Parameter Optimization Method, System and Electronic Device for Composite Patch Repair Structure

Abstract

The present invention discloses a parameter optimization method, system and electronic device for a composite patch repair structure, which relates to the technical field of computer simulation. The parameter optimization method for the composite patch repair structure provided by the present invention, after selecting the parameters to be optimized for the composite patch repair structure, takes static strength, fatigue life, patch volume, and the total thickness of the patch and the adhesive layer as objective functions to establish a multi-objective optimization mathematical model. Then, a multi-objective genetic algorithm is used to solve the multi-objective optimization mathematical model to obtain the optimal solutions of the parameters to be optimized, which can optimize the four objective functions of the patch volume, the total thickness of the patch and the adhesive layer, static strength and fatigue life of the composite patch repair structure, so as to achieve the purpose of optimizing the overall performance of the structure.

Description

Technical Field

[0001] The present invention relates to the field of computer simulation technology, and particularly to a method, a system and an electronic device for optimizing parameters of a composite patch repair structure. Background Art

[0002] During the service process of a composite laminate structure after patch repair, it not only bears static load but also fatigue load. When designing a repair scheme, optimizing the patch repair structure by considering the static strength and fatigue life of the structure can greatly improve the repair effect and enhance the repair efficiency. During the process of designing a patch repair structure, the first thing that designers need to consider is the static load-bearing capacity of the structure and mechanical properties such as fatigue resistance. However, other impacts brought by patch repair to structural components cannot be ignored. For example, during the repair process of a composite wing, too thick patches and adhesive layers will affect the aerodynamic performance of the wing surface. When repairing some components sensitive to mass change, too large patch volume will also cause problems such as mass imbalance of the components. Considering these factors, when designing a patch repair structure, designers usually need to follow the following principles: (1) Increase the static strength of the structure to reach or restore the original structure design level. (2) Improve the fatigue life of the structure and ensure the fatigue resistance performance of the patch repair structure. (3) Reduce the total thickness of the patches and the adhesive layer to prevent affecting the aerodynamic performance of the repair area. (4) Reduce the patch volume and minimize the local mass change caused by the repair structure. It can be seen from this that when optimizing the design of a patch repair structure, there is more than one variable to be optimized. However, the dimensions of the patch volume, thickness, and the static strength and fatigue life of the structure are different and conflict with each other, and it is difficult to achieve the goal of reducing the patch thickness and volume while increasing the static strength and fatigue life. Therefore, it is necessary to compromise and comprehensively consider multiple mutually conflicting sub-objective functions. Traditional single-objective algorithms are difficult to achieve good optimization effects for such problems. Summary of the Invention

[0003] To solve the above problems existing in the prior art, the present invention provides a method, a system and an electronic device for optimizing parameters of a composite patch repair structure.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A method for optimizing parameters of a composite patch repair structure includes:

[0006] Select the parameters to be optimized for the composite patch repair structure; the parameters to be optimized include the patch radius, the patch ply and the relative thickness of the adhesive layer;

[0007] Establish a multi-objective optimization mathematical model with the static strength, fatigue life, patch volume, and the total thickness of the patches and the adhesive layer as the objective functions;

[0008] The multi-objective genetic algorithm is used to solve the multi-objective optimization mathematical model to obtain the optimal solution of the parameters to be optimized.

[0009] Optionally, the parameters to be optimized are selected according to the design principles and performance requirements of the composite patch repair structure.

[0010] Optionally, using the multi-objective genetic algorithm to solve the multi-objective optimization mathematical model to obtain the optimal solution of the parameters to be optimized specifically includes:

[0011] Encoding the parameters to be optimized to form a population; each population includes multiple chromosome individuals;

[0012] Performing non-dominated sorting on the chromosome individuals to obtain non-dominated levels;

[0013] Determining the crowding distance of chromosome individuals in the same non-dominated level;

[0014] Based on the non-dominated levels and the crowding distance, performing crossover and mutation on the chromosome individuals to generate the next generation of chromosome individuals, and continuously iterating until the preset population iteration number is reached, and taking the chromosome individuals in the first non-dominated level of the population obtained in the last iteration as the optimal solution of the parameters to be optimized.

[0015] Optionally, encoding the parameters to be optimized to form a population specifically includes:

[0016] Performing binary encoding on the patch radius;

[0017] Performing integer encoding on the order of the patch ply;

[0018] Performing discrete encoding on the relative thickness of the adhesive layer.

[0019] Optionally, performing non-dominated sorting on the chromosome individuals to obtain non-dominated levels specifically includes:

[0020] Comparing any chromosome individual in the population with other chromosome individuals to determine the number of times each chromosome individual is dominated;

[0021] Completing non-dominated sorting by dividing levels according to the number of times each chromosome individual is dominated.

[0022] Optionally, determining the crowding distance of chromosome individuals in the same non-dominated level specifically includes:

[0023] Taking a certain objective function as a benchmark, sorting the chromosome individuals in the same non-dominated level to obtain an individual sequence;

[0024] Taking the sum of the differences between the objective functions of the two chromosome individuals before and after any chromosome individual in the individual sequence as the crowding distance.

[0025] Optionally, crossover and mutation are performed on chromosome individuals based on the non-dominated rank and the crowding distance, specifically including:

[0026] When two chromosome individuals are not in the same non-dominated layer, compare the ranks of the non-dominated layers where the two chromosome individuals are located, and perform crossover and mutation on the chromosome individual with the earlier rank.

