A frame multi-objective optimization method based on safety of battery pack in roll-over collision

Abstract

The present application relates to the technical field of vehicles, and in particular to a frame multi-objective optimization method based on rollover collision battery pack safety, comprising the following steps: constructing a battery homogenization mechanical model to obtain the distance between the current collectors of the positive and negative electrodes of the battery under stress; establishing an electrical model of the battery through an equivalent circuit, and determining whether the battery is short-circuited according to the distance between the current collectors of the positive and negative electrodes in the electrical model of the battery; establishing a three-dimensional model of the frame of the mobile charging vehicle to obtain the instantaneous angular velocity, and analyzing the damage degree of the battery through a coupling model; converting each unit of the finite element analysis model into an independent cell, adjusting the cell properties according to the preset rules, and performing multiple rounds of iteration on the cell properties to complete the optimization of the frame model. The present application can simultaneously achieve lightweight of the frame and improvement of the battery pack safety in view of the deficiency of the optimization of the battery pack safety during rollover collision, and effectively improves the optimization efficiency of the complex model.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and in particular to a multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions. Background Technology

[0002] To address the problem of electric vehicles running out of power and unable to move before reaching a charging station, a mobile charging vehicle has been designed. This vehicle is equipped with a large number of rechargeable batteries and a power distribution system and charging guns to form a mobile energy-carrying battery charging and power distribution system, enabling emergency charging of electric vehicles without power.

[0003] The frame of a mobile charging vehicle is a crucial load-bearing component. To meet the requirements of a compact shape and variable load conditions, the frame structure design needs to comprehensively consider factors such as strength and lightweighting, ensuring the safety of the battery pack in the event of a rollover collision. Summary of the Invention

[0004] To address the aforementioned issues, this invention provides a multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions. This method addresses the shortcomings in optimizing battery pack safety during rollover collisions and can simultaneously achieve improvements in both frame lightweighting and battery pack safety, effectively enhancing the optimization efficiency of complex models.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions includes the following steps:

[0007] S1. Establishment of the battery electromechanical coupling model: By conducting quasi-static mechanical experiments on the positive electrode, negative electrode and separator of the battery, a homogenized mechanical model of the battery is constructed to obtain the distance between the current collectors of the positive and negative electrodes under force; an electrical model of the battery is established through an equivalent circuit, and based on the distance between the current collectors of the positive and negative electrodes, it is determined whether the battery is short-circuited in the battery electrical model.

[0008] S2. Establishment of the rollover collision model: Establish a three-dimensional model of the mobile charging vehicle frame, and establish a finite element analysis model of the mobile charging vehicle frame based on the three-dimensional model to obtain the instantaneous angular velocity of the mobile charging vehicle frame at the point of contact with the cement ground. Based on the instantaneous angular velocity, and through the coupling model of the battery homogenization mechanical model and the battery electrical model, analyze the degree of battery damage.

[0009] S3. Optimization of the chassis model: Each element of the finite element analysis model is converted into an independent cell. The converted finite element analysis model is processed by the solver LS-DYNA to extract key response parameters. The cell properties are adjusted according to preset rules. Multiple iterations are performed by adjusting the cell properties to complete the optimization of the chassis model.

[0010] Further, in step S1, based on the positive and negative electrode and casing experiments, the positive and negative electrode and casing compression experiments, and the RS structure indentation experiment, the load-displacement curve of the battery is obtained. The stress-strain curves of the positive electrode, negative electrode, and separator of the battery are then calculated using the load-displacement curve.

[0011]

[0012] Where P is the load, A is the original cross-sectional area of ​​the battery, ε1 is the experimental strain, L0 is the original length of the battery, and L is the deformed length of the battery.

[0013] Furthermore, based on the relationship between stress and stress between layers in the RS structure, and by calculating the stress-strain curve of the RS structure through electrodes and diaphragms, the stress-strain curve of the RS structure is converted into a constitutive model using equivalent energy to obtain the increase in internal energy of the RS structure during the process:

[0014] σ average =σ a v a +σ c v c +σ s v s Formula (2)

[0015]

[0016] σ e =Aε1 n Formula (4)

[0017] Where, σ average The average stress of the RS structure; v a v c v s These represent the ratios of the volumes of the negative electrode, positive electrode, and separator to the total volume of the RS structure, respectively; U represents the increase in internal energy of the RS structure during the process; ε1 represents the experimental strain; and σ... e The stress represents the electrode, and A and n are both fitting parameters.

[0018] Furthermore, the failure strain of the RS structure after homogenization is ε. f If the energy absorbed by the homogenized material under failure strain is U0, then:

[0019]

[0020] Where, ε f U0 is the failure strain; ε is the failure strain. f Energy absorbed by homogenized materials;

[0021] Based on the fact that the energy absorbed by the RS structure under the same strain is the same as that absorbed by the homogenized material, the equivalent energy under the failure strain is obtained:

[0022]

[0023] L′=ε2L0 Formula (7)

[0024] Based on the value of n, each failure strain value corresponds to the original cross-sectional area A of the battery. By continuously changing the failure strain value until the load-displacement curve and failure mode of the fully homogenized model are close to the experimental results, the original cross-sectional area value is obtained. Using the original cross-sectional area of ​​the battery, the constitutive model of the homogenized material is determined. Simulations are then performed on the original cross-sectional area of ​​the battery to obtain the simulated strain ε2 under different collision conditions. Furthermore, the distance L′ between the current collectors of the positive and negative electrodes is obtained based on the simulated strain.

