Track design method and system for electromagnetic track launching device

Abstract

The invention discloses a track design method and system for an electromagnetic track launching device. The method comprises the following steps: constructing an electromagnetic track finite element model, segmenting a track, and setting the thickness of a functional contact layer as a design variable; generating a test set of the thickness of the functional contact layer; operating the electromagnetic track finite element model, establishing a batch simulation calculation process based on the design variables, extracting multi-physical field response index data under each design variable, and constructing a key performance index data set; training the key performance index data set to obtain a prediction function, and forming a mapping relation between the thickness design of the track composite structure and the key performance indexes; defining a multi-objective optimization function and constraint conditions; taking a representative solution in the Pareto optimal solution set of the initial section as the input of an electromagnetic track finite element model, obtaining corresponding state output, and taking the state output as an initial condition of acceleration section simulation to construct an inter-section state transfer mechanism; and selecting an optimal three-section track configuration combination and outputting a corresponding performance index report and a structure distribution diagram.

Description

Technical Field

[0001] The present invention belongs to the field of electromagnetic rail launch devices, and in particular relates to a track design method and system for an electromagnetic rail launch device. Background Art

[0002] Track-type electromagnetic drive technology is a new type of launch technology that relies on the Lorentz force generated by the strong magnetic field formed by the armature and the track and the armature current to accelerate the armature to ultra-high speed. Due to its simple structure, fast response speed, high current amplitude and current density, fast relative sliding speed, and suitability for small-mass projectiles, it is widely used in various fields such as naval, land, and air military weapons, pulsed high magnetic fields, etc.

[0003] A key bottleneck hindering the practical application of electromagnetic rail launch technology lies in the reliability and lifespan of the rail structure. Because the rail is subjected to extremely high current density, electromagnetic forces, and transient thermal shock during launch, it is highly susceptible to damage such as ablation, gouging, and transition. Especially during repeated launches or long-term operation, the rail's conductivity and contact degradation will severely impact system stability. Therefore, optimizing rail design has become a core challenge hindering the development of electromagnetic launch systems.

[0004] Currently, electromagnetic rails primarily utilize composite materials for structural optimization. This involves superimposing a layer of heat-resistant and wear-resistant functional material on a highly conductive metal substrate as a contact layer to enhance the rail's serviceability in high-current and high-temperature environments. However, traditional design methods typically employ an integrated design, making it difficult to achieve differentiated performance matching based on the load characteristics at different locations during launch, limiting the overall lifespan of the rail. Current composite rail design methods typically rely on empirical formulas or simple experiments, lacking a systematic multi-objective trade-off optimization mechanism. This makes it difficult to simultaneously address multiple requirements, including electrical conductivity, thermal performance, and wear control, in high-power, high-frequency launch environments. Summary of the Invention

[0005] In response to the problems existing in the prior art, the purpose of the present invention is to provide a track design method and system for an electromagnetic track launch device. On the basis of traditional composite material structures, the contact layer thickness is used as a variable, a segmented structural design concept is introduced, and a multi-objective optimization method is combined with genetic algorithms and machine learning to achieve systematic optimization of track structure performance.

[0006] The technical solution of the present invention:

[0007] A method for designing a track of an electromagnetic track launcher comprises the following steps:

[0008] Step S1: Construct a finite element model of an electromagnetic track with a parameterized structure. The track is divided into a starting section, an acceleration section, and a launch section according to the armature motion process. A double-layer structure consisting of a conductive base layer and a functional contact layer is adopted, and the thickness of the functional contact layer is set as an adjustable design variable.

[0009] Step S2: Using an experimental design method to generate a test set with different functional contact layer thicknesses; running the electromagnetic track finite element model, establishing a batch simulation calculation process based on the design variables, extracting multi-physics field response indicator data under each design variable, and constructing a key performance indicator data set; using a machine learning method to train the key performance indicator data set to obtain a prediction function, forming a mapping relationship between the track composite structure thickness design and the key performance indicators;

[0010] Step S3: Based on the key performance indicator dataset constructed in step S2, a multi-objective optimization function and constraints are defined; a non-dominated sorting genetic algorithm is used to optimize the thickness parameters of the functional contact layer of each track segment, and a Pareto optimal solution set corresponding to the composite structure of each track segment is generated;

[0011] The multi-objective function is defined as follows:

[0012] min t F(t)=[f1(t),f2(t),f3(t)] T (2)

[0013] Where f1(t) = R(t) is the resistance per unit length; f2(t) = ΔT(t) is the maximum temperature rise; f3(t) = W(t) is the wear amount;

[0014] The constraints are:

[0015] (1) Thickness constraint: Ensure that the thickness of the functional contact layer is within the constraint range;

[0016] (2) Performance constraints: The maximum temperature rise of the track does not exceed the heat resistance temperature of the material, and the armature outlet speed is not lower than the design requirement value;

[0017] Step S4: Using the representative solution in the Pareto optimal solution set of the starting segment as the input of the electromagnetic track finite element model, the corresponding state output is obtained and used as the initial condition for the acceleration segment simulation. This state output is then transferred backwards to build an inter-segment state transfer mechanism, realizing the linkage of performance states between different segments and obtaining the performance output of the entire track combination.

