A method and system for scheduling the operation and maintenance of offshore wind turbines that takes into account the use of working hours.

Abstract

This invention discloses a method and system for scheduling offshore wind turbine operations and maintenance (O&M) considering working hours, relating to the field of multi-O&M task scheduling technology for offshore wind farms. It designs an improved genetic algorithm based on the characteristics of offshore wind farm O&M operations and uses a fitness function to objectively and accurately evaluate the merits of scheduling schemes, fully leveraging the powerful advantages of the improved genetic algorithm to continuously optimize and intelligently select scheduling schemes. This invention initializes the population with a randomly generated specified number of individuals and calculates the fitness value of each individual in the population. A probability-driven roulette wheel method is used to select parent individuals, performing crossover and mutation operations with an elite retention strategy to ensure that high-quality individuals stably enter the next generation of the population. When the fitness value of the best individual in the population does not improve effectively within a preset number of generations, a catastrophe operation is initiated. By introducing new gene combinations and individuals, the diversity of the population is effectively improved, breaking the stalemate of being trapped in local optima.

Description

Technical Field

[0001] This invention relates to the field of multi-operation and maintenance task scheduling technology for offshore wind farms, specifically to an offshore wind turbine operation and maintenance scheduling method and system that takes into account the utilization of working hours. Background Technology

[0002] With the ever-increasing global demand for clean energy, offshore wind power, with its abundant resources and enormous development potential, is entering a new phase of rapid expansion. Large-scale offshore wind farms are springing up like mushrooms in my country's eastern coastal areas and the North Sea of ​​Europe. These farms often deploy a massive number of wind turbines, sometimes hundreds or even thousands, widely distributed across vast sea areas. However, the marine environment in which offshore wind turbines operate is extremely complex, changeable, and harsh. Frequent exposure to strong winds, giant waves, and high salt spray makes offshore wind turbines more prone to failure and their performance degrades faster than onshore turbines. This undoubtedly brings greater difficulty and higher costs to the operation and maintenance of offshore wind farms.

[0003] The operation and maintenance of offshore wind turbines not only heavily relies on specialized vessels and equipment but also must be carried out within limited suitable weather windows and is subject to strict constraints from various factors such as marine meteorological conditions and tides. Furthermore, offshore wind turbine maintenance requires specialized technicians, as well as specific maintenance tools and spare parts. However, in practice, the allocation and transportation of these resources face numerous challenges. For example, improper spare parts inventory management can lead to delays in maintenance work, thereby increasing operation and maintenance costs. Therefore, how to comprehensively consider various influencing factors and scientifically and rationally optimize the scheduling of offshore wind turbine operation and maintenance has become a pressing technical problem that needs to be solved in this field.

[0004] The scheduling of multiple operation and maintenance tasks in offshore wind farms currently has three main technical solutions:

[0005] (1) Scheduling methods based on experience rules; highly subjective and lacking in quantitative basis: This method relies entirely on the personal experience and subjective judgment of maintenance personnel, lacking scientific quantitative analysis and systematic planning. Due to differences in the experience and judgment standards of different maintenance personnel, this may lead to inconsistencies and irrationalities in scheduling results. Faced with situations where multiple wind turbines in a large-scale offshore wind farm fail simultaneously or have multiple complex combinations of failures, this method is difficult to quickly and accurately determine the optimal maintenance sequence and resource allocation scheme. This may lead to unreasonable resource allocation, with some wind turbines experiencing maintenance delays while other wind turbines excessively occupy resources, thereby increasing the overall downtime and maintenance costs of the wind turbines.

[0006] (2) Traditional mathematical programming model scheduling methods; Offshore wind turbine operation and maintenance involves numerous factors, including turbine operating status, marine environment, and operation and maintenance resources. Constructing an accurate mathematical programming model requires considering a large number of variables and complex constraints, making the model construction and solution process extremely complex and computationally intensive. Sometimes, there may even be no solution or the solution time may be too long, making it difficult to meet the timeliness requirements of actual operation and maintenance. Once the model is constructed, its structure and parameters are relatively fixed, making it difficult to adapt to the constantly changing conditions in offshore wind power systems. For example, new changes in turbine fault types, dynamic adjustments to operation and maintenance resources, or sudden anomalies in the marine environment all require rebuilding and solving the model, which increases the difficulty and cost of operation and maintenance management.

[0007] (3) Traditional Genetic Algorithm Scheduling Method: When dealing with the offshore wind turbine operation and maintenance scheduling problem, due to the limitations of chromosome encoding and genetic operations, ordinary genetic algorithms may require a large number of iterations to converge to a better solution. The selection, crossover, and mutation operations of ordinary genetic algorithms may lead to a continuous decrease in population diversity, causing the algorithm to converge to a local optimum prematurely and fail to find the globally optimal operation and maintenance scheduling scheme. When dealing with the offshore wind turbine operation and maintenance scheduling problem, ordinary genetic algorithms ignore more possibilities during the round trip of the maintenance vessel and lack the function of local search for the round trip of the maintenance vessel, which limits their effectiveness in practical applications. Summary of the Invention

[0008] The purpose of this invention is to propose a method and system for scheduling offshore wind turbine operation and maintenance that takes into account working hours, thereby improving the efficiency of wind farm operation and maintenance, increasing the power generation and working hours utilization efficiency of wind farms, and reducing operation and maintenance costs.

[0009] According to a first aspect of the present disclosure, a method for scheduling the operation and maintenance of offshore wind turbines that takes into account working hours is provided, comprising the following steps:

[0010] To obtain the information needed for wind farm operation and maintenance scheduling, including wind farm overview, operation and maintenance task details, and operation and maintenance resource data;

[0011] The information is preprocessed, including numbering and discretizing the maintenance vessel and the wind turbine, and variableizing the maintenance task type.

[0012] The chromosome of individual M is composed of the mission wind turbine and maintenance vessel number as its genetic content. s This constitutes population M;

[0013] The individuals in the population are decoded into their respective operation and maintenance scheduling schemes. The fitness values ​​of all individuals in the population are evaluated using the decoded sequences of each individual as objects. The lower the fitness value, the closer the individual is to the optimal solution.

