Lightweight optimization method based on multi-element combined frame

Abstract

The invention relates to the technical field of automobile design, in particular to a multi-element combination-based frame lightweight optimization method, and solves the technical problems that in existing truss-type frame lightweight design, model establishment is complex, and a traditional Harris eagle algorithm is slow in convergence speed and low in precision. Analyzing strength and rigidity performance under bending and torsion working conditions and acquiring inherent frequency; then, response surface optimization is carried out by taking the I-beam as an object, and three groups of candidate solutions are obtained by taking the maximum deformation and the maximum stress as constraints and the mass as a target; optimizing a traditional Harlisia eagle algorithm to obtain an IHHO algorithm, optimizing the cross section size of the I-shaped beam by using the IHHO algorithm to obtain three groups of optimal solutions, and determining final parameters of the beam cross section after comparison; and optimizing other beams of the frame according to the steps, and finally verifying the strength, rigidity and fatigue life of the optimized frame. The method has a good application prospect in the field of design of truss type frames with lightweight requirements.

Description

Technical Field

[0001] This invention relates to the field of automotive design technology, and more specifically to a method for lightweight optimization of multi-component combined chassis. Background Technology

[0002] Extensive research has been conducted by scholars both domestically and internationally in the field of chassis structural optimization and lightweighting. Currently, most studies achieve lightweighting goals by using numerical simulation analysis and parameter optimization to address the specific working conditions of the chassis. Some studies, based on bionics theory and combined with traditional methods such as topology optimization, have achieved significant lightweight design results. In recent years, lightweight design has primarily focused on intelligent algorithms and multi-objective optimization. By establishing an overall mathematical model and solving it using optimization algorithms, new methods for lightweight design have been provided. Furthermore, some scholars have constructed numerous frame cross-sectional parameter models, achieving component lightweighting from a multi-objective optimization perspective, thereby achieving overall chassis lightweighting and laying the foundation for future development of simplified optimization methods.

[0003] However, there is still relatively little research at home and abroad on using multi-condition numerical simulation combined with optimization algorithms for lightweighting of truss-type vehicle frames. Most of the research objects are traditional passenger car body-in-white, and there are problems such as complex model building and cumbersome optimization process. Summary of the Invention

[0004] To address the technical problems of complex model building and cumbersome optimization processes in the lightweight design of existing truss-type vehicle frames (such as snowplow frames), as well as the slow convergence speed, low convergence accuracy, and susceptibility to local optima of the traditional Harris Eagle (HHO) algorithm, this invention provides a multi-element joint vehicle frame lightweight optimization method, including:

[0005] S1: Based on the working environment and relevant regulations of the snowplow, determine the frame parameters and materials, select Q235 steel as the frame material, and establish a three-dimensional model of the frame;

[0006] S2: Based on the three-dimensional model of the frame, establish a finite element model of the frame and complete the preprocessing. Analyze the strength and stiffness performance under bending and torsional conditions. Obtain the first 6 natural frequencies of the frame through modal analysis. After confirming that the frame meets the preset requirements, proceed to the optimization process.

[0007] S3: Optimize the cross-sectional dimensions of the vehicle frame beam solid element, carry out response surface optimization with the I-beam as the research object, determine the design variables and response values ​​through sensitivity analysis, establish and verify the response surface model, take the maximum deformation and maximum stress as constraints, take the minimization of mass as the optimization objective, and obtain three sets of candidate solutions;

[0008] S4: An improved Harris Eagle algorithm is proposed based on Bernoulli chaotic mapping, log convergence factor adjustment strategy and position update strategy of vulture search algorithm;

[0009] S5: Optimize the cross-sectional dimensions of the I-beam using the improved Harris Eagle algorithm. With maximum deformation and maximum stress as constraints and mass minimization as the objective function, establish a mathematical model for the optimization of the I-beam solid element and obtain three sets of optimal solutions.

[0010] S6: Compare the three sets of candidate solutions with the three sets of optimal solutions, and determine the final parameters of the beam section based on the processing and manufacturing process;

[0011] S7: Optimize the remaining beam solid elements of the frame according to the methods of steps S3, S5 and S6 to obtain the optimized frame model;

[0012] S8: Verify the strength, stiffness, and fatigue life of the optimized chassis model.

[0013] Furthermore, the relevant parameters of the frame are: 600mm width, 1800mm length, and the main structure of the crossbeam is selected from at least one of the following: channel beam, square beam, I-beam, and tubular beam.

[0014] Furthermore, the preprocessing includes: setting the finite element mesh unit size to 5mm, adding standard Earth gravity in the Y-axis direction, selecting a dynamic coefficient of 2.5, and simultaneously applying constraints for bending and torsional conditions; wherein, under the bending condition, the left front, right front, left rear, and right rear suspensions are all constrained by U... y U z Directional degree of freedom; left front suspension under torsional conditions with U-shaped force applied. y =2000N load, remaining suspension constraints U x U y U z Degrees of freedom of direction.

