A High-Dimensional Multi-Objective Optimization Design Method for the Skeleton of an Electric Bus

Abstract

The present invention provides a high-dimensional multi-objective optimization design method for the frame of an electric bus, comprising: establishing a finite element model of the bus frame; conducting basic performance analysis, modal analysis, and side collision analysis of the whole vehicle frame on the model; dividing the bus frame into several groups; setting the acceleration at the driver's position, the acceleration at the seat near the middle door, and the low-order modal frequency during the side collision of the whole vehicle as optimization constraints, setting the vehicle mass, the maximum deformation under the ultimate torsion condition, and the side collision intrusion amount as optimization objectives, and taking the thickness of the grouped design variables as optimization variables; screening out the subsequent optimization variables; conducting experimental design; fitting an approximate model and checking the accuracy of the model; establishing a mathematical optimization model and conducting optimization; screening out a set of data meeting the requirements, re-importing it into the finite element model, and then comparing it with the initial model to judge the optimization effect. The present invention is a high-dimensional multi-objective collaborative optimization design method that takes into account multiple linear and highly non-linear responses.

Description

Technical Field

[0001] The present invention relates to the field of automobiles, and particularly to a high-dimensional multi-objective optimization design method for the skeleton of an electric bus. Background Art

[0002] With the development of computer technology and the continuous development and improvement of technologies such as numerical analysis theory and optimization algorithms, more advanced, accurate, and efficient methods have emerged in the design of the skeleton system.

[0003] Currently, there are patent-proposed multi-objective optimization design methods for automobiles:

[0004] The Chinese invention patent (application number: 202011259985.1) "A Multi-objective Optimization Design Method for the Skeleton of a Hybrid Electric Bus" discloses a multi-disciplinary collaborative optimization design method that takes into account multiple linear and highly non-linear responses such as the first-order modal frequency, ultimate torsional condition, acceleration at the driver's position during vehicle frontal collision, and energy absorption during vehicle frontal collision.

[0005] The Chinese invention patent (application number: 201910215479.3) "A Body Frame Disciplinary Collaborative Optimization Design Method and System" discloses a body frame collaborative optimization design method with multiple sub-condition models, and the multiple sub-condition models include linear conditions such as bending stiffness, torsional stiffness, and mode, as well as a non-linear condition of collision condition.

[0006] Although the above-mentioned patents involve considering collision performance, they only consider frontal collision and collision deformation, and do not take other collision situations and related collision safety data into the design scheme. At the same time, these optimization design methods consider fewer optimization objectives, all of which are two objectives. For the skeleton of a bus, its performance involves all aspects, so it is necessary to optimize with multiple performances as objectives. Summary of the Invention

[0007] In view of the deficiencies in the prior art, the present invention provides a high-dimensional multi-objective optimization design method for the skeleton of an electric bus, which takes into account high-dimensional multi-objective collaborative optimization design of multiple linear and highly non-linear responses such as the first-order modal frequency, ultimate torsional condition, side collision intrusion amount, acceleration at the driver's position during vehicle side collision, and acceleration at the seat near the middle door.

[0008] The present invention achieves the above technical objectives through the following technical means.

[0009] A high-dimensional multi-objective optimization design method for the skeleton of an electric bus, characterized by comprising:

[0010] S1: Establish a finite element model of the bus skeleton for optimization analysis;

[0011] S2: Perform basic performance analysis, modal analysis, and side collision analysis of the bus frame finite element model. The basic performance analysis includes static analysis under horizontal bending conditions, extreme torsion conditions, emergency braking conditions, and emergency turning conditions.

[0012] S3: Group the bus frame into several groups based on the position, function, and thickness of the bus frame components.

[0013] S4: Set the acceleration at the driver's position, the acceleration at the seat near the middle door, and the low-order modal frequencies during the side collision of the whole vehicle as optimization constraints. Set the vehicle mass, the maximum deformation under extreme torsion conditions, and the side collision intrusion amount as optimization objectives. Take the thickness of each component in the bus frame grouping in step S3 as the optimization variables.