[0014] The optimal individuals in the population are selected based on their fitness values, and then the hill-climbing algorithm is optimized using these individuals.

[0015] A probability-driven roulette wheel method is used to select parent individuals in the population;

[0016] For the selected parent individuals, those satisfying the crossover probability P j Individuals undergo crossover using an elite preservation strategy. After crossover, the fitness values ​​of parents and offspring are compared to select elite individuals. Individuals meeting the mutation probability P are then selected. b The individual undergoes mutation calculations using gene transposition methods;

[0017] When the best individual in the population is not optimized within the preset number of generations, a catastrophe optimization operation is initiated.

[0018] The optimal individual in the population is obtained, and the operation and maintenance scheduling scheme it represents is the final application scheme.

[0019] In one embodiment, the operation and maintenance task type is variableized, including characteristic variables related to the operation and maintenance vessel, operation and maintenance team, and task turbine. The characteristic variables related to the operation and maintenance vessel include: total voyage segment characteristic variable x. ijt The characteristic variables of the first mission segment s ijt Characteristic variable e of the mission wind turbine combination flight segment ijt Return segment characteristic variable b ijt State characteristic variables Service wind turbine state characteristic variables The characteristic variables related to the operation and maintenance team include: service turbine status characteristic variables. The characteristic variables related to the wind turbine in the task include: combined state characteristic variable r ijt The characteristic variable T of working hours consumption it .

[0020] In one embodiment, the operation and maintenance scheduling scheme is encoded using a non-repeating natural number encoding method, with the task turbine and maintenance vessel numbers forming the genetic content of the chromosome individual M. s The mission turbine and maintenance vessel numbers appear exactly once; the chromosome individual M s The representative operation and maintenance scheduling plan includes the service range of the wind turbines by the operation and maintenance vessels and the order of operation of the wind turbines. The wind turbines between two operation and maintenance vessel numbers are the service range of the operation and maintenance vessel with the first number in the order. The order of operation of the wind turbines is determined according to the coding order.

[0021] The individual decoding process within the group includes the selection of wind turbines for maintenance vessel services, the initial mission, the combination of wind turbines for the mission, and the return trip; let M be... s This is the s-th chromosome in the population; A collection of wind turbines required for the operation and maintenance of vessel k; M respectively s The set of wind turbines that the maintenance vessel k needs to serve on day t represents, the number of crews it carries, the set of wind turbines for the first mission, the optimal path for the first mission, the set of wind turbines that the crew is currently maintaining, the sequence of wind turbine combinations for the mission, the set of wind turbines for the return trip, and the optimal path for the return trip.

[0022] A branch-and-bound search method was embedded in both the initial mission and the return journey. This method is based on minimizing the flight cost. or The wind turbine element generates several initial task or return trip plan sets for the content. The above plan sets are used as search tree nodes, and the cumulative voyage distance of the maintenance ships of each plan is used as the upper bound value of each node. If the upper bound value of a node is equal to or greater than the upper bound value of the current optimal node during the calculation process, the child nodes of that node will not be generated.

[0023] Population initialization method: In one embodiment, a task combination strategy is introduced to comprehensively consider the work hour utilization efficiency of the maintenance team. The associated variable of the task combination strategy is the task wind turbine combination segment feature variable e. ijt and combined state characteristic variable r ijt The preset known quantity is the task combination threshold R. A , when e ijt =1 means that task fans i and j are combined. In the initial task set, fan number i is shifted one position to the right until the end of the initial task sequence and recorded in the maintenance schedule for day t. Otherwise, e ijt When =0, task fans i and j are not combined, and fan j is shifted one position to the last position.

[0024] In one embodiment, the fitness value of an individual is:

[0025]

[0026] Among them, D ij C represents the travel distance between mission turbines i and j, in kilometers; J This indicates the cost per unit distance traveled by maintenance vessels, expressed in yuan per kilometer; G i Indicates the fault status of task fan i; if it is a shutdown fault, then G i =1, otherwise G i =0; W represents the average output power of the wind turbine, in kilowatt-hours; C E This indicates the cost per unit of electricity, expressed in yuan per kilowatt-hour; L i This indicates the total time from the start of the maintenance cycle to the completion of maintenance on wind turbine i, in hours; C BThis indicates the daily rental cost for a single maintenance vessel, expressed in yuan / day; C R The unit represents the average daily labor cost of a single maintenance team, in yuan / team; T represents the total number of days in the maintenance cycle, in days; N represents the set of task turbine numbers N = {1, 2, ..., n}; K represents the set of maintenance vessel numbers K = {1, 2, ..., v}; M represents the set of maintenance team numbers M = {1, 2, ..., p}.

[0027] In one embodiment, a probability-driven roulette wheel method is used to select parent individuals in the population, as follows:

[0028] Obtain individual M in the population s The reciprocal of the fitness value

[0029] The ratio of the reciprocal of an individual's fitness value to the sum of the reciprocals of the fitness values ​​of all individuals in the population is taken as the value of individual M. s The probability of being selected P(M) s );

[0030] According to population order, use probability P(M) s Individual M is determined by accumulation. s The region R occupied between 0 and 1 s Generate a random number in the interval [0, 1], and select the individual corresponding to the region where the number is located as the parent individual.

[0031] In one embodiment, the elite retention strategy includes two stages: crossover operation and elite individual selection. In the crossover operation stage, a random region gene crossover method is used. Two non-repeating random numbers are generated within the interval [0, individual gene length] to determine the crossover interval, thereby dividing the parent individual's genes into genes before the crossover interval, genes within the crossover interval, and genes after the crossover interval. During the parent individual gene crossover process, genes before the crossover interval are directly inherited to the offspring individuals. Genes within the crossover interval and genes after the crossover interval are compared with the corresponding offspring's existing genes. If duplicate genes exist, they are carried forward; otherwise, they are directly inherited into the offspring individuals, thus obtaining complete offspring individuals.

