Welded beam design method based on improved badger algorithm

By mapping the honey badger algorithm to a scale-free network and dividing it into elite and non-elite populations, the improved honey badger algorithm improves search capabilities and accuracy in welding beam design, solves local optimal problems, and reduces design costs.

CN119026218APending Publication Date: 2024-11-26JIANGSU UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411133863.6
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-19
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The honey badger algorithm has problems such as low search accuracy, poor search ability and easy to fall into local optimality in the design problems of welding beams, resulting in high design costs.

Method used

Improve the honey badger algorithm by mapping honey badger populations into scaleless networks, divided into elite populations and non-elite populations, and adopting different update strategies. Elite populations use random strategies, and non-elite populations use the best neighbor strategy to avoid Early convergence is locally optimal.

Benefits of technology

It significantly improves search capability and accuracy, reduces the cost of welded beam design, avoids local optimal traps, and improves the performance of the algorithm.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119026218A_ABST
    Figure CN119026218A_ABST
Patent Text Reader

Abstract

The invention discloses a welded beam design method based on an improved badger algorithm. The welded beam design method comprises the following steps: S1, setting parameters in the improved badger algorithm based on a welded beam design problem; s2, generating an initial position of the badger population; s3, generating a scale-free network of which the scale is NP; s4, calculating the fitness value of each badger position; s5, determining three specific coefficients of the badger algorithm; s6, sorting the populations in an ascending order according to the fitness; s7, sorting the individuals of the badger algorithm according to the fitness, and mapping the individuals to a scale-free network of the same scale; s8, realizing a new mining strategy and a new honey strategy of the badgers by utilizing the topological structure of the population; s9, judging whether a termination condition is met or not, if not, skipping to S4, and otherwise, entering the next step; and S10, outputting the optimal position of the badger. The method has higher search capability and precision, avoids premature convergence to local optimum, and remarkably reduces the cost of welding beam design.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of engineering optimization, and in particular to a welding beam design method based on an improved honey badger algorithm. Background Art

[0002] Engineering optimization problems involve finding the best solution from a set of feasible solutions. These problems are very common in many fields such as mathematics, computer science and engineering. Traditionally, researchers have used mathematical analysis-based methods or gradient-based methods to solve optimization problems with well-defined mathematical forms and smooth, continuous characteristics. However, with the increasing complexity of real-world problems, especially engineering optimization problems, traditional optimization algorithms face major challenges, including being trapped in local optimality and requiring high computational costs. The welded beam design problem is a minimization problem among many engineering optimization problems, which is characterized by nonlinearity, a large number of local optimal points and multiple constraints. At the current research stage, the algorithms that have been used to solve this problem include: particle swarm optimization (PSO), differential evolution algorithm (DE), grey wolf algorithm (GWO), sparrow algorithm (SSA), dung beetle optimization algorithm (DBO) and cuckoo search algorithm (CSA), but the design solutions obtained by the optimization of the above algorithms need to be further optimized.

[0003] The Honey Badger Algorithm (HBA) is a swarm intelligence algorithm proposed by Hashim et al. in 2022. It shows great potential due to its balanced exploration and exploitation capabilities and fast convergence speed. The Honey Badger Algorithm simulates the foraging behavior of honey badgers, involving intensive local search and extensive global search, which enables the algorithm to effectively explore the search space and find the optimal solution. However, there is still room for improvement in search ability, convergence accuracy, and avoiding local optimality. These shortcomings make the Honey Badger Algorithm challenging to solve the welded beam design problem.