[0015] Furthermore, in S2, the analysis of strength and stiffness performance under bending and torsional conditions is as follows: based on the length of the transverse and longitudinal beams of the frame, the allowable deflection is calculated according to the allowable deflection formula, and the maximum deformation of the main beam is compared with the allowable deflection to confirm that the stiffness requirements are met; based on the allowable stress of Q235 steel of 156.7MPa, the maximum stress value of the frame is compared to confirm that the strength requirements are met and there is strength redundancy.

[0016] The allowable deflection formula is:

[0017]

[0018] Where [f] is the allowable deflection, L is the length of the beam, and α is the allowable deflection coefficient, which is taken as 500.

[0019] Furthermore, in S2, the modal analysis is specifically free modal analysis, extracting the first 6 modal parameters and modal shape cloud diagrams. The first 6 modal frequencies of the frame avoid the resonance zone range where the motor operating frequency of 11.70Hz and the ground feedback excitation frequency of 5.30Hz are located.

[0020] Furthermore, in S3, the response surface optimization specifically involves: generating 25 sets of design points using a central composite experimental design method; selecting four design variables that significantly affect the response value through sensitivity analysis; constructing a response surface model using the standard second-order response surface method; verifying through goodness-of-fit curves that the model's predicted values ​​are nearly equal to the actual observed values, and that the coefficient of determination R² for each response surface is greater than 0.97; taking quality as the optimization objective, with constraints of maximum deformation ≤ 0.25134 mm and maximum equivalent stress ≤ 110.4 MPa; and optimizing the design variables to obtain the three sets of candidate solutions.

[0021] Furthermore, in S4, the improved Harris Hawk algorithm specifically includes:

[0022] S41: Population initialization is performed based on the Bernoulli chaotic mapping, with the initial population size set to 100. The expression is as follows:

[0023]

[0024] Where β is the adjustment coefficient, Z k The kth individual in the population generates a more evenly distributed and more ergodic population.

[0025] S42: Introducing a logarithmic nonlinear convergence factor to optimize the escape energy E improves the convergence accuracy and speed of the algorithm. Its expression is:

[0026]

[0027] Where E is the improved escape energy, iter is the maximum number of iterations, and t is the current iteration number;

[0028] S43: Fit the position update strategy of the vulture search algorithm. When q≥0.5, a large-area search strategy is adopted, and when q<0.5, a spiral search strategy is adopted to improve the optimization ability.

[0029] Furthermore, in S5, the optimized mathematical model for the I-beam solid element is as follows:

[0030] A = 2P1P4 + (P2 - 2P4)P 3.

[0031] Among them, P1, P2, P3, and P4 are four cross-sectional dimension design variables, and A is a mass-related calculated value.

[0032] Furthermore, in S6, determining the final parameters of the beam section specifically involves: comparing the candidate solution obtained by response surface optimization with the optimal solution obtained by the improved Harris Eagle algorithm, verifying the feasibility of the algorithm by combining three sets of iterative curves, ensuring that the final parameters meet the processing and manufacturing requirements, and that the difference between the design variables P2, P3, and P4 is ≤0.5mm.

[0033] Furthermore, in S8, the verification of strength, stiffness, and fatigue life specifically involves: employing high-cycle fatigue calculations, combining Goodman's theory to correct for the influence of mean stress, and determining based on the material's SN curve that when the frame's fatigue cycle count reaches 10... 6 When the order of magnitude is reached, the material fatigue strength is ≤225MPa, which meets the design requirements for infinite life, and the overall structural fatigue strength, excluding welded joints, is ≥235MPa.

[0034] The beneficial effects of this invention are as follows: This invention utilizes a multi-factor joint optimization scheme combining finite element analysis, response surface optimization, and the improved Harris Eagle algorithm (IHHO). This achieves significant weight reduction of the chassis while ensuring that the maximum deformation of the optimized chassis is less than the allowable deflection, the maximum stress does not exceed the allowable range of the material, the first six modal frequencies avoid the resonance band, and the design requirements for high-cycle fatigue infinite life are met. Simultaneously, the IHHO algorithm effectively solves the problems of slow convergence, low accuracy, and susceptibility to local optima in the traditional HHO algorithm. The entire optimization process features convenient model establishment, clear optimization logic, and accurate and reliable results. It balances the weight reduction objective with the comprehensive performance of chassis strength, stiffness, modal characteristics, and fatigue life, possessing strong engineering practical value. Attached Figure Description

[0035] Figure 1 Here is a flowchart of a lightweight optimization method based on a multi-component combined frame;

[0036] Figure 2 This is a 3D model of the overall vehicle frame.

