Battery module heat dissipation channel optimization design method based on genetic algorithm

Abstract

The invention discloses a genetic algorithm-based battery module heat dissipation channel optimization design method. The method comprises the following steps of S1, acquiring operation data of a battery module under a plurality of dynamic load working conditions; s2, constructing an initial heat dissipation channel topological structure model based on a physical packaging structure of the battery module and a flow channel layout constraint; s3, setting a multi-objective optimization objective function set; s4, performing heat-fluid-solid coupling simulation on each heat dissipation channel structure candidate body in the first generation population structure by taking the heat generation power matrix as a heat source boundary condition; s5, performing fitness calculation on the obtained heat-flow performance index set by using a multi-objective genetic algorithm based on Pareto optimal sorting; s6, repeating S4 and S5 until a preset convergence criterion or an upper limit of the number of iterations is met; and S7, carrying out comprehensive performance verification and manufacturing feasibility evaluation on the heat dissipation channel topological structure model. The method has the advantages of being high in global search capability, high in heat source response authenticity and good in structural expression diversity.

Description

Technical Field

[0001] The present invention relates to the field of thermal management technology, and in particular to a method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm. Background Art

[0002] With the large-scale development of new energy vehicles and energy storage systems, thermal management of battery modules has become one of the key technical bottlenecks restricting their performance, safety, and lifespan. Especially under complex operating conditions such as fast charging, high-rate discharge, and operation in high-temperature environments, the heat generated by the battery accumulates rapidly. If the heat dissipation path is not designed properly, it will lead to uneven temperature distribution within the module and an increased risk of local thermal runaway, further inducing battery capacity decay, reduced system efficiency, and even thermal failure. Therefore, how to improve the overall heat dissipation capacity and heat distribution uniformity of the battery module heat dissipation channel while meeting structural constraints, manufacturing feasibility, and energy consumption control has become a major technical challenge in the current field of battery thermal management.

[0003] In the prior art, the heat dissipation structure design of battery modules mostly relies on empirical rules or parameter sweep methods, and CFD simulation is used to assist in optimizing channel size and layout. Although such methods can obtain feasible solutions within the local parameter space, due to the large number of design variables, complex heat source behavior, and strict structural constraints, traditional optimization strategies generally have the following defects. First, there is a lack of modeling and dynamic response capabilities for the time-varying characteristics of battery heat sources. Most optimization methods are based on static heat source models for channel structure design, which cannot truly reflect the heat distribution evolution of the battery under different working conditions (such as charging, discharging, and static), resulting in deviations between the optimization results and the actual thermal field and poor generalization capabilities. Second, the global search capability of the optimization algorithm is limited and it is easy to fall into local optimality. Commonly used single-objective or heuristic algorithms are difficult to fully explore the optimal solution set in high-dimensional nonlinear structural space, especially when it is necessary to take into account multiple objectives such as thermal uniformity, structural complexity, and energy loss. The optimization efficiency is low and the result stability is poor. Third, there is a lack of systematic modeling of the channel structure perturbation mechanism and population generation method. The current design process mostly uses regularized parameter perturbations or template replacements to generate structural candidates. The structural diversity is insufficient and there is a lack of adaptive control mechanisms for topological paths, flow nodes, and channel density, which limits the search capabilities of genetic algorithms at the topological level. Fourth, the optimization closed-loop process is broken, and the design results are difficult to directly map to engineering manufacturable structures. Existing optimization results are mostly output in a graphical manner, without considering CAD convertibility and simulation verification traceability, which increases the product development cycle and the difficulty of engineering implementation. Fifth, there is a lack of a unified multi-physics field performance evaluation system and fitness function construction method. The coupling relationship between the three fields of heat, flow, and solid has not formed effective feedback in the optimization framework, resulting in a disconnect between the optimization process and thermal simulation, one-sided evaluation criteria, and it is difficult to reflect the comprehensive performance of channel design at the system level.

[0004] Therefore, how to provide a battery module heat dissipation channel optimization design method based on genetic algorithm is a problem that those skilled in the art urgently need to solve. Summary of the Invention

[0005] One purpose of the present invention is to propose a method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm. The present invention integrates a multi-objective genetic algorithm with a time-varying modeling mechanism of a battery heat source. By constructing a closed loop of heat-fluid-solid coupling simulation, a topological structure perturbation generation method, and a multi-physical field fitness evaluation system, the method realizes the collaborative optimization of the heat dissipation channel structure under performance indicators such as heat flow uniformity, structural complexity, and energy loss. The method has the advantages of strong global search capability, high authenticity of heat source response, good structural expression diversity, and feasible design process engineering. It is suitable for thermal management structure optimization design scenarios of high-density battery modules such as electric vehicles and energy storage systems.

[0006] According to an embodiment of the present invention, a method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm includes the following steps:

[0007] S1. Collecting operating data of the battery module under multiple dynamic load conditions and calculating a heat generation power matrix based on the operating data;

[0008] S2. Based on the physical packaging structure and flow channel layout constraints of the battery module, an initial heat dissipation channel topology model is constructed. Several heat dissipation channel structure candidates are generated as the first generation population structure through parameter perturbation and structural combination.

[0009] S3. Setting a multi-objective optimization objective function set;

[0010] S4. Using the heat generation power matrix as the heat source boundary condition, a thermal-fluid-solid coupling simulation is performed on each heat dissipation channel structure candidate in the first generation population structure to obtain a corresponding set of thermal-fluid performance indicators;

[0011] S5. Utilize a multi-objective genetic algorithm based on Pareto optimal sorting to calculate the fitness of the obtained heat-flow performance index set, and perform crossover, mutation, and screening operations based on the multi-objective optimization objective function set to generate a new generation of population structure candidates;

[0012] S6. Repeat the process described in step S4 and step S5 until a preset convergence criterion or an upper limit of the number of iterations is met, and output a heat dissipation channel topology structure model with an optimal trade-off solution on the multi-objective optimization objective function set;

[0013] S7. Convert the heat dissipation channel topology structure model with the optimal trade-off solution into a manufacturable CAD geometric model and import it into the computational fluid dynamics simulation platform for comprehensive performance verification and manufacturing feasibility assessment.