[0004] Therefore, it is necessary to improve the honey badger algorithm in order to maintain a better balance between exploration and exploitation, enhance the diversity of the honey badger population, and significantly improve the performance of the honey badger algorithm in solving the welding beam design problem. Summary of the invention

[0005] In order to solve the above technical problems, the present invention proposes a welding beam design method based on an improved honey badger algorithm. In the method, it is intended to solve the problems of low search accuracy, poor search ability and easy to fall into local optimum in the honey badger algorithm in the welding beam design problem. The improved honey badger algorithm sorts the individuals of the honey badger algorithm according to fitness and maps them to a scale-free network of the same size. Different from the original honey badger algorithm that updates the population with the best position as a reference, the improved algorithm first divides the population mapped to the scale-free network into two parts: elite population and non-elite population; the reference position of the elite population is a randomly selected neighbor, while the reference position of the non-elite population is the best neighbor. Experimental results show that the improved honey badger algorithm has higher search ability and accuracy when solving the welding beam design problem, avoids premature convergence to the local optimum, and significantly reduces the cost of welding beam design.

[0006] In order to achieve the above object, the technical solution of the present invention is as follows:

[0007] The welding beam design method based on the improved honey badger algorithm includes the following steps:

[0008] Step S1: Setting parameters in the improved honey badger algorithm based on the welding beam design problem, wherein the parameters include the population size NP, the problem dimension D, the maximum number of iterations T max , search the upper bound UB and search the lower bound LB; the honey badger position is represented by a D-dimensional vector X i =(x1,x2,…,x D )express;

[0009] Step S2: Initialize the population stage to generate the initial position of the honey badger population;

[0010] Step S3: Generate a scale-free network of size NP;

[0011] Step S4: Population evaluation phase, calculate the fitness value f for each honey badger position i , i = 1, 2, ..., NP;

[0012] Step S5: Specific coefficient calculation stage, determining three specific coefficients of the honey badger algorithm, the specific coefficients including odor intensity coefficient, density coefficient and direction sign coefficient;

[0013] Step S6: In the population sorting stage, the population is sorted in ascending order according to fitness and numbered from 1 to NP;

[0014] Step S7: In the population topology construction phase, the positions of honey badgers in the population are matched one by one with the nodes in the scale-free network to form a population structure with a scale-free topology structure;

[0015] Step S8: In the population update phase, the topological structure of the population is used to implement the new mining strategy and new honey strategy of the honey badger;

[0016] Step S9: Determine whether the termination condition is met, if not, jump to step S4, otherwise go to the next step;

[0017] Step S10: Output the best honey badger position.

[0018] Preferably, in step S2, the initial position of the honey badger population is generated randomly, and the position generation formula of the honey badger is as follows:

[0019] x j =LB j +r×(UB j -LB j ),j=1,2,…D

[0020] Among them, LB and UB are the lower bound and upper bound of the problem to be optimized, respectively, and r represents a random number between 0 and 1.

[0021] Preferably, in step S3, the step of generating a scale-free network with a size of NP is as follows:

[0022] Step S3.1: construct a fully connected network with M nodes;

[0023] Step S3.2: Calculate the probability values ​​of all nodes in the current network, where the probability of node i is P i Calculated as P i =d(i) / ∑ j d(i);

[0024] Step S3.3: Add a new node to the network, with P i The probability of choosing its attachment point;

[0025] Step S3.4: Repeat steps S3.2 and S3.3 until the NPth node is added to the network.

[0026] Preferably, the D-dimensional vectors correspond to the weld width x1, weld length x2, beam section height x3 and beam section width x4 in the welded beam design problem, respectively, where x1, x4∈[0.1, 2], x2, x3∈[0.1, 10], i.e. UB=[2, 10, 10, 2], LB=[0.1, 0.1, 0.1, 0.1],

[0027] Determine the seven constraints as follows:

[0028] Shear stress τ constraint of the beam: G1 = τ(X)-τ max ≤0;

[0029] Bending stress σ constraint of beam: G2=σ(X)-σ max ≤0;

[0030] Beam end deviation σ constraint: G3 = δ(X)-δ max ≤0;

[0031] The weld width is less than the beam section width constraint: G4 = x1-x4≤0;

[0032] The moment P of the beam is less than the adjacent load (P c (X)) Constraint: G5 = PP c (X)≤0;

