Multi-target genetic algorithm-based layout optimization method for marine launch support equipment

Abstract

The invention discloses a multi-target genetic algorithm-based layout optimization method for maritime launching support equipment, which comprises the following steps of: S1, acquiring multi-source data of a maritime launching platform, support equipment and a launching task, and processing the acquired data into a four-dimensional associated data set in real time; s2, constructing a layout optimization model of the marine launch support equipment; s3, a multi-target genetic algorithm is adopted to carry out iterative optimization on the layout optimization model of the marine launch support equipment, and an optimized parameter combination is obtained; s4, applying the optimized parameter combination to the layout optimization model of the maritime launching support equipment, and outputting a Pareto optimal solution set; s5, selecting an optimal layout scheme from the Pareto solution set based on the weight, and outputting a coordinate table, a list and a tradeoff map; s6, monitoring four-dimensional parameters in real time, triggering dynamic optimization when the four-dimensional parameters exceed a threshold, quickly restarting optimization based on a historical solution set, and outputting a dynamic adjustment scheme; by constructing a multi-objective optimization model including total cost minimization, task response time minimization and safety redundancy maximization and combining a Pareto optimal solution set and an analytic hierarchy process, multi-objective tradeoff and decision under complex constraints are realized, and the overall guarantee capability of the marine launch task is remarkably improved.

Description

Technical Field

[0001] This invention relates to the interdisciplinary fields of maritime space launch support, intelligent optimization algorithms, and equipment layout planning, specifically a method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm. Background Technology

[0002] With the routine advancement of maritime space launch missions, the rationality of the layout of support equipment directly determines the safety, efficiency, and economy of launch missions. Maritime launch scenarios are characterized by limited platform space, a wide variety of equipment types (including high-risk fuel equipment, precision measurement and control equipment, emergency rescue equipment, etc.), high coupling of operational processes, and dynamic environmental interference (sea turbulence, weather changes), which impose stringent requirements on equipment layout.

[0003] Current mainstream methods for deploying maritime launch support equipment rely heavily on engineering experience or single-objective optimization models, which have significant drawbacks: First, traditional experience-based deployments do not quantitatively consider multi-objective constraints, easily leading to contradictory scenarios such as low cost despite response delays or excessive safety redundancy despite insufficient efficiency. Second, existing single-objective optimization methods (such as those that only aim to minimize response time) ignore cost controllability and operational safety, making them difficult to adapt to the complex constraints of maritime launches. Third, some optimization algorithms (such as single genetic algorithms and particle swarm optimization algorithms) do not have adaptation mechanisms designed for the multi-constraint characteristics of maritime deployments, resulting in problems such as convergence getting stuck in local optima and insufficient diversity of solution sets.

[0004] To address this, a method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide a method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm, comprising the following steps;

[0007] S1: Collects multi-source data from the sea-based launch platform, support equipment, and launch mission. The collected data is then processed in real-time using moving average filtering to eliminate short-term noise from the dynamic environment. The criteria remove outliers from equipment attribute data and linearly normalize data with different dimensions, then use Bayesian networks to generate a four-dimensional association dataset of "platform-equipment-mission-environment". ;

[0008] S2: Construct an optimization model for the layout of maritime launch support equipment, and associate the aforementioned four-dimensional dataset of "platform-equipment-mission-environment". As input, the layout optimization model for maritime launch support equipment includes a decision variable vector definition module, a constraint set setting module, and a multi-objective optimization function construction module;

[0009] S3: A multi-objective genetic algorithm is used to iteratively optimize the layout model of maritime launch support equipment. Each individual in the population is encoded as a set of parameters including equipment quantity ratio, spatial coordinate threshold, and constraint weight coefficient to obtain the optimized parameter combination.

[0010] S4: Apply the optimized parameter combination to the layout optimization model of maritime launch support equipment, and use the optimized model to analyze the four-dimensional correlation dataset of "platform-equipment-mission-environment". Perform the solution processing and output the Pareto optimal solution set. ;

[0011] S5: Combining the target weights calculated by the analytic hierarchy process, select the candidate equipment layout scheme with the best overall performance from the Pareto optimal solution set, and simultaneously output the equipment coordinate configuration table, quantity list and four-dimensional weighted visualization map.

[0012] S6: Real-time monitoring of four dimensions of platform spatial attitude, dynamic environmental parameters, equipment operation and deployment status, and task flow and priority parameters; when the change of any dimension parameter exceeds its preset threshold, the population is immediately initialized based on the historical Pareto optimal solution set, and the current real-time data is injected to perform iterative optimization, thereby outputting a dynamic adjustment scheme for equipment layout.