[0025] Furthermore, in step S1, the potential drop between the current collector nodes of the positive and negative electrodes is:

[0026]

[0027] in, The local potential of the negative current collector node N1; The local potential of the positive current collector node N2 is obtained from the distance between the current collectors of the positive and negative electrodes. and I is the local current; R0 is the internal resistance; U is the local open-circuit voltage; V D This is a localized diffusion overpotential;

[0028] Obtained from the distance between the current collectors of the positive and negative electrodes. The presence of a short circuit in a battery can be determined by assessing the localized overpotential.

[0029] Furthermore, based on the composition of the equivalent circuit, and with the nodes corresponding to the current collectors of the positive and negative terminals connected through the equivalent circuit, the voltage across the RC pair is obtained as follows:

[0030]

[0031] Among them, C D R is the capacitance of the RC pair in the equivalent circuit; D The resistance of the RC pair in the equivalent circuit;

[0032] The local current I is obtained by calculating the state of charge (SOC) of the local equivalent circuit:

[0033]

[0034] Among them, c Q The conversion factor is R0; Q is the electrical quantity; R0 and R1 are also present. D C D The voltage was obtained by pulse discharge experimental curves at different temperatures and SOCs; the local open-circuit voltage U was determined by SOC.

[0035] Furthermore, in step S2, a test bench is set up, the vehicle frame can be placed on the test bench, and the test bench is provided with a pivot so that the vehicle frame can be tilted along the pivot.

[0036] The chassis underwent a rollover simulation on the test bench;

[0037] The initial velocity at the critical rollover point CG' of the chassis when it begins to rollover is zero, and the initial energy is:

[0038]

[0039] The instantaneous energy at the point of impact is:

[0040]

[0041] Based on the initial energy and the instantaneous energy at the point of impact, the instantaneous angular velocity of the chassis at the point of contact with the concrete surface is obtained as follows:

[0042]

[0043] Where M is the total mass of the mobile charging vehicle; ω is the angular velocity; and J is the moment of inertia, which is obtained by assigning a random angular velocity in the LS-dyna software. The values ​​are: h0 is the height between the critical rollover point and the bottom of the chassis when the chassis is on the test bench; h1 is the height between the critical rollover point and the bottom of the chassis when the chassis is on the test bench at the critical rollover point; h2 is the height between the critical rollover point and the bottom of the chassis after the chassis rolls over; and d is the height between the test bench and the concrete ground.

[0044] The failure strain is obtained by angular velocity simulation of the mobile charging vehicle frame rollover collision test and modal simulation analysis. The degree of battery damage is analyzed by the coupled model of the battery homogenized mechanical model and the battery electrical model through the failure strain calculation.

[0045] Further, in step S3, the density parameter of the element is mapped to the elastic modulus of the element, and the elastic modulus is calculated as follows:

[0046] El (x l )=x l p E0 formula (14)

[0047] Where l is a unit; x l E is the density parameter of the element; E0 is the initial elastic modulus of the element; E l The elastic modulus of the element;

[0048] In the elastoplastic material constitutive model of the frame structure, initial yield stress and strain hardening modulus are introduced to obtain the yield stress and strain hardening modulus of the elements:

[0049] σ yl (x l )=x l p σ y0 Formula (15)

[0050] E hl (x l )=x l p E h0 Formula (16)

[0051] Where, σ y0 σ is the initial yield stress; yl E represents the yield stress of the element; h0 E represents the initial strain hardening modulus. hl The strain hardening modulus of the element;

[0052] With the optimization objectives of maximizing frame stiffness and minimizing volume, strain energy density is used as a local stiffness index, and stiffness is quantified through strain energy. Furthermore, maximizing stiffness is equivalent to minimizing strain energy. Therefore:

[0053] C(X)=F T KF formula (17)

[0054]

[0055] Where C(X) is the structural strain energy function; X = {x l} represents the element density vector; K represents the overall structural stiffness matrix; F represents the structural nodal displacement vector; V(X) represents the structural volume function; v l The volume of the unit;

[0056] The structural strain energy and structural volume are constrained by the constraints of the unit density meeting the limit value and mechanical performance requirements, and the maximum mechanical abuse that the battery pack must withstand must not cause internal short circuit failure, so as to obtain the constrained optimization objective.

[0057] Further, in step S3, each element of the finite element analysis model is converted into an independent cell, and the cell state is as follows:

[0058] S i =[x i U i S n ] Formula (19)

[0059] Among them, S i Cellular state; x i For x l The optimization variable is the density parameter of the cell; U i S is the strain energy; n This refers to the domain state; the calculation method is as follows:

[0060]

[0061] Where n is the number of cells in the neighborhood; η i is the weighting factor of cell i in the neighborhood, and the neighborhood strain energy of a cell is the average value of the strain energies of all cells in the neighborhood.