[0033] Weld width lower limit constraint: G5 = 0.125-x1≤0;

[0034] The total cost upper limit constraint excluding welding labor cost is:

[0035] in, τ”=MRJ -1 , M=P(L+0.5x2), τ max =13600psi,σ max =30000psi,δ max =0.25in; P=6000lb, L=14in, E=30×10 6 psi,G=12×10 6 psi,

[0036] For each i-th constraint Gi, the penalty function is:

[0037] P(Gi)=[max(0,i)] 2 *10 10 ,

[0038] The evaluation function f(X) is:

[0039]

[0040] Preferably, the step S7 specifically includes the following steps:

[0041] The nodes in a scale-free network are numbered from 1 to NP according to the order in which they join the network;

[0042] The fitness values ​​corresponding to the honey badger positions in step 4 are sorted and matched one-to-one with the nodes with the same numbers in the scale-free network to form an adjacency structure with a scale-free topology for the honey badger positions in the population.

[0043] Preferably, the calculation formula of the odor intensity coefficient I is as follows:

[0044]

[0045] Among them, r2 is a random number between 0 and 1, S represents the concentration of the prey, and d i represents the distance between the prey and the i-th honey badger;

[0046] The calculation formula of the density coefficient α is as follows:

[0047]

[0048] Among them, K is a constant, T represents the current number of iterations;

[0049] The calculation formula of the direction sign coefficient F is as follows:

[0050]

[0051] Where r3 is a random number between 0 and 1.

[0052] Preferably, in step S8, the first M honey badger positions in the honey badger group are divided into an elite group, and the honey badger positions corresponding to numbers M+1 to NP are divided into a non-elite group. A roulette wheel is used to randomly select a neighbor for the honey badger neighbors of the elite group, while a neighbor with the best fitness is selected for the non-elite group; one of the new mining strategy and the new honey strategy is executed with a probability of 50%, wherein,

[0053] The calculation formula of the new mining strategy is as follows:

[0054]

[0055] Among them, X rn Indicates the current position X i A randomly selected neighbor, X bn is the current position X i The neighbor with the best fitness among the neighbors, r4, r5 and r6 are all random numbers between 0 and 1, and β is a constant;

[0056] The new honey strategy adds an offset item of the search area to optimize the search area. The calculation formula of the new honey strategy is as follows:

[0057]

[0058] where μ is the scaling factor, Represents the center of the first M honey badgers in the population, r7 and r8 are random numbers between 0 and 1, Represents the shift of the overall search area toward the collective center of the elite group.

[0059] Based on the above technical solution, the beneficial effects of the present invention are:

[0060] 1) The present invention introduces a scale-free network topology into the honey badger algorithm population, thereby enhancing the information exchange between honey badger locations and improving the search capability of the honey badger algorithm;

[0061] 2) The present invention divides the population into two sub-populations, elite and non-elite, according to the topological structure, and adopts different update strategies for different sub-populations. For example, the random strategy is used in the elite population to improve the population diversity and avoid premature convergence to the local optimum; the best neighbor strategy is used in the non-elite population to enhance the non-elite population to learn useful information from topological neighbors, thereby improving the search efficiency and speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 is a schematic flow chart of a welding beam design method based on an improved honey badger algorithm in an embodiment;

[0063] Figure 2 is a schematic diagram of a welding beam in a welding beam design method based on an improved honey badger algorithm in an embodiment;

[0064] Figure 3 is a schematic diagram of a search area offset of an improved honey badger algorithm in a welding beam design method based on an improved honey badger algorithm in an embodiment;

[0065] Figure 4 is a convergence comparison diagram of the improved honey badger algorithm and other existing algorithms in optimizing welded beams in an embodiment;

[0066] Figure 5 The figure is a box line comparison diagram of the improved honey badger algorithm and other existing algorithms in optimizing welded beams in an embodiment. DETAILED DESCRIPTION