[0013] Preferably, S1 specifically includes:

[0014] S11. Collect platform functional zone boundary parameters through a sensor network deployed in key areas of the offshore launch platform. Carrying capacity distribution matrix and dangerous area location set ;

[0015] S12. Collect dynamic environmental parameters, including wind speed. Wave height and temperature Noise removal is achieved through moving average filtering, as shown in the following formula:

[0016]

[0017] S13. Geometric Dimensions of Data Acquisition Support Equipment ,weight ,cost and based on The criteria include outlier detection and removal;

[0018] S14. Acquire Launch Mission Operation Parameters This includes task priority, operation time window, and equipment coordination relationships;

[0019] S15. Perform linear normalization on multi-source heterogeneous data, using the following formula:

[0020]

[0021] S16. Construct a four-dimensional association model based on Bayesian networks to generate a four-dimensional association dataset of "platform-equipment-mission-environment". .

[0022] Preferably, S2 specifically includes:

[0023] S21. Set up a four-dimensional association dataset of "platform-equipment-mission-environment". For the input matrix ,in, For the set of real numbers, Indicates the number of samples. This represents the feature dimension of each sample;

[0024] S22. Construct a decision variable vector definition module, adopting a multi-dimensional coding structure, including equipment quantity ratio vector. Spatial coordinate threshold vector and constraint weight coefficient vector ,in Indicates the quantity of equipment types. Indicates the number of equipment layout positions. The number of constraints is represented by the decision variable vector, which is defined as follows: ;

[0025] S23. Construct a constraint set setting module, defining a set of hard constraints based on platform space constraints, equipment physical constraints, mission safety constraints, and environmental dynamic constraints. and soft constraint set The hard constraints include the platform boundary insurmountability condition and the minimum safe distance condition between equipment, while the soft constraints include the cost ceiling condition and the mission response time condition, and are further constrained by the constraint relaxation factor. Dynamically adjust the tolerance for violations of soft constraints;

[0026] S24. Construct a multi-objective optimization function construction module and define a multi-objective function vector. ,in This represents the objective of minimizing total cost. This indicates the objective of minimizing task response time. This represents the objective of maximizing safety redundancy, and introduces an objective weight vector. The relative importance of each objective is dynamically adjusted based on the analytic hierarchy process (AHP).

[0027] S25. An adaptive constraint coordination unit is introduced between the decision variable vector definition module and the constraint condition set setting module to monitor constraint conflicts in real time and dynamically adjust the constraint weight coefficient vector according to changes in platform spatial attitude. Ensure the feasibility of constraints in dynamic environments;

[0028] S26. Integrate the Pareto dominance evaluation mechanism into the multi-objective optimization function construction module. Generate the initial Pareto solution set by calculating the non-dominated order and crowding distance of the solution set, and provide a diversity preservation strategy for the genetic algorithm.

[0029] S27. Define the set of structural parameters for the layout optimization model of maritime launch support equipment, including the dimensions of decision variables. Number of constraints Number of objective functions Constraint relaxation factor and target weight vector ;

[0030] S28. Complete the structural initialization of the layout optimization model of the maritime launch support equipment, and load the current parameter configuration to perform model verification and pre-optimization.

[0031] Preferably, S3 specifically includes:

[0032] S31. Initialize the population size for the multi-objective genetic algorithm. Set the maximum number of iterations. Crossover probability With the probability of mutation ;

[0033] S32. Each genetic individual is represented using a structure-parameter joint encoding method, encoding the genetic individual as... The specific formula is as follows:

[0034]

[0035] in, Represents a vector of decision variables , This represents the corresponding combination of algorithm parameters, including crossover probabilities. Probability of mutation Selecting stress factors and constraint relaxation factor ;

[0036] S33. The decision variable vector corresponding to each genetic individual. Input the layout optimization model of maritime launch support equipment and calculate the multi-objective function vector. And assess the degree of constraint violation. ;

[0037] S34. Define the fitness evaluation function for genetic individuals. Combining multi-objective performance indicators Degree of violation of constraints With solution set distribution index The specific formula is as follows:

[0038]

[0039] in, For non-dominated ranking levels based on Pareto dominance, , , All are weighted coefficients;

[0040] S35. Calculate the solution set distribution entropy of the current generation population. The Pareto front is used to measure the uniformity of its distribution. The specific formula is as follows:

[0041]

[0042] in: For the first The non-dominated solution set of the generation, To solve Crowding distance between adjacent solutions;

[0043] S36. An adaptive crossover mutation adjustment mechanism based on distribution entropy is adopted, according to... Dynamically adjust crossover probability With the probability of mutation The specific formula is as follows:

[0044]

[0045]

[0046] in, , As the baseline probability, , For adjustment coefficients, This is a reference value for the distribution entropy.

[0047] S37. Perform a tournament selection operation based on crowding order on the population, and give priority to individuals with high non-dominance level and large crowding distance to enter the mating pool.