[0018] Step S5: Perform normalized analysis and ranking of the full track performance indicators of each combined solution. Use a decision-making method based on the target weight priority to perform a multi-objective compromise evaluation on all candidate solutions. Finally, select the optimal three-segment track configuration combination and output the corresponding performance indicator report and structure distribution diagram.

[0019] Preferably, S1 comprises the following steps:

[0020] Step S11: determining the geometric parameters of the electromagnetic track finite element model, dividing the track into a starting section, an acceleration section, and a launch section, and determining the geometric dimensions of the starting section, the acceleration section, and the launch section respectively;

[0021] Step S12: defining the material properties of the rail composite structure in the finite element simulation software, and setting the material properties of the armature guide rail portion of the electromagnetic rail launcher;

[0022] Step S13: performing finite element mesh division and optimization, performing mesh refinement processing on key areas of the track composite structure, and inspecting and optimizing the mesh quality; wherein the key areas include contact interfaces and material transition areas;

[0023] Step S14: setting multiple physical fields in the electromagnetic track finite element model, including electromagnetic field, temperature field and structural field, and defining the coupling relationship between the physical fields;

[0024] Step S15: Establish a parameterized interface program, set the thickness of the functional contact layer as the main design variable, and realize the automated process of geometric parameters, meshing, solving and post-processing.

[0025] Preferably, in step S12, the material properties include electrical, thermal and mechanical parameters.

[0026] Preferably, in step S14, the electromagnetic field model is constructed. Since the conduction current of the conductor in the electromagnetic emission system is generated by the magnetic field, the displacement current can be ignored. The Maxwell equations based on which the model is constructed are:

[0027]

[0028] Where B is the magnetic induction intensity; E is the electric field intensity; J is the conduction current density; v is the movement speed of the conductor; μ is the magnetic permeability; and t is time.

[0029] Preferably, S2 includes the following steps:

[0030] Step S21: Determine the design variable range of each track section, establish the constraint boundary of the functional contact layer thickness of each section, and use the experimental design method to generate a test set with different functional contact layer thicknesses;

[0031] Step S22: Establishing a batch simulation calculation process, inputting the functional contact layer thickness in each test set into the electromagnetic track finite element model established in step S1, and controlling the entire process of geometric parameter update, mesh regeneration, and multi-physics field solution calculation of the electromagnetic track finite element model through an automated script;

[0032] Step S23: Based on the calculation results of step S22, extract the multi-physics field response index data in the finite element simulation and build a key performance index data set;

[0033] Step S24: Using a machine learning method to train the key performance indicator data set to obtain a prediction function, and establish a mapping relationship between the design variables and the multi-physics field key performance indicators.

[0034] Preferably, in step S21, the experimental design method adopts the central composite design (CCD) method.

[0035] Preferably, in step S24, the machine learning method adopts a support vector machine (SVR) method.

[0036] Preferably, in step S3, the multi-objective objective function is defined as follows:

[0037] min t F(t)=[f1(t),f2(t),f3(t)] T (2)

[0038] Where f1(t)=R(t) is the resistance per unit length; f2(t)=ΔT(t) is the maximum temperature rise; f3(t)=W(t) is the wear amount.

[0039] The constraints are:

[0040] Thickness constraint: Ensure that the thickness of the functional contact layer is within the constraint range.

[0041] Performance constraints: The maximum temperature rise shall not exceed the heat resistance temperature of the material, and the armature outlet speed shall not be lower than the design requirement value.

[0042] Preferably, S4 includes the following steps:

[0043] Step S41: Select a representative solution from the Pareto optimal solution set of the initial stage, and extract the corresponding state output data as the initial condition of the acceleration stage simulation;

[0044] Step S42: Based on the initial conditions delivered in step S41, perform a multi-physics field coupling simulation of the acceleration section, then perform the key performance indicator data set construction in step S2 to establish an acceleration section prediction function model;

[0045] Step S43: Execute the multi-objective optimization of step S3 based on the acceleration segment prediction function model of S42, use the multi-objective optimization algorithm to generate the Pareto optimal solution set of the acceleration segment, select a representative solution from the Pareto optimal solution set, and extract the corresponding state output data as the initial conditions of the launch segment simulation;

[0046] Step S44: Based on the initial conditions passed in step S43, perform multi-physics field coupling simulation of the launch segment, execute the key performance indicator data set construction of step S2, and establish the launch segment prediction function model; execute the multi-objective optimization of step S3 to generate the Pareto solution set of the launch segment;

[0047] Step S45: Establish a combination model, systematically combine the geometric parameters, material properties and thickness distribution of the starting segment, acceleration segment and launch segment to form a complete full-track thickness combination solution, perform full-track multi-physics field simulation, and calculate and obtain the performance index data of the full-track thickness combination.