[0067] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0068] like Figure 1 As shown, this embodiment provides a welding beam design method based on an improved honey badger algorithm, comprising the following steps:

[0069] Step S1: Parameter setting stage, set the algorithm parameters, including population size NP, problem dimension D, maximum number of iterations T max , search for upper and lower bounds UB and LB; the schematic diagram of the welded beam design problem is as follows Figure 2 As shown in the figure, there are 4 parameters to be optimized, namely: weld width (h), weld length (l), beam section height (t) and beam section width (b). Therefore, the problem dimension is 4. Here, the population size NP is set to 50, and the maximum number of iterations is T.max is 500. The design goal of this problem is to reduce the manufacturing cost as much as possible under specific constraints, which corresponds to a minimization problem. In the welding beam design, the position of the honey badger is defined as the vector X i =(x1,x2,x3,x4), where each dimension corresponds to h, l, t and b respectively. The boundaries of each dimension are set as: x1,x4∈[0.1,2], x2,x3∈[0.1,10], i.e. UB=[2,10,10,2], LB=[0.1,0.1,0.1,0.1].

[0070] Step S2: Initialize the population stage. The initial position of the honey badger population is generated randomly. The honey badger position generation formula is as follows:

[0071] x j =LB j +r×(UB j -LB j ),j=1,2,…D

[0072] Among them, LB and UB are the lower bound and upper bound of the problem to be optimized, respectively, and r represents a random number between 0 and 1.

[0073] Step S3: Generate a scale-free network of size NP:

[0074] The values ​​of M and NP are 3 and 50 respectively. The steps to generate a scale-free network of size NP are as follows:

[0075] Step S3.1: construct a fully connected network with M nodes;

[0076] Step S3.2: Calculate the probability values ​​of all nodes in the current network, where the probability Pi of node i is calculated as P i =d(i) / ∑ j d(i);

[0077] Step S3.3: Add a new node to the network and select its attachment point with probability Pi;

[0078] Step S3.4: Repeat steps S3.2 and S3.3 until the NPth node is added to the network.

[0079] Step S4: Population evaluation phase, calculate the fitness value f for each honey badger position i , i = 1, 2, ..., NP:

[0080] Before population evaluation, the objective function of the welded beam needs to be defined. The cost function is as follows:

[0081]

[0082] The second step is to determine the constraints, which involve a total of 7 constraints:

[0083] Shear stress (τ) constraint of the beam: G1 = τ(X) - τ max ≤0;

[0084] Bending stress (σ) constraint of beam: G2 = σ(X)-σ max ≤0;

[0085] Beam end deviation (σ) constraint: G3 = δ(X)-δ max ≤0;

[0086] The weld width is less than the beam section width constraint: G4 = x1-x4≤0;

[0087] The moment of the beam (P) is less than the adjacent load (P c (X)) Constraint: G5 = PP c (X)≤0;

[0088] Weld width lower limit constraint: G5 = 0.125-x1≤0;

[0089] The total cost upper limit constraint excluding welding labor cost is:

[0090] in, τ”=MRJ -1 , M=P(L+0.5x2), τ max =13600psi,σ max =30000psi,δ max =0.25in; P=6000lb, L=14in, E=30×10 6 psi,G=12×10 6 psi.

[0091] Since the welding beam design problem is a constrained problem, the target cannot be optimized directly. In this specific implementation, the final evaluation function is constructed by a penalty function. For each i-th constraint Gi, the penalty function is:

[0092] P(Gi)=[max(0,i)] 2 *10 10 ,

[0093] The final evaluation function is:

[0094]

[0095] Step S5: Specific coefficient calculation stage, determine the three specific coefficients of the honey badger algorithm, including the odor intensity coefficient (I), density coefficient (α) and direction sign coefficient (F), the calculation formula is as follows:

[0096] Odor intensity coefficient:

[0097]

[0098] Among them, r2 is a random number between 0 and 1, S represents the concentration of the prey, and d i Represents the distance between the prey and the i-th honey badger.