[0014] S5: Based on the design variables of the bus frame grouping in step S3, draw the linear main effect diagrams of the five optimization responses, namely the side collision acceleration at the driver's position, the side collision intrusion amount of the whole vehicle, the first-order modal frequency, the vehicle mass, and the maximum deformation of the whole vehicle static analysis. Select the variables with sensitivity values exceeding a certain value to the response as the subsequent optimization variables.

[0015] S6: Conduct experimental design on the selected variables through the experimental design method.

[0016] S7: For the sampling data in S6, fit the approximate models of the quality of the approximate model fitting, the stress under torsion conditions, the first-order modal frequency, the acceleration at the driver's position, and the acceleration at the middle door seat. Check the accuracy of the model through the coefficient of determination R 2 Check the accuracy of the model.

[0017] S8: Establish a mathematical optimization model and perform final optimization using a multi-objective optimization algorithm.

[0018] S9: Select a set of data that meets the requirements from the optimized data. Round the optimized variables and re-import them into the finite element model, and then compare with the original model to judge the optimization effect.

[0019] Furthermore, the material used for the body outer frame skeleton in the bus frame finite element model is Q235 structural steel, and the materials used for the frame and the floor skeleton are Qste700tm structural steel.

[0020] Furthermore, the speed of the oncoming vehicle in the side collision simulation analysis in step S2 is 40 km / h.

[0021] Furthermore, in step S3, the outer frame skeleton, the frame, and the floor skeleton of the bus are divided into 50 groups, and the corresponding optimization variables in step S4 are 50.

[0022] Furthermore, the experimental design method used in step S6 is the Latin hypercube experimental design method.

[0023] Further, the approximation model fitting method adopted in step S7 is to fit with a radial basis neural network.

[0024] Further, the specific steps of step S7 are as follows:

[0025] The acceleration at the driver's position in a frontal collision, the centroid acceleration in a rollover, the frontal collision energy absorption of the whole vehicle, the rollover intrusion amount, and the mass of the whole vehicle are fitted by the radial basis neural network (RBF) method. Through the coefficient of determination R 2 Check the accuracy of the RBF model, that is

[0026]

[0027] In the formula, y i is the true response value of the design space, is the sum of squares of the response mean differences, is the calculated value of the surrogate model.

[0028] Further, the multi-objective optimization algorithm adopted in step S8 is the NSGA-III algorithm.

[0029] Further, the optimization mathematical model established in S8 is:

[0030]

[0031] In the formula, x is the design variable, m is the mass of the whole vehicle, S is the maximum value of the frame deformation, w is the side collision intrusion amount, f 1 , f 2 are the first and second order modes of the bus frame, G(x) is the maximum acceleration at the middle door seat in a side collision, U(x) is the maximum acceleration at the driver's seat in a side collision, G 0 (x), U 0 (x) are the initial values of the maximum accelerations of the middle door seat and the driver's seat in a side collision.

[0032] Advantages of the present invention:

[0033] The present invention overcomes the deficiencies in the optimization of the side collision of a bus skeleton, provides a high-dimensional multi-objective optimization design method for the skeleton of an electric bus, determines the side collision of the bus skeleton based on CAE technology and the finite element method, takes the plate thickness of the skeleton as the design variable, takes the acceleration at the driver's position and the acceleration at the middle door seat during side collision as the optimization constraints, and takes the vehicle mass and the rollover intrusion amount as the corresponding objectives. Based on the Hyperstudy integration platform, the optimal Latin hypercube method is used to conduct experimental design on each design variable, an approximate model is established accordingly, and then the NSGA-III multi-objective optimization algorithm is used to optimize the plate thickness of the skeleton, and finally the optimal design parameters are obtained. Therefore, the present invention is a high-dimensional multi-objective collaborative optimization design method that takes into account multiple linear and highly non-linear responses such as the first-order modal frequency, the ultimate torsion condition, the side collision intrusion amount, the acceleration at the driver's position during the vehicle's side collision, and the acceleration at the seat near the middle door, provides a reliable analysis method for the comprehensive performance and lightweight design of the skeleton, and thus effectively improves the product development efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a flowchart of a high-dimensional multi-objective optimization design method for the skeleton of an electric bus according to an embodiment of the present invention;

[0035] Figure 2 It is a finite element schematic diagram of the outer frame skeleton of the bus according to an embodiment of the present invention;