[0037] Figure 3 A finite element mesh model of the vehicle frame;

[0038] Figure 4 This is a diagram showing the load distribution on the vehicle frame.

[0039] Figure 5 Diagram showing the deformation of the vehicle frame under bending conditions;

[0040] Figure 6 Diagram showing the deformation of the vehicle frame under torsional conditions;

[0041] Figure 7 Stress cloud diagram of the vehicle frame under bending conditions;

[0042] Figure 8 Stress cloud diagram of the chassis under torsional conditions;

[0043] Figure 9 The first six modal shapes of the chassis are shown in cloud diagrams.

[0044] Figure 10 For I-beam parameters and model

[0045] Figure 11 Local sensitivity plots showing the response values ​​corresponding to each design variable of the I-beam;

[0046] Figure 12 For response surface model diagram;

[0047] Figure 13 To fit the goodness-of-fit curve;

[0048] Figure 14 A comparison diagram showing the distribution evenness of the population before and after improvement.

[0049] Figure 15 To improve the comparison diagram of the escape energy change process before and after;

[0050] Figure 16 Here is the flowchart for the IHHO algorithm;

[0051] Figure 17 Optimize the beam section iteration curve for IHHO;

[0052] Figure 18 For the SN curve;

[0053] Figure 19 A cloud map showing the fatigue life distribution of the chassis under bending conditions;

[0054] Figure 20 To reverse the fatigue life distribution cloud map of the chassis under working conditions. Detailed Implementation

[0055] The technical solution of the present invention will be further described below with reference to embodiments, but it is not limited thereto. Any modifications or equivalent substitutions to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention should be covered within the protection scope of the present invention. In the following embodiments, process equipment or devices not specifically specified are all conventional equipment or devices in the art. Unless specifically specified, the technical means used in the embodiments of the present invention are all conventional means well known to those skilled in the art.

[0056] Example 1, combined with Figure 1 This embodiment describes a lightweight optimization method based on a multi-component composite frame, comprising:

[0057] S1: Based on the working environment and relevant regulations of the snowplow, determine the frame parameters and materials, select Q235 steel as the frame material, and establish a three-dimensional model of the frame;

[0058] S2: Based on the three-dimensional model of the frame, establish a finite element model of the frame and complete the preprocessing. Analyze the strength and stiffness performance under bending and torsional conditions. Obtain the first 6 natural frequencies of the frame through modal analysis. After confirming that the frame meets the preset requirements, proceed to the optimization process.

[0059] S3: Optimize the cross-sectional dimensions of the vehicle frame beam solid element, carry out response surface optimization with the I-beam as the research object, determine the design variables and response values ​​through sensitivity analysis, establish and verify the response surface model, take the maximum deformation and maximum stress as constraints, take the minimization of mass as the optimization objective, and obtain three sets of candidate solutions;

[0060] S4: An improved Harris Eagle algorithm is proposed based on Bernoulli chaotic mapping, log convergence factor adjustment strategy and position update strategy of vulture search algorithm;

[0061] S5: Optimize the cross-sectional dimensions of the I-beam using the improved Harris Eagle algorithm. With maximum deformation and maximum stress as constraints and mass minimization as the objective function, establish a mathematical model for the optimization of the I-beam solid element and obtain three sets of optimal solutions.

[0062] S6: Compare the three sets of candidate solutions with the three sets of optimal solutions, and determine the final parameters of the beam section based on the processing and manufacturing process;

[0063] S7: Optimize the remaining beam solid elements of the frame according to the methods of steps S3, S5 and S6 to obtain the optimized frame model;

[0064] S8: Verify the strength, stiffness, and fatigue life of the optimized chassis model.

[0065] Specifically, the overall process of this invention is as follows: Figure 1 As shown, the frame parameters were first determined and 3D modeling was completed in Solidworks software to obtain the overall 3D model of the frame (as shown in Figure 2). Then, this 3D model was imported into Ansys software to establish a finite element simulation model of the frame. Bending and torsional conditions were selected to verify the frame stiffness and strength, and the natural frequencies under free modes were used as indicators to evaluate modal performance (Figure 3 shows the finite element mesh model of the frame). In the preprocessing stage of finite element analysis, the mesh unit size was set to 5mm, and standard Earth gravity and a dynamic load factor of 2.5 were added in the negative Y-axis direction. Loads were applied to the corresponding action surfaces according to the frame load parameters shown in Table 1 (frame load distribution is shown in Figure 4), and suspension constraints for bending and torsional conditions were set according to Table 2.

[0066] Table 1

[0067]

[0068] Table 2

[0069]

[0070] The deformation cloud diagrams of the frame under bending (Fig. 5) and torsion (Fig. 6), the stress cloud diagrams of the frame under bending (Fig. 7) and torsion (Fig. 8), and the mode shape cloud diagrams of the first six orders (Fig. 9) were obtained by finite element analysis.