[0062] By quantifying the center distance between neighboring cells and the central cell, a cell whose center is within the radius of a certain cell's neighborhood can be considered as the neighborhood of that cell. According to the von Neumann model, the neighborhood radius is r = a, and the side length of the cell is equivalent to the diameter of a circle with equal area or a sphere with equal volume.

[0063]

[0064] Based on the material properties of the chassis, when the strain energy density of a cell rises to the peak value allowed by the material properties, the density of the cell will tend to stabilize. At this point, the relative density of the cell is locked at 1, and the optimization objective function is expressed as:

[0065]

[0066]

[0067] in, Here, represents the average strain energy density and the target strain energy density of element i, respectively; N is the total number of elements. To minimize the design variable, a value of 0.001 is used to avoid singular matrices;

[0068] The agent's action is defined as an adjustment amount of the cell density value x, where the adjustment value is the agent's action space. The action space is set as an array containing n elements, where each element represents a possible adjustment amount. In each decision cycle, the agent selects a specific action 'a' from the preset action space to perform the adjustment of the cell density.

[0069] x i+1 =x i +a formula (24)

[0070] a = [a1 a2 a3 …a] n ] Formula (25)

[0071] Where, x i Design variables for the cell, where 'a' represents the action chosen by the cell, and 'x' represents the action chosen by the cell. i +1 represents the state of the cell in the next iteration step;

[0072] With the optimization objectives of maximizing frame stiffness and minimizing volume, the optimization variable x is... i Strain energy U i Domain State S n Construct the sub-reward function:

[0073]

[0074] Where, x i ',U i ',S' n These represent the cell states after the action is performed, U max ,S max These are the maximum values ​​of the strain energy of the structural element in the initial state and the maximum values ​​in the neighborhood state, respectively.

[0075] Based on the optimization objective, each reward is negativeed, and the reward functions for each item are unified to obtain the reward R(S,a) of a cell performing action a in state S:

[0076] R(S,a)=αR1+βR2+γR .3 Formula (27)

[0077] The agent's learning strategy employs single-step temporal difference to obtain:

[0078]

[0079] Among them, Q(S) i ,a i ) represents the action value function of the agent; α is the learning factor; ε is the discount factor with a value of 0.4;

[0080] The agent learns based on the action value function, and uses the learned action value function of the agent to optimize the frame. The cells undergo multiple iterations through the action value function of the agent until the action value function of the agent converges to obtain the optimal cell state. The elastic modulus is optimized by optimizing the cell state to calculate the optimization target of the frame.

[0081] The beneficial effects of this invention are as follows: By constructing a homogenized mechanical model of the battery, the distance between the current collectors of the positive and negative electrodes after a collision is obtained. Using this distance, the battery electrical model is used to determine if a short circuit exists. The coupling of the homogenized mechanical model and the battery electrical model ensures close interaction between the models, enabling a comprehensive and accurate analysis of the battery's safety and performance under complex operating conditions. Furthermore, by establishing a three-dimensional model of the mobile charging vehicle frame, finite element analysis of the frame's rollover collision safety has been performed, achieving simulation and modal analysis of the mobile charging vehicle frame's rollover collision test. Simulation analysis, based on data from the rollover collision model, and the ability to evaluate battery damage through a battery electromechanical coupling model, facilitates subsequent model optimization. In the optimization process, finite element units are used as cells, elastic modulus is the optimization object, and the optimization objective is to maximize frame stiffness while minimizing volume. The process employs a nonlinear topology optimization with constraints including preventing internal short-circuit failure of the battery pack, the frame structural state, and maximum stress. This invention addresses the shortcomings in battery pack safety optimization during rollover collisions, simultaneously achieving frame lightweighting and improved battery pack safety, effectively enhancing the optimization efficiency of complex models. Attached Figure Description

[0082] Figure 1 This is a flowchart of a preferred embodiment of the present invention, which describes a multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions.

[0083] Figure 2 This is a schematic diagram of the equivalent circuit model structure of a vehicle frame multi-objective optimization method based on the safety of battery packs in rollover collisions, according to a preferred embodiment of the present invention.

[0084] Figure 3 This is a side-roll collision schematic diagram of a vehicle frame multi-objective optimization method based on battery pack safety in a side-roll collision, according to a preferred embodiment of the present invention. Detailed Implementation

[0085] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0086] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0087] Please also see Figure 1 and Figure 3 A preferred embodiment of the present invention provides a multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions, comprising the following steps:

[0088] S1. Establishment of the battery electromechanical coupling model: By conducting quasi-static mechanical experiments on the positive electrode, negative electrode and separator of the battery, a homogenized mechanical model of the battery is constructed to obtain the distance between the current collectors of the positive and negative electrodes under force; the battery electrical model is established through an equivalent circuit, and the battery short circuit is determined based on the distance between the current collectors of the positive and negative electrodes in the battery electrical model.