[0048] S38. Perform adaptive arithmetic crossover and Gaussian mutation operations on individuals in the mating pool to generate the offspring population, and then adjust the offspring population according to the constraint relaxation factor. Dynamically adjust the magnitude of variation;

[0049] S39. Merge the parent and offspring populations, perform fast non-dominated sorting and crowding calculation, and select the top... Individuals constitute a new generation of population;

[0050] S310. Determine whether the iteration termination conditions are met, including reaching the maximum number of iterations or the Pareto front convergence index being lower than the set threshold.

[0051] S311. If the termination condition is not met, return to step S33.

[0052] S312. If the termination condition is met, output the set of optimal parameter combinations. ,in: The optimized decision variable vector, This represents the corresponding optimal combination of algorithm parameters.

[0053] Preferably, S4 specifically includes:

[0054] S41. The optimal parameter combination set obtained in claim 4. Applied to the layout optimization model of maritime launch support equipment, among which The optimized decision variable vector, This represents the corresponding optimal combination of algorithm parameters;

[0055] S42, Based on the optimized decision variable vector Deconstructing the equipment quantity ratio vector Spatial coordinate threshold vector and constraint weight coefficient vector Generate the basic configuration parameters for the equipment layout;

[0056] S43. Link the four-dimensional dataset of "platform-equipment-mission-environment". Input a configured layout optimization model of maritime launch support equipment, perform multi-objective optimization, and output a Pareto optimal solution set containing multiple non-dominated solutions. Each solution This represents a feasible equipment layout scheme.

[0057] Preferably, S5 specifically includes:

[0058] S51. Construct a target weight decision matrix using the analytic hierarchy process (AHP), and determine the relative importance weights of each objective function based on expert evaluation. ,in: Corresponding to the total cost weight, Corresponding response time weights, Corresponding safety redundancy weight;

[0059] S52, Based on the target weight vector Calculate the Pareto optimal solution set Each solution Comprehensive evaluation value The specific formula is as follows:

[0060]

[0061] in, , and All are normalized objective function values;

[0062] S53. Select the solution with the highest comprehensive evaluation value. The specific formula for the candidate equipment layout scheme with the best overall performance is as follows:

[0063]

[0064] S54, Based on the optimal layout scheme The system generates an equipment coordinate configuration table, an equipment quantity list, and a four-dimensional trade-off visualization map. The equipment coordinate configuration table specifies the exact coordinates of each piece of equipment in the platform space, the equipment quantity list details the optimal configuration quantity of each type of equipment, and the four-dimensional trade-off visualization map intuitively displays the trade-off relationship between different objective functions.

[0065] Preferably, S6 specifically includes:

[0066] S61. Real-time monitoring of four-dimensional parameters: platform spatial attitude, dynamic environmental parameters, equipment operation and deployment status, and task flow and priority; and establishment of a parameter change assessment mechanism.

[0067] S62. Set dynamic thresholds for parameters in each dimension. When the change in any dimension parameter exceeds its preset threshold, the dynamic optimization process is triggered.

[0068] S63, Based on Historical Pareto Optimal Solution Sets Quickly initialize the genetic algorithm population and inject current real-time data to perform iterative optimization;

[0069] S64. Based on the dynamic optimization results, output a dynamic adjustment scheme for equipment layout to ensure that the system response time meets the real-time requirements under dynamic conditions.

[0070] Compared with the prior art, the beneficial effects of the present invention are:

[0071] 1. This invention constructs a multi-objective optimization model that includes minimizing total cost, minimizing mission response time, and maximizing safety redundancy. Combining Pareto optimal solution sets and the analytic hierarchy process (AHP), it achieves multi-objective trade-offs and decision-making under complex constraints. Compared to traditional single-objective optimization or empirical layout methods, it can systematically balance economy, efficiency, and safety, avoiding the layout defects of "choosing one thing at the expense of another," and significantly improving the overall support capability of maritime launch missions.

[0072] 2. By introducing a real-time monitoring mechanism and a dynamic optimization process, this invention can quickly trigger layout re-optimization when platform attitude, environmental parameters, and mission status change. Based on the historical Pareto solution set for initializing the population, the optimization convergence time is significantly shortened, ensuring that feasible equipment layout adjustment schemes can still be quickly generated under unforeseen circumstances such as severe sea conditions or mission changes, thus guaranteeing the efficient and safe execution of launch missions.

[0073] 3. This invention employs advanced mechanisms such as structure-parameter joint encoding, adaptive crossover and mutation adjustment based on distribution entropy, and crowding-based sorting selection to effectively enhance the global search capability and solution set uniformity of the multi-objective genetic algorithm, avoiding premature convergence and local optima problems. Furthermore, by using constraint relaxation factors and adaptive coordination units, the feasibility and robustness of the algorithm under complex constraints are further improved.