[0036] Figure 3 It is a finite element schematic diagram of the roof of the bus according to an embodiment of the present invention;

[0037] Figure 4 It is a schematic diagram of bus parameter screening according to an embodiment of the present invention

[0038] Figure 5 It is an acceleration-time curve diagram at the driver's seat according to an embodiment of the present invention;

[0039] Figure 6 It is an acceleration-time curve diagram at the middle door seat according to an embodiment of the present invention;

[0040] Figure 7 It is a side collision intrusion amount diagram according to an embodiment of the present invention;

[0041] Figure 8 It is a stress nephogram under the torsion condition according to an embodiment of the present invention;

[0042] Figure 9 It is a finite element model diagram of the side collision of the bus according to an embodiment of the invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific embodiments of the present invention in conjunction with the accompanying drawings. Many specific details are set forth in the following description to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0044] The following describes a lightweight design method for the frame of a pure electric bus in conjunction with examples. The specific implementation steps are as follows:

[0045] S1. Based on CAE technology and the finite element method, a finite element model of the vehicle frame for optimization analysis is established. The element size is set to 10 mm. The whole vehicle has a total of 2,400,554 elements and 2,511,692 nodes. Two materials are used for the whole vehicle. The material used for the outer frame of the body is Q235 structural steel, and the materials used for the vehicle frame and the floor frame are Qste700tm structural steel. The material properties are shown in Table 1;

[0046] Table 1 Material Properties

[0047]

[0048] S2. Perform basic performance analysis, unconstrained modal analysis, and side collision simulation at 40 km / h on the finite element model of the bus frame. A collision vehicle model is set on the left side of the finite element model of the bus frame, and loads such as the battery, occupants, glass, and engine on the vehicle are applied in the form of mass points. Among them, the basic performance analysis is the static analysis of four working conditions: horizontal bending condition, ultimate torsion condition, emergency braking condition, and emergency turning condition. Specifically as follows:

[0049] In the ultimate torsion condition, the YZ degrees of freedom of the left front wheel and the XYZ degrees of freedom of the left rear wheel are constrained, and the XZ degrees of freedom of the right rear wheel are constrained; in the horizontal bending condition, the XYZ degrees of freedom of the left front wheel, the XZ degrees of freedom of the right front wheel, the YZ degrees of freedom of the left rear wheel, and the Z degrees of freedom of the right rear wheel are constrained. In the emergency braking condition, the XYZ degrees of freedom of the left front wheel, the XZ degrees of freedom of the right front wheel, the YZ degrees of freedom of the left rear wheel, and the Z degrees of freedom of the right rear wheel are constrained; in the emergency turning condition, the XYZ degrees of freedom of the left front wheel, the XZ degrees of freedom of the right front wheel, the YZ degrees of freedom of the left rear wheel, and the Z degrees of freedom of the right rear wheel are constrained. The maximum stresses of the bus frame under the four road conditions are shown in Table 2:

[0050] Table 2 Maximum Stresses of the Bus Frame under Four Working Conditions

[0051]

[0052] The stress result in the ultimate torsion condition is as Figure 2As shown in the figure, the maximum stress position is at the connection between the bottom of the bus frame and the powertrain. Since the material at the maximum stress point of the frame is Qste700tm and the yield limit of this material is 650 MPa, the bus frame has sufficient safety margin and also leaves some room for lightweight design;

[0053] Modal analysis:

[0054] Nowadays, the requirements for bus comfort are getting higher and higher. To ensure comfort, modal analysis needs to be carried out on the bus frame. The first-order modal frequency is as Figure 3 shown; during the driving process of the bus on the road surface, the body structure will vibrate due to the excitation of various vibration sources, thus affecting the riding experience; when the natural frequency of the bus frame is close to the road surface vibration frequency, resonance will occur, which will not only generate intense vibration and noise but also affect the service life of the bus frame; through modal analysis of the bus frame, the frequency range of the frame can be clearly known and it can be judged whether resonance will occur; the modal of the bus frame in the free state, the first 6 natural frequencies are shown in Table 3:

[0055] Table 3 The first six-order modes of the bus frame

[0056]