[0032] After the crossover operation phase, the elite individual selection stage uses a method of comparing the fitness of all individuals in the crossover process. This involves comparing the fitness of all parents and offspring, and selecting the two best individuals to replace the offspring individuals in the next generation, thereby ensuring the inheritance of high-quality genes in the population.

[0033] In one embodiment, the mutation operation is performed for conditions satisfying the mutation probability P. b For the parent individuals, the gene transposition method is used to generate two random numbers as mutation sites in the interval [0, individual gene length], and their gene positions are swapped.

[0034] The catastrophic optimization operation involves selecting a certain proportion of non-repeating superior and inferior individuals from the population, while simultaneously randomly generating the same number of new individuals as the unselected individuals. These new individuals are then combined with the selected individuals to form a new population that replaces the original population for continued evolution and iteration.

[0035] According to a second aspect of the present disclosure, an offshore wind turbine operation and maintenance scheduling system that takes into account working hours is provided, comprising:

[0036] A method for scheduling the operation and maintenance of offshore wind turbines that takes into account working hours includes the following steps:

[0037] The information acquisition module acquires the information required for the wind farm's operation and maintenance scheduling, including an overview of the wind farm, details of operation and maintenance tasks, and operation and maintenance resource data.

[0038] The preprocessing module preprocesses the information, including numbering and discretizing the maintenance vessel and the task turbine, and variableizing the maintenance task type.

[0039] The population construction module uses the mission turbine and maintenance vessel numbers as the genetic content to form chromosome individuals M. s This constitutes population M;

[0040] The encoding and decoding module decodes the individuals in the population into their respective operation and maintenance scheduling schemes. It also evaluates the fitness values ​​of all individuals in the population based on the decoded sequences of each individual. The lower the fitness value, the closer the individual is to the optimal solution.

[0041] The optimization module selects the best individuals in the population based on their fitness values ​​and uses them as the target for hill climbing algorithm optimization.

[0042] The selection module uses a probability-driven roulette wheel method to select parent individuals in the population;

[0043] The elite retention strategy module, for selected parent individuals, selects individuals that satisfy the crossover probability P. j Individuals undergo crossover using an elite preservation strategy. After crossover, the fitness values ​​of parents and offspring are compared to select elite individuals. Individuals meeting the mutation probability P are then selected. b The individual undergoes mutation calculations using gene transposition methods;

[0044] The disaster module initiates a disaster optimization operation when the best individual in the population is not optimized within a preset number of generations.

[0045] The scheme acquisition module obtains the optimal individual in the population, and the operation and maintenance scheduling scheme it represents is the final application scheme.

[0046] According to a third aspect of the present disclosure, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and running on the memory, wherein the processor executes the program to implement the aforementioned method for scheduling offshore wind turbine operations and maintenance that takes into account working hours.

[0047] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the aforementioned method for scheduling offshore wind turbine operations and maintenance that takes into account working hours.

[0048] Compared with the prior art, the above technical solutions adopted in this invention have the following advantages: (1) This invention transforms the scheduling problem of multiple operation and maintenance tasks in offshore wind farms into the task selection problem of operation and maintenance teams in the state of executable tasks (idle), and uses an improved genetic algorithm that comprehensively considers the total operation and maintenance cost and the working hours utilization of the operation and maintenance teams to select and optimize the operation and maintenance scheduling scheme of offshore wind farms. Specifically, an improved genetic algorithm framework is constructed in combination with the actual operation of offshore wind farms to solve the scheduling problem of multiple operation and maintenance tasks in offshore wind farms. The total operation and maintenance cost calculation function is set as the fitness function, and the scheduling scheme is evaluated based on this. The improved genetic algorithm is used to complete the optimization selection of the scheduling scheme. The improved genetic algorithm encodes each scheduling scheme into population individuals and performs cross-genetic operations. After completing the corresponding number of evolutions, the population represents the non-dominated solution set of the current scheduling scheme. The scheduling scheme is selected according to the total operation and maintenance cost and the working hours utilization efficiency index of the operation and maintenance teams, so as to achieve the scheduling optimization goal of multiple operation and maintenance tasks in offshore wind farms.

[0049] (2) In order to quantify the order of operation and maintenance, the number of the operation and maintenance vessel and the task wind turbine is used as the content and a non-repeating natural number encoding method is adopted. Each number element appears only once in the scheduling scheme sequence, and the appearance of all elements represents the encoding sequence of the current individual scheduling scheme. To enhance the algorithm's local search capability, a branch-and-bound search method is embedded in the improved genetic algorithm decoding process. For the initial mission and return journey phases, the cumulative voyage distance of the maintenance ship is used as the upper bound value for each node, continuously updating the optimal node. To improve the quality of the population's genes, an elite retention strategy is implemented after the crossover genetic operation. Fitness values ​​of the parent and offspring individuals after the crossover operation are compared, and the two best individuals are selected to enter the population for subsequent operations. To improve the algorithm's convergence efficiency, a hill-climbing optimization algorithm is added. Optimization is performed on the best individuals in the population after each evolution, and selection is based on the fitness values ​​before and after optimization. To enhance population diversity, a catastrophe optimization operation is introduced. For populations where the best individuals within a preset number of generations have not undergone optimization, a certain proportion of individuals with non-repeating fitness values ​​are combined with randomly generated new individuals to form a new population for subsequent genetic operations.

[0050] (3) Based on the characteristics of offshore wind farm operation and maintenance, this invention designs targeted operation and maintenance scheduling feature variables, including: total voyage feature variables, initial mission voyage feature variables, mission turbine combination voyage feature variables, return voyage feature variables, status feature variables, operation and maintenance vessel service turbine status feature variables, operation and maintenance team service turbine status feature variables, combination status feature variables, and man-hour consumption feature variables. These feature variables quantify the operation and maintenance vessel route, man-hour consumption characteristics of each mission turbine, service turbine status, and turbine combination status characteristics; they contribute to the accurate solution of the algorithm and improve its practicality.

[0051] (4) In order to comprehensively consider the impact of the operation and maintenance team's working hours on the operation and maintenance cost, the present invention introduces a task turbine combination strategy to determine the daily task turbine combination part and form a daily operation and maintenance schedule plan. Attached Figure Description

[0052] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an undue limitation of this application.