[0027] When two chromosome individuals are in the same non-dominated layer, select the chromosome individual with the larger crowding distance among the two chromosome individuals for crossover and mutation.

[0028] Optionally, the crossover operation is: two parent chromosome individuals exchange some segments on their respective chromosomes with each other; the mutation operation is: mutation is performed respectively at the coding positions of the patch radius, the patch thickness, and the patch ply.

[0029] According to the specific embodiments provided by the present invention, the following technical effects are disclosed by the present invention:

[0030] The parameter optimization method for the composite patch repair structure provided by the present invention, after selecting the parameters to be optimized for the composite patch repair structure, takes the static strength, fatigue life, patch volume, and the total thickness of the patch and the adhesive layer as objective functions, establishes a multi-objective optimization mathematical model, and then uses a multi-objective genetic algorithm to solve the multi-objective optimization mathematical model to obtain the optimal solutions of the parameters to be optimized, which can optimize the four objective functions of the patch volume, the total thickness of the patch and the adhesive layer, the static strength, and the fatigue life of the composite patch repair structure, so as to achieve the purpose of optimizing the overall performance of the structure.

[0031] To implement the above-mentioned parameter optimization method for the composite patch repair structure, the present invention also provides the following implementation structure:

[0032] A parameter optimization system for a composite patch repair structure, which is applied to the above-mentioned parameter optimization method for the composite patch repair structure; the system includes:

[0033] A parameter selection module, which is used to select the parameters to be optimized for the composite patch repair structure; the parameters to be optimized include the patch radius, the patch ply, and the relative thickness of the adhesive layer.

[0034] A model construction module, which is used to establish a multi-objective optimization mathematical model with the static strength, fatigue life, patch volume, and the total thickness of the patch and the adhesive layer as objective functions.

[0035] A parameter optimization module, which is used to use a multi-objective genetic algorithm to solve the multi-objective optimization mathematical model to obtain the optimal solutions of the parameters to be optimized.

[0036] An electronic device, including:

[0037] A memory for storing a computer program;

[0038] A processor, connected to the memory, for retrieving and executing the computer program to implement the parameter optimization method for the composite patch repair structure described above.

[0039] Since the technical effects achieved by the two implementation structures provided by the present invention are the same as those achieved by the parameter optimization method for the composite patch repair structure provided by the present invention, they will not be elaborated here. Description of the Drawings

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0041] Figure 1 It is a flowchart of the parameter optimization method for the composite patch repair structure provided by the present invention;

[0042] Figure 2 It is an implementation flowchart of the parameter optimization method for the composite patch repair structure provided by the present invention;

[0043] Figure 3 It is a schematic diagram of the composite patch repair structure provided by the present invention; wherein, Figure 3 (a) of it is a top view of the composite patch repair structure, Figure 3 and (b) of it is a side view of the composite patch repair structure;

[0044] Figure 4 It is a schematic diagram of the hybrid coding of the patch repair structure provided by the present invention;

[0045] Figure 5 It is a schematic diagram of the coding of the patch radius D provided by the present invention;

[0046] Figure 6 It is a schematic diagram of the patch ply coding provided by the present invention;

[0047] Figure 7 It is a schematic diagram of the congestion degree provided by the present invention;

[0048] Figure 8 It is a schematic diagram of chromosome crossover and mutation provided by the present invention; wherein, Figure 8 (a) of it is a schematic diagram of chromosome crossover, Figure 8 and (b) of it is a schematic diagram of chromosome mutation;

[0049] Figure 9The structural diagram of the optimization program provided by the present invention;

[0050] Figure 10 The schematic diagram of the optimization process provided by the present invention;

[0051] Figure 11 The flowchart of the next-generation population generation provided by the present invention;

[0052] Figure 12 The comparison diagram of the static strength before and after optimization provided by the present invention;

[0053] Figure 13 The comparison diagram of the fatigue life before and after optimization provided by the present invention;

[0054] Figure 14 The comparison diagram of the total thickness of the patch and the adhesive layer before and after optimization provided by the present invention;

[0055] Figure 15 The comparison diagram of the patch volume before and after optimization provided by the present invention;

[0056] Figure 16 The comparison diagram of the patch diameter before and after optimization provided by the present invention. Specific embodiments

[0057] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0058] The purpose of the present invention is to provide a parameter optimization method, system and electronic device for a composite patch repair structure, which can optimize four objective functions of the patch volume, the total thickness of the patch and the adhesive layer, the static strength and the fatigue life of the composite patch repair structure, so as to achieve the purpose of optimal overall performance of the structure.

[0059] To make the above objects, features and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0060] As Figure 1 shown, the parameter optimization method for the composite patch repair structure provided by the present invention includes:

[0061] Step 100: Select the parameters to be optimized for the composite patch repair structure.

[0062] In the actual application process, the patch radius, patch ply, and relative thickness of the adhesive layer of the composite patch repair structure can be selected as the parameters to be optimized according to the design principles and performance requirements of the composite patch repair structure, and their value ranges can be determined.

[0063] Step 101: Establish a multi-objective optimization mathematical model with static strength, fatigue life, patch volume, and the total thickness of the patch and adhesive layer as the objective functions. Among them, determine that the objective functions of the optimization design are static strength, fatigue life, patch volume, and the total thickness of the patch and adhesive layer, and optimize the objective functions. Considering the design principles of the patch repair structure, maximize some sub-objective functions (such as static strength and fatigue life), and minimize some sub-objective functions (such as patch volume and the total thickness of the patch and adhesive layer).

