidate base stations, wind turbine positions, and base station types, comprehensively considering the base station coverage requirements and deployment costs, the optimal balance between the communication network coverage effect and the deployment cost is achieved, providing a reference for reducing the 5G base station deployment cost and optimizing the base station site selection;

[0065] 2) By constructing a coverage model based on a bipartite graph, the coverage relationship between candidate base station locations and wind turbines is clarified, effectively ensuring that each wind turbine is covered by at least one base station, thus meeting the communication coverage requirements of the wind turbines. At the same time, by comprehensively considering the coverage range and deployment cost of different types of base stations, the proposed base station deployment plan is more in line with the requirements of the actual application scenario, ensuring the effectiveness of communication coverage while achieving reasonable cost control;

[0066] 3) The greedy algorithm is used to quickly generate an initial deployment plan to ensure a feasible solution that meets the coverage requirements within a short time. Next, the initial plan is iteratively optimized through a heuristic optimization algorithm to gradually change the deployment plan, reduce redundant deployment and deployment costs, and finally find a deployment plan that can meet the requirements of covering all wind turbines and has a relatively low cost;

[0067] The present invention can, while meeting the 5G communication requirements of an offshore wind farm, find a deployment plan that can significantly reduce the deployment and operation and maintenance costs of base stations, thereby providing efficient and reliable communication guarantee for the intelligent operation and maintenance of the offshore wind farm. Brief Description of the Drawings

[0068] Figure 1 It is a flowchart of the 5G base station deployment decision method provided by the present invention for an offshore wind farm.

[0069] Figure 2 It is a flowchart of the greedy algorithm for finding an initial solution.

[0070] Figure 3 It is a flowchart of the composition of a new population. Detailed Embodiment

[0071] The present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0072] As Figure 1 shown, a 5G base station deployment decision method for an offshore wind farm specifically includes the following steps:

[0073] S1. Construct a wind turbine set M, a candidate deployment point set N, and a base station set K, and generate a coverage relationship matrix R, a deployment cost matrix C, and a unit cost coverage rate matrix W. The attributes of the elements m in the set M, n in the set N, and k in the set K can be respectively expressed as: m(L m ), n(L n , T n ), k(T k , A k ), where L m and L n respectively represent the location information of the wind turbine and the candidate deployment point, T n represents the type of base station that can be deployed at this candidate deployment point, Tk Indicates the type of base station, A k Indicates the coverage range that this type of base station can cover:

[0074] The coverage relationship matrix R is used to describe the coverage relationships between different base stations deployed at each candidate deployment point and the wind turbines. The coverage relationship matrix R is a three-dimensional matrix. The three dimensions respectively represent wind turbines, candidate deployment points, and base station types, with a size of |M|×|N|×|K|. Its element R m,n,k Indicates the coverage relationship of the k-th type of base station deployed at the n-th candidate deployment point to the m-th wind turbine. 0 indicates non-coverage, and 1 indicates coverage;

[0075] The deployment cost matrix C is used to describe the deployment costs of different base stations at each candidate deployment point. The deployment cost matrix C is a two-dimensional matrix. The two dimensions respectively represent candidate deployment points and base station types, with a size of |N|×|K|. Its element C n,k Indicates the cost of deploying the k-th type of base station at the n-th candidate deployment point;

[0076] The unit cost coverage rate matrix W is used to describe the number of wind turbines that can be covered per unit cost. The unit cost coverage rate matrix W is a two-dimensional matrix. The two dimensions respectively represent candidate deployment points and base station types, with a size of |N|×|K|. Its element W n,k Indicates the number of wind turbines that can be covered per unit cost when the k-th type of base station is deployed at the n-th candidate deployment point. The formula is as follows:

[0077]

[0078] S2. Based on the coverage relationship matrix R generated in step S1, establish a bipartite graph model G, and model the base station deployment problem as a bipartite graph coverage problem:

[0079] The bipartite graph model G is constructed according to the coverage relationship matrix R and can be represented by an undirected bipartite graph G=(V, E). According to the characteristics of the bipartite graph, its vertex set V can be divided into two non-overlapping subsets V1 and V2, and V1 and V2 satisfy and V1∪V2 = V. Therefore, the undirected bipartite graph G can also be represented as G=(V1, V2, E). Its two types of vertex subsets V1 and V2 respectively correspond to the wind turbine node m and the combined node (n, k) of the candidate deployment point and the base station. And only when there is a coverable wind turbine m for the (n, k) combination, that is, it satisfies in the coverage relationship matrix R will the combined node (n, k) be added to V2;

[0080] For the edge set E of the undirected bipartite graph G=(V1, V2, E), when the (n, k) combination can cover the wind turbine m, that is, it satisfies R m,n,k = 1 in the coverage relationship matrix R, then a corresponding edge is added to the edge set E.

