Ship structure parameter optimization method based on goodness sorting

Abstract

The invention belongs to the field of ship structure design, and particularly relates to a ship structure parameter optimization method based on goodness ranking, which comprises the following steps: setting an optimization objective function, a constraint calculation formula function and a constraint range of ship structure parameter optimization; the user inputs the variable upper bound of the target structure; searching a feasible lower bound for each variable on the basis of the upper bound by using a dichotomy method; randomly arranging a chromosome solution in a feasible lower bound and upper bound range to serve as an initial population; performing constraint calculation on the chromosomes to obtain qualified chromosomes and retain part of unqualified chromosomes; carrying out constraint multiplexing on unqualified chromosomes in a regression optimization mode, and trying to enable the unqualified chromosomes to meet constraints; carrying out parameter perturbation on the chromosomes, then carrying out constraint calculation, and reserving part of unqualified chromosomes; carrying out goodness sorting on all chromosomes, and selecting excellent individuals ranked in the top and a part of unqualified individuals; new individuals are generated through parameter exchange; and gradually narrowing the fine tuning constraint range, carrying out periodic optimization of chromosomes until the goodness change is smaller than a threshold value, obtaining a parameter optimal solution as input of a ship parameterization model, and realizing optimization of a design target structure.

Description

Technical Field

[0001] The invention relates to the field of ship structure design, and in particular to a ship structure parameter optimization method based on excellence ranking. Background Art

[0002] Ship structure design is the core link in the shipbuilding process. When designing ship structures, designers first need to have a deep understanding of the functional needs and safety requirements of the ship to ensure that the designed ship can not only meet the basic functions such as navigation and transportation, but also ensure safety performance under various complex sea conditions. During the design process, designers will calculate multiple sets of parameter groups that can meet all structural strength specification constraints based on multi-dimensional ship structure strength calculations, including strength verification, modulus calculations, buckling analysis and other calculation steps. Then, calculate the ship structure weight corresponding to each set of parameter groups and find the lowest weight among them, aiming to ensure that the ship meets the strength and stability under various working conditions while minimizing material consumption.

[0003] In the parameter design of structures such as the overall profile of a ship, by fine-tuning parameters through optimization methods, more optimized use of materials can be achieved while ensuring that the design requirements of the specifications are met. For example, for a ship with a deadweight of 10,000 tons, optimizing the total weight by 1% can reduce 100 tons of materials, thereby significantly reducing production costs. Traditional ship design usually relies on manual setting of parameters and fine-tuning based on past design experience. The number of iterations is small, and it is difficult to achieve optimal design performance. Therefore, a ship structure parameter optimization method based on excellence sorting is involved, which can achieve multiple rounds of rapid optimization of structural parameters, and it is easier to find a better solution than manual optimization, so as to greatly improve the design optimization capabilities of shipyards and design manufacturers. Summary of the invention

[0004] (I) Purpose of the invention

[0005] In order to solve the technical problems existing in the background technology, the present invention proposes a ship structure parameter optimization method based on goodness sorting, which can quickly iterate and obtain the optimization solution by inputting initial values ​​and constraints, and can also obtain the optimization solution by referring to past design parameters.

[0006] (II) Technical solution

[0007] In order to solve the above technical problems, the present invention provides a ship structure parameter optimization method based on excellence ranking, comprising:

[0008] S1. Set the optimization objective function, constraint calculation formula function and constraint range for ship structure parameter optimization. The user refers to the structural variables and design experience of historical ships and inputs the upper bound of the variable of the target structure.

[0009] S2. Use the binary search method to find a feasible lower bound for each variable based on the upper bound;

[0010] S3, randomly arrange chromosome solutions within the feasible lower bound and upper bound as the initial population;

[0011] S4, performing constraint calculation on chromosomes, obtaining qualified chromosomes and retaining some unqualified chromosomes;

[0012] S5. For chromosomes that do not meet the constraints, regression optimization is performed to try to make them meet the constraints;

[0013] S6, perform parameter perturbation on the chromosome, and then perform constraint calculation to obtain qualified chromosomes and retain some unqualified chromosomes;

[0014] S7, sort all chromosomes by their merits, and select the top-ranked excellent individuals and a portion of unqualified individuals;

[0015] S8, generate new individuals through parameter exchange operation, and retain all chromosomes before and after the exchange;

[0016] S9, return to S4 to iterate until the change in goodness is less than the threshold and the convergence condition is reached;

[0017] S10. The optimal chromosome parameters are output to the ship structure model as the optimal solution of the input parameters.