[0064] Step 102: Solve the multi-objective optimization mathematical model using the multi-objective genetic algorithm to obtain the optimal solutions of the parameters to be optimized. In the actual application process, the implementation process of this step can be as follows:

[0065] Step 102-1: Encode the parameters to be optimized to form a population. Each population includes multiple chromosome individuals. In the actual application process, this step can select the hybrid encoding method, that is, combine binary encoding and integer encoding. Among them, for continuous optimization variables: binary encode the patch radius. Integer encode the order of the patch plies. Discretely encode the relative thickness of the adhesive layer. Based on this, the chromosome encoding of the patch repair structure consists of three parts: patch radius encoding, patch ply encoding, and adhesive layer thickness encoding.

[0066] For example, for the adhesive layer thickness, since it has little influence on the static strength and fatigue life of the patch repair structure, to improve the optimization efficiency, its value is only taken as 0.2, 0.4, 0.6, 0.8, 1.0 times the patch thickness value for calculation, so the adhesive layer thickness adopts discrete encoding. For the patch radius D, its search space is [a, b], and the number A in the search space is converted into the decimal number B in [0, 2 n -1], and the conversion formula is as follows:

[0067]

[0068] Finally, convert the decimal number B into an n-bit binary chromosome segment. To prevent the chromosome from being too long and affecting the optimization efficiency, n is taken as 6, that is, the binary chromosome segment corresponding to the radius consists of 6 binary numbers.

[0069] For the patch ply, according to the constraint conditions, there are mainly four ply cases for each layer: 0°, 45°, -45° and 90°. To achieve the optimization goal of reducing the patch thickness and the number of patch plies, in addition to the four ply angle selection cases for each ply, another case of ply deletion is introduced, that is, each ply is represented by 5 integers: 0 represents -45°, 1 represents 0°, 2 represents 45°, 3 represents 90°, and 4 represents the deletion of this ply. The maximum number of patch plies is 12 and the minimum is 4. Therefore, the ply segment on the chromosome is mainly composed of 12 integers encoded from 0 to 4, where the code 4 representing the deleted ply cannot exceed 8, and the code 2 representing 45° and the code 0 representing -45° must be adjacent, and the number of adjacent identical codes cannot exceed 3. For the adhesive layer thickness, its value is taken as 0.2, 0.4, 0.6, 0.8, 1.0 times of the patch thickness. Therefore, 5 integers are used to represent its relative thickness: 1, 2, 3, 4, 5, which respectively represent 0.2, 0.4, 0.6, 0.8, 1.0 times of the patch thickness.

[0070] Step 102-2: Perform non-dominated sorting on the chromosome individuals to obtain non-dominated levels. Among them, in multi-objective optimization, a new relationship is defined to compare the advantages and disadvantages of two chromosome individuals: domination and being dominated.

[0071] For example, for any two chromosome individuals A' and B' in the population, if all the sub-objective function values of chromosome individual A' are greater than or equal to the sub-objective function values of chromosome individual B', and A' has at least one sub-objective function value greater than that of B', then it is considered that A' dominates B', and chromosome individual A' is superior to chromosome individual B'. Compare each chromosome individual in the population with all other chromosome individuals in the population, record the number of times it is dominated, and finally divide the population into levels according to the number of times each chromosome individual is dominated, that is, the fast non-dominated sorting of the population is completed. Among them, the chromosome individual with fewer times of being dominated is better and has a greater probability of entering the next generation population. All chromosome individuals at the level with 0 times of being dominated are called non-dominated solutions, and the non-dominated solutions of each generation of population can be used as the Pareto optimal solution set of the contemporary population. Based on this, the calculation process of fast non-dominated sorting is as follows:

[0072] Step 1: Let chromosome individual i = 1.

[0073] Step 2: Compare the i-th chromosome individual with each other chromosome individual in the population to record the number of times the chromosome individual i is dominated.

[0074] Step 3: Judge whether the number of times the chromosome individual i is dominated is 0. If so, the chromosome individual i is not dominated by any other chromosome individual, is marked as a non-dominated individual, and the chromosome individual i is stored in the non-dominated solution set.

[0075] Step 4: Let i' = i + 1, and repeat Step 2 until the entire population is traversed.

[0076] The first layer of non-dominated sorting of the population is obtained through the above 4 steps. Then, delete the chromosome individuals that have been marked as non-dominated, repeat Steps 1 to 4, and obtain the second layer of non-dominated sorting. Repeat this screening process until all individuals in the entire population are marked with non-dominated levels, and the non-dominated sorting is completed.

[0077] Step 102-3: Determine the crowding distance of chromosome individuals in the same non-dominated layer.

[0078] Among them, the crowding distance is an index to evaluate the density of other individuals around a chromosome individual. It is proposed to improve the diversity of the population and avoid the disadvantage of the non-dominated sorting genetic algorithm (NSGA) relying on the sharing radius to maintain population diversity. Using it as the selection criterion for chromosome individuals in the population iteration process can make the optimal solutions evenly distributed. The determination of the crowding distance is mainly through calculating the sum of all distance differences between chromosome individual i - 1 and chromosome individual i + 1 around chromosome individual i in the space of each optimization sub-objective function. Its calculation formula is:

[0079]

[0080] In the formula, d i is the crowding distance of chromosome individual i, M is the number of optimization objective functions, is the value of the m-th sub-objective function of chromosome individual i + 1, is the value of the m-th sub-objective function of chromosome individual i - 1, is the maximum value of the m-th sub-objective function, is the minimum value of the m-th sub-objective function. For the individuals at both ends of the non-dominated sorting layer, their crowding degrees are set to infinity.