[0074] 4. This invention constructs an integrated dataset encompassing "platform-equipment-mission-environment" through multi-source data acquisition, Bayesian network fusion, and four-dimensional correlation modeling, providing high-quality input for the optimization model. The final output not only includes the specific coordinates and quantity list of the optimal layout scheme but also provides a four-dimensional trade-off visualization map, intuitively displaying the trade-off relationships between various objectives and assisting decision-makers in making comprehensive judgments and selecting solutions. Attached Figure Description

[0075] Figure 1 This is a flowchart of the layout optimization method for maritime launch support equipment based on a multi-objective genetic algorithm proposed in this invention. Detailed Implementation

[0076] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0077] Please see Figure 1 A method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm includes the following steps;

[0078] S1: Collects multi-source data from the sea-based launch platform, support equipment, and launch mission. The collected data is then processed in real-time using moving average filtering to eliminate short-term noise from the dynamic environment. The criteria remove outliers from equipment attribute data and linearly normalize data with different dimensions, then use Bayesian networks to generate a four-dimensional association dataset of "platform-equipment-mission-environment". ;

[0079] S2: Construct an optimization model for the layout of maritime launch support equipment, and associate the four-dimensional dataset of "platform-equipment-mission-environment". As input, the layout optimization model for maritime launch support equipment includes a decision variable vector definition module, a constraint set setting module, and a multi-objective optimization function construction module.

[0080] S3: A multi-objective genetic algorithm is used to iteratively optimize the layout model of maritime launch support equipment. Each individual in the population is encoded as a set of parameters including equipment quantity ratio, spatial coordinate threshold, and constraint weight coefficient to obtain the optimized parameter combination.

[0081] S4: Apply the optimized parameter combination to the layout optimization model of maritime launch support equipment, and use the optimized model to analyze the four-dimensional correlation dataset of "platform-equipment-mission-environment". Perform the solution processing and output the Pareto optimal solution set. ;

[0082] S5: Combining the target weights calculated by the analytic hierarchy process, select the candidate equipment layout scheme with the best overall performance from the Pareto optimal solution set, and simultaneously output the equipment coordinate configuration table, quantity list and four-dimensional weighted visualization map.

[0083] S6: Real-time monitoring of four dimensions of platform spatial attitude, dynamic environmental parameters, equipment operation and deployment status, and task flow and priority parameters; when the change of any dimension parameter exceeds its preset threshold, the population is immediately initialized based on the historical Pareto optimal solution set, and the current real-time data is injected to perform iterative optimization, thereby outputting a dynamic adjustment scheme for equipment layout.

[0084] In this embodiment, S1 specifically includes:

[0085] S11. Collect platform functional zone boundary parameters through a sensor network deployed in key areas of the offshore launch platform. Carrying capacity distribution matrix and dangerous area location set ;

[0086] S12. Collect dynamic environmental parameters, including wind speed. Wave height and temperature To eliminate short-term fluctuation noise, a moving average filter is used for noise removal. The specific formula is as follows:

[0087]

[0088] in, This indicates the size of the sliding window, which is dynamically set according to the sensor sampling frequency. The wave meter used for sampling, took To effectively smooth out random fluctuations within a 10-second period, Indicates the first At this current moment, Indicating a historical moment The collected raw data values, The coefficient representing the average value. This represents the index variable of the summation loop.

[0089] S13. Collect the static attributes of the data collection support equipment, including geometric dimensions. ,weight ,cost To ensure data quality, based on The criteria include outlier detection and removal;

[0090] S14. Acquire Launch Mission Operation Parameters This includes task priority, operation time window, and equipment coordination relationships;

[0091] S15. Perform linear normalization on multi-source heterogeneous data, using the following formula:

[0092]

[0093] in, This represents the normalized data value. Represents the original data value. This represents the minimum value in the original data sequence. This represents the maximum value in the original data sequence.

[0094] S16. Construct a four-dimensional association model based on Bayesian networks, integrate all preprocessed data, and generate a unified, intrinsically linked four-dimensional association dataset of "platform-equipment-mission-environment". This serves as the input for subsequent optimization models.

[0095] In this embodiment, S2 specifically includes:

[0096] S21. Set up a four-dimensional association dataset of "platform-equipment-mission-environment". For the input matrix ,in, For the set of real numbers, Indicates the number of samples. This represents the feature dimension of each sample;

[0097] S22. Construct a decision variable vector definition module, adopting a multi-dimensional coding structure, including equipment quantity ratio vector. Spatial coordinate threshold vector and constraint weight coefficient vector ,in Indicates the quantity of equipment types. Indicates the number of equipment layout positions. The decision variable vector represents the number of constraints. ;

[0098] S23. Construct a constraint set setting module, defining a set of hard constraints based on platform space constraints, equipment physical constraints, mission safety constraints, and environmental dynamic constraints. and soft constraint set Among them, hard constraints include the condition that the platform boundary cannot be crossed (the equipment projection does not exceed the limit). Defined physical boundaries) and minimum safe distance conditions between equipment (Euclidean distance between the center points of any two pieces of equipment) Soft constraints include cost ceiling conditions and task response time conditions, and are constrained by slack factors. Dynamically adjust the tolerance for violations of soft constraints;

[0099] S24. Construct a multi-objective optimization function construction module and define a multi-objective function vector. ,in This represents the objective of minimizing total cost. This indicates the objective of minimizing task response time. This represents the objective of maximizing safety redundancy, and introduces an objective weight vector. The relative importance of each objective is dynamically adjusted based on the analytic hierarchy process (AHP).