[0018] Preferably, in S1, the optimization target and constraint calculation formula of the structure are derived from the structural design specification; the upper bound of the variable is input by the user, and the constraint range is the range between the upper and lower bounds of the structure's variable.

[0019] Preferably, in S2, the input structural variable upper bound includes a critical basic lower bound value and a collection of upper bound ranges less than the critical basic upper bound value;

[0020] When the structural variable is the upper limit value of the critical basis, the absolute value of the difference between the corresponding lower limit value of the critical basis is the maximum constraint range;

[0021] The input structural variable is within the upper bound range collection, and the corresponding lower bound is greater than the critical basic lower bound value. According to the input structural variable, the lower bound range collection of 1 greater than the critical basic lower bound is confirmed.

[0022] Preferably, in S3, any value is selected as a reference value in the collection of upper bounds less than the upper bound of the critical basis and the collection of lower bounds greater than the lower bound of the critical basis, and the initial chromosome population is generated by taking different distributions of the upper and lower bounds in the two collections.

[0023] Preferably, in S4, the generated chromosomes are calculated by the constraint calculation formula, and those that do not meet the constraint requirements are unqualified chromosomes. In order to maintain the diversity of chromosome solutions, a portion of unqualified chromosomes are retained while screening out qualified chromosomes.

[0024] Preferably, in S5, regression optimization is performed on the chromosomes that do not meet the constraints, any value of the upper and lower bounds of the chromosomes that do not meet the constraints is set to a fixed value, and any value of the corresponding reference upper and lower bounds is obtained, and the changed two values ​​meet the constraint range.

[0025] Preferably, in S6, the chromosome parameters are perturbed to narrow the constraint range, the chromosome parameters are adjusted within the constraint range, the constraint calculation is performed again, and the chromosomes are reselected to screen out qualified chromosomes while retaining some unqualified chromosomes.

[0026] Preferably, in S7, all chromosomes are ranked in ascending order according to the optimization target value, and the top-ranked qualified chromosome set and a portion of unqualified chromosomes are selected, with the ratio of the two being 8:2.

[0027] Preferably, in S8, the parameters of the chromosomes are exchanged, the critical upper and lower bounds of the structural variables are changed again, and all chromosomes before and after the exchange are retained to calculate the constraints and target values ​​again.

[0028] Preferably, in S9, as the number of chromosome parameter exchanges increases, until the goodness change value of the chromosome with the top optimization target value is less than a threshold, the convergence condition is reached, and the top-ranked chromosome is output as the optimal chromosome.

[0029] The above technical scheme of the present invention has the following beneficial technical effects: through any numerical value of the upper and lower bounds of the structure variable, another value is obtained by combining the dichotomy method, and a number of chromosomes are arranged within the constraint range, and the chromosomes are calculated to obtain qualified chromosomes and unqualified chromosomes. Unqualified chromosomes are constrained and reused by regression optimization, and the sufficient amount of chromosomes is ensured under the condition of satisfying the constraints. The fine-tuning constraint range is gradually narrowed, and the chromosomes are periodically optimized. The qualified chromosome set is screened by goodness sorting until the goodness change is less than the threshold, and the optimal chromosome is obtained for ship design. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the method of the present invention;

[0031] Figure 2 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0032] In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with specific embodiments and with reference to the accompanying drawings. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0033] like Figure 1-2 As shown, the present invention proposes a ship structure parameter optimization method based on goodness ranking, comprising:

[0034] S1. Set the optimization objective function, constraint calculation formula function and constraint range for ship structure parameter optimization. The user refers to the structural variables and design experience of historical ships and inputs the upper bound of the variable of the target structure.