[0089] In step S1, based on the positive and negative electrode and casing experiments, the positive and negative electrode and casing compression experiments, and the RS structure indentation experiment, the load-displacement curve of the battery is obtained. The stress-strain curves of the positive electrode, negative electrode, and separator of the battery are then calculated using the load-displacement curve.

[0090]

[0091] Where P is the load, A is the original cross-sectional area of ​​the battery, ε1 is the experimental strain, L0 is the original length of the battery, and L is the deformed length of the battery. The experimental strain is obtained by simulating the original cross-sectional area of ​​the battery, and the distance between the current collectors of the positive and negative electrodes is obtained based on the experimental strain, L and L0.

[0092] Based on the relationship between stress and stress between layers in the RS structure, and by calculating the stress-strain curve of the RS structure through electrodes and diaphragms, the stress-strain curve of the RS structure is converted into a constitutive model using equivalent energy to obtain the increase in internal energy of the RS structure during the process:

[0093] σ average =σ a v a +σ c v c +σ s v s Formula (2)

[0094]

[0095] σ e =Aε1n Formula (4)

[0096] Where, σ average The average stress of the RS structure; v a v c v s These represent the ratios of the volumes of the negative electrode, positive electrode, and separator to the total volume of the RS structure, respectively; U represents the increase in internal energy of the RS structure during the process; ε² represents the experimental strain; and σ... e The stress represents the electrode, and A and n are both fitting parameters.

[0097] The failure strain of the RS structure after homogenization is ε. f If the energy absorbed by the homogenized material under failure strain is U0, then:

[0098]

[0099] Where, ε f U0 is the failure strain; ε is the failure strain. f Energy absorbed by homogenized materials;

[0100] Based on the fact that the energy absorbed by the RS structure under the same strain is the same as that absorbed by the homogenized material, the equivalent energy under the failure strain is obtained:

[0101]

[0102] L′=ε2L0 Formula (7)

[0103] Based on the value of n, each failure strain value corresponds to the original cross-sectional area A of the battery. By continuously changing the failure strain value until the load-displacement curve and failure mode of the fully homogenized model are close to the experimental results, the original cross-sectional area value is obtained. Using the original cross-sectional area of ​​the battery, the constitutive model of the homogenized material is determined. Simulations are then performed on the original cross-sectional area of ​​the battery to obtain the simulated strain ε2 under different collision conditions. Furthermore, the distance L′ between the current collectors of the positive and negative electrodes is obtained based on the simulated strain.

[0104] In step S1, the potential drop between the current collector nodes of the positive and negative electrodes is:

[0105]

[0106] in, The local potential of the negative current collector node N1; The local potential of the positive current collector node N2 is obtained from the distance between the current collectors of the positive and negative electrodes. and I is the local current; R0 is the internal resistance; U is the local open-circuit voltage; V DThis is a localized diffusion overpotential;

[0107] Obtained from the distance between the current collectors of the positive and negative electrodes. The presence of a short circuit in a battery can be determined by assessing the localized overpotential.

[0108] Based on the composition of the equivalent circuit, and with the nodes corresponding to the current collectors of the positive and negative terminals connected through the equivalent circuit, the voltage across the RC pair is obtained as follows:

[0109]

[0110] Among them, C D R is the capacitance of the RC pair in the equivalent circuit; D The resistance of the RC pair in the equivalent circuit;

[0111] The local current I is obtained by calculating the state of charge (SOC) of the local equivalent circuit:

[0112]

[0113] Among them, c Q The conversion factor is R0; Q is the electrical quantity; R0 and R1 are also present. D C D The voltage was obtained by pulse discharge experimental curves at different temperatures and SOCs; the local open-circuit voltage U was determined by SOC.

[0114] In the battery electromechanical coupling model, the internal short circuit phenomenon is simulated by introducing a local short-circuit resistor. When the battery is deformed by external force, the distance between the positive and negative current collectors decreases accordingly. Once this distance shrinks below a pre-set threshold, a short-circuit resistor is used to replace the current collector. Figure 1 middle and An equivalent circuit is established between the electrodes to accurately simulate the internal short-circuit state of the battery. The battery's homogenized mechanical model accurately calculates the deformation of the battery under stress and transmits this deformation information to the electrical model. Using the electrical model, based on the transmitted deformation data, especially the actual distance between the positive and negative current collectors, the risk of an internal short circuit is determined. If a short circuit is detected, the electrical model further calculates relevant electrical parameters, such as current and voltage, to comprehensively assess the impact of the short circuit on battery performance.

[0115] The coupling between the mechanical and electrical models is unidirectional; information flows only from the mechanical model to the electrical model. The distance between the positive and negative current collectors serves as a crucial bridge connecting these two models, reflecting not only the degree of physical deformation of the battery but also directly determining the short-circuit risk assessment result in the electrical model. The mechanical model calculates displacement information and transmits the decisive parameter—the actual distance between the positive and negative current collectors—to the electrical model. The latter, based on this actual distance, uses an internal short-circuit model to determine whether a short circuit has occurred in the battery.