[0100] S25. An adaptive constraint coordination unit is introduced between the decision variable vector definition module and the constraint condition set setting module to monitor constraint conflicts in real time and dynamically adjust the constraint weight coefficient vector according to changes in platform spatial attitude. Ensure the feasibility of constraints in dynamic environments;

[0101] S26. Integrate the Pareto dominance evaluation mechanism into the multi-objective optimization function construction module. Generate the initial Pareto solution set by calculating the non-dominated order and crowding distance of the solution set, and provide a diversity preservation strategy for the genetic algorithm.

[0102] S27. Define the set of structural parameters for the layout optimization model of maritime launch support equipment, including the dimensions of decision variables. Number of constraints Number of objective functions Constraint relaxation factor and target weight vector ;

[0103] S28. Complete the structural initialization of the layout optimization model of the maritime launch support equipment, and load the current parameter configuration to perform model verification and pre-optimization.

[0104] In this embodiment, S3 specifically includes:

[0105] S31. Initialize the population size for the multi-objective genetic algorithm. Set the maximum number of iterations. Crossover probability With the probability of mutation ;

[0106] S32. Each genetic individual is represented using a structure-parameter joint encoding method, encoding the genetic individual as... The specific formula is as follows:

[0107]

[0108] in, Represents a vector of decision variables , This represents the corresponding combination of algorithm parameters, including crossover probabilities. Probability of mutation Selecting stress factors and constraint relaxation factor ;

[0109] S33. The decision variable vector corresponding to each genetic individual. Input the layout optimization model of maritime launch support equipment and calculate the multi-objective function vector. And assess the degree of constraint violation. ;

[0110] S34. Define the fitness evaluation function for genetic individuals. Combining multi-objective performance indicators Degree of violation of constraints With solution set distribution index The specific formula is as follows:

[0111]

[0112] in, For non-dominated ranking levels based on Pareto dominance, , , All are weighted coefficients;

[0113] S35. Calculate the solution set distribution entropy of the current generation population. The Pareto front is used to measure the uniformity of its distribution. The specific formula is as follows:

[0114]

[0115] in: For the first The non-dominated solution set of the generation, To solve Crowding distance between adjacent solutions;

[0116] S36. An adaptive crossover mutation adjustment mechanism based on distribution entropy is adopted, according to... Dynamically adjust crossover probability With the probability of mutation The specific formula is as follows:

[0117]

[0118]

[0119] in, , As the baseline probability, , For adjustment coefficients, This is a reference value for the distribution entropy.

[0120] S37. Perform a tournament selection operation based on crowding order on the population, and give priority to individuals with high non-dominance level and large crowding distance to enter the mating pool.

[0121] S38. Perform adaptive arithmetic crossover and Gaussian mutation operations on the individuals in the mating pool to generate the offspring population, and then adjust the offspring population according to the constraint relaxation factor. Dynamically adjust the magnitude of variation;

[0122] S39. Merge the parent and offspring populations, perform fast non-dominated sorting and crowding calculation, and select the top... Individuals constitute a new generation of population;

[0123] S310. Determine whether the iteration termination conditions are met, including reaching the maximum number of iterations or the Pareto front convergence index being lower than the set threshold.

[0124] S311. If the termination condition is not met, return to step S33.

[0125] S312. If the termination condition is met, output the set of optimal parameter combinations. ,in: The optimized decision variable vector, This represents the corresponding optimal combination of algorithm parameters.

[0126] In this embodiment, S4 specifically includes:

[0127] S41. The optimal parameter combination set obtained in claim 4. Applied to the layout optimization model of maritime launch support equipment, among which The optimized decision variable vector, This represents the corresponding optimal combination of algorithm parameters;

[0128] S42, Based on the optimized decision variable vector Deconstructing the equipment quantity ratio vector Spatial coordinate threshold vector and constraint weight coefficient vector Generate the basic configuration parameters for the equipment layout;

[0129] S43. Link the four-dimensional dataset of "platform-equipment-mission-environment". Input a configured layout optimization model of maritime launch support equipment, perform multi-objective optimization, and output a Pareto optimal solution set containing multiple non-dominated solutions. Each solution This represents a feasible equipment layout scheme.