[0053] Figure 1 Flowchart of offshore wind turbine operation and maintenance scheduling method considering working hours;

[0054] Figure 2 This is a schematic diagram of the encoding and decoding process in this invention;

[0055] Figure 3 This is a flowchart of the branch delimitation search method in this invention;

[0056] Figure 4 This is a flowchart of the improved genetic algorithm in this invention;

[0057] Figure 5 This is a distribution diagram of wind turbine units in the embodiment. Detailed Implementation

[0058] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0059] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0060] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0061] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and systems according to various embodiments of this disclosure. It should be noted that each block in a flowchart or block diagram may represent a module, segment, or portion of code, which may include one or more executable instructions for implementing the logical functions specified in the various embodiments. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutively represented blocks may actually be executed substantially in parallel, or they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, may be implemented using a dedicated hardware-based system that performs the specified functions or operations, or using a combination of dedicated hardware and computer instructions.

[0062] Offshore wind farms typically have a unified operation and maintenance (O&M) center, equipped with one or more O&M vessels and multiple O&M teams, all under the unified command of the O&M center. Given the specific locations of the wind turbines, the consumables required for the O&M work of each turbine, and the estimated man-hours, and considering constraints such as the maximum daily man-hours of the O&M teams, the maximum sailing distance of the vessels, and the maximum load capacity, the O&M scheduling plan should be optimized by comprehensively considering the sailing costs and leasing costs of the O&M vessels, the labor costs of the O&M teams, and the economic losses from turbine failures.

[0063] Example 1:

[0064] This embodiment provides a method for scheduling offshore wind turbine operations and maintenance that considers working hours. It employs an improved genetic algorithm to solve the offshore wind turbine scheduling problem, optimizing the selection of task-specific wind turbine operation and maintenance scheduling schemes. The process is shown below. Figure 1 As shown; the core idea is to design an improved genetic algorithm based on the characteristics of offshore wind farm operation and maintenance, and to use a fitness function to objectively and accurately evaluate the merits of scheduling schemes, fully leveraging the powerful advantages of the improved genetic algorithm to continuously optimize and intelligently select scheduling schemes. This includes the following steps:

[0065] S1. Obtain the information required for the wind farm operation and maintenance scheduling, including wind farm overview, operation and maintenance task details, and operation and maintenance resource data;

[0066] In this embodiment, there is one operation and maintenance center, numbered 0, with a fixed location; the operation and maintenance vessels are of the same type and return to the operation and maintenance center after completing the transportation task; the operation and maintenance work begins as soon as the operation and maintenance vessel arrives at the wind turbine location; only one operation and maintenance team is allowed to work at the same time for each wind turbine; one operation and maintenance team can operate and maintain multiple wind turbines per day, with limits on the maximum number of wind turbines that the operation and maintenance team can operate and maintain per day and the maximum number of working hours per day.

[0067] S2. Preprocess the information, including numbering and discretizing the maintenance vessel and the task wind turbine, and variableizing the maintenance task type;

[0068] In this embodiment, the operation and maintenance task type is variableized, including characteristic variables related to the operation and maintenance vessel, operation and maintenance team, and task turbine. The characteristic variables related to the operation and maintenance vessel include: total voyage segment characteristic variable x. ijt The characteristic variables of the first mission segment s ijt Characteristic variable e of the mission wind turbine combination flight segment ijt Return segment characteristic variable b ijt State characteristic variables Service wind turbine state characteristic variables The characteristic variables related to the operation and maintenance team include: service turbine status characteristic variables. The characteristic variables related to the wind turbine in the task include: combined state characteristic variable r ijt The characteristic variable T of working hours consumption it .

[0069] The specific descriptions of the above variables are as follows:

[0070] Total flight segment characteristic variable x ijt This characterizes the number of voyages the maintenance vessel undertakes from wind turbine i to wind turbine j on day t, and x ijt =s ijt +e ijt +b ijt .

[0071] First mission segment feature variables s ijt This characterizes the segment type in which the maintenance vessel travels from wind turbine i to wind turbine j during the initial mission phase on day t, and the existence of time s. ijt =1, otherwise s ijt =0.

[0072] Mission wind turbine combined flight segment characteristic variable e ijt This characterizes the segment type of the maintenance vessel's journey from wind turbine i to wind turbine j during the wind turbine assembly phase on day t, where e exists. ijt =1, otherwise e ijt =0.

[0073] Return segment characteristic variable b ijt This characterizes the segment type of the maintenance vessel's journey from wind turbine i to wind turbine j during the return trip on day t, where b exists. ijt =1, otherwise b ijt =0.

[0074] State characteristic variables This represents the charter status of the maintenance vessel k on day t; if it is chartered, then... on the contrary

[0075] Service wind turbine state characteristic variables This characterizes the service status of maintenance vessel k on day t for the task wind turbine i, and the planned service... on the contrary

[0076] Service wind turbine state characteristic variables This characterizes the service status of maintenance team m on day t for task wind turbine i, and the planned service... on the contrary

[0077] Combined state characteristic variable r ijt The combined state of wind turbines i and j on day t is represented by the characteristic variable e of the combined wind turbine flight segment. ijt They are related. The correlation formula is:

[0078] r ijt =(T M -T it )÷T jM

[0079] M + e ijt ≥r ijt -R A

[0080] R A e ijt ≤r ijt

[0081] Among them, T i The required working time for completing the maintenance of the wind turbine i task; T M This represents the maximum daily working hours for the maintenance team; T jM For T i Divide by T M The remainder; M + R is a sufficiently large constant; A The set threshold parameters for the task fan combination.

[0082] Working hours consumption characteristic variable T itThe value represents the length of maintenance time consumed by wind turbine i on day t, and its value ranges from 0 to T. M .