[0081] The bipartite graph covering problem is to find an edge set such that all vertices v ∈ V1 in the subset V1 belong to at least one edge in the set X.

[0082] S3. For the bipartite graph model G, find an initial solution that can meet the requirement of covering all wind turbines based on the greedy algorithm, as Figure 2 shown below:

[0083] S31. Initialization: construct empty sets S, Covered, E', and E", where the set S is used as the solution set, the set Covered is used to record the coverage status of wind turbines, and the sets E' and E" are used to record the edges that change during the process;

[0084] S32. Find all vertices v1 with degree 1 in the subset V1, find their adjacent vertices u1 ∈ V2, add all adjacent nodes N1(v) = {v ∈ V1|(u1, v) ∈ E} of vertex u1 to the set Covered, and add vertex u1 to the solution set S;

[0085] S33. Delete the edges in the bipartite graph G, that is, delete the associated edges between vertex u1 and its adjacent nodes N1(v);

[0086] S34. Delete all associated edges of vertices N2(u) that represent the same candidate deployment point as vertex u1 in the bipartite graph Meanwhile, let

[0087] S35. Find the vertex u2 with the largest degree in the vertex subset V2 at present, add all adjacent nodes N3(v) = {v ∈ V1|(u2, v) ∈ E} of vertex u2 to the set Covered, and add vertex u2 to the solution set S;

[0088] S36. Delete the edges

[0089] S37. Delete all associated edges of vertices N4(u) that represent the same candidate deployment point as vertex u2 in the bipartite graph Meanwhile, let

[0090] S38. Judge whether the condition Covered = V1 is satisfied. If it is satisfied, go to step S39; otherwise, go to step S32;

[0091] S39. Construct a matrix Y1 of size |N|×|K|, whose elements Indicate whether to deploy the k-th type of base station at the n-th candidate deployment point, and set the elements corresponding to the vertices in the set S to 1, and the values of the remaining elements to 0. Use the matrix Y1 as the initial solution of the deployment decision. The value is set to 1, and the values of the remaining elements are set to 0. Use the matrix Y1 as the initial solution of the deployment decision.

[0092] S4. Generate a diverse initial population by perturbing the initial solution, and calculate the deployment cost of individuals in the population using the deployment cost matrix C. Use the initial population as the input of the genetic algorithm:

[0093] Perturbing the initial solution means randomly adjusting the deployment plans for some candidate deployment points, including adding new base station deployments at candidate deployment points where no base station is deployed, removing deployed base stations, or replacing the types of deployed base stations, so as to generate diverse individuals Y. The generated individuals need to meet the constraints That is, at most one base station is deployed at each candidate deployment point;

[0094] The initial population allows the existence of a certain proportion of deployment plans that do not fully cover all wind turbines, increasing the diversity of solutions, thereby enhancing the ability of the algorithm to jump out of local optimal solutions;

[0095] The initial population allows the existence of a certain proportion of deployment plans that do not fully cover all wind turbines, increasing the diversity of solutions, thereby enhancing the ability of the algorithm to jump out of local optimal solutions;

[0096] The calculation of the deployment cost of an individual is obtained by multiplying the deployment plan matrix Y by the transpose of the deployment cost matrix C and then calculating the trace of the resulting matrix. The formula is as follows:

[0097] cost = Tr(Y · C T )

[0098] Among them, Y is the deployment plan matrix corresponding to the individual in the population, and C T is the transpose of the deployment cost matrix C. Since there is at most one non-zero element in each row of the matrix Y and its value is 1, that is, it satisfies Therefore, multiply the deployment plan matrix Y by the transpose C of the deployment cost matrix C T and then calculate the trace of the result. The obtained result is the deployment cost cost of the corresponding individual.

[0099] S5. Use the genetic algorithm to perform crossover operations and mutation operations on the population to obtain new individuals, and calculate the deployment cost of the new individuals using the deployment cost matrix C:

[0100] Performing a crossover operation on the population through the genetic algorithm means randomly selecting two individuals in the population as parents, and then assigning values to each row of the offspring matrix. Among them, the individual with a lower deployment cost has a higher probability of being selected as a parent. The i-th row of the offspring matrix is determined by the element Y with a value of 1 in the i-th row of the parent matrix. Select Y n,k to determine, select Yn,k In the unit cost coverage rate matrix W, the parent with a larger corresponding element W n,k is used as the source of the i-th row of the offspring matrix;

[0101] Performing a mutation operation on the population through a genetic algorithm means that for the offspring matrix Y generated through the crossover operation, on the premise of satisfying the constraint randomly change the element distribution in the offspring matrix according to a certain probability.