[0071] The allowable deflection for the seven horizontal beams (L=591mm) is 1.182mm, and the allowable deflection for the two longitudinal beams (L=1800mm) is 3.600mm. The maximum deformation of each beam is shown in Table 3.

[0072] Table 3

[0073]

[0074] Strength assessment was based on the allowable stress of 156.7 MPa for Q235 steel. The maximum stress under bending conditions was 83.63 MPa, and the maximum stress under torsion conditions was 142.66 MPa, both meeting the requirements. Modal analysis employed multi-degree-of-freedom structural motion differential equations:

[0075] in, The structural mass matrix, For nodal velocity vectors, Here is the structural damping matrix. Here is the structural stiffness matrix. For nodal displacement vectors, It is an external load function;

[0076] Ignoring damping and external loads, the simplified result is:

[0077]

[0078] The extracted parameters for the first six modalities are shown in Table 4.

[0079] Table 4

[0080]

[0081] Combining the motor frequency formula and the ground excitation frequency formula, the calculated motor frequency is 11.70Hz and the ground excitation frequency is 5.30Hz, ensuring the frame modal frequencies avoid the resonance band. The motor frequency formula is:

[0082]

[0083] The formula for ground excitation frequency is:

[0084]

[0085] Where f1 is the frequency generated by the motor; P is the number of pole pairs of the motor; n is the rotational speed of the motor; f2 is the excitation frequency generated by the ground per revolution of the tire; v is the maximum speed of the vehicle; and D is the tire diameter.

[0086] After verifying that the frame meets the initial performance requirements, response surface optimization is carried out using the I-beam as the research object. The three-dimensional model and cross-sectional dimensions of the I-beam are as follows: Figure 10 As shown, through sensitivity analysis (such as...) Figure 11 As shown, a standard second-order response surface model is constructed, and its expression is:

[0087]

[0088] Where y(x) is the fitting function, x i Let i = 1,2,…,n, a0, a i a ii a ij The coefficients are undetermined; ε is the accuracy error.

[0089] Taking a partial response surface model as an example, such as Figure 12 As shown; export the generated response surface model and goodness-of-fit curve, as follows. Figure 13 As shown; in engineering, the coefficient of determination R² is commonly used to reflect the fitting accuracy of the response surface, and its calculation formula is:

[0090]

[0091] Among them, R 2 As the coefficient of determination, y i For the i-th observation, The average of the observed values. This is the model's predicted value for the i-th observation;

[0092] The determination coefficient R of each response value was obtained by calculation. 2 As shown in Table 5, the determination coefficients for each response value are all greater than 0.97.

[0093] Table 5 shows the coefficient of determination R for each response value. 2

[0094] Table 5

[0095]

[0096] After obtaining three sets of candidate solutions, the HHO algorithm is improved to obtain the IHHO algorithm. Its core improved formulas include Bernoulli chaotic mapping, escape energy formula optimized by log convergence factor, and vulture search position update formula.

[0097] The HHO algorithm consists of three phases: search, judgment, and hunting. It has the ability to perform global search and local exploitation. In the search phase, the HHO algorithm uses an equal chance strategy to simulate the process of a flock of eagles searching for prey. Its expression is:

[0098]

[0099]

[0100] Where t is the iteration number; X(t) and X(t+1) are the current and subsequent positions of the Harris Eagle individual, respectively; X rand (t) represents the randomly selected individual position; X rabbit (t) represents the location of the hare; r n (n=1, 2, 3, 4) represents the probability of the hare escaping, r ≥ 0.5 indicates a successful escape, and r < 0.5 indicates a failed escape; q is a random number between [0, 1]; u b l b X represents the upper and lower bounds of the search space; m (t) represents the average position of the Harris Eagle population;

[0101] During the judgment phase, the HHO algorithm uses the rabbit's escape energy E to implement the conversion, which decreases as the number of iterations increases. The expression for the rabbit's escape energy behavior is:

[0102]

[0103] Where E0 is a random number between [-1, 1], representing the initial energy of the rabbit; t is the number of iterations; iter is the maximum number of iterations;

[0104] During the decay of E, when |E| ≥ 1, it indicates that the probability of the hare escaping the ambush increases; when |E| < 1, it indicates that the probability of the hare escaping the ambush decreases, and the eagle flock enters the hunting phase.

[0105] During the hunting phase, the eagle flock chooses four hunting methods based on E and r: soft encirclement, hard encirclement, soft encirclement with gradual and rapid dive, and hard encirclement with gradual and rapid dive.