[0014] Optionally, the S1 specifically includes:

[0015] S11. Collecting operating data of the battery module under multiple dynamic load conditions, wherein the operating data includes time series, current density, and battery internal resistance;

[0016] S12. Calculate the thermal power generation at each moment based on the current density and battery internal resistance at different time series;

[0017] S13. Discretize the heat generation power Q(t) in the time dimension and construct the heat generation power matrix Q = {Q(t1), Q(t2), ..., Q(t n )}, where t1, t2, …, t n is the sampling time node.

[0018] Optionally, the S2 specifically includes:

[0019] S21. Obtaining the physical packaging structure of the battery module, wherein the physical packaging structure includes the arrangement of battery cells, the boundary dimensions of the battery module, and the positions and dimensions of the cooling medium inlet and outlet;

[0020] S22. Based on the physical packaging structure and flow channel layout constraints of the battery module, determine the initial topological construction area of ​​the heat dissipation channel and construct an initial heat dissipation channel topological structure model. The initial topological construction area is the remaining layout space within the battery module boundary, excluding the area occupied by the battery cells and meeting the accessibility requirements of the inlet and outlet connection paths. The flow channel layout constraints include that the channel width is not less than a lower limit, the channel centerline does not cross the module boundary, the minimum spacing between channels is not less than a threshold, the inlet and outlet positions are fixed, and the overall structure maintains bilateral symmetry.

[0021] S23, setting the perturbation parameter set of the structure Θ = {θ1, θ2, θ3, θ4} based on the geometric adjustable parameters in the initial heat dissipation channel topology model, where θ1 represents the channel segment width, θ2 represents the node connection angle, θ3 represents the number of channel branches, and θ4 represents the distance between adjacent channels;

[0022] S24, introducing a structural perturbation operation on the initial heat dissipation channel topology model based on the perturbation parameter set Θ, for each parameter θ in the perturbation parameter set i Construct the perturbation function f d (θ i ):

[0023] f d (θ i )=θ i +η i ·sin((ω i ·r i );

[0024] Among them, η i is the parameter θ i The maximum disturbance amplitude, ω i is the disturbance frequency coefficient, r i ∈[0,1] is a normalized random variable;

[0025] S25. Based on the parameter combination after the structural perturbation, generate multiple heat dissipation channel structure candidates with different channel topology characteristics, and assemble all the heat dissipation channel structure candidates into a first-generation population structure.

[0026] Optionally, the S3 specifically includes:

[0027] Construct a multi-objective optimization objective function set F = {f1, f2, f3}, where:

[0028] f1 is the heat flux density distribution uniformity evaluation function, which is used to measure the balance of temperature distribution in different areas of the heat dissipation channel:

[0029]

[0030] in, is the gradient operator, T(x,y) is the steady-state temperature distribution at point (x,y), Q(t) is the heat generation power per unit time, Ω is the heat source area, k is the thermal conductivity, A is the cross-sectional area of ​​the channel structure, x0 and y0 are the horizontal and vertical coordinate positions of the center of the heat source area, respectively;

[0031] f2 is the channel structure complexity evaluation function, which is used to comprehensively describe the geometric complexity of the channel layout:

[0032]

[0033] Where L is the total length of the channel path, A c is the actual occupied area of ​​the channel, κ i is the local curvature of the i-th channel segment, n is the number of structural branch nodes, N max is the maximum number of nodes allowed, γ1, γ2, γ3 are the preset weight coefficients;

[0034] f3 is the flow pressure energy loss function, which is used to evaluate the pressure loss caused by structural resistance during the flow of cooling medium in the channel:

[0035]

[0036] Among them, p in (l) is the inlet pressure at the path position l, p out (l) is the corresponding outlet pressure, Q(l) is the volume flow rate per unit length, and L is the total length of the channel path.

[0037] Optionally, the S4 specifically includes:

[0038] S41. Each heat dissipation channel structure candidate in the first generation population structure is used as the structure input of the thermal-fluid-solid coupling simulation. At the same time, the heat generation power matrix Q = {Q(t1), Q(t2), ..., Q(t n )} as the heat source boundary condition input;

[0039] S42. performing a thermal-fluid-solid coupling simulation on each heat dissipation channel structure candidate in a three-dimensional finite volume modeling environment;

[0040] S43. After each thermal-fluid-solid coupling simulation is completed, a thermal-fluid performance index set P corresponding to each heat dissipation channel structure candidate is extracted from the thermal field simulation, the flow field simulation, and the solid structure field simulation. max ,T avg ,Δp max ,V,φ avg ,δ max}, where T max is the maximum temperature in the temperature field, T avg is the average temperature of the channel area, which is calculated by the temperature distribution function T(x,y,z,t), Δp max is the maximum pressure difference generated by the cooling medium on the flow path, which is calculated by the pressure field p(x, y, z, t), V is the total volume of the heat dissipation channel structure, which is calculated by the structural model geometry, and φ avg is the average heat flux per unit volume, and the heat flow vector After taking the modular integral and normalizing it, δ max It is the maximum displacement value that occurs in the structural response, which is extracted from the displacement function δ(x, y, z, t) of the solid structure field.

[0041] Optionally, the thermal field simulation in the heat-fluid-solid coupling simulation specifically includes:

[0042] Perform thermal field simulation on the candidate heat dissipation channel structure and establish the governing equation of the temperature field T(x,y,z,t):

[0043]

[0044] in, is the gradient operator, T represents the temperature field distribution function, represents the partial derivative of the temperature field distribution over time, ρ represents the channel material density, c p represents specific heat capacity, k represents thermal conductivity, and Q(t) is the heat generation power;

[0045] During the simulation process, the boundary conditions include the convection heat transfer boundary of the channel outer wall and the temperature continuity boundary of the cooling medium interface, and the initial condition is the battery module ambient temperature.

[0046] Optionally, the flow field simulation in the thermal-fluid-solid coupling simulation specifically includes:

[0047] Perform flow field simulation on the candidate heat dissipation channel structure to establish the flow velocity field The governing equations are:

[0048]

[0049] in, is the gradient operator, represents the cooling medium velocity field distribution, ρ f represents the cooling medium density, p represents the pressure field, μ f Indicates the dynamic viscosity of the cooling medium, represents the partial derivative of the cooling medium velocity field with time;

[0050] During the simulation, the inlet boundary is set as a constant velocity boundary, the outlet boundary is set as a constant pressure boundary, the inner wall of the channel is set as a no-slip boundary, and the initial condition is that the cooling medium is stationary.