[0130] In this embodiment, S5 specifically includes:

[0131] S51. Construct a target weight decision matrix using the analytic hierarchy process (AHP), and determine the relative importance weights of each objective function based on expert evaluation. ,in: Corresponding to the total cost weight, Corresponding response time weights, Corresponding safety redundancy weight;

[0132] S52, Based on the target weight vector Calculate the Pareto optimal solution set Each solution Comprehensive evaluation value The specific formula is as follows:

[0133]

[0134] in, , and All are normalized objective function values;

[0135] S53. Select the solution with the highest comprehensive evaluation value. The specific formula for the candidate equipment layout scheme with the best overall performance is as follows:

[0136]

[0137] S54, Based on the optimal layout scheme Generate an equipment coordinate configuration table, an equipment quantity list, and a four-dimensional trade-off visualization map. The equipment coordinate configuration table specifies the exact coordinates of each piece of equipment in the platform space, the equipment quantity list details the optimal configuration quantity of each type of equipment, and the four-dimensional trade-off visualization map intuitively displays the trade-off relationship between different objective functions.

[0138] In this embodiment, S6 specifically includes:

[0139] S61. Real-time monitoring of four-dimensional parameters: platform spatial attitude, dynamic environmental parameters, equipment operation and deployment status, and task flow and priority; and establishment of a parameter change assessment mechanism.

[0140] S62. Set dynamic thresholds for parameters in each dimension. When the change in any dimension parameter exceeds its preset threshold, the dynamic optimization process is triggered.

[0141] S63, Based on Historical Pareto Optimal Solution Sets Quickly initialize the genetic algorithm population and inject current real-time data to perform iterative optimization;

[0142] S64. Based on the dynamic optimization results, output a dynamic adjustment scheme for equipment layout to ensure that the system response time meets the real-time requirements under dynamic conditions.

[0143] Example 1:

[0144] The method of this invention was applied in a certain sea launch mission of the "Zhuque" satellite. The launch platform was a semi-submersible platform, requiring the deployment of a total of 8 types of support equipment, including fuel loading vehicles, telemetry and control antenna vehicles, fire support vehicles, and hoisting equipment.

[0145] The system collected the platform deck partition map ( ), load-bearing capacity of each area ( Equipment dimensions () Equipment cost ) and task flow ( Data such as [data missing] were preprocessed and fused using a Bayesian network to construct an optimization model. The objective was set as the total cost ([data missing]). ), average task response time ( ) and safety clearance redundancy ( Three-objective optimization;

[0146] A multi-objective genetic algorithm (population size 100, 500 iterations) was used for optimization, ultimately obtaining a Pareto front containing 22 non-dominated solutions. The weight vector was set using the analytic hierarchy process (AHP). The optimal solution was selected through calculation. The plan places the telemetry and control antenna vehicle at a high, unobstructed point on the platform, positions the fuel refueling vehicle downwind and away from other high-risk equipment, and provides a precise list of coordinates for all equipment.

[0147] Three hours before launch, the monitoring system detected the height of the waves. Exceeding the preset threshold (wave height suddenly increases from 1.5 meters to 3.0 meters, with a threshold of 2.5 meters) could affect the stability of the lifting operation. The system immediately triggered dynamic optimization, quickly generating a new layout adjustment scheme based on historical solutions within 95 seconds. This moved the lifting equipment 5 meters closer to the platform's center of gravity and fine-tuned the positions of surrounding equipment, ensuring the safe execution of the mission under adverse sea conditions.

[0148] This invention constructs a multi-objective optimization model that includes minimizing total cost, minimizing mission response time, and maximizing safety redundancy. Combining Pareto optimal solutions and the analytic hierarchy process (AHP), it achieves multi-objective trade-offs and decision-making under complex constraints. Compared to traditional single-objective optimization or empirical layout methods, it can systematically balance economy, efficiency, and safety, avoiding the layout defects of "choosing one thing at the expense of another," and significantly improving the overall support capability of maritime launch missions.

[0149] This invention, by introducing a real-time monitoring mechanism and a dynamic optimization process, can quickly trigger layout re-optimization when platform attitude, environmental parameters, and mission status change. By initializing the population based on historical Pareto solutions, the optimization convergence time is significantly shortened, ensuring that feasible equipment layout adjustment schemes can be quickly generated even in adverse sea conditions or under unforeseen circumstances such as mission changes, thus guaranteeing the efficient and safe execution of launch missions.

[0150] This invention employs advanced mechanisms such as structure-parameter joint encoding, adaptive crossover and mutation adjustment based on distribution entropy, and crowding-based sorting selection to effectively enhance the global search capability and solution set uniformity of the multi-objective genetic algorithm, avoiding premature convergence and local optima problems. Furthermore, by using constraint relaxation factors and adaptive coordination units, the feasibility and robustness of the algorithm under complex constraints are further improved.