[0081] Step 102-4: Based on the non-dominated level and crowding distance, perform crossover and mutation on chromosome individuals to generate the next generation of chromosome individuals, and continuously iterate until the preset population iteration number is reached. Then, use the chromosome individuals in the first non-dominated layer of the population obtained in the last iteration as the optimal solutions of the parameters to be optimized.

[0082] In order to select the chromosome individuals from the parent population to generate the offspring population, in the multi-objective optimization algorithm, the tournament selection method is generally used to select chromosome individuals, that is, randomly select two chromosome individuals from the population, and by comparing their non-dominated sorting ranks or crowding degrees, the winner can be used to generate the next generation of the population. The relevant comparison process is as follows:

[0083] Step 1: Compare the non-dominated levels of two individual chromosomes. The chromosome individual with a higher level undergoes crossover and mutation to generate the next generation of individuals. If the two chromosome individuals are in the same dominance level, proceed to Step 2.

[0084] Step 2: Compare the crowding degrees (i.e., crowding distances) of the two chromosome individuals. The chromosome individual with a larger crowding degree is selected for crossover and mutation to generate the next generation of individuals.

[0085] Furthermore, in the actual application process, a parametric modeling script is written in Python to quickly generate simulation models with different patch radii, patch layups, and relative thicknesses of the adhesive layer, and calculate their static strength and fatigue life.

[0086] As a preferred embodiment, the multi-objective optimization process for the patch repair structure is as follows:

[0087] Step 1: Randomly generate the first generation population of N chromosome individuals.

[0088] Step 2: Determine whether the current population is the first generation of parent population. If so, decode it. If not, proceed to the next step.

[0089] Step 3: Combine the parent population and its fitness for non-dominated sorting. On this basis, generate the offspring population through selection, crossover, and mutation.

[0090] Step 4: Perform decoding operations on the newly generated offspring population.

[0091] Step 5: Combine the parent and offspring populations to generate the next generation population. First, merge the parent and offspring populations to form a combined population of 2N chromosome individuals. Then, perform non-dominated sorting on the combined population. Finally, starting from the first non-dominated level of the combined population, take out each chromosome individual layer by layer and store it in the next generation population. When the individual scale of the new generation population needs to reach N, but only some chromosome individuals in a certain dominance level are required, only select the chromosome individuals with a larger crowding degree in that dominance level and store them in the next generation population, and eliminate the remaining chromosome individuals and levels. Through this process, excellent individuals are promoted into the next generation population, thus enabling the algorithm to converge faster.

[0092] Step 6: Determine whether the iteration number of the current population exceeds 40. If not, use the newly formed next generation population as the new parent population and repeat Steps 2 to 5. If it exceeds, output the results of the first non-dominated level of the population as the Pareto optimal solution set.

[0093] Step 7: Analyze and compare the Pareto optimal solution sets of the first generation population and the last generation to verify the effectiveness of the optimization algorithm.

[0094] Based on the above description, compared with the prior art, the present invention has the following advantages:

[0095] First, in view of the defects of the existing composite patch repair structure design research with fewer parameters and lower optimization efficiency, the present invention realizes the genetic algorithm optimization of various parameters of the patch repair structure. While considering multiple optimization objectives, through the optimization method provided by the present invention, a patch repair structure that meets the comprehensive performance requirements and has a good simulation effect is finally successfully obtained.

[0096] Second, the present invention proposes a hybrid coding method for the chromosome composed of multiple parameters such as the patch radius, patch ply, and relative thickness of the adhesive layer of the composite patch repair structure, which simply and clearly defines the complex parameters of each individual and can improve the optimization efficiency.

[0097] Third, the present invention proposes a performance evaluation method that simultaneously considers the strength and fatigue life of the composite patch repair structure, which is convenient for systematically comparing the advantages and disadvantages of different patch repair structures.

[0098] Fourth, in the optimization process of the composite patch repair structure, the present invention uses parametric modeling to quickly and conveniently establish simulation models of different patch repair structures.

[0099] It can be seen that the purpose of the present invention is to overcome the technical defects existing in the prior art, optimize the four objective functions of the patch volume, the total thickness of the patch and the adhesive layer, the static strength and the fatigue life of the composite patch repair structure by means of a multi-objective genetic algorithm, and propose an evaluation method that comprehensively considers the mechanical properties, aerodynamic properties and control of quality change of the composite patch repair structure, so as to achieve the design purpose of the optimal overall performance of the structure.

[0100] A specific embodiment is provided below to illustrate the specific implementation process of the above-mentioned parameter optimization method for the composite patch repair structure provided by the present invention. As Figure 2 shown, the implementation process of the parameter optimization method for the composite patch repair structure in this embodiment includes:

[0101] Step SS1: Design a composite patch repair structure for optimal design, as Figure 3 shown. Determine the overall optimization objectives of the composite patch repair structure. According to the design principles and performance requirements of the composite patch repair structure, select the parameters to be optimized as the patch radius, patch ply, and relative thickness of the adhesive layer, and determine their value ranges. Taking the static strength, fatigue life, patch volume, and total thickness of the patch and the adhesive layer as the objective functions, establish a multi-objective optimization data model, and perform optimization processing on the multi-objective optimization data model.