[0051] Optionally, the solid structure field simulation in the thermal-fluid-solid coupling simulation specifically includes:

[0052] Perform solid structural field simulation on the candidate heat dissipation channel structure and establish the governing equations for the structural response displacement field δ(x, y, z, t):

[0053]

[0054] in, is the gradient operator, δ represents the solid structure displacement vector field, σ represents the stress tensor, and f b represents the volume force density, ρ s represents the density of the channel wall material;

[0055] The stress tensor σ and the strain tensor ε satisfy the following thermal stress relationship:

[0056] σ=D:(ε-α s ΔTI);

[0057] Where D represents the material elastic matrix, α s represents the thermal expansion coefficient of the channel material, ΔT represents the temperature rise obtained from the thermal field simulation, and I is the unit tensor;

[0058] During the simulation, the solid structure boundary is set as a combination of fixed boundary and symmetric boundary, and the initial condition is zero displacement state.

[0059] Optionally, the S5 specifically includes:

[0060] S51. Using a multi-objective genetic algorithm based on Pareto optimal sorting, normalize the thermal-fluid performance index set corresponding to each heat dissipation channel structure candidate in the first generation population structure, and calculate the heat flux density distribution uniformity index, structural complexity index, and pressure energy loss index based on the multi-objective optimization objective function set, and perform non-dominated level division and crowding distance calculation in the objective function space;

[0061] S52. Construct a ranking mapping relationship based on the non-dominated rank and crowding distance results, and use a tournament selection mechanism to select heat dissipation channel structure candidates in the Pareto front region from the current population as parent individuals;

[0062] S53, inputting the selected parent structure into the genetic operation module in the form of structural parameter encoding, using a uniform crossover operator to combine structural parameters among multiple parents, constructing a new topological structure perturbation combination, and performing a mutation operation based on the probability density function set by the perturbation parameter set to improve structural diversity;

[0063] S54. After the generation structure candidate is constructed, the heat generation power matrix is ​​used as the heat source boundary condition, and a thermal-fluid-solid coupling simulation process is performed to extract its temperature field, flow velocity field, and structural response index to form a corresponding thermal-fluid performance index set;

[0064] S55. The offspring structural candidates and their thermal-fluid performance indicators are merged with the previous generation non-dominated structure set to form an intermediate population structure. The solution set distribution state of the population structure in the objective function space is optimized and screened again based on the Pareto ranking and crowding distance calculation. Several structural candidates with higher ranking levels and good solution set diversity are selected to form a new generation of population structure.

[0065] Optionally, the S6 specifically includes:

[0066] S61, set the iteration termination criterion, the iteration termination criterion includes the maximum number of iterations G max and the convergence threshold δ conv ;

[0067] S62. After each generation of new population structure is generated, perform non-dominated sorting on all heat dissipation channel structure candidates in the new generation population structure based on the multi-objective optimization objective function set, and extract the current generation non-dominated structure set. And compared with the non-dominated structure set saved in the previous iteration Compare and terminate the iteration process if any of the following conditions is met:

[0068] Current iteration number g≥Gmax ;

[0069] The change in the distance between the non-dominated structure sets of the current generation and the previous generation

[0070] S63. After the iteration termination criterion is met, a set of heat dissipation channel structure candidates with the highest non-dominated level is selected from the final generation population structure and output as a heat dissipation channel topology structure model with an optimal trade-off solution.

[0071] The beneficial effects of the present invention are:

[0072] (1) The present invention introduces an optimization framework that integrates time-varying heat source modeling and multi-objective genetic algorithm. By constructing a heat generation power matrix under the dynamic working conditions of the battery module, the authenticity and dynamic response capability of the heat source boundary conditions are significantly improved. This breaks through the limitation of traditional static thermal modeling methods that cannot adapt to changes in actual working conditions, and provides more accurate input support for thermal field distribution prediction and channel structure optimization.

[0073] (2) The present invention constructs a fitness evaluation mechanism for multi-physical field coupling, combines performance objective functions such as heat flux density distribution gradient, channel structure complexity and flow pressure loss, establishes a globally searchable multi-objective optimization system, and controls population evolution through non-dominated sorting and crowding distance, significantly enhancing the ability of optimization results to balance multiple objectives and avoiding the problems of traditional optimization methods that are prone to falling into local optimality and single evaluation dimension.

[0074] (3) The present invention proposes a parametric generation method based on topological structure perturbation modeling, combines the perturbation function with a set of geometric adjustable parameters to construct a population of structural candidates, and performs feedback iteration based on the results of thermal-fluid-solid simulation. While completing the search for the optimal topological structure, it outputs a CAD geometric model, taking into account both structural performance optimization and manufacturing feasibility, and constructs an integrated closed-loop design process of optimization-simulation-output, thereby improving the automation level of structural optimization design and the engineering implementation capability. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0076] Figure 1 This is the overall flow chart of the battery module heat dissipation channel optimization design method based on genetic algorithm proposed in the present invention;

[0077] Figure 2 Schematic diagram of the process of constructing the initial heat dissipation channel topology model and generating structural disturbances proposed in the present invention;

[0078] Figure 3 This is a flowchart of the linked execution of the multi-objective genetic algorithm optimization and thermal-fluid-solid coupling simulation proposed in the present invention. DETAILED DESCRIPTION

[0079] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0080] refer to Figure 1-Figure 3 A method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm comprises the following steps:

[0081] S1. Collecting operating data of the battery module under multiple dynamic load conditions and calculating a heat generation power matrix based on the operating data;

[0082] S2. Based on the physical packaging structure and flow channel layout constraints of the battery module, an initial heat dissipation channel topology model is constructed. Several heat dissipation channel structure candidates are generated as the first generation population structure through parameter perturbation and structural combination.

[0083] S3. Setting a multi-objective optimization objective function set;

[0084] S4. Using the heat generation power matrix as the heat source boundary condition, a thermal-fluid-solid coupling simulation is performed on each heat dissipation channel structure candidate in the first generation population structure to obtain a corresponding set of thermal-fluid performance indicators;

[0085] S5. Utilize a multi-objective genetic algorithm based on Pareto optimal sorting to calculate the fitness of the obtained heat-flow performance index set, and perform crossover, mutation, and screening operations based on the multi-objective optimization objective function set to generate a new generation of population structure candidates;

[0086] S6. Repeat the process described in step S4 and step S5 until a preset convergence criterion or an upper limit of the number of iterations is met, and output a heat dissipation channel topology structure model with an optimal trade-off solution on the multi-objective optimization objective function set;

[0087] S7. Convert the heat dissipation channel topology structure model with the optimal trade-off solution into a manufacturable CAD geometric model and import it into the computational fluid dynamics simulation platform for comprehensive performance verification and manufacturing feasibility assessment.