[0083] S3. Chromosomes M are composed of mission turbine and maintenance vessel numbers as genetic content. s This constitutes population M;

[0084] In this embodiment, a non-repeating natural number encoding method is used to encode the operation and maintenance scheduling scheme, with the task wind turbine and operation and maintenance vessel numbers as the genetic content to form the chromosome individual M. s The mission turbine and maintenance vessel numbers appear exactly once; the chromosome individual M s The representative operation and maintenance scheduling plan includes the service range of the wind turbines by the operation and maintenance vessels and the order of operation of the wind turbines. The wind turbines between two operation and maintenance vessel numbers are the service range of the operation and maintenance vessel with the first number in the order. The order of operation of the wind turbines is determined according to the coding order.

[0085] The individual decoding process within the group includes the selection of wind turbines for the maintenance vessel, the initial mission, the combination of wind turbines for the mission, and the return trip. The encoding and decoding process is illustrated as follows: Figure 2 As shown; let M s This is the s-th chromosome in the population; A collection of wind turbines required for the operation and maintenance of vessel k; M respectively s The set of wind turbines that the maintenance vessel k needs to serve on day t represents, the number of crews it carries, the set of wind turbines for the first mission, the optimal path for the first mission, the set of wind turbines that the crew is currently maintaining, the sequence of wind turbine combinations for the mission, the set of wind turbines for the return trip, and the optimal path for the return trip.

[0086] To enhance the algorithm's local search capability, a branch-and-bound search method is embedded in the initial task and return journey phases of the decoding process. This method is based on minimizing the travel cost. or The wind turbine element generates several initial or return trip plans for the content. These plans are used as nodes in a search tree, with the cumulative voyage distance of the maintenance vessels for each plan serving as the upper bound for each node. If, during the calculation, the upper bound of a node is equal to or greater than the upper bound of the current optimal node, then no more child nodes are generated for that node (pruning). Figure 3 As shown.

[0087] Population initialization method: Let the population size be A. Merge the set of maintenance vessel numbers K and the set of task wind turbine numbers N into a set S. Randomly select j∈S and generate a sequence of non-repeating elements containing the number of elements in S according to the selection order. This sequence constitutes the encoded sequence (individual) M. s s∈{1,2,…,A}; repeating this operation generates A encoded sequences (individuals), forming a population M.

[0088] S4. Decode the individuals in the population into their respective operation and maintenance scheduling schemes, and evaluate the fitness values ​​of all individuals in the population using the decoded sequences of each individual as objects. The lower the fitness value, the closer the individual is to the optimal solution.

[0089] In this embodiment, a task combination strategy is introduced to comprehensively consider the working hour utilization efficiency of the maintenance team. The associated variable of the task combination strategy is the task wind turbine combination segment characteristic variable e. ijt and combined state characteristic variable r ijt The preset known quantity is the task combination threshold R. A . When e ijt =1 represents task fan i, which is combined with the first task set. The fan numbered i in the initial task set is shifted one position to the end of the initial task sequence and recorded in the maintenance schedule for day t. Conversely, e represents the fan numbered i in the initial task set. ijt When =0, task fans i and j are not combined, and fan j is shifted one position to the last position.

[0090] The fitness value of an individual is:

[0091]

[0092] Among them, D ij C represents the travel distance between mission turbines i and j, in kilometers; J This indicates the cost per unit distance traveled by maintenance vessels, expressed in yuan per kilometer; G i Indicates the fault status of task fan i; if it is a shutdown fault, then G i =1, otherwise G i =0; W represents the average output power of the wind turbine, in kilowatt-hours; C E This indicates the cost per unit of electricity, expressed in yuan per kilowatt-hour; L i This indicates the total time from the start of the maintenance cycle to the completion of maintenance on wind turbine i, in hours; C B This indicates the daily rental cost for a single maintenance vessel, expressed in yuan / day; C R The unit represents the average daily labor cost of a single maintenance team, in yuan / team; T represents the total number of days in the maintenance cycle, in days; N represents the set of task turbine numbers N = {1, 2, ..., n}; K represents the set of maintenance vessel numbers K = {1, 2, ..., v}; M represents the set of maintenance team numbers M = {1, 2, ..., p}.

[0093] S5. Select the best individual in the population based on the individual fitness value, and use it as the object to optimize the hill climbing algorithm;

[0094] In this embodiment, the hill-climbing optimization algorithm uses the best individual in the population after each evolutionary iteration as the optimization target. It employs a gene translocation method to change the gene composition of the target individual to generate a new individual. If the new individual is superior to the original individual, it replaces the original individual as the best individual in the population; otherwise, the original individual is retained. The optimization algorithm stops after reaching a preset number of operations.

[0095] S6. A probability-driven roulette wheel method is used to select parent individuals in the population;

[0096] S61. Obtain individual M from the population. s The reciprocal of the fitness value

[0097] S62. The ratio of the reciprocal of an individual's fitness value to the sum of the reciprocals of the fitness values ​​of all individuals in the population is taken as the value of individual M. s The probability of being selected P(M) s );

[0098] S63. According to the population order, use the probability P(M) s Individual M is determined by accumulation. s The region R occupied between 0 and 1 s Generate a random number in the interval [0, 1], and select the individual corresponding to the region where the number is located as the parent individual.

[0099] Where the selection probability P(M) s ) and individual roulette area R s The calculation method is as follows:

[0100]

[0101]

[0102] S7. For the selected parent individuals, for those satisfying the crossover probability P j Individuals undergo crossover using an elite preservation strategy. After crossover, the fitness values ​​of parents and offspring are compared to select elite individuals. Individuals meeting the mutation probability P are then selected. b The individual undergoes mutation calculations using gene transposition methods;

[0103] In this embodiment, the elite retention strategy includes two stages: crossover operation and elite individual selection. In the crossover operation stage, a randomized regional gene crossover method is used, with a crossover probability P. jThe recommended value range is 0.6–0.95. Two unique random numbers within the interval [0, individual gene length] are generated to determine the crossover interval, thus dividing the parent individual's genes into genes before, after, and at the end of the crossover interval. During the parent individual's gene crossover process, genes before the crossover interval are directly inherited by the offspring. Genes at the end of the crossover interval are compared with the corresponding existing genes in the offspring. If duplicate genes exist, they are carried forward; otherwise, they are directly inherited into the offspring, resulting in a complete offspring individual.