[0035] S2. Use the binary search method to find a feasible lower bound for each variable based on the upper bound;

[0036] S3, randomly arrange chromosome solutions within the feasible lower bound and upper bound as the initial population;

[0037] S4, performing constraint calculation on chromosomes, obtaining qualified chromosomes and retaining some unqualified chromosomes;

[0038] S5. For chromosomes that do not meet the constraints, regression optimization is performed to try to make them meet the constraints;

[0039] S6, perform parameter perturbation on the chromosome, and then perform constraint calculation to obtain qualified chromosomes and retain some unqualified chromosomes;

[0040] S7, sort all chromosomes by their merits, and select the top-ranked excellent individuals and a portion of unqualified individuals;

[0041] S8, generate new individuals through parameter exchange operation, and retain all chromosomes before and after the exchange;

[0042] S9, return to S4 to iterate until the change in goodness is less than the threshold and the convergence condition is reached;

[0043] S10. The optimal chromosome parameters are output to the ship structure model as the optimal solution of the input parameters.

[0044] It should be noted that the optimization objectives and constraint calculation formulas of the structure are derived from the structural design specifications; the upper bound of the structural variables comes from user input, and the constraint range is the range between the upper and lower bounds of the structural variables.

[0045] In this embodiment, arbitrary values ​​of the upper and lower bounds of the variables of the structure are used to obtain another value in conjunction with the dichotomy method, and several chromosomes are laid out within the constraint range. The chromosomes are calculated to obtain qualified chromosomes and unqualified chromosomes. Unqualified chromosomes are constrained and reused through regression optimization. The sufficient amount of chromosomes is ensured under the condition that the constraints are met. The fine-tuning constraint range is gradually narrowed, and periodic optimization of chromosomes is performed. The qualified chromosome set is screened by goodness sorting until the goodness change is less than the threshold, and the optimal chromosome is obtained for ship design.

[0046] It can be understood that the upper bound of the input structural variable includes the critical basic lower bound value and the upper bound range collection that is less than the critical basic upper bound;

[0047] When the structural variable is the upper limit value of the critical basis, the absolute value of the difference between the corresponding lower limit value of the critical basis is the maximum constraint range;

[0048] The input structural variable is within the upper bound range collection, and the corresponding lower bound is greater than the critical basic lower bound value. The lower bound range collection greater than the critical basic lower bound is confirmed based on the input structural variable.

[0049] Among them, the critical basis upper limit value and the critical basis lower limit value are the boundaries of the two range sets, that is, the initial constraint range.

[0050] Furthermore, any value is selected as a reference value in the collection of upper bounds less than the upper bound of the critical basis and the collection of lower bounds greater than the lower bound of the critical basis, and chromosomes are generated by taking different distributions of the upper and lower bounds in the two collections.

[0051] The generated chromosomes are calculated by the constraint calculation formula. The chromosomes that do not meet the constraint requirements are unqualified chromosomes, and qualified chromosomes are directly saved to the qualified chromosome set. Unqualified chromosomes can be added to the qualified chromosome set to ensure that the number of chromosomes is sufficient to cope with periodic screening.

[0052] As a regression optimization example, regression optimization is performed on chromosomes that do not meet the constraints. Any value of the upper and lower bounds of the chromosomes that do not meet the constraints is set to a fixed value, and any value of the corresponding reference upper and lower bounds is obtained. The two values ​​after the change meet the constraint range.

[0053] In one embodiment, the chromosome parameters are perturbed to narrow the constraint range, and the chromosome parameters are adjusted within the constraint range, so as to perform a secondary screening of the chromosomes and obtain the optimization target value again. Unqualified chromosomes are also saved and wait for recycling.

[0054] All chromosomes are sorted in ascending order according to the optimization target value. When sorting, some chromosomes with high optimization target values ​​are selected for preservation, and the ratio of qualified chromosome sets to unqualified chromosome sets is always 8:2 to screen the number of chromosomes.

[0055] The parameter exchange of chromosomes changes the critical upper and lower bounds of the structural variables again, and retains all chromosomes before and after the exchange to calculate the optimization target value again. As the number of chromosome parameter exchanges increases, the goodness change value of the chromosome with the highest optimization target value is less than the threshold, and the convergence condition is reached.

[0056] It should be understood that the above specific embodiments of the present invention are only used to illustrate or explain the principles of the present invention, and do not constitute a limitation of the present invention. Therefore, any modifications, equivalent substitutions, improvements, etc. made without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. In addition, the appended claims of the present invention are intended to cover all changes and modifications that fall within the scope and boundaries of the appended claims, or the equivalent forms of such scope and boundaries.