[0106] When |E| ≥ 0.5 and r ≥ 0.5, the Harris Eagle uses a soft encirclement hunting strategy to ambush a hare, the expression of which is:

[0107]

[0108]

[0109]

[0110] Where ∆X(t) is the difference between the rabbit's current position and its current position; J is a random number between [0,2], representing the rabbit's jump energy; and r5 is a random number between [0,1].

[0111] When |E| < 0.5 and r ≥ 0.5, the eagle flock uses a hard-swarm hunting strategy to launch a surprise attack, the expression of which is:

[0112] When |E| ≥ 0.5 and r < 0.5, a Levy flight function is introduced to deplete the rabbit's stamina. A soft encirclement strategy is used to prevent the rabbit from escaping. When the rabbit's stamina is exhausted, a surprise attack is launched. Its expression is:

[0113]

[0114]

[0115]

[0116] Where f(~) is the fitness function, D is the dimension of the problem to be solved, S is a 1×D random vector; LF(~) is the Levy flight function; β is a constant of 1.5; u and v are random variables between [0,1].

[0117] When |E| < 0.5 and r < 0.5, the eagle flock will form a hard encirclement and reduce the distance between the eagles and the hare. When the hare is exhausted, it will launch a surprise attack. The expression for this is:

[0118]

[0119]

[0120]

[0121] The HHO optimization algorithm has advantages such as simple principle and few parameter adjustments, but it has also revealed many shortcomings in practical engineering applications, such as slow convergence speed, low convergence accuracy, easy getting trapped in local optima, and difficulty in balancing search and hunting. This paper proposes an improvement based on its algorithm principle to address the above problems.

[0122] Population initialization based on Bernoulli chaotic mapping has the characteristics of uniform distribution, good ergodicity, and high convergence accuracy, which can effectively improve the performance of HHO algorithm. Introducing chaotic mapping can improve the fitness of function, and replacing the conventional uniformly distributed random number generator can yield better results, especially when there are multiple local solutions in the search space, it is easier to find the global optimum.

[0123] To verify the performance improvement effect of Bernoulli chaotic mapping on population initialization, the population size was set to 100. The population was generated using both the Random random number generator and Bernoulli chaotic mapping. The distribution uniformity of the two methods was compared, and the results are as follows: Figure 14 As shown;

[0124] Observations show that the population generated by the Bernoulli chaotic mapping has a more uniform distribution, fewer overlaps, higher search traversal, and wider coverage area.

[0125] The escape energy E is a crucial parameter for determining the hunting method. In practice, the HHO algorithm is prone to getting trapped in local optima in the later stages of computation, leading to decreased accuracy. Therefore, a logarithmic nonlinear convergence factor is introduced to improve convergence accuracy, reduce convergence time, and enhance optimization performance. Its expression is:

[0126] The change in escape energy before and after the improvement was depicted using Matlab's plotting function, verifying the performance improvement of escape energy E during the iteration process. The comparison results are as follows: Figure 15 As shown; through comparison, it can be seen that the improved algorithm has a faster convergence speed in the early stage of the calculation and can enter the convergence state earlier in the later stage of the calculation, thus improving the overall convergence speed and optimization ability of the algorithm.

[0127] A vulture search algorithm is fitted for position updates, and the position update strategy in the search phase of the BES algorithm is combined to improve the optimization ability and convergence accuracy of the HHO algorithm. When the eagle flock searches for hares, q ≥ 0.5 indicates that no target has been found, and a large-area search strategy is adopted. At this time, the large-area search strategy in the BES algorithm is introduced, randomly selecting an area to search within a wide range. By judging the number of hares in the area, an optimal search position is determined; its expression is:

[0128]

[0129] Among them, P best (t) represents the optimal search position in the current iteration, α is the control position update parameter, r is a random number in [0,1], and P m P(t) represents the average position of the eagle flock after the previous iteration, and P(t) represents the position of the t-th Harris eagle. q < 0.5 indicates that the rabbit's location has been found, reducing the search area. At this point, the spiral search strategy from the BES algorithm is introduced to accelerate the search process and find the optimal hunting position. Its expression is as follows:

[0130]

[0131]

[0132]

[0133]

[0134]

[0135]

[0136] Where θ(t) is the polar angle of the spiral equation, r(t) is the polar radius of the spiral equation, α and R are the control parameters with ranges of [0,5] and [0.5,2] respectively, rand is the random variable range (0,1), and x(t) and y(t) are the positions of the Harris Eagle in polar coordinates;

[0137] The expression for updating the position during a search is as follows:

[0138]

[0139] In summary, this paper proposes an improved Harris Hawk algorithm, the IHHO algorithm, based on the optimized HHO algorithm. The algorithm flow is as follows: Figure 16 As shown;

[0140] The frame optimization application uses the four cross-sectional dimensions of the crossbeam as design variables, the maximum deformation and maximum stress as constraints, and the minimum mass as the objective function to establish a mathematical model for the optimization of the I-beam solid element.