[0088] This implementation combines dynamic heat source modeling with a multi-objective genetic algorithm optimization process to construct a closed-loop design path based on thermal-fluid-solid coupling simulation feedback. This not only enables the intelligent generation of the heat dissipation channel topology of the battery module under multiple working conditions, but also establishes a multi-objective evaluation system based on multiple indicators such as temperature distribution, pressure loss, and structural response. This enables the channel design process to have global search capabilities and physical constraint perception capabilities, thereby improving the accuracy, responsiveness, and manufacturability of structural optimization. It combines the thermal response authenticity of physical modeling with the optimization depth of intelligent algorithms, providing a systematic design tool for high-performance battery module thermal management.

[0089] In this embodiment, S1 specifically includes:

[0090] S11. Collecting operating data of the battery module under multiple dynamic load conditions, wherein the operating data includes time series, current density, and battery internal resistance;

[0091] S12. Calculate the thermal power generation at each moment based on the current density and battery internal resistance at different time series;

[0092] S13. Discretize the heat generation power Q(t) in the time dimension and construct the heat generation power matrix Q = {Q(t1), Q(t2), ..., Q(t n )}, where t1, t2, …, t n is the sampling time node.

[0093] This implementation method is based on the operating status of the battery module under multiple dynamic load conditions. It collects core variables such as time series, current density, and battery internal resistance, constructs a heat generation power function, and discretizes it in the time dimension to form a heat generation power matrix. It provides accurate modeling capabilities for the evolution of heat source distribution with operating conditions. It not only retains the timing characteristics of the heat source, but also provides controllable and adjustable boundary inputs for subsequent simulations, improves the physical accuracy and real-time response of thermal modeling, and combines the dynamic expression ability of time series with the quantitative descriptive power of heat calculation to provide high-quality heat source data support for the optimization algorithm.

[0094] In this embodiment, S2 specifically includes:

[0095] S21. Obtaining the physical packaging structure of the battery module, wherein the physical packaging structure includes the arrangement of battery cells, the boundary dimensions of the battery module, and the positions and dimensions of the cooling medium inlet and outlet;

[0096] S22. Based on the physical packaging structure and flow channel layout constraints of the battery module, determine the initial topological construction area of ​​the heat dissipation channel and construct an initial heat dissipation channel topological structure model. The initial topological construction area is the remaining layout space within the battery module boundary, excluding the area occupied by the battery cells and meeting the accessibility requirements of the inlet and outlet connection paths. The flow channel layout constraints include that the channel width is not less than a lower limit, the channel centerline does not cross the module boundary, the minimum spacing between channels is not less than a threshold, the inlet and outlet positions are fixed, and the overall structure maintains bilateral symmetry.

[0097] S23, setting the perturbation parameter set of the structure Θ = {θ1, θ2, θ3, θ4} based on the geometric adjustable parameters in the initial heat dissipation channel topology model, where θ1 represents the channel segment width, θ2 represents the node connection angle, θ3 represents the number of channel branches, and θ4 represents the distance between adjacent channels;

[0098] S24, introducing a structural perturbation operation on the initial heat dissipation channel topology model based on the perturbation parameter set Θ, for each parameter θ in the perturbation parameter set i Construct the perturbation function f d (θ i ):

[0099] f d (θ i )=θ i +η i ·sin((ω i ·r i );

[0100] Among them, η i is the parameter θ i The maximum disturbance amplitude, ω i is the disturbance frequency coefficient, r i ∈[0,1] is a normalized random variable;

[0101] S25. Based on the parameter combination after the structural perturbation, generate multiple heat dissipation channel structure candidates with different channel topology characteristics, and assemble all the heat dissipation channel structure candidates into a first-generation population structure.

[0102] This embodiment obtains the physical packaging structure information of the battery module, combines the battery cell arrangement method and the cooling interface boundary, and clearly defines the layout area of ​​the heat dissipation channel topology structure. On this basis, adjustable parameters such as channel width, connection angle, number of branches and spacing are set to construct a perturbation parameter set. The sinusoidal perturbation function is further introduced to realize the systematic perturbation operation of multi-dimensional structural parameters, thereby generating multiple groups of topological structure candidates with structural differences. This not only improves the expressive power and diversity of topological modeling, but also enhances the diverse coverage of the initial population of the genetic algorithm, provides a richer structural solution space for optimization search, and improves the controllability and global search capability of the channel design process.

[0103] In this embodiment, S3 specifically includes:

[0104] Construct a multi-objective optimization objective function set F = {f1, f2, f3}, where:

[0105] f1 is the heat flux density distribution uniformity evaluation function, which is used to measure the balance of temperature distribution in different areas of the heat dissipation channel:

[0106]

[0107] in, is the gradient operator, T(x,y) is the steady-state temperature distribution at point (x,y), Q(t) is the heat generation power per unit time, Ω is the heat source area, k is the thermal conductivity, A is the cross-sectional area of ​​the channel structure, x0 and y0 are the horizontal and vertical coordinate positions of the center of the heat source area, respectively;

[0108] f2 is the channel structure complexity evaluation function, which is used to comprehensively describe the geometric complexity of the channel layout:

[0109]

[0110] Where L is the total length of the channel path, A c is the actual occupied area of ​​the channel, κ i is the local curvature of the i-th channel segment, n is the number of structural branch nodes, N max is the maximum number of nodes allowed, γ1, γ2, γ3 are the preset weight coefficients;

[0111] f3 is the flow pressure energy loss function, which is used to evaluate the pressure loss caused by structural resistance during the flow of cooling medium in the channel:

[0112]

[0113] Among them, p in (l) is the inlet pressure at the path position l, p out(l) is the corresponding outlet pressure, Q(l) is the volume flow rate per unit length, and L is the total length of the channel path.