[0057] When the bus is driving on the road, it will be affected by external excitation and the vibration of its own wheels, engine, air conditioner, transmission system, etc.; the excitation frequency of the road surface is less than 3 Hz, the resonance frequency of the body and suspension is 2.0 Hz - 3.6 Hz, and the idle speed frequency of the engine is about 40 Hz; from the results of the first 6 modal frequencies obtained by modal analysis, the modal frequencies of the bus frame are distributed between 7 Hz and 25 Hz, which can effectively avoid the vibration frequencies of the road surface and the bus itself;

[0058] Side collision analysis of the bus frame finite element model:

[0059] The acceleration values at the driver's position and the center of mass, whether the energy of the whole vehicle is conserved before and after the collision, and the displacement curve after the collision when the bus frame has a frontal collision are important reference data for judging the quality of the collision result; in the initial model during the collision simulation, the vehicle speed is 40 km / h and the collision calculation time is 0.2 seconds.

[0060] Within 0 - 50 ms, the speed and acceleration of the moving bus do not change because the moving trolley and the bus have not collided during this period. At the 80 ms moment, the speed of the bus reaches a peak state; at about 120 ms and 200 ms, the acceleration of the moving bus changes to reach a peak, but at the same time there is a rebound, and no deformation occurs at 200 ms, so the acceleration is the largest.

[0061] S3. To improve the optimization calculation efficiency, the outer frame skeleton, vehicle frame, and floor skeleton of the passenger car are divided into 50 groups according to characteristics such as function, thickness, and shape. Among them: T1 - T7 are the body roof skeletons, T8 - T25 are the left and right side wall skeletons, and T26 - T50 are the floor and bottom skeletons.

[0062] S4. Take the first - order modal frequency, second - order modal frequency, limit torsion condition, side - impact intrusion amount, acceleration at the driver's position during the vehicle's side - impact, and acceleration at the seat near the middle door obtained from the previous simulation as the optimization responses, and the thicknesses of the 50 groups of variables as the optimization variables.

[0063] S5. First, use the optimal Latin hypercube method to generate preliminary DOE test data. According to the linear main - effect diagram method of the hyperstudy software, select the variables with a response sensitivity value exceeding 0.3, and take the union of the variables selected for different responses as the final variables for the next optimization, as Figure 4 shown.

[0064] S6. Conduct an optimal Latin hypercube experimental design based on the design variables selected according to sensitivity, and perform a complete experimental design for the optimization variables;

[0065] S7. Then, establish an approximate model for the DOE data. Fit the acceleration at the driver's position during the frontal impact, the centroid acceleration during rollover, the energy absorption during the vehicle's frontal impact, the rollover intrusion amount, and the vehicle mass with the radial basis function neural network (RBF) method, and test the accuracy of the RBF model through the coefficient of determination, that is

[0066]

[0067] where y i is the true response value in the design space, is the sum of squares of the response mean differences, is the calculated value of the surrogate model. The multiple correlation coefficient R2 is a value that varies in the range of 0 - 1. The closer the value is to 1, the higher the accuracy of the surrogate model.

[0068] Table 3 Approximate model error analysis

[0069]

[0070] S8. Establish an optimization mathematical model

[0071]

[0072] where x is the design variable, m is the vehicle mass, S is the maximum value of the skeleton deformation, w is the side - impact intrusion amount, f 1 , f 2For the first - order and second - order modes of the passenger car frame, G(x) is the maximum acceleration at the middle - door seat during side - impact, and U(x) is the maximum acceleration at the driver's seat during side - impact. G 0 (x), U 0 (x) is the initial value of the maximum accelerations of the middle - door seat and the driver's seat during side - impact. The NSGA - III algorithm is used for the final optimization.

[0073] S9. Finally, after weighing the optimization results, a set of variable values that meet the requirements are rounded and re - imported into the finite - element model. Compare with the maximum deformation of the pillar before optimization, the acceleration at the driver's position during side - impact, the centroid acceleration during roll - over, the energy absorption of the whole vehicle, and the intrusion amount during roll - over to judge the optimization effect.

[0074] The final optimization effect is shown in Table 4:

[0075] Table 4 Comparison of responses before and after optimization

[0076]

[0077] The above - described embodiments merely represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the appended claims.