[0099] Density coefficient:

[0100]

[0101] Where K is a constant (usually 2) and T represents the current iteration number.

[0102] Direction sign coefficient:

[0103]

[0104] Where r3 is a random number between 0 and 1.

[0105] Step S6: In the population sorting stage, the population is sorted in ascending order according to fitness and numbered from 1 to NP; the welding beam design problem is a cost minimization problem, so the population is sorted from small to large.

[0106] Step S7: In the population topology construction phase, individuals in the population are matched one-to-one with nodes in the scale-free network to form a population structure with a scale-free topology:

[0107] The nodes in a scale-free network are numbered from 1 to NP according to the order in which they join the network; then, the numbered individuals in the population are matched one-to-one with the nodes with the same number, thus forming an adjacency structure in which the individuals in the population have a scale-free topology.

[0108] Step S8: In the population update phase, the topological structure of the population is used to implement the new mining strategy and new honey strategy of the honey badger:

[0109] The new mining strategy and the new honey strategy execute one of the strategies with a 50% chance. In the new mining strategy, the update of the honey badger position no longer refers to the best honey badger position in the population, but is determined based on the honey badger neighbors that form a scale-free network topology. Before determining the honey badger neighbors, the honey badger population is divided into two parts: the elite group and the non-elite group. The elite group corresponds to the first M honey badger positions, which have better fitness values ​​and relatively high degrees. The non-elite group corresponds to the honey badger positions numbered from M+1 to NP. The fitness of these honey badger positions is relatively poor, and it is more necessary to learn the good location information of the elite group. For the selection of honey badger neighbors of the elite group, a roulette wheel method is used to randomly select a neighbor, while the non-elite group selects the neighbor with the best fitness. The random selection strategy between elite groups can share beneficial information between strong elite groups while maintaining a certain population diversity. The selection of the best neighbors for the non-elite group can accelerate the convergence process, enhance the search ability, and maintain population diversity. The calculation formula of the new mining strategy is as follows:

[0110]

[0111] Among them, X rn Indicates the current position X i A randomly selected neighbor, X bn is the current position X i The neighbor with the best fitness among the neighbors of . r4, r5 and r6 are all random numbers between 0 and 1. β is a constant with a value of 6.

[0112] The neighbor selection method in the new honey strategy is the same as that in the mining strategy. However, the offset term of the search area is added to the separation strategy formula of the original honey badger algorithm to optimize the search area. The calculation formula of the new honey strategy is as follows:

[0113]

[0114] Where μ is the scaling factor and its value is 0.2. Represents the center of the first M honey badgers in the population. r7 and r8 are random numbers between 0 and 1. represents the offset of the search area to the geometric center of the elite group. See Figure 3 .

[0115] Step S9: Determine whether the termination condition is met, if not, jump to step S4, otherwise go to the next step;

[0116] Step S10: Output the best honey badger position.

[0117] It should be understood that, although the various steps in the above-mentioned flow chart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the above-mentioned flow chart may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0118] experiment

[0119] This example uses the improved honey badger algorithm (SFHBA) to design welded beams, and compares the design effects with the honey badger algorithm (HBA), particle swarm algorithm (PSO), differential evolution algorithm (DE), gray wolf algorithm (GWO), sparrow algorithm (SSA), dung beetle optimization algorithm (DBO) and cuckoo search algorithm (CSA). The population size of each algorithm is NP = 50, and it is repeated 30 times. In order to make a fair comparison and avoid some algorithms having more function evaluation times in one iteration, the stopping condition of the algorithm is no longer T max , but set the maximum number of function evaluations FE max = 10000. The statistical results of each algorithm on the welded beam design problem are shown in Table 1.