[0104] After the crossover operation phase, the elite individual selection stage adopts a method of comparing the fitness of all individuals in the crossover process. The fitness of all parents and offspring is compared, and the two best individuals are selected to replace the offspring individuals and be passed on to the next generation. This ensures the inheritance of high-quality genes in the population and improves the convergence efficiency of the algorithm.

[0105] Mutation operation, for conditions satisfying the mutation probability P b (P b For parent individuals with suggested values ​​of 0.01 to 0.18, the gene transposition method is used to generate two random numbers as mutation sites within the interval [0, individual gene length], and their gene positions are swapped.

[0106] S8. When the best individual in the population is not optimized within the preset number of generations, the catastrophe optimization operation is initiated;

[0107] To enhance population diversity, a catastrophe optimization operation is introduced. This operation selects a certain proportion of non-repeating superior and inferior individuals from the population, while simultaneously generating the same number of new individuals as the unselected individuals. These new individuals are then combined with the selected individuals to form a new population, replacing the original population for continued evolutionary iteration. The trigger condition is that the fitness value of the best individual in the population has not been optimized within a preset number of generations.

[0108] S9. Obtain the optimal individual in the population, and the operation and maintenance scheduling scheme it represents is the final application scheme.

[0109] In the scheduling algorithm application phase, the improved genetic algorithm uses previously collected and preprocessed information on operation and maintenance tasks, resources, and wind farms (such as task types and number of maintenance vessels) as input data. Based on this improved genetic algorithm, the operation and maintenance scheduling scheme is updated, achieving an optimized scheme that minimizes operation and maintenance costs while also considering labor utilization. Figure 4 As shown.

[0110] The experimental scheme of this invention was designed based on data from an offshore wind farm in Jiangsu Province. There are 50 wind turbines, each with a capacity of 1MW. The operation and maintenance center is located 14km west of wind turbine X-01. The spacing between wind turbine rows is 0.6km, and the spacing between turbine columns is 1.4km. The turbine distribution is as follows: Figure 5 As shown. Wind turbines are categorized based on whether they are shut down due to malfunction, and a combined maintenance and condition-based inspection approach is adopted. Maintenance teams and materials depart from the maintenance center, which is equipped with one maintenance vessel and six maintenance teams. The maximum daily working hours per team are 8 hours. The maximum vessel load capacity is 35 tons, the sailing cost is 180 yuan / km, and the vessel rental fee is 30,000 yuan / day. The working hour requirements for the seven types of routine inspections and wind turbine malfunction maintenance tasks are detailed in Table 1. The offshore wind power grid connection price is referenced, with an electricity price of 0.85 yuan / kW·h. There are 15 wind turbines on the task; detailed information on each turbine is shown in Table 2. The experimental population size is 60, the maximum number of evolutions is 1000, and the crossover probability P0 is... j =0.9, mutation probability P b =0.1 and the task wind turbine combination threshold R A =0.94.

[0111] Table 1. Maintenance Resources Required for Wind Turbine Units

[0112]

[0113]

[0114] Table 2 Detailed Information on the Task Fan

[0115] Task fan number Fault type Repair time XG-01 shutdown 13h XG-05 Non-stop 19.3h XG-11 shutdown 23h XG-13 Non-stop 10.5h XG-19 shutdown 11.3h XG-29 Non-stop 13.8h XG-31 shutdown 7.5h XG-37 Non-stop 17.5h XG-38 shutdown 19.3h XG-39 shutdown 17.3h XG-41 Non-stop 7.3h XG-45 shutdown 20h XG-46 Non-stop 18h XG-48 shutdown 15.5h XG-50 Non-stop 14h

[0116] Experiments show that the multi-task scheduling method for offshore wind farms proposed in this invention can effectively solve the corresponding scheduling schemes. Table 3 shows the time utilization efficiency and total operation and maintenance cost of this invention and the reference method, where the reference method is a classic genetic algorithm. It can be seen that compared with the reference method, the total operation and maintenance cost is reduced by more than 19%, the time utilization rate is increased by more than 11%, the operation and maintenance cost of wind farms is reduced, and the utilization efficiency of operation and maintenance teams is enhanced.

[0117] Table 3 Comparison of experimental results

[0118]

[0119]

[0120] Example 2:

[0121] This embodiment provides an offshore wind turbine operation and maintenance scheduling system that takes into account working hours, including:

[0122] The information acquisition module acquires the information required for the wind farm's operation and maintenance scheduling, including an overview of the wind farm, details of operation and maintenance tasks, and operation and maintenance resource data.

[0123] The preprocessing module preprocesses the information, including numbering and discretizing the maintenance vessel and the task turbine, and variableizing the maintenance task type.

[0124] The population construction module uses the mission turbine and maintenance vessel numbers as the genetic content to form chromosome individuals M. s This constitutes population M;

[0125] The encoding and decoding module decodes the individuals in the population into their respective operation and maintenance scheduling schemes. It also evaluates the fitness values ​​of all individuals in the population based on the decoded sequences of each individual. The lower the fitness value, the closer the individual is to the optimal solution.

[0126] The optimization module selects the best individuals in the population based on their fitness values ​​and uses them as the target for hill climbing algorithm optimization.

[0127] The selection module uses a probability-driven roulette wheel method to select parent individuals in the population;

[0128] The elite retention strategy module, for selected parent individuals, selects individuals that satisfy the crossover probability P. j Individuals undergo crossover using an elite preservation strategy. After crossover, the fitness values ​​of parents and offspring are compared to select elite individuals. Individuals meeting the mutation probability P are then selected. b The individual undergoes mutation calculations using gene transposition methods;

[0129] The disaster module initiates a disaster optimization operation when the best individual in the population is not optimized within a preset number of generations.

[0130] The scheme acquisition module obtains the optimal individual in the population, and the operation and maintenance scheduling scheme it represents is the final application scheme.