[0116] S2. Establishment of the rollover collision model: Establish a three-dimensional model of the mobile charging vehicle frame, and establish a finite element analysis model of the mobile charging vehicle frame based on the three-dimensional model to obtain the instantaneous angular velocity of the mobile charging vehicle frame at the point of contact with the cement ground. Based on the instantaneous angular velocity, and through the coupling model of the battery homogenization mechanical model and the battery electrical model, analyze the degree of battery damage.

[0117] like Figure 3 As shown, in step S2, a test bench is set up, the vehicle frame can be placed on the test bench, and the test bench is provided with a pivot so that the vehicle frame can be tilted along the pivot.

[0118] The chassis underwent rollover simulation on the test bench;

[0119] The initial velocity at the critical rollover point CG' of the chassis when it begins to rollover is zero, and the initial energy is:

[0120]

[0121] The instantaneous energy at the point of impact is:

[0122]

[0123] Based on the initial energy and the instantaneous energy at the point of impact, the instantaneous angular velocity of the chassis at the point of contact with the concrete surface is obtained as follows:

[0124]

[0125] Where M is the total mass of the mobile charging vehicle; ω is the angular velocity; and J is the moment of inertia, which is obtained by assigning a random angular velocity in the LS-dyna software. The values ​​are: h0 is the height between the critical rollover point and the bottom of the chassis when the chassis is on the test bench; h1 is the height between the critical rollover point and the bottom of the chassis when the chassis is on the test bench at the critical rollover point; h2 is the height between the critical rollover point and the bottom of the chassis after the chassis rolls over; and d is the height between the test bench and the concrete ground.

[0126] The failure strain was obtained by angular velocity simulation of the rollover collision test of the mobile charging vehicle frame and modal simulation analysis. Through failure strain calculation, the degree of battery damage was analyzed by a coupled model of the battery homogenization mechanical model and the battery electrical model.

[0127] S3. Optimization of the chassis model: Each element of the finite element analysis model is converted into an independent cell. The converted finite element analysis model is then processed by the solver LS-DYNA to extract key response parameters. The cell properties are adjusted according to preset rules. Multiple iterations are performed by adjusting the cell properties to complete the optimization of the chassis model.

[0128] In step S3, the density parameter of the element is mapped to the elastic modulus of the element. The elastic modulus is calculated as follows:

[0129] E l (x l )=x l p E0 formula (14)

[0130] Where l is a unit; x l E is the density parameter of the element; E0 is the initial elastic modulus of the element; E l The elastic modulus of the element;

[0131] In the elastoplastic material constitutive model of the frame structure, initial yield stress and strain hardening modulus are introduced to obtain the yield stress and strain hardening modulus of the elements:

[0132] σ yl (x l )=x l p σ y0 Formula (15)

[0133] E hl (x l )=x l p E h0 Formula (16)

[0134] Where, σ y0 σ is the initial yield stress; yl E represents the yield stress of the element; h0 E represents the initial strain hardening modulus. hl The strain hardening modulus of the element;

[0135] With the optimization objectives of maximizing frame stiffness and minimizing volume, strain energy density is used as a local stiffness index, and stiffness is quantified through strain energy. Furthermore, maximizing stiffness is equivalent to minimizing strain energy. Therefore:

[0136] C(X)=F T KF formula (17)

[0137]

[0138] Where C(X) is the structural strain energy function; X = {x l} represents the element density vector; K represents the overall structural stiffness matrix; F represents the structural nodal displacement vector; V(X) represents the structural volume function; v l The volume of the unit;

[0139] The structural strain energy and structural volume are constrained by the constraints of the unit density meeting the limit value and mechanical performance requirements, and the maximum mechanical abuse that the battery pack must withstand must not cause internal short circuit failure. The optimization objective is obtained after the constraints are obtained.

[0140] Two parameters are selected to evaluate whether the structure meets the performance requirements: the structural morphological stability and the maximum stress level. The structural condition and the maximum stress level are selected for evaluation based on the element density x. l Whether the value limits and mechanical performance requirements are met, morphological stability is assessed by monitoring changes in the structure during the optimization process, observing whether the structure evolves from its original rigid body state to a mechanical state. Once the structure transforms into a mechanical state, it means that it has lost its original load-bearing capacity and is therefore considered a failure state. This transformation can be visually reflected by monitoring the maximum displacement of the structure.

[0141] The maximum stress on the structure must be controlled within the material's allowable stress range to ensure the structure's safety and reliability. Considering the stresses the structure may experience in different directions, the impact of these stresses must be assessed to ensure they do not exceed the material's limits. The maximum mechanical abuse the battery pack can withstand must not cause internal short-circuit failure.