Claims

1. A high-dimensional multi-objective optimization design method for the skeleton of an electric bus, characterized in that, it includes: S1: Establish a finite element model of the bus skeleton for optimization analysis; S2: Conduct basic performance analysis, modal analysis, and side collision analysis of the whole vehicle skeleton on the finite element model of the bus skeleton. The basic performance analysis includes static analysis under horizontal bending conditions, ultimate torsion conditions, emergency braking conditions, and emergency turning conditions; S3: Group the bus skeleton into several groups according to the position, function, and thickness of the bus skeleton components; S4: Set the acceleration at the driver's position, the acceleration at the seat near the middle door, and the low-order modal frequency during the side collision of the whole vehicle as optimization constraints, set the vehicle mass, the maximum deformation under the ultimate torsion condition, and the side collision intrusion amount as optimization objectives, and use the thickness of the design variables for grouping the bus skeleton in step S3 as optimization variables; S5: According to the linear main effect diagrams of the five optimization responses of the side collision acceleration at the driver's position, the side collision intrusion amount of the whole vehicle, the first-order modal frequency, the vehicle mass, and the maximum deformation of the whole vehicle static analysis with respect to the design variables for grouping the bus skeleton in step S3, select the variables whose sensitivity values to the response exceed a certain value as the subsequent optimization variables; S6: Conduct experimental design on the selected variables through the experimental design method; S7: For the sampled data in S6, approximate models for the quality fitted by the approximate model, the stress under the torsional condition, the first-order modal frequency, the acceleration at the driver's position, and the acceleration at the middle-door seat are used to check the accuracy of the model by the coefficient of determination R 2 The specific steps are as follows: The acceleration at the driver's position during a frontal collision, the centroid acceleration during a rollover, the energy absorption during a frontal collision of the whole vehicle, the intrusion amount during a rollover, and the mass of the whole vehicle are fitted using the radial basis function neural network (RBF) method. The accuracy of the RBF model is verified by the coefficient of determination R 2 to verify the accuracy of the RBF model, that is where y i is the true response value of the design space, is the sum of squares of the response mean differences, is the calculated value of the surrogate model; S8: Establish a mathematical optimization model and use a multi-objective optimization algorithm for final optimization. The optimization mathematical model is: where x is the design variable, m is the vehicle mass, S is the maximum value of the frame deformation, w is the side impact intrusion, f 1 , f 2 are the first and second order modes of the bus frame, G(x) is the maximum acceleration at the middle door seat during side impact, U(x) is the maximum acceleration at the driver's seat during side impact, G 0 (x), U 0 (x) are the initial values of the maximum accelerations of the middle door seat and the driver's seat during side impact; S9: Select a set of data that meets the requirements from the optimized data, round the optimized variables and re-import them into the finite element model, and then compare with the initial model to judge the optimization effect.

2. The high-dimensional multi-objective optimization design method for the skeleton of an electric bus according to claim 1, characterized in that: In the finite element model of the bus skeleton, the material used for the body outer frame skeleton is Q235 structural steel, and the materials used for the frame and the floor skeleton are Qste700tm structural steel.

3. The high-dimensional multi-objective optimization design method for the skeleton of an electric bus according to claim 1, characterized in that: In the side collision simulation analysis in step S2, the speed of the oncoming vehicle is 40 km / h.

4. The high-dimensional multi-objective optimization design method for the skeleton of an electric bus according to claim 1, characterized in that: In step S3, the outer frame skeleton, the frame, and the floor skeleton of the bus are divided into 50 groups, and the corresponding optimization variables in step S4 are 50.

5. The high-dimensional multi-objective optimization design method for the skeleton of an electric bus according to claim 1, characterized in that: The experimental design method used in step S6 is the Latin hypercube experimental design method.

6. The high-dimensional multi-objective optimization design method for the skeleton of an electric bus according to claim 1, characterized in that: The approximate model fitting method used in step S7 is to fit with a radial basis neural network.

7. The multi-objective optimization method for the car skeleton according to claim 1, characterized in that: The multi-objective optimization algorithm used in step S8 is the NSGA-III algorithm.

Images