[0151] This invention constructs an integrated dataset encompassing "platform-equipment-mission-environment" through multi-source data acquisition, Bayesian network fusion, and four-dimensional correlation modeling, providing high-quality input for the optimization model. The final output includes not only the specific coordinates and quantity list of the optimal layout scheme but also a four-dimensional trade-off visualization map, intuitively displaying the trade-off relationships between various objectives and assisting decision-makers in making comprehensive judgments and selecting solutions.

[0152] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing the layout of marine launch support equipment based on a multi-objective genetic algorithm, characterized in that: Includes the following steps; S1: Collects multi-source data from the sea-based launch platform, support equipment, and launch mission. The collected data is then sequentially subjected to moving average filtering to eliminate short-term noise in dynamic environmental data. The criteria remove outliers from equipment attribute data, process heterogeneous data through linear normalization, and finally generate a four-dimensional association dataset based on Bayesian network fusion. ; S2: Construct an optimization model for the layout of maritime launch support equipment, and integrate the aforementioned four-dimensional relational dataset. As input, the layout optimization model for maritime launch support equipment includes a decision variable vector definition module, a constraint set setting module, and a multi-objective optimization function construction module; S3: A multi-objective genetic algorithm is used to iteratively optimize the layout model of maritime launch support equipment. Each individual in the population is encoded as a set of parameters including equipment quantity ratio, spatial coordinate threshold, and constraint weight coefficient to obtain the optimized parameter combination. S4: Apply the optimized parameter combination to the layout optimization model of maritime launch support equipment, and use the optimized model to analyze the four-dimensional correlation dataset. Perform the solution processing and output the Pareto optimal solution set. ; S5: Combining the target weights calculated by the analytic hierarchy process, select the candidate equipment layout scheme with the best overall performance from the Pareto optimal solution set, and simultaneously output the equipment coordinate configuration table, quantity list and four-dimensional weighted visualization map. S6: Real-time monitoring of four dimensions of platform spatial attitude, dynamic environmental parameters, equipment operation and deployment status, and task flow and priority parameters; when the change of any dimension parameter exceeds its preset threshold, it immediately initializes the population based on the historical Pareto optimal solution set, injects the current real-time data to perform iterative optimization, and outputs a dynamic adjustment scheme for equipment layout.

2. The method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm according to claim 1, characterized in that: S1 specifically includes: S11. Collect platform functional zone boundary parameters through a sensor network deployed in key areas of the offshore launch platform. Carrying capacity distribution matrix and dangerous area location set ; S12. Collect dynamic environmental parameters, including wind speed. Wave height and temperature Noise removal is achieved through moving average filtering, as shown in the following formula: ； S13. Geometric Dimensions of Data Acquisition Support Equipment ,weight ,cost and based on The criteria include outlier detection and removal; S14. Acquire Launch Mission Operation Parameters This includes task priority, operation time window, and equipment coordination relationships; S15. Perform linear normalization on multi-source heterogeneous data, using the following formula: ； S16. Construct a four-dimensional association model based on Bayesian networks to generate a four-dimensional association dataset of "platform-equipment-mission-environment". .

3. The method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm according to claim 1, characterized in that: S2 specifically includes: S21. Setting up a four-dimensional relational dataset For the input matrix ,in, For the set of real numbers, Indicates the number of samples. This represents the feature dimension of each sample; S22. Construct a decision variable vector definition module, adopting a multi-dimensional coding structure, including equipment quantity ratio vector. Spatial coordinate threshold vector and constraint weight coefficient vector ,in Indicates the quantity of equipment types. Indicates the number of equipment layout positions. The number of constraints is represented by the decision variable vector, which is defined as follows: ; S23. Construct a constraint set setting module, defining a set of hard constraints based on platform space constraints, equipment physical constraints, mission safety constraints, and environmental dynamic constraints. and soft constraint set The hard constraints include the platform boundary insurmountability condition and the minimum safe distance condition between equipment, while the soft constraints include the cost ceiling condition and the mission response time condition, and are further constrained by the constraint relaxation factor. Dynamically adjust the tolerance for violations of soft constraints; S24. Construct a multi-objective optimization function construction module and define a multi-objective function vector. ,in: This represents the objective of minimizing total cost. This represents the objective of minimizing task response time. This represents the objective of maximizing safety redundancy, and introduces an objective weight vector. ; S25. An adaptive constraint coordination unit is introduced between the decision variable vector definition module and the constraint condition set setting module to monitor constraint conflicts in real time and dynamically adjust the constraint weight coefficient vector according to changes in platform spatial attitude. ; S26. Integrate the Pareto dominance evaluation mechanism into the multi-objective optimization function construction module. Generate the initial Pareto solution set by calculating the non-dominated order and crowding distance of the solution set, and provide a diversity preservation strategy for the genetic algorithm. S27. Define the set of structural parameters for the layout optimization model of maritime launch support equipment, including the dimensions of decision variables. Number of constraints Number of objective functions Constraint relaxation factor and target weight vector ; S28. Complete the structural initialization of the layout optimization model of the maritime launch support equipment, and load the current parameter configuration to perform model verification and pre-optimization.