[0142] In step S3, each element of the finite element analysis model is converted into an independent cell. The state of a cell should include a record of its feedback to its surrounding environment and the state information of its neighboring regions. This complex set of information can be organized and stored in a multidimensional array for subsequent analysis and processing. In this way, each cell can not only perceive changes in its own state but also capture the influence of the environment on it and the state of its neighboring cells, thereby realizing a more refined and dynamic behavior pattern in the cellular automata system. The cell state is as follows:

[0143] S i =[x i U i S n ] Formula (19)

[0144] Among them, S i Cellular state; xi For x l The optimization variable is the density parameter of the cell; U i S is the strain energy; n This refers to the domain state; the calculation method is as follows:

[0145]

[0146] Where n is the number of cells in the neighborhood; η i is the weighting factor of cell i in the neighborhood, and the neighborhood strain energy of a cell is the average value of the strain energies of all cells in the neighborhood.

[0147] By quantifying the center distance between neighboring cells and the central cell, a cell whose center is within the radius of a certain cell's neighborhood can be considered as the neighborhood of that cell. According to the von Neumann model, the neighborhood radius is r = a, and the side length of the cell is equivalent to the diameter of a circle with equal area or a sphere with equal volume.

[0148]

[0149] For each cell in the system, regardless of its specific location, the update rule is uniform. The core of this rule lies in collecting and analyzing the state information of other cells within each cell's neighborhood, allowing the cells to adjust their own states accordingly. The concept of a von Neumann neighborhood is used for collecting and analyzing neighborhood states. By quantifying the center distance between cells within a neighborhood and the central cell, cells whose center is within the neighborhood radius of a given cell can be considered its neighborhood. The neighborhood radius corresponding to the von Neumann type is r = a, and the equivalent side length of the cell is represented by the diameter of a circle with equal area or a sphere with equal volume.

[0150] Based on the material properties of the chassis, when the strain energy density of a cell rises to the peak value allowed by the material properties, the density of the cell will tend to stabilize. At this point, the relative density of the cell is locked at 1, and the optimization objective function is expressed as:

[0151]

[0152]

[0153] in, Here, represents the average strain energy density and the target strain energy density of element i, respectively; N is the total number of elements. To minimize the design variable, a value of 0.001 is used to avoid singular matrices;

[0154] The agent's action is defined as the adjustment amount of the cell density value x, where the adjustment value is the agent's action space. The action space is set as an array containing n elements, where each element represents a possible adjustment amount. In each decision cycle, the agent selects a specific action 'a' from the preset action space to perform the adjustment of the cell density.

[0155] x i+1 =x i +a formula (24)

[0156] a = [a1 a2 a3 …a] n ] Formula (25)

[0157] Where, x i Design variables for the cell, where 'a' represents the action chosen by the cell, and 'x' represents the action chosen by the cell. i +1 represents the state of the cell in the next iteration step;

[0158] The agent's reward function serves as a bridge connecting the optimization objective and the agent's behavioral decisions. To ensure that the agent's decisions accurately reflect the intention of topology optimization, both the primary objective and the constraints must be explicitly encoded in the agent's reward function. The goal is to minimize material usage while maximizing structural stiffness, i.e., minimizing volume.

[0159] With the optimization objectives of maximizing frame stiffness and minimizing volume, the optimization variable x is... i Strain energy U i Domain State S n Construct the sub-reward function:

[0160]

[0161] Where, x i ',U i ',S' n These represent the cell states after the action is performed, U max ,S max These are the maximum values ​​of the strain energy of the structural element in the initial state and the maximum values ​​in the neighborhood state, respectively.

[0162] Based on the optimization objective, all rewards are negativeed, and the reward functions for each item are standardized to obtain the reward R(S,a) after a cell performs action a in state S:

[0163] R(S,a)=αR1+βR2+γR .3 Formula (27)

[0164] The agent's learning strategy employs single-step temporal difference to obtain:

[0165]

[0166] Among them, Q(S) i ,a i ) represents the action value function of the agent; α is the learning factor; ε is the discount factor with a value of 0.4;

[0167] The agent learns based on the action value function, and uses the learned action value function of the agent to optimize the frame. The cells undergo multiple iterations through the action value function of the agent until the action value function of the agent converges to obtain the optimal cell state. The elastic modulus is optimized by optimizing the cell state to calculate the optimization target of the frame.

[0168] In topology optimization of the structure, the original finite element model is first initialized. This process involves converting each element in the model into an independent cell, each with unique properties, material, and identification number. The transformed model is then submitted to the solver LS-DYNA for computational analysis, focusing on extracting key response data such as strain energy density. To further optimize the structure, the relative density of each cell is dynamically adjusted based on preset neighborhood definitions and cell evolution criteria. This process is achieved through multiple iterations until the entire system reaches a stable state, thus completing one effective structural topology optimization cycle. By initializing the finite element model, decomposing it into independently operable cells, analyzing and extracting important information using the solver, adjusting cell properties according to specific rules, and iterating repeatedly until the optimization objective is achieved, precise topology optimization of the structure is realized.

[0169] During the learning process, each cell first perceives its own state and the situation of its surrounding neighborhood, interacting with the agent module. The agent makes corresponding decisions based on the cell's current state and obtains feedback from the environment, thereby acquiring a large amount of data for learning and updating. This cyclical process makes the entire training more efficient.