[0102] S6. Considering the deployment cost and coverage of individuals comprehensively, select more superior individuals to form a new population, as Figure 3 shown:

[0103] S61. Set the maximum population size |P max |, the distribution coefficient α, divide the population into two parts and calculate the population sizes of the two parts respectively. The formula is as follows:

[0104] |P a | = α|P max |

[0105] |P b | = (1 - α)|P max |

[0106] S62. Mix the parent population and the offspring population to obtain a mixed population P0, and divide the individuals in P0 that can meet the requirement of covering all wind turbines and those that cannot meet the requirement into populations P1 and P2;

[0107] S63. For the individuals in population P1, sort them in ascending order according to the deployment cost cost. If |P1| > P a |, then select the first |P’1| = |P a | individuals according to the sorting result and add them to the new population. If not, add all the individuals in P1 to the new population, and the insufficient part is supplemented by the individuals in P2;

[0108] S64. For the individuals in population P2, define the coverage rate as Then, according to the deployment cost index and the coverage rate index φ, perform a weighted sum to obtain its corresponding comprehensive index and calculate its comprehensive index γ,

[0109] The formula for the deployment cost index is as follows:

[0110]

[0111] The formula for the coverage rate index φ is as follows:

[0112]

[0113] The formula for the comprehensive index γ is as follows:

[0114]

[0115] where cost i , ρ i respectively represent the deployment cost and coverage rate of the i-th individual, represents the minimum deployment cost and maximum coverage rate of each individual in the population P2, and λ is a non-negative constant less than 1;

[0116] S65. Sort the individuals in the population P2 in descending order of the comprehensive index. If the condition |P1| > |P a | is satisfied, then select the first |P’2| = |P b | individuals according to the sorting result and add them to the new population; if the condition is not satisfied, then select the first |P’2| = |P max | - |P1| individuals according to the sorting result and add them to the new population.

[0117] S7. Verify whether the algorithm reaches the termination condition. If the termination condition is satisfied, end the algorithm and output the deployment decision; otherwise, go to step S5:

[0118] Verifying whether the termination condition is reached includes verifying whether there is an individual whose deployment plan meets the standard and whether the algorithm reaches the maximum number of iterations. If either of them is satisfied, it is considered that the termination condition is reached;

[0119] Set the expected deployment cost ε. If there appears an individual in the population that can simultaneously meet the requirements of covering all wind turbines and having a deployment cost not higher than the expected deployment cost ε, then it is considered that this deployment plan meets the standard;

[0120] Set the maximum number of iterations G of the algorithm max , when the number of iterations G of the algorithm = G max but the algorithm has not ended yet, it is considered that the deployment cost cannot be reduced to not higher than the expected deployment cost ε. At this time, select the individual with the minimum cost among those that can cover all wind turbines as the deployment plan for output.

[0121] The present invention also provides an offshore wind farm 5G base station deployment decision-making system, including the following modules:

[0122] A data model construction module for constructing a wind turbine set, a candidate deployment point set, and a base station set, and generating a coverage relationship matrix, a deployment cost matrix, and a unit cost coverage rate matrix;

[0123] A bipartite graph modeling module for modeling the deployment problem of the base station as a bipartite graph coverage problem;

[0124] An initial solution generation module, which is used to find an initial solution that can meet the requirements of covering all wind turbines based on a greedy algorithm;

[0125] A genetic algorithm optimization module, which is used to optimize the genetic algorithm;

[0126] An output module, which is used to find a coverage plan that meets the expected cost, or when the maximum number of iterations is reached, output a deployment matrix that covers all wind turbines and has the lowest cost.

[0127] The present invention also provides an electronic device. The electronic device includes a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory complete mutual communication through the communication bus.

[0128] The memory is used to store a computer program;

[0129] The processor is used to execute the computer program stored on the memory to implement the steps of the method for making a decision on the deployment of 5G base stations in an offshore wind farm as described above.

[0130] The present invention also provides a non-volatile storage medium. The non-volatile storage medium stores an executable program. When the executable program is executed by a processor, the steps of the method for making a decision on the deployment of 5G base stations in an offshore wind farm as described above are implemented.