[0114] This implementation comprehensively characterizes the thermal, flow, and geometric performance of the heat dissipation channel structure by constructing a set of multi-objective optimization objective functions encompassing heat flux density distribution uniformity, structural complexity, and flow pressure energy loss. The proposed heat flux uniformity index function accurately quantifies the temperature gradient variation pattern, the structural complexity function combines path length, curvature, and branch density for geometric modeling, and the pressure loss function expresses the actual flow resistance process based on path integrals. Together, these three constitute a multi-dimensional fitness evaluation system that not only enhances the genetic algorithm's discriminative ability during structural screening but also ensures the balance of optimization results across multiple performance objectives, providing a systematic optimization basis for channel structure design under complex and multi-working conditions.

[0115] In this embodiment, the S4 specifically includes:

[0116] S41. Each heat dissipation channel structure candidate in the first generation population structure is used as the structure input of the thermal-fluid-solid coupling simulation. At the same time, the heat generation power matrix Q = {Q(t1), Q(t2), ..., Q(t n )} as the heat source boundary condition input;

[0117] S42. performing a thermal-fluid-solid coupling simulation on each heat dissipation channel structure candidate in a three-dimensional finite volume modeling environment;

[0118] S43. After each thermal-fluid-solid coupling simulation is completed, a thermal-fluid performance index set P corresponding to each heat dissipation channel structure candidate is extracted from the thermal field simulation, the flow field simulation, and the solid structure field simulation. max ,T avg ,Δp max ,V,φ avg ,δ max}, where T max is the maximum temperature in the temperature field, T avg is the average temperature of the channel area, which is calculated by the temperature distribution function T(x,y,z,t), Δp max is the maximum pressure difference generated by the cooling medium on the flow path, which is calculated by the pressure field p(x, y, z, t), V is the total volume of the heat dissipation channel structure, which is calculated by the structural model geometry, and φ avg is the average heat flux per unit volume, and the heat flow vector After taking the modular integral and normalizing it, δ max It is the maximum displacement value that occurs in the structural response, which is extracted from the displacement function δ(x, y, z, t) of the solid structure field.

[0119] In this embodiment, the thermal field simulation in the thermal-fluid-solid coupling simulation specifically includes:

[0120] Perform thermal field simulation on the candidate heat dissipation channel structure and establish the governing equation of the temperature field T(x,y,z,t):

[0121]

[0122] in, is the gradient operator, T represents the temperature field distribution function, represents the partial derivative of the temperature field distribution over time, ρ represents the channel material density, c p represents specific heat capacity, k represents thermal conductivity, and Q(t) is the heat generation power;

[0123] During the simulation process, the boundary conditions include the convection heat transfer boundary of the channel outer wall and the temperature continuity boundary of the cooling medium interface, and the initial condition is the battery module ambient temperature.

[0124] In this embodiment, the flow field simulation in the thermal-fluid-solid coupling simulation specifically includes:

[0125] Perform flow field simulation on the candidate heat dissipation channel structure to establish the flow velocity field The governing equations are:

[0126]

[0127] in, is the gradient operator, represents the cooling medium velocity field distribution, ρ f represents the cooling medium density, p represents the pressure field, μ f Indicates the dynamic viscosity of the cooling medium, represents the partial derivative of the cooling medium velocity field with time;

[0128] During the simulation, the inlet boundary is set as a constant velocity boundary, the outlet boundary is set as a constant pressure boundary, the inner wall of the channel is set as a no-slip boundary, and the initial condition is that the cooling medium is stationary.

[0129] In this embodiment, the solid structure field simulation in the thermal-fluid-solid coupling simulation specifically includes:

[0130] Perform solid structural field simulation on the candidate heat dissipation channel structure and establish the governing equations for the structural response displacement field δ(x, y, z, t):

[0131]

[0132] in, is the gradient operator, δ represents the solid structure displacement vector field, σ represents the stress tensor, and fb represents the volume force density, ρ s represents the density of the channel wall material;

[0133] The stress tensor σ and the strain tensor ε satisfy the following thermal stress relationship:

[0134] σ=D:(ε-α s ΔTI);

[0135] Where D represents the material elastic matrix, α s represents the thermal expansion coefficient of the channel material, ΔT represents the temperature rise obtained from the thermal field simulation, and I is the unit tensor;

[0136] During the simulation, the solid structure boundary is set as a combination of fixed boundary and symmetric boundary, and the initial condition is zero displacement state.

[0137] This implementation simulates the realistic physical response of each candidate structure under dynamic thermal load by inputting first-generation heat dissipation channel structure candidates into a three-dimensional coupled thermal-fluid-solid simulation environment and introducing the heat generation power matrix as a time-varying heat source boundary condition. Key metrics such as maximum temperature, average temperature, heat flux uniformity, pressure difference, and structural displacement are extracted from the simulation results to establish a multi-dimensional thermal-fluid performance index set. This fully reflects the thermal management effectiveness and structural stability of the channel structure in actual operation. This not only strengthens the physical foundation of fitness assessment but also improves the accuracy of the objective function response during the optimization process, providing refined and quantified physical support for the structural optimization results.

[0138] In this embodiment, the S5 specifically includes:

[0139] S51. Using a multi-objective genetic algorithm based on Pareto optimal sorting, normalize the thermal-fluid performance index set corresponding to each heat dissipation channel structure candidate in the first generation population structure, and calculate the heat flux density distribution uniformity index, structural complexity index, and pressure energy loss index based on the multi-objective optimization objective function set, and perform non-dominated level division and crowding distance calculation in the objective function space;

[0140] S52. Construct a ranking mapping relationship based on the non-dominated rank and crowding distance results, and use a tournament selection mechanism to select heat dissipation channel structure candidates in the Pareto front region from the current population as parent individuals;

[0141] S53, inputting the selected parent structure into the genetic operation module in the form of structural parameter encoding, using a uniform crossover operator to combine structural parameters among multiple parents, constructing a new topological structure perturbation combination, and performing a mutation operation based on the probability density function set by the perturbation parameter set to improve structural diversity;

[0142] S54. After the generation structure candidate is constructed, the heat generation power matrix is ​​used as the heat source boundary condition, and a thermal-fluid-solid coupling simulation process is performed to extract its temperature field, flow velocity field, and structural response index to form a corresponding thermal-fluid performance index set;

[0143] S55. The offspring structural candidates and their thermal-fluid performance indicators are merged with the previous generation non-dominated structure set to form an intermediate population structure. The solution set distribution state of the population structure in the objective function space is optimized and screened again based on the Pareto ranking and crowding distance calculation. Several structural candidates with higher ranking levels and good solution set diversity are selected to form a new generation of population structure.