[0170] SIMP and BESO methods, as two common topology optimization methods, both rely on sensitivity analysis to evaluate cell utilization. However, sensitivity calculation is directly based on the mathematical model of the problem being solved, meaning that sensitivity needs to be recalculated when the problem model changes. This limits the flexibility of these algorithms in dealing with diverse problems. The proposed method establishes a connection between discrete individuals and the whole system. When performing discrete calculations, the agent's decisions are more global, converge faster, and have a stronger ability to solve nonlinear problems.

Claims

1. A multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions, characterized in that, Includes the following steps: S1. Establishment of the battery electromechanical coupling model: By conducting quasi-static mechanical experiments on the positive electrode, negative electrode and separator of the battery, a homogenized mechanical model of the battery is constructed to obtain the distance between the current collectors of the positive and negative electrodes under force; an electrical model of the battery is established through an equivalent circuit, and based on the distance between the current collectors of the positive and negative electrodes, it is determined whether the battery is short-circuited in the battery electrical model. S2. Establishment of the rollover collision model: Establish a three-dimensional model of the mobile charging vehicle frame, and establish a finite element analysis model of the mobile charging vehicle frame based on the three-dimensional model to obtain the instantaneous angular velocity of the mobile charging vehicle frame at the point of contact with the cement ground. Based on the instantaneous angular velocity, and through the coupling model of the battery homogenization mechanical model and the battery electrical model, analyze the degree of battery damage. S3. Optimization of the chassis model: Each element of the finite element analysis model is converted into an independent cell. The converted finite element analysis model is processed by the solver LS-DYNA to extract key response parameters. The cell properties are adjusted according to preset rules. Multiple iterations are performed on the cell properties to complete the optimization of the chassis model. In step S3, each element of the finite element analysis model is converted into an independent cell, and the cell state is as follows: Official (19); in, It is in the cell state; for x l The optimization variable is the density parameter of the cell; For strain energy; This refers to the domain state; the calculation method is as follows: Official (20); in, n The number of cells in the neighborhood; For cells within the neighborhood i The weighting factor is such that the neighborhood strain energy of a cell is the average value of the strain energies of all cells in the neighborhood. By quantifying the center distance between neighboring cells and the central cell, a cell whose center is within the radius of a certain cell's neighborhood can be considered as the neighborhood of that cell. According to the von Neumann model, the neighborhood radius is r=a, and the side length of the cell is equivalent to that of a circle with equal area or a sphere with equal volume. Official (21); Based on the material properties of the chassis, when the strain energy density of a cell rises to the peak value allowed by the material properties, the density of the cell will tend to stabilize. At this point, the relative density of the cell is locked at 1, and the optimization objective function is expressed as: Official (22); Official (23); in, Units i The average strain energy density and the target value of strain energy density; N The total number of units; To minimize the design variable, a value of 0.001 is used to avoid singular matrices; The action of an agent is defined as the cell density value. x The adjustment amount, the adjustment value is the action space of the agent, and the action space is set to a space containing... n An array of n elements, where each element represents a possible adjustment amount, is used. In each decision cycle, the agent selects a specific action from the preset action space. a To perform adjustments to the cell density: Official (24) Official (25); in, x i Design variables for cells. a The action chosen by the cell, x i +1 represents the state of the cell in the next iteration step; With the optimization objectives of maximizing frame stiffness and minimizing volume, the optimization variables are... x i strain energy U i Domain status S n Construct the sub-reward function: Official (26); in, These represent the cell states after the action is performed. These are the maximum values ​​of the strain energy of the structural element in the initial state and the maximum values ​​in the neighborhood state, respectively. Based on the optimization objective, each reward is negativeed, and the reward functions for each item are standardized to obtain the cell in... S Perform actions in a certain state a The reward after : Official (27); The agent's learning strategy employs single-step temporal difference to obtain: Official (28); in, The action value function of the intelligent agent; For learning factors; This is a discount factor with a value of 0.4; The agent learns based on the action value function, and uses the learned action value function of the agent to optimize the frame. The cells undergo multiple iterations through the action value function of the agent until the action value function of the agent converges to obtain the optimal cell state. The elastic modulus is optimized by optimizing the cell state to calculate the optimization target of the frame.

2. The multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions according to claim 1, characterized in that: In step S1, based on the positive and negative electrode and casing experiments, the positive and negative electrode and casing compression experiments, and the RS structure indentation experiment, the load-displacement curve of the battery is obtained. The stress-strain curves of the positive electrode, negative electrode, and separator of the battery are then calculated using the load-displacement curve. Official (1); in, P For load, A This represents the original cross-sectional area of ​​the battery. For experimental strain; L 0 represents the original length of the battery; L This is the length of the battery after deformation.

3. The multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions according to claim 2, characterized in that: Based on the relationship between stress and stress between layers in the RS structure, and by calculating the stress-strain curve of the RS structure through electrodes and diaphragms, the stress-strain curve of the RS structure is converted into a constitutive model using equivalent energy to obtain the increase in internal energy of the RS structure during the process: Official (2); Official (3); Official (4); in, The average stress of the RS structure; These are the ratios of the volume of the negative electrode, positive electrode, and separator to the volume of the RS structure, respectively. U This represents the increase in internal energy of the RS structure during the process. Represents experimental strain. The stress representing the electrode, A and n All of these are fitting parameters.