[0102] As a preferred embodiment, step SS1 specifically includes: according to the design principle and performance requirements, the patch radius D, the patch ply X = (θ, x), and the relative thickness h of the adhesive layer are selected as the parameters to be optimized. Among them, θ is the ply angle, selected from 0°, 45°, -45°, 90°, that is: θ ∈ (0°, 45°, -45°, 90°), and x is the number of patch plies. The adjacent plies with the same orientation cannot be laid more than 3 layers, and 45° and -45° need to be laid adjacent to each other. The minimum number of patch layers in the present invention is 4 layers, and the maximum is half of the thickness of the mother board, that is, 12 layers. Therefore, x ∈ [4, 12], and the patch radius is selected as D ∈ [6, 18.6].

[0103] The objective functions mainly include: static strength σ b (X, D, h), fatigue life N f (X, D, h), patch volume V(D, x), and the total thickness H(h, X) of the patch and the adhesive layer, a total of four. Considering the design principle of the patch repair structure, in this embodiment, some sub-objective functions are maximized, and some sub-objective functions are minimized. Among them, through the optimization process, the static strength and fatigue life are made as large as possible, and the patch volume and the total thickness of the patch and the adhesive layer are made as small as possible.

[0104] The fatigue life of the patch repair structure is negatively correlated with the maximum stress σ 11 value in the fiber direction of the 0° ply of the mother board. Calculating the fatigue life of each patch repair structure member requires thousands of iterations. To save calculation time and improve the optimization efficiency, for the part of the optimization objective that involves increasing the fatigue life, it is replaced by reducing the maximum stress σ 11 level in the fiber direction of the 0° ply of the mother board.

[0105] In summary, the multi-objective optimization of the patch repair structure can be described as a mathematical model in the following form:

[0106] Max σ b (X, D, h)

[0107] Min V(D, x)

[0108] Min H(h, x)

[0109] Min σ 11 (X, D, h)

[0110] St. X = (θ, x), θ ∈ (0°, 45°, -45°, 90°), 4 ≤ x ≤ 12

[0111] 6 ≤ D ≤ 18.6, 0.2 ≤ h ≤ 1

[0112] Step SS2: Use a multi-objective genetic algorithm to optimize the parameters of the patch repair structure. This genetic algorithm mainly consists of five parts: fast non-dominated sorting, crowding distance, chromosome coding, tournament selection, and crossover and mutation. This step introduces the chromosome coding, non-dominated sorting, and crowding distance parts. The combinations of various parameters of the patch and the adhesive layer are encoded as a chromosome through a hybrid coding method. Non-dominated sorting is performed on the generated numerous chromosome individuals, and their crowding distances are calculated to evaluate the quality of the individuals for the next selection.

[0113] Based on this, step SS2 specifically includes:

[0114] Process and encode the 3 parameters to be optimized (patch radius D, patch ply X = (θ, x), and relative adhesive layer thickness h) so that they form a chromosome representing a composite patch repair structure with certain characteristics. Select the hybrid coding method, that is, combine binary coding and integer coding. As Figure 4 shown, the chromosome coding of the patch repair structure consists of three parts: patch radius coding, patch ply coding, and adhesive layer thickness coding. The processing process for the parameters to be optimized includes: binary coding is used for the patch radius D. Integer coding is used for the patch ply sequence. For the adhesive layer thickness, since it has little effect on the static strength and fatigue life of the patch repair structure, to improve the optimization efficiency, its value is only taken as 0.2, 0.4, 0.6, 0.8, 1.0 times the patch thickness value for calculation. Therefore, the adhesive layer thickness also adopts discrete coding.

[0115] The coding method for the parameters to be optimized includes: for the patch radius D, its search space is [a, b]. According to the Figure 5 process shown, the number A in the search space is now converted into the decimal number B in [0, 2 n -1]. Finally, the decimal number B is converted into an n-bit binary chromosome segment. To prevent the chromosome from being too long and affecting the optimization efficiency, n is taken as 6, that is, the binary chromosome segment corresponding to the radius consists of 6 binary numbers.

[0116] For the patch ply, according to the constraint conditions, there are mainly four ply cases for each layer: 0°, 45°, -45°, and 90°. To achieve the optimization goal of reducing the patch thickness and the number of patch plies, in addition to the four ply angle selection cases for each ply, one more case of ply deletion is introduced, that is, each ply is represented by 5 integers: 0 represents -45°, 1 represents 0°, 2 represents 45°, 3 represents 90°, and 4 represents the deletion of this ply. The maximum number of patch plies is 12 and the minimum is 4. Therefore, the ply segment on the chromosome is mainly composed of 12 integers encoded from 0 to 4, where the code 4 representing the deleted ply cannot exceed 8, the code 2 representing 45° and the code 0 representing -45° must be adjacent, and the number of adjacent identical codes cannot exceed 3. The ply codes are as Figure 6 shown:

[0117] For the adhesive layer thickness, its value is taken as 0.2, 0.4, 0.6, 0.8, 1.0 times the patch thickness. Therefore, 5 integers are used to represent its relative thickness: 1, 2, 3, 4, 5, which respectively represent 0.2, 0.4, 0.6, 0.8, 1.0 times the patch thickness.

[0118] Perform non-dominated sorting on many chromosome individuals encoded for the three parameters using the above method. For example, for any two individuals Y and Z in the population, if all the sub-objective function values of individual Y, σ b (X, D, h), σ 11 (X, D, h), V(D, x), and H(h, X) are all greater than or equal to the sub-objective function values of individual Z, and at least one sub-objective function value of individual Y is greater than that of individual Z, then individual Y is considered to dominate individual Z and individual Y is superior to individual Z. Compare each individual in the population with all other individuals in the population, record the number of times it is dominated, and finally divide the population into levels according to the number of times each individual is dominated, that is, the fast non-dominated sorting of the population is completed. Among them, the fewer the number of times an individual is dominated, the better it is, and there is a greater probability of entering the next generation population. All individuals at the level with the number of times dominated being 0 are called non-dominated solutions, and the non-dominated solutions of each generation of the population can be used as the Pareto optimal solution set of the contemporary population. The calculation process of fast non-dominated sorting is as follows:

[0119] Step 1: Let individual i = 1.