[0131] So far, the technical solution of the present invention has been described in combination with the specific experimental process shown in the drawings. However, the protection scope of the present invention is not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A decision-making method for 5G base station deployment in an offshore wind farm, characterized in that, It includes the following steps: S1. Construct a set of wind turbines \(M\), a set of candidate deployment points \(N\), and a set of base stations \(K\), and generate a coverage relationship matrix \(R\), a deployment cost matrix \(C\), and a unit cost coverage rate matrix \(W\); S2. Based on the coverage relationship matrix \(R\) generated in step S1, establish a bipartite graph model \(G\), and model the deployment problem of base stations as a bipartite graph coverage problem; S3. For the bipartite graph model \(G\), find an initial solution that can meet the requirement of covering all wind turbines based on the greedy algorithm; S4. Generate a diverse initial population by perturbing the initial solution, and calculate the deployment cost of individuals in the population using the deployment cost matrix \(C\), and use the initial population as the input of the genetic algorithm; S5. Use the genetic algorithm to perform crossover operations and mutation operations on the population to obtain new individuals, and calculate the deployment cost of the new individuals using the deployment cost matrix \(C\); S6. Comprehensively consider the deployment cost and coverage of individuals, and select more superior individuals to form a new population; S7. Verify whether the algorithm reaches the termination condition. If the termination condition is met, end the algorithm and output the deployment decision. Otherwise, go to step S5.

2. The method according to claim 1, wherein: In step S1, The coverage relationship matrix R is used to describe the coverage relationships between different base stations and wind turbines deployed at each candidate deployment point. The coverage relationship matrix R is a three-dimensional matrix, and the three dimensions respectively represent wind turbines, candidate deployment points, and base station types, with a size of |M|×|N|×|K|. Its element R m,n,k represents the coverage relationship of the m-th wind turbine after the k-th type of base station is deployed at the n-th candidate deployment point. 0 indicates non-coverage, and 1 indicates coverage; The deployment cost matrix C is used to describe the costs of deploying different base stations at each candidate deployment point. The deployment cost matrix C is a two-dimensional matrix, and the two dimensions respectively represent the candidate deployment points and the base station types, with a size of |N|×|K|. Its element C n,k represents the cost of deploying the k-th type of base station at the n-th candidate deployment point; The unit cost coverage matrix W is used to describe the number of wind turbines that can be covered per unit cost. The unit cost coverage matrix W is a two-dimensional matrix, and the two dimensions represent the candidate deployment points and the base station types respectively, with a size of |N|×|K|. Its element W n,k represents the number of wind turbines that can be covered per unit cost when the k-th type of base station is deployed at the n-th candidate deployment point. The formula is as follows:

3. The method according to claim 1, wherein: In step S2, the bipartite graph model G is constructed according to the covering relation matrix R, where G = (V, E), and its vertex set V can be partitioned into two disjoint subsets V1 and V2, and V1 and V2 satisfy and V1 ∪ V2 = V; Among them, the vertices in the subset V1 correspond one-to-one with the fan dimensions in the coverage relationship matrix R. For the subset V2, if the elements in the coverage relationship matrix R satisfy then add a vertex to the subset V2, indicating that the k-th type of base station is deployed at the n-th candidate deployment point; If the element R in the coverage relationship matrix R m,n,k = 1, then add a corresponding edge to the edge set E of the bipartite graph model G, indicating that when deploying the k-th type of base station at the n-th candidate deployment point, the fan m can be covered.

4. The method according to claim 1 or 3, characterized in that: In step S2, the bipartite graph covering problem is to find an edge set such that all vertices v ∈ V1 in subset V1 belong to at least one edge in set X.