[0141] In optimization algorithms, setting the range of design variables is crucial. The expression for the constraint range of design variables is as follows:

[0142]

[0143]

[0144]

[0145]

[0146] The expression for the constraint condition of the maximum deflection of the beam is:

[0147]

[0148]

[0149]

[0150] Among them l maxLet ω be the maximum deflection, ω be the uniformly distributed load, L be the beam length, E be the material's elastic modulus, and I be the beam's moment of inertia. After simplification, we get:

[0151]

[0152] The maximum stress on the frame during use should be less than the allowable stress of the material, expressed as:

[0153]

[0154]

[0155] Where σ is the maximum stress value and M is the bending moment on the beam, after simplification we get:

[0156]

[0157] The expressions for the critical yield stress of each beam are as follows:

[0158]

[0159]

[0160]

[0161] Where σ cr Let be the critical yield stress of the beam, and A be the cross-sectional area of ​​the beam. The maximum working stress of the beam is given by E, which is the elastic modulus of the material. Let be the critical yield load of the beam. Let be the moment of inertia of the beam section. Let L be the cross-sectional shape factor and L be the calculated length of the beam. After simplification, we get:

[0162]

[0163] The objective function is as follows:

[0164]

[0165]

[0166]

[0167] Where M is mass, For the goal of minimizing quality, For material density, Let be the cross-sectional area of ​​the beam. The flange width of the I-beam. The total height of the I-beam. The web thickness of the I-beam. Let be the flange thickness of the I-beam, and min size be the optimization direction. Let X be the quality function, find be the optimization object, i be the index, and X be the index. i The cross-sectional dimensions are defined by `subjectto`, which is a constraint prefix.

[0168] The optimal general solution generated by the algorithm is random. Therefore, three sets of optimal general solutions are generated for reference and compared with the three sets of candidate points generated by response surface optimization in the previous text to obtain the final optimization result.

[0169] The three optimal solutions are shown in Table 6, and the IHHO iteration curves are shown in Table 6. Figure 17 As shown, all three sets of iterative curves gradually approach a certain stable value, indicating that the algorithm converges to a solution or a local optimum, and demonstrating the feasibility of the algorithm in practical applications.

[0170] Table 6

[0171]

[0172] By comparing the candidate solutions with the optimal solution and combining the processing technology to determine the final parameters, and after optimizing the remaining beam solid elements, the strength, stiffness and fatigue life of the frame are verified to ensure that the design requirements are met.

[0173] The optimized frame was tested for strength, stiffness, and fatigue life. The three candidate points obtained by the response surface methodology were compared with the three optimal general solutions obtained by the IHHO algorithm. It was found that the differences between the P2, P3, and P4 values ​​obtained by the two methods were all less than 0.5 mm, but P1 showed a large difference.

[0174] The remaining beam solid elements of the frame are optimized in the same way to obtain the overall optimized structure of the frame. Finite element simulation is then performed on the modified frame to verify the final performance indicators of the frame.

[0175] Table 7 shows a comparison of performance indicators before and after frame optimization.

[0176] Table 7

[0177]

[0178] Fatigue failure is the main failure mode of a structure. Fatigue life analysis of the optimized chassis can better reflect its application value. This invention uses high-cycle fatigue for chassis calculations, and considering the influence of mean stress on the simulation results, Goodman's theory is used for correction. The resulting fatigue life distribution cloud map of the chassis is shown below. Figure 19 and Figure 20As shown, under two specific working conditions, the minimum fatigue life cycles in the frame were 487,411 and 183,066, respectively. These weak points mainly occurred at the connection points between beams. The main reason for this phenomenon is that the beams are fixed together by welding, and the strength of the welded parts is significantly lower than the strength of the beam structure itself. The fatigue life cycles of other parts of the frame all reached 10. 6 Order of magnitude Figure 18 The figure shows the SN curve of the material. When the number of fatigue cycles reaches 10... 6 After several orders of magnitude, the fatigue strength of the material approaches 225 MPa, while the fatigue strength of the frame studied in this paper is approximately 235 MPa, meaning it reaches 1000 cycles. 6 The fatigue strength of the material is orders of magnitude less than that of the frame, so this level can be considered as having an infinite lifespan. Therefore, except for the welded joints, the overall structure of the frame performs well in terms of fatigue strength and meets the design requirements.

[0179] Furthermore, the relevant parameters of the frame are: 600mm width, 1800mm length, and the main structure of the crossbeam is selected from at least one of the following: channel beam, square beam, I-beam, and tubular beam.

[0180] Furthermore, the preprocessing includes: setting the finite element mesh unit size to 5mm, adding standard Earth gravity in the Y-axis direction, selecting a dynamic coefficient of 2.5, and simultaneously applying constraints for bending and torsional conditions; wherein, under the bending condition, the left front, right front, left rear, and right rear suspensions are all constrained by U... y U z Directional degree of freedom; left front suspension under torsional conditions with U-shaped force applied. y =2000N load, remaining suspension constraints U x U y U z Degrees of freedom of direction.