[0120] Table 1

[0121]

[0122] As shown in Table 1, the improved honey badger algorithm (SFHBA) is not only better than the honey badger algorithm (HBA) in terms of the five indicators of worst value (Worst), best value (Best), standard deviation (Std), mean (Mean) and median (median), but also better than other compared algorithms (particle swarm algorithm (PSO), differential evolution algorithm (DE), grey wolf algorithm (GWO), sparrow algorithm (SSA), dung beetle optimization algorithm (DBO) and cuckoo search algorithm (CSA)). Table 2 shows the optimal parameter combination and corresponding welding beam cost found by each algorithm in 30 runs. In order to further illustrate the superiority of the scheme proposed in the present invention, Figure 4 The average convergence curve of 30 experiments for each algorithm is provided. Figure 5 The box plots of the cost of 30 welded beams for each algorithm are shown. Figure 4It can be seen that the convergence speed and final welding beam cost of the improved honey badger algorithm (SFHBA) are significantly better than those of the honey badger algorithm (HBA); compared with other algorithms, the improved honey badger algorithm (SFHBA) still maintains its advantage in convergence speed, and the final welding beam cost is better than all compared algorithms. Figure 5 Among them, the boxes of the improved honey badger algorithm (SFHBA), particle swarm algorithm (PSO) and honey badger algorithm (HBA) are obviously lower, which shows that they have better search performance in the welding beam design problem. Figure 5 It can also be seen that the improved honey badger algorithm (SFHBA) not only has the narrowest interquartile range (IQR), but also has the least outliers, which shows that the improved honey badger algorithm (SFHBA) has extremely strong stability and robustness in the optimization problem of welded beams.

[0123] Table 2

[0124]

[0125] Based on the above analysis, it can be concluded that the improved honey badger algorithm (SFHBA) not only greatly reduces the cost of welding beams compared with the honey badger algorithm (HBA), but also has great advantages over other common algorithms.

[0126] The above description is only a preferred implementation of the welding beam design method based on the improved honey badger algorithm disclosed in the present invention, and is not intended to limit the protection scope of the embodiments of this specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the embodiments of this specification shall be included in the protection scope of the embodiments of this specification.

[0127] The above are only preferred embodiments of the embodiments of the present application and are not intended to limit the embodiments of the present application. For those skilled in the art, the embodiments of the present application may have various modifications and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the embodiments of the present application shall be included in the protection scope of the embodiments of the present application.

Claims

1. A welding beam design method based on an improved honey badger algorithm, characterized in that: The steps include: Step S1: Setting parameters in the improved honey badger algorithm based on the welding beam design problem, wherein the parameters include the population size NP, the problem dimension D, the maximum number of iterations T max , search the upper bound UB and search the lower bound LB; the honey badger position is represented by a D-dimensional vector X i =(x1,x2,…,x D )express; Step S2: Initialize the population stage to generate the initial position of the honey badger population; Step S3: Generate a scale-free network of size NP; Step S4: Population evaluation phase, calculate the fitness value f for each honey badger position i , i = 1, 2, ..., NP; Step S5: Specific coefficient calculation stage, determining three specific coefficients of the honey badger algorithm, the specific coefficients including odor intensity coefficient, density coefficient and direction sign coefficient; Step S6: In the population sorting stage, the population is sorted in ascending order according to fitness and numbered from 1 to NP; Step S7: In the population topology construction phase, the positions of honey badgers in the population are matched one by one with the nodes in the scale-free network to form a population structure with a scale-free topology structure; Step S8: In the population update phase, the topological structure of the population is used to implement the new mining strategy and new honey strategy of the honey badger; Step S9: Determine whether the termination condition is met, if not, jump to step S4, otherwise go to the next step; Step S10: Output the best honey badger position.

2. The welding beam design method based on the improved honey badger algorithm according to claim 1 is characterized in that: In step S2, the initial position of the honey badger population is generated randomly, and the position generation formula of the honey badger is as follows: x j =LB j +r×(UB j -LB j ),j=1,2,…D Among them, LB and UB are the lower bound and upper bound of the problem to be optimized, respectively, and r represents a random number between 0 and 1.