4. The method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm according to claim 1, characterized in that: S3 specifically includes: S31. Initialize the population size for the multi-objective genetic algorithm. Set the maximum number of iterations. Crossover probability With the probability of mutation ; S32. Each genetic individual is represented using a structure-parameter joint encoding method, encoding the genetic individual as... The specific formula is as follows: in, Represents a vector of decision variables , This represents the corresponding combination of algorithm parameters, including crossover probabilities. Probability of mutation Selecting stress factors and constraint relaxation factor ; S33. The decision variable vector corresponding to each genetic individual. Input the layout optimization model of maritime launch support equipment and calculate the multi-objective function vector. And assess the degree of constraint violation. ; S34. Define the fitness evaluation function for genetic individuals. Combining multi-objective performance indicators Degree of violation of constraints With solution set distribution index The specific formula is as follows: in, For non-dominated ranking levels based on Pareto dominance, , , All are weighted coefficients; S35. Calculate the solution set distribution entropy of the current generation population. The Pareto front is used to measure the uniformity of its distribution. The specific formula is as follows: in: For the first The non-dominated solution set of the generation, To solve Crowding distance between adjacent solutions; S36. An adaptive crossover mutation adjustment mechanism based on distribution entropy is adopted, according to... Dynamically adjust crossover probability With the probability of mutation The specific formula is as follows: in, , As the baseline probability, , For adjustment coefficients, This is a reference value for the distribution entropy. S37. Perform a tournament selection operation based on crowding order on the population, and give priority to individuals with high non-dominance level and large crowding distance to enter the mating pool. S38. Perform adaptive arithmetic crossover and Gaussian mutation operations on individuals in the mating pool to generate the offspring population, and then adjust the offspring population according to the constraint relaxation factor. Dynamically adjust the magnitude of variation; S39. Merge the parent and offspring populations, perform fast non-dominated sorting and crowding calculation, and select the top... Individuals constitute a new generation of population; S310. Determine whether the iteration termination conditions are met, including reaching the maximum number of iterations or the Pareto front convergence index being lower than the set threshold. S311. If the termination condition is not met, return to step S33. S312. If the termination condition is met, output the set of optimal parameter combinations. ,in: The optimized decision variable vector, This represents the corresponding optimal combination of algorithm parameters.

5. The method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm according to claim 1, characterized in that: S4 specifically includes: S41. The optimal parameter combination set obtained in claim 4. Applied to the layout optimization model of maritime launch support equipment, among which The optimized decision variable vector, This represents the corresponding optimal combination of algorithm parameters; S42, Based on the optimized decision variable vector Deconstructing the equipment quantity ratio vector Spatial coordinate threshold vector and constraint weight coefficient vector Generate the basic configuration parameters for the equipment layout; S43. Connect the four-dimensional dataset Input a configured layout optimization model of maritime launch support equipment, perform multi-objective optimization, and output a Pareto optimal solution set containing multiple non-dominated solutions. Each solution This represents a feasible equipment layout scheme.

6. The method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm according to claim 1, characterized in that: S5 specifically includes: S51. Construct a target weight decision matrix using the analytic hierarchy process (AHP), and determine the relative importance weights of each objective function based on expert evaluation. ,in: Corresponding to the total cost weight, Corresponding response time weight, Corresponding safety redundancy weight; S52, Based on the target weight vector Calculate the Pareto optimal solution set Each solution Comprehensive evaluation value The specific formula is as follows: in, , and All are normalized objective function values; S53. Select the solution with the highest comprehensive evaluation value. The specific formula for the candidate equipment layout scheme with the best overall performance is as follows: S54, Based on the optimal layout scheme Generate an equipment coordinate configuration table, an equipment quantity list, and a four-dimensional trade-off visualization map. The equipment coordinate configuration table specifies the exact location coordinates of each piece of equipment in the platform space, and the equipment quantity list details the optimal configuration quantity of each type of equipment.

7. The method for optimizing the layout of maritime launch support equipment based on a multi-objective genetic algorithm according to claim 1, characterized in that: S6 specifically includes: S61. Real-time monitoring of four-dimensional parameters: platform spatial attitude, dynamic environmental parameters, equipment operation and deployment status, and task flow and priority; and establishment of a parameter change assessment mechanism. S62. Set dynamic thresholds for parameters in each dimension. When the change in any dimension parameter exceeds its preset threshold, the dynamic optimization process is triggered. S63, Based on Historical Pareto Optimal Solution Sets Quickly initialize the genetic algorithm population and inject current real-time data to perform iterative optimization; S64. Based on the dynamic optimization results, output a dynamic adjustment scheme for equipment layout.

Images