[0144] This implementation introduces a multi-objective genetic algorithm to normalize and non-dominatedly sort the thermal-fluid performance index set, and combines it with crowding distance to assess individual performance, effectively constructing an evaluation mechanism capable of balancing multiple performance metrics. Furthermore, a tournament selection strategy is employed to enhance the probability of retaining high-quality structures. A new generation of structural candidates is generated through uniform crossover of structural parameters and perturbation mutation. Simulation feedback is used to establish an intermediate population structure. Finally, a Pareto ranking is performed using a multi-objective function set to select the optimal structure. This significantly enhances the global search capability and multi-objective coordinated control capabilities of the optimization process, improving the diversity, stability, and physical feasibility of the optimization results.

[0145] In this embodiment, S6 specifically includes:

[0146] S61, set the iteration termination criterion, the iteration termination criterion includes the maximum number of iterations G max and the convergence threshold δ conv ;

[0147] S62. After each generation of new population structure is generated, perform non-dominated sorting on all heat dissipation channel structure candidates in the new generation population structure based on the multi-objective optimization objective function set, and extract the current generation non-dominated structure set. And compared with the non-dominated structure set saved in the previous iteration Compare and terminate the iteration process if any of the following conditions is met:

[0148] Current iteration number g≥G max ;

[0149] The change in the distance between the non-dominated structure sets of the current generation and the previous generation

[0150] S63. After the iteration termination criterion is met, a set of heat dissipation channel structure candidates with the highest non-dominated level is selected from the final generation population structure and output as a heat dissipation channel topology structure model with an optimal trade-off solution.

[0151] This implementation introduces a phased convergence judgment mechanism for each round of population structure evolution by setting a dual iterative termination criterion consisting of a maximum number of iterations and a convergence threshold. This mechanism dynamically monitors the distance between the current and previous generation's non-dominated structure sets, automatically terminating the evolution process when the number of iterations reaches its upper limit or when the optimal structure set stabilizes. This ensures that the optimization algorithm achieves a sufficient search while avoiding invalid iterations and wasted computational resources. Furthermore, after the iterations terminate, the optimal non-dominated structure is extracted from the final population as the output, enhancing the adaptive control capabilities of the optimization process and the stability of the final structure, ensuring the reliability of the optimization design results in terms of physical feasibility and multi-objective compromise performance.

[0152] Example 1:

[0153] To verify the engineering feasibility and actual performance improvement capabilities of this invention in the design of complex battery module heat dissipation channel structures, the research team, in collaboration with a large new energy vehicle manufacturer in East China, completed a complete optimization design practice on its thermal management simulation platform and production test line from August to September 2024. The company selected a 72S12P liquid-cooled module platform for the test, which uses high-energy-density square cells and has a system capacity of 83.2kWh. The module cooling structure is a composite layout of parallel channels at the bottom and auxiliary heat exchange paths on both sides. The initial design had multiple thermal management shortcomings, especially under high-temperature and high-rate conditions, which showed typical problems such as uneven heat distribution, excessive pressure drop, and complex structural layout.

[0154] The original design approach was based primarily on experience, with parameters adjusted through CFD simulation and relying on engineers to gradually adjust parameters. Actual evaluations showed that under high-rate 1.5C discharge, the core temperature of the module was significantly higher than the edges, with a maximum temperature difference of 13°C. Sudden changes in pressure drop occurred in localized channels, resulting in dead zones and a sharp drop in heat dissipation efficiency at the edges. Furthermore, on the manufacturing side, this irregular structure placed high demands on molds, extending the manufacturing cycle and increasing costs.

[0155] After introducing the proposed method, the module's operating data was first collected under four typical dynamic operating conditions: stable discharge at room temperature, high-rate rapid discharge, low-speed operation in urban congestion, and restart after high summer sun exposure. Under these conditions, time, current density, and cell internal resistance were measured to form a time-series thermal power function. This discretized thermal power matrix was then constructed to serve as the thermal boundary condition input for the optimization process.

[0156] Subsequently, a topological construction region was constructed based on the module's geometric boundary conditions and flow channel interface constraints. A perturbed structural template was established using parameters such as channel width, number of branch nodes, connection angle, and minimum spacing. A sinusoidal perturbation function was used to generate the initial population, and the thermal-fluid-solid simulation process was then initiated. During the simulation, the temperature distribution, pressure drop distribution, heat flux field, and structural response of each structure were extracted for the corresponding heat source input. Multiple indicators, such as heat flux uniformity, structural complexity, and flow pressure loss, were incorporated into the objective function set to form the optimization direction.

[0157] The NSGA-II multi-objective genetic algorithm was used for optimization. Within each generation, non-dominated sorting, crowding calculation, tournament selection, uniform crossover, and perturbation mutation were performed, ultimately selecting the optimal structural topology as the multi-objective Pareto front solution. The structure was modeled using CAD and imported into a CFD and structural strength simulation platform, forming a complete verification process.

[0158] The optimized structure shows stable improvement in both actual measurements and simulations. The following is a performance comparison between the present invention and the traditional design under different typical working conditions:

[0159] Table 1 Performance comparison between the present invention and the traditional structure under typical working conditions

[0160]

[0161]

[0162] As can be seen from the table above, in the normal temperature discharge scenario, the present invention controls the maximum module temperature within 43.1°C, reduces the pressure drop to 18.3kPa, and improves the heat flow uniformity index to 0.82, indicating that the cooling channel has formed a more balanced heat exchange capacity after structural perturbation optimization. Under high-rate discharge, the structure of the present invention effectively avoids heat accumulation, and the maximum temperature drops by nearly 7°C compared to the traditional structure, and the pressure drop is also reduced by 4.3kPa. The heat flow is more evenly distributed in the channel, and the structural complexity score drops to 0.62. The manufacturing end does not need to use special molds to quickly process, reducing costs.