3. The welding beam design method based on the improved honey badger algorithm according to claim 1 is characterized in that: In step S3, the steps of generating a scale-free network with a size of NP are as follows: Step S3.1: construct a fully connected network with M nodes; Step S3.2: Calculate the probability values ​​of all nodes in the current network, where the probability of node i is P i Calculated as P i =d(i) / ∑ j d(i); Step S3.3: Add a new node to the network, with P i The probability of choosing its attachment point; Step S3.4: Repeat steps S3.2 and S3.3 until the NPth node is added to the network.

4. The welding beam design method based on the improved honey badger algorithm according to claim 3 is characterized in that: The D-dimensional vectors correspond to the weld width x1, weld length x2, beam section height x3 and beam section width x4 in the welded beam design problem, where x1, x4∈[0.1, 2], x2, x3∈[0.1, 10], i.e. UB=[2, 10, 10, 2], LB=[0.1, 0.1, 0.1, 0.1], Determine the seven constraints as follows: Shear stress τ constraint of the beam: G1 = τ(X)-τ max ≤0; Bending stress σ constraint of beam: G2=σ(X)-σ max ≤0; Beam end deviation σ constraint: G3 = δ(X)-δ max ≤0; The weld width is less than the beam section width constraint: G4 = x1-x4≤0; The moment P of the beam is less than the adjacent load (P c (X)) Constraint: G5 = PP c (X)≤0; Weld width lower limit constraint: G5 = 0.125-x1≤0; The total cost upper limit constraint excluding welding labor cost is: Among them, t max =13600psi,s max =30000psi,δ max =0.25in;P=6000lb,L=14in,E=30×10 6 psi,G=12×10 6 psi, For each i-th constraint Gi, the penalty function is: P(Gi)=[max(0,i)] 2 *10 10 , The evaluation function f(X) is:

5. The welding beam design method based on the improved honey badger algorithm according to claim 4 is characterized in that: The step S7 specifically includes the following steps: The nodes in a scale-free network are numbered from 1 to NP according to the order in which they join the network; The fitness values ​​corresponding to the honey badger positions in step 4 are sorted and matched one-to-one with the nodes with the same numbers in the scale-free network to form an adjacency structure with a scale-free topology for the honey badger positions in the population.

6. The welding beam design method based on the improved honey badger algorithm according to claim 1 is characterized in that: The calculation formula of the odor intensity coefficient I is as follows: Among them, r2 is a random number between 0 and 1, S represents the concentration of the prey, and d i represents the distance between the prey and the i-th honey badger; The calculation formula of the density coefficient α is as follows: Among them, K is a constant, T represents the current number of iterations; The calculation formula of the direction sign coefficient F is as follows: Where r3 is a random number between 0 and 1.

7. The welding beam design method based on the improved honey badger algorithm according to claim 6 is characterized in that: In the step S8, the first M honey badger positions in the honey badger group are divided into an elite group, and the honey badger positions corresponding to numbers M+1 to NP are divided into a non-elite group. A roulette wheel is used to randomly select a neighbor for the honey badger neighbors of the elite group, and a neighbor with the best fitness is selected for the non-elite group; one of the new mining strategy and the new honey strategy is executed with a probability of 50%, wherein, The calculation formula of the new mining strategy is as follows: Among them, X rn Indicates the current position X i A randomly selected neighbor, X bn is the current position X i The neighbor with the best fitness among the neighbors, r4, r5 and r6 are all random numbers between 0 and 1, and β is a constant; The new honey strategy adds an offset item of the search area to optimize the search area. The calculation formula of the new honey strategy is as follows: where μ is the scaling factor, Represents the center of the first M honey badgers in the population, r7 and r8 are random numbers between 0 and 1, Represents the shift of the overall search area toward the collective center of the elite group.