[0120] Step 2: Compare individual i with each other individual in the population for dominance and record the number of times individual i is dominated.

[0121] Step 3: Determine whether the number of times individual i is dominated is 0. If so, individual i is not dominated by any other individual, is marked as a non-dominated individual, and individual i is stored in the non-dominated solution set.

[0122] Step 4: Let i' = i + 1, and repeat Step 2 until the entire population is traversed.

[0123] The first layer of the non - dominated sorting of the population is obtained through the above 4 steps.

[0124] Then, delete the individuals that have been marked as non - dominated, and repeat Steps 1 to 4 to obtain the second layer of non - dominated sorting.

[0125] Repeat this screening process until all individuals in the entire population are marked with non - dominated levels.

[0126] Crowding distance is an index to evaluate the density of other individuals around an individual, as Figure 7 shown. Among the individuals at the same level, the individual with a larger crowding distance is more likely to be selected to enter the next iteration. The determination of the crowding distance is mainly by calculating the sum of all distance differences between individuals i - 1 and i + 1 around individual i in the space of each optimization sub - objective function.

[0127] Step SS3: This step includes two parts: tournament selection and crossover - mutation of chromosomes in the multi - objective genetic algorithm. The exchange of chromosome coding segments and the change of individual numbers are used as chromosome crossover and mutation. Consider the non - dominated level and crowding degree to select individuals from each generation of the population for crossover - mutation to generate the next - generation individuals.

[0128] This step specifically includes: The main role of selection is to select individuals from the parent population to generate the offspring population. In the multi - objective optimization algorithm, the tournament selection method is generally used, that is, randomly select two individuals from the population, and by comparing their non - dominated sorting levels or crowding distances, the winner can be used to generate the next - generation population. The relevant comparison process is as follows:

[0129] Step 1: According to the non - dominated ranking method in Step SS2, compare the non - dominated levels of two individuals, and the individual with a higher level undergoes crossover - mutation to generate the next - generation individuals. If the two individuals are in the same domination layer, then go to Step 2.

[0130] Step 2: According to the definition of the crowding distance in Step SS2, compare the crowding distance sizes of two individuals, and the individual with a larger crowding distance is selected for crossover - mutation to generate the next - generation individuals.

[0131] The crossover and mutation of chromosome coding are as Figure 8 shown. For crossover, mainly adopt as Figure 8For the shown crossover method, two parental chromosomes exchange some segments of each other's chromosomes. For mutation, mutation is performed at the coding positions of the patch radius, patch thickness, and patch ply. During the chromosome mutation process, a bit is randomly selected in the coding part of the patch radius. For example, change the 1 at this bit to 0 and the 0 to 1. For the coding parts of the adhesive layer thickness and patch ply, replace the number at this position with other numbers within the coding range during mutation.

[0132] Step SS4: Write a multi-objective genetic algorithm optimization program to output the optimal solution set. For example, use the Python language to write the multi-objective genetic algorithm optimization program for the above steps, and write a parametric modeling script to implement the parametric modeling process of the composite patch repair structure corresponding to the chromosome individuals, and use numerical simulation methods to predict the static strength and fatigue life of the model respectively to obtain the optimal solution set.

[0133] As a preferred embodiment, this step specifically includes: writing a program to implement the genetic algorithm optimization main program of the above steps, combining the static strength calculation program and fatigue life calculation program of the simulation model to form an optimized complete process, as Figure 9 shown. Use the Python language to write a parametric modeling script to quickly generate simulation models with different patch radii, patch plies, and relative thicknesses of the adhesive layer, and calculate their static strength and fatigue life.

[0134] Step SS5: When the population iteration number in the genetic algorithm exceeds 40, output the chromosomes of the results in the first non-dominated layer of the population as the Pareto optimal solution set. Compare the optimal solution sets of the first-generation population and the last-generation optimal solution set to verify the effectiveness.

[0135] The multi-objective optimization process of the patch repair structure in this step is as Figure 10 shown, including:

[0136] Step 1: Randomly generate the first-generation population of N individuals (that is, generate the initial population and parental population).

[0137] Step 2: Determine whether the current population is the first-generation parental population. If so, decode it, that is, run the parametric modeling script. The execution process of the parametric modeling script is geometric modeling - load application - mesh division - job submission. If not, execute Step 3.

[0138] Step 3: Combine the parental population and its fitness for non-dominated sorting, and on this basis, generate the offspring population through selection, crossover, and mutation.

[0139] Step 4: Perform decoding operations on the newly generated offspring population.

[0140] Step 5: According to Figure 11Perform non-dominated sorting on the shown process, and generate the next generation population (i.e., the new generation population) by combining the parent population and the offspring population. First, merge the parent population and the offspring population to form a combined population of 2N individuals. Then, perform non-dominated sorting on the combined population. Finally, starting from the first non-dominated level of the combined population, take out each individual layer by layer and store them in the next generation population. The individual scale of the new generation population should reach N. However, when only some individuals in a certain dominated level are needed, only select the individuals with a larger crowding degree in that dominated level and store them in the next generation population, and eliminate the remaining individuals and levels. Through such a process, excellent individuals are promoted to enter the next generation population, thereby enabling the algorithm to converge faster.