5. The method according to claim 1, characterized in that: Step S3 also includes the following sub-steps: S31. Initialize, construct empty sets \(S\), \(Covered\), \(E'\) and \(E''\), where the set \(S\) is used as the solution set, the set \(Covered\) is used to record the coverage status of wind turbines, and the sets \(E'\) and \(E''\) are used to record the edges that change during the process; S32. Find all vertices \(v1\) with degree 1 in the subset \(V1\), find their adjacent vertices \(u1\in V2\), add all adjacent nodes \(N1(v)=\{v\in v1|(u1,v)\in E\}\) of vertex \(u1\) to the set \(Covered\), and add vertex \(u1\) to the solution set \(S\); S33. Delete the edge in the bipartite graph G That is, delete the incident edge between the vertex u1 and its adjacent node N1(v); S34. Delete all the incident edges of the vertices N2(u) that are the same as the candidate deployment points represented by the vertex u1 in the bipartite graph. Meanwhile, let S35. Find the vertex \(u2\) with the largest degree in the vertex subset \(V2\) currently, and add all adjacent nodes \(N3(v)=\{v\in V1|(u2,v)\in E\}\) of vertex \(u2\) to the set \(Covered\), and add vertex \(u2\) to the solution set \(S\); S36. Delete the edge in the bipartite graph G S37. Delete all the associated edges of the vertices N4(u) that are the same as the candidate deployment points represented by the vertex u2 in the bipartite graph. Meanwhile, let S38. Judge whether the condition \(Convered = V1\) is satisfied. If it is satisfied, go to step S39. Otherwise, go to step S32; S39. Construct a matrix Y1 of size |N|×|K|, whose elements indicate whether to deploy the k-th type of base station at the n-th candidate deployment point. Set the elements corresponding to the vertices in the set S to 1 and the values of the remaining elements to 0. Use the matrix Y1 as the initial solution for the deployment decision.

6. The method according to claim 1, wherein: In step S4, the calculation of the deployment cost of an individual is obtained by multiplying the deployment plan matrix \(Y\) by the transpose of the deployment cost matrix \(C\) and then calculating the trace of the result matrix. The formula is as follows: cost = Tr(Y·C T )。 7. The method according to claim 1, wherein: Step S6 also includes the following sub-steps: S61. Set the maximum population size | P max |, the distribution coefficient α, divide the population into two parts and calculate the population sizes of the two parts respectively. The formula is as follows: |P a | = α|P max | |P b | = (1 - α)|P max | S62. Mix the parent population and the offspring population to obtain a mixed population \(P0\), and divide the individuals in \(P0\) that can meet the requirement of covering all wind turbines and those that cannot meet the requirement into populations \(P1\) and \(P2\); S63. For the individuals in population P1, sort them in ascending order according to the deployment cost cost. If |P1| > |P a |, then select the first |P’1| = |P a | individuals according to the sorting result and add them to the new population. If not, add all the individuals in P1 to the new population, and the insufficient part is supplemented by the individuals in P2; S64. For the individuals in population P2, the coverage rate is defined as Then, according to the deployment cost index of each individual and the coverage rate index φ, the corresponding comprehensive index γ is obtained by weighted summation to calculate its comprehensive index γ. Deployment Cost Index The formula is as follows: The formula for the coverage index \(\varphi\) is as follows: The formula for the comprehensive index \(\gamma\) is as follows: where cost i and ρ i represent the deployment cost and coverage rate of the i-th individual respectively, represents the minimum deployment cost and maximum coverage rate of each individual in population P2, and λ is a non-negative constant less than 1; S65. Sort the individuals in population P2 in descending order according to the comprehensive index. If the condition |P1| > |P a | is satisfied, then select the top |P’2| = |P b | individuals according to the sorting result and add them to the new population; if the condition is not satisfied, then select the top |P’2| = |P max | - |P1| individuals according to the sorting result and add them to the new population.

8. A 5G base station deployment decision-making system for an offshore wind farm, characterized in that, It includes the following modules: A data model construction module, which is used to construct a set of wind turbines, a set of candidate deployment points, and a set of base stations, and generate a coverage relationship matrix, a deployment cost matrix, and a unit cost coverage rate matrix; A bipartite graph modeling module, which is used to model the deployment problem of base stations as a bipartite graph coverage problem; An initial solution generation module, which is used to find an initial solution that can meet the requirement of covering all wind turbines based on the greedy algorithm; A genetic algorithm optimization module, which is used to optimize the genetic algorithm; An output module, configured to find a coverage plan that meets the expected cost or reach the maximum number of iterations, and output a deployment matrix that covers all wind turbines with the lowest cost.

9. An electronic device, comprising a processor, a communication interface, a memory, and a communication bus. The processor, the communication interface, and the memory communicate with each other through the communication bus. The electronic device is characterized in that: A memory, configured to store a computer program; A processor, configured to execute the computer program stored on the memory to implement the steps of the method for making a decision on the deployment of 5G base stations in an offshore wind farm according to any one of claims 1-7.

10. A non-volatile storage medium, characterized in that: The non-volatile storage medium stores an executable program. When the executable program is executed by the processor, the steps of the method for making a decision on the deployment of 5G base stations in an offshore wind farm according to any one of claims 1-7 are implemented.

Images