[0048] Preferably, in step S5, the decision-making method adopts the entropy weight TOPSIS method, which is used to perform comprehensive evaluation and decision analysis on the full track combination solution in step S4 to select the ideal solution, which includes:

[0049] (1) Constructing the original evaluation matrix

[0050] The performance index data of all track thickness combinations are obtained from step S4 to form an original evaluation matrix, which can be expressed as:

[0051]

[0052] Where: X is the original evaluation matrix; x ij represents the value of the i-th optimization solution on the j-th performance index; m is the number of optimization solutions; n is the number of performance indicators contained in each solution.

[0053] (2) Normalization of the range method

[0054] Since the performance indicators are all inverse indicators, the range method is used to normalize the original evaluation matrix. The range method formula can be expressed as:

[0055]

[0056] Where: is the maximum value of the jth index; is the minimum value of the jth index; r ij ∈[0,1] is an element in the standardized decision matrix.

[0057] (3) Determine indicator weights using entropy weight method

[0058] The entropy weight method indicator weight is expressed as:

[0059]

[0060] Where: w j is the weight of the j-th indicator; e j is the entropy value of the j-th indicator; is the weight of the j-th indicator.

[0061] (4) Constructing a weighted normalization matrix

[0062] Use weight w j Weighting the standardized decision matrix gives the weighted decision matrix:

[0063] v ij =w j ×r ij (6)

[0064] Where: v ij is an element in the weighted decision matrix.

[0065] (5) Determine the positive ideal solution and the negative ideal solution

[0066]

[0067] Where: A + represents the positive ideal solution set; A - represents the set of positive ideal solutions; represents a positive ideal solution; represents a negative ideal solution.

[0068] (6) Calculate the distance from the ideal solution

[0069] The Euclidean distance between the i-th optimization solution and the positive ideal solution is:

[0070]

[0071] The Euclidean distance between the i-th optimization solution and the negative ideal solution is:

[0072]

[0073] (7) Calculate proximity

[0074] The closeness coefficient of the i-th optimization solution:

[0075]

[0076] Among them C i ∈[0,1],C i The larger the value, the closer the optimization solution is to the ideal solution, and the performance is the best. Select C i The largest solution is regarded as the optimal combination.

[0077] The present application further provides an electromagnetic rail launcher track design system for executing the above-mentioned electromagnetic rail launcher track design method. The system includes:

[0078] A model construction module is used to construct a finite element model of an electromagnetic track with a parameterized structure. The track is divided into three parts: a starting section, an acceleration section, and a launch section according to the armature motion process. A double-layer structure consisting of a conductive base layer and a functional contact layer is adopted, and the thickness of the contact layer is set as an adjustable design variable. The model construction includes determining geometric parameters, defining the material properties of the track composite structure, performing finite element meshing and optimization, defining multi-physics field coupling, and establishing a parameterized interface program.

[0079] The data acquisition module is used to generate a test set with different contact layer thicknesses using an experimental design method, run the electromagnetic track finite element model to establish a batch simulation calculation process, extract multi-physics field response indicator data under various design variable combinations, and use machine learning methods to train the key performance indicator data set to obtain a prediction function, thereby forming a mapping relationship between the track structure thickness design and performance indicators;

[0080] The optimization solution module defines the multi-objective optimization function and constraints based on the data set and prediction function built by the data construction module. It uses the non-dominated sorting genetic algorithm to optimize the contact layer thickness parameters of each stage of the track and generate the Pareto optimal solution set corresponding to each section of the track structure.

[0081] The decision-making evaluation module is used to perform normalized analysis and ranking of the full track performance indicators of each segment combination solution. It uses the entropy weight TOPSIS decision-making method to perform multi-objective compromise evaluation on all candidate solutions, and finally selects the optimal three-segment track configuration combination and outputs the corresponding performance indicator report and structure distribution diagram.