4. The multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions according to claim 3, characterized in that: After homogenization, the failure strain of the RS structure is: The energy absorbed by the homogenized material under failure strain is Then we have: Official (5); in, For failure strain; For failure strain Energy absorbed by homogenized materials; Based on the fact that the energy absorbed by the RS structure under the same strain is the same as that absorbed by the homogenized material, the equivalent energy under the failure strain is obtained: Official (6); Official (7); according to n The value of each failure strain corresponds to the value of the original cross-sectional area A of the battery. By continuously changing the value of the failure strain, the load-displacement curve and failure mode of the fully homogenized model are obtained until they are close to the experimental results, thus obtaining the value of the original cross-sectional area. Using the original cross-sectional area of ​​the battery, the constitutive model of the homogenized material is determined, and the original cross-sectional area of ​​the battery is simulated to obtain the simulated strain under different collision conditions. Furthermore, the distance between the current collectors of the positive and negative electrodes was obtained based on the simulated strain. .

5. The multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions according to claim 1, characterized in that: In step S1, the potential drop between the current collector nodes of the positive and negative electrodes is: Official (8); in, negative collector node N The local potential of 1; Positive current collector node N The local potential of 2 is obtained based on the distance between the current collectors of the positive and negative electrodes. and ; I For local current; R 0 represents internal resistance; U This is a local open-circuit voltage; V D This is a localized diffusion overpotential; Obtained from the distance between the current collectors of the positive and negative electrodes. This allows for the determination of whether a battery is short-circuited by assessing the localized overpotential.

6. The multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions according to claim 5, characterized in that: Based on the composition of the equivalent circuit, and with the nodes corresponding to the current collectors of the positive and negative terminals connected through the equivalent circuit, the voltage across the RC pair is obtained as follows: Official (9); in, The capacitance of the RC pair in the equivalent circuit; The resistance of the RC pair in the equivalent circuit; The local current is obtained by calculating the state of charge (SOC) of the local equivalent circuit. I : Official (10); in, The conversion factor; Q For electricity; R 0、 R D , C D The partial open-circuit voltage was identified through pulse discharge experimental curves at different temperatures and SOCs. U Determined by SOC.

7. A multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions according to claim 5, characterized in that: In step S2, a test bench is set up, the vehicle frame can be placed on the test bench, and the test bench is provided with a pivot so that the vehicle frame can be tilted along the pivot. The chassis underwent a rollover simulation on the test bench; Critical rollover point of the frame CG' The initial velocity at the point where the rollover begins is zero, and the initial energy is: Official (11); The instantaneous energy at the point of impact is: Official (12); Based on the initial energy and the instantaneous energy at the point of impact, the instantaneous angular velocity of the chassis at the point of contact with the concrete surface is obtained as follows: Official (13); in, M The total mass of the mobile charging vehicle; Angular velocity is J; moment of inertia is assigned by random angular velocity in LS-dyna software. get; h 0 represents the height of the critical rollover point from the bottom of the chassis when the chassis is on the test bench; h 1 represents the height between the critical rollover point and the bottom of the chassis when the chassis is at the critical rollover point on the test bench; h 2 represents the height between the critical rollover point and the bottom of the frame after the frame has rolled over. d The height of the test bench relative to the concrete floor; The failure strain is obtained by angular velocity simulation of the mobile charging vehicle frame rollover collision test and modal simulation analysis. The degree of battery damage is analyzed by the coupled model of the battery homogenized mechanical model and the battery electrical model through the failure strain calculation.

8. The multi-objective optimization method for vehicle frame based on battery pack safety in rollover collisions according to claim 1, characterized in that: In step S3, the density parameter of the element is mapped to the elastic modulus of the element. The elastic modulus is calculated as follows: Official (14); in, l As a unit; x l The density parameter of the element; E 0 represents the initial elastic modulus of the element; E l The elastic modulus of the element; In the elastoplastic material constitutive model of the frame structure, initial yield stress and strain hardening modulus are introduced to obtain the yield stress and strain hardening modulus of the elements: Official (15); Official (16); in, This is the initial yield stress; The yield stress of the element; This is the initial strain hardening modulus; The strain hardening modulus of the element; With the optimization objectives of maximizing frame stiffness and minimizing volume, strain energy density is used as a local stiffness index, and stiffness is quantified through strain energy. Furthermore, maximizing stiffness is equivalent to minimizing strain energy. Therefore: Official (17); Official (18); in, The structural strain energy function; The unit density vector; K Here is the overall stiffness matrix of the structure; F The displacement vector of the structural nodes; It is a structural volume function; v l The volume of the unit; The structural strain energy and structural volume are constrained by the constraints of the unit density meeting the limit value and mechanical performance requirements, and the maximum mechanical abuse that the battery pack must withstand must not cause internal short circuit failure, so as to obtain the constrained optimization objective.

Images