[0163] Urban congestion scenarios are particularly challenging for the responsiveness of the heat dissipation structure under conditions of low flow rates and frequent thermal shocks. Due to the optimization of channel branches and symmetrical path control, the structure of the present invention achieves a more sufficient redistribution of the coolant, reduces the maximum temperature by 5.9°C, reduces the pressure drop by 2.7kPa, and the heat flow distribution is displayed as a continuous gradient change area in the simulation diagram. Hot spot faults rarely occur, and the system stability is significantly enhanced. Under the most challenging high-temperature exposure conditions, the structure of the present invention controls the extreme temperature of the traditional structure of 57.6°C to 50.8°C, and the pressure drop is controlled within 22.7kPa, effectively avoiding BMS alarms and performance degradation.

[0164] In terms of manufacturing feasibility, the proposed structure can be directly converted into a standard STEP format and imported into the company's existing integrated CAD-CAM tool chain. Compared to traditional structures, the number of path nodes is reduced by approximately 36%, resulting in a simpler CNC path, shortening the machining cycle from 13.4 days to 7.2 days, and reducing the process scrap rate from 7.3% to 1.6%. Furthermore, engineers stated that the proposed structure demonstrated good interface compatibility in 3D pre-assembly simulations, improving overall process stability.

[0165] Based on this data, the project's technical lead stated in their final report that this method not only optimizes the thermal performance of the heat dissipation channel but also significantly improves structural design efficiency and manufacturing feasibility, representing a promising technology path for intelligent design of thermal management structures. The company's management has decided to integrate this method into the standard process of its 2025 thermal system simulation platform, ensuring long-term use as a core module for channel structure optimization.

[0166] In summary, this embodiment clearly verifies the systematic optimization capability of the present invention under the constraints of multiple working conditions, dynamic thermal loads, and battery structure complexity. It has the advantages of realistic heat source modeling, efficient search path, intuitive and manufacturable structural output, and smooth engineering process docking. It is an innovative structural optimization technology with potential for actual engineering implementation.

[0167] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A battery module heat dissipation channel optimization design method based on genetic algorithm, characterized in that: The steps include: S1. Collecting operating data of the battery module under multiple dynamic load conditions and calculating a heat generation power matrix based on the operating data; S2. Based on the physical packaging structure and flow channel layout constraints of the battery module, an initial heat dissipation channel topology model is constructed. Several heat dissipation channel structure candidates are generated as the first generation population structure through parameter perturbation and structural combination. S3. Setting a multi-objective optimization objective function set; S4. Using the heat generation power matrix as the heat source boundary condition, a thermal-fluid-solid coupling simulation is performed on each heat dissipation channel structure candidate in the first generation population structure to obtain a corresponding set of thermal-fluid performance indicators; S5. Utilize a multi-objective genetic algorithm based on Pareto optimal sorting to calculate the fitness of the obtained heat-flow performance index set, and perform crossover, mutation, and screening operations based on the multi-objective optimization objective function set to generate a new generation of population structure candidates; S6. Repeat the process described in step S4 and step S5 until a preset convergence criterion or an upper limit of the number of iterations is met, and output a heat dissipation channel topology structure model with an optimal trade-off solution on the multi-objective optimization objective function set; S7. Convert the heat dissipation channel topology structure model with the optimal trade-off solution into a manufacturable CAD geometric model and import it into the computational fluid dynamics simulation platform for comprehensive performance verification and manufacturing feasibility assessment.

2. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 1, characterized in that: Said S1 specifically includes: S11. Collecting operating data of the battery module under multiple dynamic load conditions, wherein the operating data includes time series, current density, and battery internal resistance; S12. Calculate the thermal power generation at each moment based on the current density and battery internal resistance at different time series; S13. Discretize the heat generation power Q(t) in the time dimension and construct the heat generation power matrix Q = {Q(t1), Q(t2), ..., Q(t n )}, where t1, t2, …, t n is the sampling time node.

3. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 1, characterized in that: The S2 specifically includes: S21. Obtaining the physical packaging structure of the battery module, wherein the physical packaging structure includes the arrangement of battery cells, the boundary dimensions of the battery module, and the positions and dimensions of the cooling medium inlet and outlet; S22. Based on the physical packaging structure and flow channel layout constraints of the battery module, determine the initial topological construction area of ​​the heat dissipation channel and construct an initial heat dissipation channel topological structure model. The initial topological construction area is the remaining layout space within the battery module boundary, excluding the area occupied by the battery cells and meeting the accessibility requirements of the inlet and outlet connection paths. The flow channel layout constraints include that the channel width is not less than a lower limit, the channel centerline does not cross the module boundary, the minimum spacing between channels is not less than a threshold, the inlet and outlet positions are fixed, and the overall structure maintains bilateral symmetry. S23, setting the perturbation parameter set of the structure Θ = {θ1, θ2, θ3, θ4} based on the geometric adjustable parameters in the initial heat dissipation channel topology model, where θ1 represents the channel segment width, θ2 represents the node connection angle, θ3 represents the number of channel branches, and θ4 represents the distance between adjacent channels; S24, introducing a structural perturbation operation on the initial heat dissipation channel topology model based on the perturbation parameter set Θ, for each parameter θ in the perturbation parameter set i Construct the perturbation function f d (θ i ): f d (i i )=θ i +n i ·sin((ω i ·r i ); Among them, η i is the parameter θ i The maximum disturbance amplitude, ω i is the disturbance frequency coefficient, r i ∈[0,1] is a normalized random variable; S25. Based on the parameter combination after the structural perturbation, generate multiple heat dissipation channel structure candidates with different channel topology characteristics, and assemble all the heat dissipation channel structure candidates into a first-generation population structure.

4. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 1, characterized in that: The S3 specifically includes: Construct a multi-objective optimization objective function set F = {f1, f2, f3}, where: f1 is the heat flux density distribution uniformity evaluation function, which is used to measure the balance of temperature distribution in different areas of the heat dissipation channel: Where T(x,y) is the steady-state temperature distribution at point (x,y), Q(t) is the heat generation power per unit time, Ω is the heat source area, k is the thermal conductivity, A is the cross-sectional area of ​​the channel structure, and x0 and y0 represent the horizontal and vertical coordinate positions of the center of the heat source area, respectively. f2 is the channel structure complexity evaluation function, which is used to comprehensively describe the geometric complexity of the channel layout: Where L is the total length of the channel path, A c is the actual occupied area of ​​the channel, κ i is the local curvature of the i-th channel segment, n is the number of structural branch nodes, N max is the maximum number of nodes allowed, γ1, γ2, γ3 are the preset weight coefficients; f3 is the flow pressure energy loss function, which is used to evaluate the pressure loss caused by structural resistance during the flow of cooling medium in the channel: Among them, p in (l) is the inlet pressure at the path position l, p out (l) is the corresponding outlet pressure, Q(l) is the volume flow rate per unit length, and L is the total length of the channel path.

5. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 1, characterized in that: The S4 specifically includes: S41. Each heat dissipation channel structure candidate in the first generation population structure is used as the structure input of the thermal-fluid-solid coupling simulation. At the same time, the heat generation power matrix Q = {Q(t1), Q(t2), ..., Q(t n )} as the heat source boundary condition input; S42. performing a thermal-fluid-solid coupling simulation on each heat dissipation channel structure candidate in a three-dimensional finite volume modeling environment; S43. After each thermal-fluid-solid coupling simulation is completed, a thermal-fluid performance index set P corresponding to each heat dissipation channel structure candidate is extracted from the thermal field simulation, the flow field simulation, and the solid structure field simulation. max ,T avg ,Δp max ,V,φ avg ,δ max }, where T max is the maximum temperature in the temperature field, T avg is the average temperature of the channel area, which is calculated by the temperature distribution function T(x,y,z,t), Δp max is the maximum pressure difference generated by the cooling medium on the flow path, which is calculated by the pressure field p(x, y, z, t), V is the total volume of the heat dissipation channel structure, which is calculated by the structural model geometry, and φ avg is the average heat flux per unit volume, which is calculated by normalizing the heat flux vector modulo integral, δ max It is the maximum displacement value that occurs in the structural response, which is extracted from the displacement function δ(x, y, z, t) of the solid structure field.

6. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 5, characterized in that: The thermal field simulation in the thermal-fluid-solid coupling simulation specifically includes: Perform thermal field simulation on the candidate heat dissipation channel structure and establish the governing equation of the temperature field T(x,y,z,t): in, is the gradient operator, T represents the temperature field distribution function, represents the partial derivative of the temperature field distribution over time, ρ represents the channel material density, c p represents specific heat capacity, k represents thermal conductivity, and Q(t) is the heat generation power; During the simulation process, the boundary conditions include the convection heat transfer boundary of the channel outer wall and the temperature continuity boundary of the cooling medium interface, and the initial condition is the battery module ambient temperature.

7. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 5, characterized in that: The flow field simulation in the thermal-fluid-solid coupling simulation specifically includes: Perform flow field simulation on the candidate heat dissipation channel structure to establish the flow velocity field The governing equations are: in, is the gradient operator, represents the cooling medium velocity field distribution, ρ f represents the cooling medium density, p represents the pressure field, μ f Indicates the dynamic viscosity of the cooling medium, represents the partial derivative of the cooling medium velocity field with time; During the simulation, the inlet boundary is set as a constant velocity boundary, the outlet boundary is set as a constant pressure boundary, the inner wall of the channel is set as a no-slip boundary, and the initial condition is that the cooling medium is stationary.

8. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 5, characterized in that: The solid structure field simulation in the thermal-fluid-solid coupling simulation specifically includes: Perform solid structural field simulation on the candidate heat dissipation channel structure and establish the governing equations for the structural response displacement field δ(x, y, z, t): in, is the gradient operator, δ represents the solid structure displacement vector field, σ represents the stress tensor, and f b represents the volume force density, ρ s represents the density of the channel wall material; The stress tensor σ and the strain tensor ε satisfy the following thermal stress relationship: σ=D:(ε-α s ΔTI); Where D represents the material elastic matrix, α s represents the thermal expansion coefficient of the channel material, ΔT represents the temperature rise obtained from the thermal field simulation, and I is the unit tensor; During the simulation, the solid structure boundary is set as a combination of fixed boundary and symmetric boundary, and the initial condition is zero displacement state.

9. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 1, characterized in that: The S5 specifically includes: S51. Using a multi-objective genetic algorithm based on Pareto optimal sorting, normalize the thermal-fluid performance index set corresponding to each heat dissipation channel structure candidate in the first generation population structure, and calculate the heat flux density distribution uniformity index, structural complexity index, and pressure energy loss index based on the multi-objective optimization objective function set, and perform non-dominated level division and crowding distance calculation in the objective function space; S52. Construct a ranking mapping relationship based on the non-dominated rank and crowding distance results, and use a tournament selection mechanism to select heat dissipation channel structure candidates in the Pareto front region from the current population as parent individuals; S53, inputting the selected parent structure into the genetic operation module in the form of structural parameter encoding, using a uniform crossover operator to combine structural parameters among multiple parents, constructing a new topological structure perturbation combination, and performing a mutation operation based on the probability density function set by the perturbation parameter set to improve structural diversity; S54. After the generation structure candidate is constructed, the heat generation power matrix is ​​used as the heat source boundary condition, and a thermal-fluid-solid coupling simulation process is performed to extract its temperature field, flow velocity field, and structural response index to form a corresponding thermal-fluid performance index set; S55. The offspring structural candidates and their thermal-fluid performance indicators are merged with the previous generation non-dominated structure set to form an intermediate population structure. The solution set distribution state of the population structure in the objective function space is optimized and screened again based on the Pareto ranking and crowding distance calculation. Several structural candidates with higher ranking levels and good solution set diversity are selected to form a new generation of population structure.

10. The method for optimizing the heat dissipation channel of a battery module based on a genetic algorithm according to claim 1, characterized in that: The S6 specifically includes: S61, set the iteration termination criterion, the iteration termination criterion includes the maximum number of iterations G max and the convergence threshold δ conv ; S62. After each generation of new population structure is generated, perform non-dominated sorting on all heat dissipation channel structure candidates in the new generation population structure based on the multi-objective optimization objective function set, and extract the current generation non-dominated structure set. And compared with the non-dominated structure set saved in the previous iteration Compare and terminate the iteration process if any of the following conditions is met: Current iteration number g≥G max ; The change in the distance between the non-dominated structure sets of the current generation and the previous generation S63. After the iteration termination criterion is met, a set of heat dissipation channel structure candidates with the highest non-dominated level is selected from the final generation population structure and output as a heat dissipation channel topology structure model with an optimal trade-off solution.

Images