[0141] Step 6: Determine whether the current population iteration number exceeds 40. If not, use the newly formed next generation population as the new parent population and repeat Steps 2 to 5. If it exceeds, output the result of the first non-dominated level of the population as the Pareto optimal solution set. Analyze and compare the Pareto optimal solution sets of the first generation population and the last generation to verify the effectiveness of the optimization algorithm.

[0142] Based on this, in this embodiment, taking the static strength, fatigue life, total thickness of the patch and the adhesive layer, and the patch volume as the optimization objectives, the parameters of the patch and the adhesive layer of the patch repair structure are optimized by combining the multi-objective genetic algorithm. The specific content of its optimization process is as follows: Analyze the mechanical properties and design principles of the composite patch repair structure in practical applications to determine the overall optimization objectives. Determine the parameters to be optimized and their value ranges for the composite patch repair structure. Use the multi-objective optimization algorithm, taking the static strength, fatigue life, patch volume, and total thickness of the patch and the adhesive layer as the four objective functions, and establish a multi-objective optimization mathematical model. Write the main program for optimizing the multi-objective genetic algorithm, and combine the parametric modeling scripts for fatigue life prediction and static strength prediction to form the optimization process and output the optimal solution set. Analyze and compare the optimal solution sets of the first generation population and the last generation to verify the effectiveness of the optimization algorithm.

[0143] Furthermore, in this embodiment, under the consideration of the four optimization objectives of the fatigue life, static strength, total thickness of the patch and the adhesive layer, and the patch volume of the patch repair structure, an optimization design of the patch repair structure is carried out. The population size of the genetic algorithm is set to 40, and the maximum number of iterations is 40. During the calculation of the static strength, the left end of the structure is fixed, and a displacement load of 1.75 mm is applied to the right end. When calculating the fatigue part, the left end of the structure is fixed, and the maximum fatigue load is applied to the right end, and its value is taken as 80% of the static strength of the intact plate. The optimization results are as Figures 12 to 16 and Table 1 shows.

[0144] Table 1 Comparison table of patch parameters before and after optimization

[0145]

[0146] Comparing the above optimization results, it can be found that:

[0147] (1) For static strength, observing Figure 12 it can be found that the range of the static strength values of the individuals in the first-generation Pareto optimal solution set is not much different from that of the fortieth generation, both being between 430 MPa and 590 MPa. However, compared with the first-generation Pareto optimal solution set, the distribution of the static strength of the individuals in the fortieth-generation Pareto optimal solution set shows a polarized distribution.

[0148] (2) For fatigue life, through Figure 13 it can be seen that after forty generations of optimization, the fatigue life of the individuals in the fortieth-generation Pareto optimal solution set also shows a polarized distribution, that is, most individuals have infinite fatigue life, and the life values of some individuals are very small, and the patch diameter also shows a certain degree of bipolar distribution.

[0149] (3) For the thickness of the patch and the adhesive layer, combining Figure 14 and Table 1, it can be found that after optimization, the thicknesses of the patch and the adhesive layer are mostly reduced to below 1.25 mm, with an average reduction of about 26%. The number of patch layers has also been reduced from the original 12 - 6 layers to 10 - 4 layers, successfully achieving the optimization goal of reducing the total thickness of the patch and the adhesive layer.

[0150] (4) For the volume of the patch, from Figure 15 it can be seen that the volume of the individuals in the fortieth-generation Pareto optimal solution set after optimization is generally reduced below that of the first-generation Pareto optimal solution set, with an average reduction of about 41%.

[0151] (5) For the patch diameter, reference can be made to Figure 16 and will not be elaborated here.

[0152] In summary, through the optimization method established in the present invention, the goals of reducing the thickness of the patch and the adhesive layer, reducing the volume of the patch, improving the static strength of the structure, and fatigue life have been successfully achieved. The optimization results show that the static strength and fatigue life of the individuals in the Pareto optimal solution set show a polarized distribution, and designers can select the most suitable repair scheme for the actual needs in the Pareto optimal solution set according to the actual requirements of the static strength and anti-fatigue performance of the engineering structure.

[0153] To implement the above-mentioned parameter optimization method for the composite patch repair structure, the present invention also provides the following implementation structure:

[0154] A parameter optimization system for a composite patch repair structure, which is applied to the above-mentioned parameter optimization method for the composite patch repair structure. The system includes:

[0155] A parameter selection module for selecting parameters to be optimized for a composite patch repair structure. The parameters to be optimized include the patch radius, patch ply, and relative thickness of the adhesive layer.

[0156] A model construction module for establishing a multi-objective optimization mathematical model with static strength, fatigue life, patch volume, and the total thickness of the patch and adhesive layer as objective functions.

[0157] A parameter optimization module for solving the multi-objective optimization mathematical model using a multi-objective genetic algorithm to obtain the optimal solutions of the parameters to be optimized.

[0158] An electronic device, comprising:

[0159] A memory for storing a computer program.

[0160] A processor connected to the memory for retrieving and executing the computer program to implement the parameter optimization method for the above-mentioned composite patch repair structure.

[0161] In addition, when the computer program in the above-mentioned memory is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories, random access memories, magnetic disks, or optical discs that can store program codes.

[0162] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same or similar parts among the various embodiments can be referred to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple. For the relevant parts, reference can be made to the description in the method part.