[0131] Example 3:

[0132] An electronic device includes a memory, a processor, and a computer program stored in the memory and running thereon, wherein the processor, when executing the program, implements the aforementioned method for scheduling offshore wind turbine operations and maintenance considering working hours, comprising:

[0133] To obtain the information needed for wind farm operation and maintenance scheduling, including wind farm overview, operation and maintenance task details, and operation and maintenance resource data;

[0134] The information is preprocessed, including numbering and discretizing the maintenance vessel and the wind turbine, and variableizing the maintenance task type.

[0135] The chromosome of individual M is composed of the mission wind turbine and maintenance vessel number as its genetic content. s This constitutes population M;

[0136] The individuals in the population are decoded into their respective operation and maintenance scheduling schemes. The fitness values ​​of all individuals in the population are evaluated using the decoded sequences of each individual as objects. The lower the fitness value, the closer the individual is to the optimal solution.

[0137] The optimal individuals in the population are selected based on their fitness values, and then the hill-climbing algorithm is optimized using these individuals.

[0138] A probability-driven roulette wheel method is used to select parent individuals in the population;

[0139] For the selected parent individuals, those satisfying the crossover probability P j Individuals undergo crossover using an elite preservation strategy. After crossover, the fitness values ​​of parents and offspring are compared to select elite individuals. Individuals meeting the mutation probability P are then selected. b The individual undergoes mutation calculations using gene transposition methods;

[0140] When the best individual in the population is not optimized within the preset number of generations, a catastrophe optimization operation is initiated.

[0141] The optimal individual in the population is obtained, and the operation and maintenance scheduling scheme it represents is the final application scheme.

[0142] Example 4:

[0143] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned method for scheduling offshore wind turbine operations and maintenance considering working hours, comprising:

[0144] To obtain the information needed for wind farm operation and maintenance scheduling, including wind farm overview, operation and maintenance task details, and operation and maintenance resource data;

[0145] The information is preprocessed, including numbering and discretizing the maintenance vessel and the wind turbine, and variableizing the maintenance task type.

[0146] The chromosome of individual M is composed of the mission wind turbine and maintenance vessel number as its genetic content. s This constitutes population M;

[0147] The individuals in the population are decoded into their respective operation and maintenance scheduling schemes. The fitness values ​​of all individuals in the population are evaluated using the decoded sequences of each individual as objects. The lower the fitness value, the closer the individual is to the optimal solution.

[0148] The optimal individuals in the population are selected based on their fitness values, and then the hill-climbing algorithm is optimized using these individuals.

[0149] A probability-driven roulette wheel method is used to select parent individuals in the population;

[0150] For the selected parent individuals, those satisfying the crossover probability P j Individuals undergo crossover using an elite preservation strategy. After crossover, the fitness values ​​of parents and offspring are compared to select elite individuals. Individuals meeting the mutation probability P are then selected. b The individual undergoes mutation calculations using gene transposition methods;

[0151] When the best individual in the population is not optimized within the preset number of generations, a catastrophe optimization operation is initiated.

[0152] The optimal individual in the population is obtained, and the operation and maintenance scheduling scheme it represents is the final application scheme.

[0153] Those skilled in the art will understand that the modules or steps described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, which can then be stored in a storage device for execution by a computer device. Alternatively, they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. This disclosure is not limited to any particular combination of hardware and software.

[0154] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0155] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this disclosure are still within the scope of protection of this disclosure.

Claims

1. A method for scheduling the operation and maintenance of offshore wind turbines that considers working hours, characterized in that, Includes the following steps: To obtain the information needed for wind farm operation and maintenance scheduling, including wind farm overview, operation and maintenance task details, and operation and maintenance resource data; The information is preprocessed, including numbering and discretizing the maintenance vessel and the task wind turbine, and variableizing the maintenance task type. Chromosomes are composed of mission wind turbine and maintenance vessel numbers as genetic content. , constitute a population ; The individuals in the population are decoded into their respective operation and maintenance scheduling schemes. The fitness values ​​of all individuals in the population are evaluated using the decoded sequences of each individual as objects. The lower the fitness value, the closer the individual is to the optimal solution. The optimal individuals in the population are selected based on their fitness values, and then the hill-climbing algorithm is optimized using these individuals. A probability-driven roulette wheel method is used to select parent individuals in the population; For the selected parent individuals, those that satisfy the crossover probability Individuals undergo crossover using an elite preservation strategy. After crossover, the fitness values ​​of parents and offspring are compared to select elite individuals. Individuals meeting the mutation probability are then selected. The individual undergoes mutation calculations using gene transposition methods; When the best individual in the population is not optimized within the preset number of generations, a catastrophe optimization operation is initiated. The optimal individual in the population is obtained, and the operation and maintenance scheduling scheme it represents is the final application scheme. The operation and maintenance scheduling scheme is encoded using a non-repeating natural number encoding method, with the task turbine and maintenance vessel numbers forming the chromosome individual as the genetic content. The mission turbine and maintenance vessel numbers appear exactly once; the chromosome individuals The representative operation and maintenance scheduling plan includes the service range of the wind turbines by the operation and maintenance vessels and the order of operation of the wind turbines. The wind turbines between two operation and maintenance vessel numbers are the service range of the operation and maintenance vessel with the first number in the order. The order of operation of the wind turbines is determined according to the coding order. The decoding process for individuals within a population includes the selection of wind turbines for maintenance vessel services, the initial mission, the combination of wind turbines for the mission, and the return journey; [The remaining text appears to be incomplete and requires further context.] The first in the population s One chromosome; For maintenance ships k Required service fan collection; , , , , , , , They are respectively The number represented t Tianyun Vessel k The set of wind turbines to be serviced, the number of work teams to be transported, the set of wind turbines for the first mission, the optimal path for the first mission, the set of wind turbines currently being maintained by the work teams, the sequence of wind turbine combinations for the mission, the set of wind turbines for the return trip, and the optimal return path. A branch-and-bound search method was embedded in both the initial mission and the return journey. This method is based on minimizing the flight cost. or The wind turbine element generates several initial task or return trip plan sets for the content. The above plan sets are used as search tree nodes, and the cumulative voyage distance of the maintenance ships of each plan is used as the upper bound value of each node. If the upper bound value of a node is equal to or greater than the upper bound value of the current optimal node during the calculation process, the child nodes of that node will no longer be generated. Population initialization method: Let the population size be... A Set the maintenance vessel numbers and task fan number set Merge into a set S Random selection Generate according to the selection order, containing S The sequence of unique elements constitutes the encoded sequence. Repeat this operation to generate A A population of encoded sequences. M .