Claims

1. A ship structure parameter optimization method based on goodness ranking, characterized in that: include: S1. Set the optimization objective function, constraint calculation formula function and constraint range for ship structure parameter optimization. The user refers to the structural variables and design experience of historical ships and inputs the upper bound of the variable of the target structure. S2. Use the binary search method to find a feasible lower bound for each variable based on the upper bound; S3, randomly arrange chromosome solutions within the feasible lower bound and upper bound as the initial population; S4, performing constraint calculation on chromosomes, obtaining qualified chromosomes and retaining some unqualified chromosomes; S5. For chromosomes that do not meet the constraints, regression optimization is performed to try to make them meet the constraints; S6, perform parameter perturbation on the chromosome, and then perform constraint calculation to obtain qualified chromosomes and retain some unqualified chromosomes; S7, sort all chromosomes by their merits, and select the top-ranked excellent individuals and a portion of unqualified individuals; S8, generate new individuals through parameter exchange operation, and retain all chromosomes before and after the exchange; S9, return to S4 to iterate until the change in goodness is less than the threshold and the convergence condition is reached; S10. The optimal chromosome parameters are output to the ship structure model as the optimal solution of the input parameters.

2. A ship structure parameter optimization method based on goodness ranking according to claim 1, characterized in that: In SI, the optimization objectives and constraint calculation formulas of the structure are derived from the structural design specifications; the upper bound of the variable comes from user input, and the constraint range is the range between the upper and lower bounds of the structure's variables.

3. A ship structure parameter optimization method based on merit ranking according to claim 1, characterized in that: In S2, the upper bound of the input structural variable includes the critical basic lower bound value and the upper bound range that is less than the critical basic upper bound; When the structural variable is the upper limit value of the critical basis, the absolute value of the difference between the corresponding lower limit value of the critical basis is the maximum constraint range; The input structural variable is within the upper bound range collection, and the corresponding lower bound is greater than the critical basic lower bound value. The lower bound range collection greater than the critical basic lower bound is confirmed based on the input structural variable.

4. A ship structure parameter optimization method based on excellence ranking according to claim 1, characterized in that: In S3, any value is selected as a reference value in the collection of upper bounds less than the upper bound of the critical basis and the collection of lower bounds greater than the lower bound of the critical basis, and the initial chromosome population is generated by taking different distributions of the upper and lower bounds in the two collections.

5. The ship structure parameter optimization method based on merit ranking according to claim 1 is characterized in that: In S4, the generated chromosomes are calculated using the constraint calculation formula, and those that do not meet the constraint requirements are considered unqualified chromosomes.

6. A ship structure parameter optimization method based on excellence ranking according to claim 1, characterized in that: In S5, regression optimization is performed on the chromosomes that do not meet the constraints, any value of the upper and lower bounds of the chromosomes that do not meet the constraints is set to a fixed value, and any value of the corresponding reference upper and lower bounds is obtained, and the two values ​​after the change meet the constraint range.

7. A ship structure parameter optimization method based on excellence ranking according to claim 1, characterized in that: In S6, the chromosome parameters are perturbed to narrow the constraint range, the chromosome parameters are adjusted within the constraint range, the constraint calculation is performed again, and the chromosomes are reselected to screen out qualified chromosomes while retaining some unqualified chromosomes.

8. A ship structure parameter optimization method based on excellence ranking according to claim 1, characterized in that: In S7, all chromosomes are ranked in ascending order according to the optimization target value. The top-ranked qualified chromosome set and a portion of unqualified chromosomes are selected, with a ratio of 8:

2.

9. The ship structure parameter optimization method based on merit ranking according to claim 1 is characterized in that: The parameters of the chromosomes are exchanged, and the critical upper and lower bound constraints of the structural variables are changed again. All chromosomes before and after the exchange are retained to calculate the constraints and target values ​​again.

10. A ship structure parameter optimization method based on excellence ranking according to claim 1, characterized in that: As the number of chromosome parameter exchanges increases, the optimality change value of the chromosome with the highest optimization target value is less than the threshold, the convergence condition is reached, and the chromosome with the highest ranking is output as the optimal chromosome.

Images