[0163] Specific examples are used in this article to elaborate on the principles and implementation manners of the present invention. The descriptions of the above embodiments are only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A parameter optimization method for a composite patch repair structure, characterized in that Including: Select the parameters to be optimized for the composite patch repair structure; the parameters to be optimized include the patch radius, patch ply, and relative thickness of the adhesive layer; Establish a multi-objective optimization mathematical model with static strength, fatigue life, patch volume, and the total thickness of the patch and adhesive layer as the objective functions; Use a multi-objective genetic algorithm to solve the multi-objective optimization mathematical model to obtain the optimal solutions of the parameters to be optimized; Among them, the multi-objective optimization process for the patch repair structure includes: Encode the parameters to be optimized into a chromosome by means of hybrid encoding, and randomly generate the first-generation population of N chromosome individuals; perform binary encoding on the patch radius, integer encoding on the order of patch plies, and discrete encoding on the relative thickness of the adhesive layer; Judge whether the current population is the first-generation parent population. If so, decode the first-generation parent population and run the parametric modeling script; if not, perform non-dominated sorting on the parent population and its fitness to obtain non-dominated levels. On this basis, generate the offspring population through selection, crossover, and mutation; write a multi-objective genetic algorithm optimization program in Python language and write a parametric modeling script; Perform a decoding operation on the newly generated offspring population; Generate the next-generation population by combining the parent and offspring populations; first, merge the parent and offspring populations to form a combined population of 2N chromosome individuals; then, perform non-dominated sorting on the combined population; finally, starting from the first non-dominated level of the combined population, take out each chromosome individual layer by layer and store it in the next-generation population; when the individual scale of the new-generation population needs to reach N, but only some chromosome individuals in a certain dominant level are required, only select the chromosome individuals with a large crowding distance in that dominant level and store them in the next-generation population, and eliminate the remaining chromosome individuals and levels; through such a process, promote excellent individuals to enter the next-generation population, so that the algorithm converges faster; Judge whether the current population iteration number exceeds the preset population iteration number. If not, use the newly formed next-generation population as the new parent population and return to execute the step of judging whether the current population is the first-generation parent population; if so, output the results of the first non-dominated level of the population as the Pareto optimal solution set.

2. The parameter optimization method of the composite patch repair structure according to claim 1, characterized in that Select the parameters to be optimized according to the design principles and performance requirements of the composite patch repair structure.

3. The parameter optimization method for the composite patch repair structure according to claim 1, characterized in that, Using a multi-objective genetic algorithm to solve the multi-objective optimization mathematical model to obtain the optimal solutions of the parameters to be optimized, specifically including: Encode the parameters to be optimized to form a population; each population includes multiple chromosome individuals; Perform non-dominated sorting on the chromosome individuals to obtain non-dominated levels; Determine the crowding distance of chromosome individuals in the same non-dominated level; Based on the non-dominated levels and the crowding distance, perform crossover and mutation on the chromosome individuals to generate the next-generation chromosome individuals, and continuously iterate until the preset population iteration number is reached. Then, use the chromosome individuals in the first non-dominated level of the population obtained in the last iteration as the optimal solutions of the parameters to be optimized.

4. The parameter optimization method for the composite patch repair structure according to claim 3, characterized in that Performing non-dominated sorting on the chromosome individuals to obtain non-dominated levels, specifically including: Compare any chromosome individual in the population with other chromosome individuals to determine the number of times each chromosome individual is dominated; Complete non-dominated sorting of the divided levels according to the number of times each chromosome individual is dominated.

5. The parameter optimization method for the composite patch repair structure according to claim 3, characterized in that Determine the crowding distance of chromosome individuals in the same non-dominated level, specifically including: Taking a certain objective function as a benchmark, sort the chromosome individuals in the same non-dominated level to obtain an individual sequence; Take the sum of the differences in the objective functions of the two chromosome individuals before and after any chromosome individual in the individual sequence as the crowding distance.

6. The parameter optimization method for the composite patch repair structure according to claim 3, characterized in that, Perform crossover and mutation on chromosome individuals based on the non-dominated level and the crowding distance, specifically including: When two chromosome individuals are not in the same non-dominated level, compare the levels of the non-dominated levels where the two chromosome individuals are located, and perform crossover and mutation on the chromosome individual with the earlier level; When two chromosome individuals are in the same non-dominated level, select the chromosome individual with the larger crowding distance among the two chromosome individuals for crossover and mutation.

7. The parameter optimization method for the composite patch repair structure according to claim 6, characterized in that The crossover operation is: two parent chromosome individuals exchange some fragments on their respective chromosomes with each other; the mutation operation is: perform mutation at the coding positions of the patch radius, the patch thickness, and the patch ply.

8. A parameter optimization system for a composite patch repair structure, characterized in that, Applied to the parameter optimization method of the composite patch repair structure as described in any one of claims 1-7; the system includes: A parameter selection module for selecting the parameters to be optimized of the composite patch repair structure; the parameters to be optimized include the patch radius, the patch ply, and the relative thickness of the adhesive layer; A model construction module for establishing a multi-objective optimization mathematical model with static strength, fatigue life, patch volume, and the total thickness of the patch and the adhesive layer as objective functions; A parameter optimization module for solving the multi-objective optimization mathematical model by using a multi-objective genetic algorithm to obtain the optimal solution of the parameters to be optimized.

9. An electronic device, characterized in that, Including: A memory for storing computer programs; A processor, connected to the memory, for retrieving and executing the computer program to implement the parameter optimization method of the composite patch repair structure as described in any one of claims 1-7.

Images