2. The offshore wind turbine operation and maintenance scheduling method considering working hours as described in claim 1, characterized in that, The operation and maintenance task types are variably defined, including characteristic variables related to the operation and maintenance vessel, operation and maintenance team, and wind turbine. The characteristic variables related to the operation and maintenance vessel include: total voyage segment characteristic variables. First mission segment characteristic variables Characteristic variables of the mission wind turbine combination flight segment Return segment characteristic variables State characteristic variables Service wind turbine state characteristic variables The characteristic variables related to the operation and maintenance team include: service turbine status characteristic variables. The characteristic variables related to the task wind turbine include: combined state characteristic variables. Working hours consumption characteristic variables .

3. The offshore wind turbine operation and maintenance scheduling method considering working hours as described in claim 2, characterized in that, The introduction of a task combination strategy comprehensively considers the work hour utilization efficiency of the maintenance team. The related variable of the task combination strategy is the characteristic variable of the wind turbine combination flight segment. and combined state characteristic variables The preset known quantity is the task combination threshold. ,when It is a mission fan i and j Combine them, the initial task set i The number of the fan is shifted one position to the last position of the initial mission, and recorded as the number of the fan. t In the daily maintenance schedule, on the contrary... At that time, the mission wind turbine i and j Without combination j The numbered fans are moved one position forward to the last position.

4. The offshore wind turbine operation and maintenance scheduling method considering working hours as described in claim 2, characterized in that, The fitness value of an individual is: in, Indicates the task fan i and j The distance between them, in kilometers; This indicates the cost per unit distance traveled by maintenance vessels, expressed in yuan per kilometer. Indicates the task fan i The fault status, if it is a shutdown fault, then ,on the contrary ; This indicates the average output power of the wind turbine, expressed in kilowatt-hours. This indicates the cost per unit of electricity, expressed in yuan per kilowatt-hour. Indicates the start of the operation and maintenance cycle to the wind turbine. Total time for operation and maintenance to be completed, in hours; This indicates the daily rental cost for a single maintenance vessel, expressed in yuan / day. This represents the average daily labor cost of a single maintenance team, expressed in yuan per team. This indicates the total number of days in the maintenance cycle, in days. Represents the set of task fan numbers ; Represents the set of maintenance vessel numbers ; Represents the set of maintenance team numbers .

5. The offshore wind turbine operation and maintenance scheduling method considering working hours as described in claim 1, characterized in that, The probability-driven roulette wheel method is used to select parent individuals in the population, as follows: Obtain individuals from the population The reciprocal of the fitness value ; The ratio of the reciprocal of an individual's fitness value to the sum of the reciprocals of the fitness values ​​of all individuals in the population is taken as the individual's fitness value. Probability of being selected ; According to population order, use probability Individuals are determined by accumulation. The area occupied between 0 and 1 Generate a random number within the interval [0,1], and select the individual corresponding to the region where the number is located as the parent individual.

6. The offshore wind turbine operation and maintenance scheduling method considering working hours as described in claim 1, characterized in that, The elite retention strategy comprises two stages: crossover operation and elite individual selection. In the crossover operation stage, a random region gene crossover method is adopted. Two non-repeating random numbers within the interval [0, individual gene length] are generated to confirm the crossover interval, thereby dividing the parent individual's genes into genes before the crossover interval, genes within the crossover interval, and genes after the crossover interval. During the parent individual gene crossover process, genes before the crossover interval are directly inherited to the offspring individuals. Genes within the crossover interval and genes after the crossover interval are compared with the corresponding offspring's existing genes. If duplicate genes exist, they are carried forward; otherwise, they are directly inherited into the offspring individuals, thus obtaining complete offspring individuals. After the crossover operation phase, the elite individual selection stage uses a method of comparing the fitness of all individuals in the crossover process. This involves comparing the fitness of all parents and offspring, and selecting the two best individuals to replace the offspring individuals in the next generation, thereby ensuring the inheritance of high-quality genes in the population.

7. An offshore wind turbine operation and maintenance scheduling system that considers working hours, for implementing the method described in any one of claims 1-6, characterized in that, include: The information acquisition module acquires the information required for the wind farm's operation and maintenance scheduling, including an overview of the wind farm, details of operation and maintenance tasks, and operation and maintenance resource data. The preprocessing module preprocesses the information, including numbering and discretizing the maintenance vessel and the task turbine, and variableizing the maintenance task type. The population construction module uses the mission turbine and maintenance vessel numbers as the genetic content to form chromosome individuals. , constitute a population ; The encoding and decoding module decodes the individuals in the population into their respective operation and maintenance scheduling schemes. It also evaluates the fitness values ​​of all individuals in the population based on the decoded sequences of each individual. The lower the fitness value, the closer the individual is to the optimal solution. The optimization module selects the best individuals in the population based on their fitness values ​​and uses them as the target for hill climbing algorithm optimization. The selection module uses a probability-driven roulette wheel method to select parent individuals in the population; The elite retention strategy module, for selected parent individuals, selects individuals that satisfy the crossover probability. Individuals undergo crossover using an elite preservation strategy. After crossover, the fitness values ​​of parents and offspring are compared to select elite individuals. Individuals meeting the mutation probability are then selected. The individual undergoes mutation calculations using gene transposition methods; The disaster module initiates a disaster optimization operation when the best individual in the population is not optimized within a preset number of generations. The scheme acquisition module obtains the optimal individual in the population, and the operation and maintenance scheduling scheme represented by it is the final application scheme.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running thereon, characterized in that, When the processor executes the program, it implements the offshore wind turbine operation and maintenance scheduling method that takes into account working hours as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the offshore wind turbine operation and maintenance scheduling method that takes into account working hours utilization, as described in any one of claims 1-6.

Images