[0181] Furthermore, in S2, the analysis of strength and stiffness performance under bending and torsional conditions is as follows: based on the length of the transverse and longitudinal beams of the frame, the allowable deflection is calculated according to the allowable deflection formula, and the maximum deformation of the main beam is compared with the allowable deflection to confirm that the stiffness requirements are met; based on the allowable stress of Q235 steel of 156.7MPa, the maximum stress value of the frame is compared to confirm that the strength requirements are met and there is strength redundancy.

[0182] The allowable deflection formula is:

[0183]

[0184] Where [f] is the allowable deflection, L is the length of the beam, and α is the allowable deflection coefficient, which is taken as 500.

[0185] Furthermore, in S2, the modal analysis is specifically free modal analysis, extracting the first 6 modal parameters and modal shape cloud diagrams. The first 6 modal frequencies of the frame avoid the resonance zone range where the motor operating frequency of 11.70Hz and the ground feedback excitation frequency of 5.30Hz are located.

[0186] Furthermore, in S3, the response surface optimization specifically involves: generating 25 sets of design points using a central composite experimental design method; selecting four design variables that significantly affect the response value through sensitivity analysis; constructing a response surface model using the standard second-order response surface method; verifying through goodness-of-fit curves that the model's predicted values ​​are nearly equal to the actual observed values, and that the coefficient of determination R² for each response surface is greater than 0.97; taking quality as the optimization objective, with constraints of maximum deformation ≤ 0.25134 mm and maximum equivalent stress ≤ 110.4 MPa; and optimizing the design variables to obtain the three sets of candidate solutions.

[0187] Furthermore, in S4, the improved Harris Hawk algorithm specifically includes:

[0188] S41: Population initialization is performed based on the Bernoulli chaotic mapping, with the initial population size set to 100. The expression is as follows:

[0189]

[0190] Where β is the adjustment coefficient, Z k The kth individual in the population generates a more evenly distributed and more ergodic population.

[0191] S42: Introducing a logarithmic nonlinear convergence factor to optimize the escape energy E improves the convergence accuracy and speed of the algorithm. Its expression is:

[0192]

[0193] Where E is the improved escape energy, iter is the maximum number of iterations, and t is the current iteration number;

[0194] S43: Fit the position update strategy of the vulture search algorithm. When q≥0.5, a large-area search strategy is adopted, and when q<0.5, a spiral search strategy is adopted to improve the optimization ability.

[0195] Furthermore, in S5, the optimized mathematical model for the I-beam solid element is as follows:

[0196] A = 2P1P4 + (P2 - 2P4)P 3.

[0197] Among them, P1, P2, P3, and P4 are four cross-sectional dimension design variables, and A is a mass-related calculated value.

[0198] Furthermore, in S6, determining the final parameters of the beam section specifically involves: comparing the candidate solution obtained by response surface optimization with the optimal solution obtained by the improved Harris Eagle algorithm, verifying the feasibility of the algorithm by combining three sets of iterative curves, ensuring that the final parameters meet the processing and manufacturing requirements, and that the difference between the design variables P2, P3, and P4 is ≤0.5mm.

[0199] Furthermore, in S8, the verification of strength, stiffness, and fatigue life specifically involves: employing high-cycle fatigue calculations, combining Goodman's theory to correct for the influence of mean stress, and determining based on the material's SN curve that when the frame's fatigue cycle count reaches 10... 6 When the order of magnitude is reached, the material fatigue strength is ≤225MPa, which meets the design requirements for infinite life, and the overall structural fatigue strength, excluding welded joints, is ≥235MPa.

Claims

1. A lightweight optimization method based on a multi-component composite frame, characterized in that, include: S1: Based on the working environment and relevant regulations of the snowplow, determine the frame parameters and materials, select Q235 steel as the frame material, and establish a three-dimensional model of the frame; S2: Based on the three-dimensional model of the frame, establish a finite element model of the frame and complete the preprocessing. Analyze the strength and stiffness performance under bending and torsional conditions. Obtain the first 6 natural frequencies of the frame through modal analysis. After confirming that the frame meets the preset requirements, proceed to the optimization process. S3: Optimize the cross-sectional dimensions of the vehicle frame beam solid element, carry out response surface optimization with the I-beam as the research object, determine the design variables and response values ​​through sensitivity analysis, establish and verify the response surface model, take the maximum deformation and maximum stress as constraints, take the minimization of mass as the optimization objective, and obtain three sets of candidate solutions; S4: An improved Harris Eagle algorithm is proposed based on Bernoulli chaotic mapping, log convergence factor adjustment strategy and position update strategy of vulture search algorithm; S5: Optimize the cross-sectional dimensions of the I-beam using the improved Harris Eagle algorithm. With maximum deformation and maximum stress as constraints and mass minimization as the objective function, establish a mathematical model for the optimization of the I-beam solid element and obtain three sets of optimal solutions. S6: Compare the three sets of candidate solutions with the three sets of optimal solutions, and determine the final parameters of the beam section based on the processing and manufacturing process; S7: Optimize the remaining beam solid elements of the frame according to the methods of steps S3, S5 and S6 to obtain the optimized frame model; S8: Verify the strength, stiffness, and fatigue life of the optimized chassis model.

2. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S1, the relevant parameters of the frame are: width 600mm, length 1800mm, and the main structure of the crossbeam is selected from at least one of the following: channel beam, square beam, I-beam, and tubular beam.

3. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S2, the preprocessing includes: setting the finite element mesh unit size to 5mm, adding standard Earth gravity in the Y-axis direction, selecting a dynamic coefficient of 2.5, and simultaneously applying constraints for bending and torsional conditions; wherein, under the bending condition, the left front, right front, left rear, and right rear suspensions are all constrained by U. y U z Directional degree of freedom; left front suspension under torsional conditions with U-shaped force applied. y =2000N load, remaining suspension constraints U x U y U z Degrees of freedom of direction.

4. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S2, the analysis of strength and stiffness performance under bending and torsion conditions is as follows: Based on the length of the transverse and longitudinal beams of the frame, the allowable deflection is calculated according to the allowable deflection formula. The maximum deformation of the main beam is compared with the allowable deflection to confirm that the stiffness requirement is met. Based on the allowable stress of Q235 steel of 156.7MPa, the maximum stress value of the frame is compared to confirm that the strength requirement is met and there is strength redundancy. The allowable deflection formula is: ； Where [f] is the allowable deflection, L is the length of the beam, and α is the allowable deflection coefficient, which is taken as 500.

5. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S2, the modal analysis is specifically free modal analysis, which extracts the first 6 modal parameters and modal shape cloud diagrams. The first 6 modal frequencies of the frame avoid the resonance zone range where the motor operating frequency of 11.70Hz and the ground feedback excitation frequency of 5.30Hz are located.

6. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S3, the response surface optimization specifically involves: generating 25 design points using a central composite experimental design method; selecting four design variables that significantly affect the response value through sensitivity analysis; constructing a response surface model using the standard second-order response surface method; verifying that the model's predicted values ​​are nearly equal to the actual observed values ​​through goodness-of-fit curves; and ensuring that the coefficient of determination R² for each response surface is greater than 0.97; and using quality as the optimization objective, with constraints of maximum deformation ≤ 0.25134 mm and maximum equivalent stress ≤ 110.4 MPa, optimizing the design variables to obtain the three sets of candidate solutions.

7. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S4, the improved Harris Hawk algorithm specifically includes: S41: Population initialization is performed based on the Bernoulli chaotic mapping, with the initial population size set to 100. The expression is as follows: ； Where β is the adjustment coefficient, Z k The kth individual in the population generates a more evenly distributed and more ergodic population. S42: Introducing a logarithmic nonlinear convergence factor to optimize the escape energy E improves the convergence accuracy and speed of the algorithm. Its expression is: ； Where E is the improved escape energy, iter is the maximum number of iterations, and t is the current iteration number; S43: Fit the position update strategy of the vulture search algorithm. When q≥0.5, a large-area search strategy is adopted, and when q<0.5, a spiral search strategy is adopted to improve the optimization ability.

8. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S5, the optimized mathematical model for the I-beam solid element is as follows: A= 2P1P4+ (P2- 2P4)P 3. ; Among them, P1, P2, P3, and P4 are four cross-sectional dimension design variables, and A is a mass-related calculated value.

9. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S6, the determination of the final parameters of the beam section specifically involves: comparing the candidate solution obtained by response surface optimization with the optimal solution obtained by the improved Harris Eagle algorithm, and verifying the feasibility of the algorithm by combining three sets of iterative curves, ensuring that the final parameters meet the processing and manufacturing process requirements, and that the difference between the design variables P2, P3, and P4 is ≤0.5mm.

10. The lightweight optimization method based on a multi-component composite frame according to claim 1, characterized in that, In S8, the verification of strength, stiffness, and fatigue life specifically involves: using high-cycle fatigue calculations, combining Goodman's theory to correct for the influence of mean stress, and determining based on the material's SN curve that when the frame's fatigue cycle count reaches 10... 6 When the order of magnitude is reached, the material fatigue strength is ≤225MPa, which meets the design requirements for infinite life, and the overall structural fatigue strength, excluding welded joints, is ≥235MPa.

Images