[0082] Beneficial effects of the present invention:

[0083] 1. The design concept of segmented optimization is adopted to divide the electromagnetic track into three functional segments. Different designs are carried out according to the characteristic requirements of different stages. The different physical characteristics of the track in the starting, acceleration and launch stages are fully considered, so that the performance of each segment can be adjusted according to its specific working conditions. Compared with the traditional integrated design, this invention achieves the best matching of each track segment under different load conditions, thereby effectively extending the service life of the track and reducing the occurrence of damage such as ablation and wear.

[0084] 2. During the simulation process, an inter-segment state transfer mechanism is established, and the boundary conditions at the end of the current stage are used as the initial boundary conditions for the next stage for simulation. Compared with traditional methods, this reduces the simulation workload and ensures the continuity and coordination between different sections of the track.

[0085] 3. This invention combines genetic algorithms and machine learning models in the multi-objective optimization process of electromagnetic tracks, effectively improving optimization efficiency and accuracy. Firstly, a non-dominated sorting genetic algorithm is used to achieve decoupled optimization of multiple track performance indicators, ensuring that the optimization results possess good multi-objective balance. Secondly, a support vector machine machine learning method is introduced to train and model finite element simulation data, establishing a nonlinear mapping relationship between design parameters and key response variables, significantly reducing the computational cost of repeated simulations. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 The present invention is a flow chart of a method for designing a track of an electromagnetic track launcher;

[0087] Figure 2 Detailed step flow diagram of step S1 in the present invention;

[0088] Figure 3 Detailed step flow diagram of step S2 in the present invention;

[0089] Figure 4 Detailed step flow diagram of step S4 in the present invention;

[0090] Figure 5 A module diagram of a track design system for an electromagnetic rail launcher. DETAILED DESCRIPTION

[0091] The specific implementation of the present invention is further described below in conjunction with the accompanying drawings and technical solutions.

[0092] Example 1

[0093] like Figure 1 FIG. 1 is a flow chart of a method for designing a track for an electromagnetic track launcher, the method comprising:

[0094] Step S1: Construct a finite element model of an electromagnetic track with a parameterized structure. The track is divided into a starting section, an acceleration section, and a launch section according to the armature motion process. A double-layer structure consisting of a conductive base layer and a functional contact layer is adopted, and the thickness of the functional contact layer is set as an adjustable design variable.

[0095] Step S2: Using an experimental design method to generate a test set with different functional contact layer thicknesses; running the electromagnetic track finite element model, establishing a batch simulation calculation process based on the design variables, extracting multi-physics field response indicator data under each design variable, and constructing a key performance indicator data set; using a machine learning method to train the key performance indicator data set to obtain a prediction function, forming a mapping relationship between the track composite structure thickness design and the key performance indicators;

[0096] Step S3: Based on the data set constructed in step S2, define the multi-objective optimization function and constraints; use the non-dominated sorting genetic algorithm to optimize the contact layer thickness parameters of each stage track to generate the Pareto optimal solution set corresponding to each track structure;

[0097] Step S4: Based on the representative solution in the optimization solution set of the starting segment, the corresponding state output is obtained and used as the initial condition for the acceleration segment simulation. This state output is then transferred backwards to build an inter-segment state transfer mechanism to achieve the linkage of performance states between different segments and obtain the performance output of the entire track combination.

[0098] Step S5: Perform normalized analysis and ranking of the full track performance indicators of each segment combination solution, use a decision-making method to perform a multi-objective compromise evaluation on all candidate solutions based on the target weight priority, and finally select the optimal three-segment track configuration combination; output the corresponding performance indicator report and structure distribution diagram.

[0099] The embodiment of the present invention introduces a segmented structural design concept, combines machine learning and genetic algorithms to perform multi-objective optimization of the electromagnetic track, and achieves systematic optimization of the track structure performance.

[0100] In this embodiment, COMSOL Multiphysics software is selected to perform finite element simulation analysis.

[0101] As an example of the present invention, refer to Figure 2 As shown, in this example, step S1 includes:

[0102] Step S11: determining the geometric parameters of the electromagnetic track finite element model, dividing the electromagnetic launch track into a starting section, an acceleration section, and a launch section, and determining the geometric dimensions of the starting section, the acceleration section, and the launch section respectively;

[0103] In an embodiment of the present invention, a three-dimensional geometric model of the electromagnetic track is constructed using modeling software (such as SolidWorks or AutoCAD) to determine key geometric parameters such as the track's total length, width, rail spacing, and rail height. The entire track is divided into a starting segment, an acceleration segment, and a launch segment. Based on the projectile's motion patterns and the electromagnetic force distribution characteristics, the length ratios of each segment are appropriately set. In this case, the starting segment accounts for 15% of the total length, the acceleration segment accounts for 65%, and the launch segment accounts for 20%.

[0104] Step S12: defining the material properties of the rail composite structure in the finite element simulation software, and setting the material properties of the armature guide rail of the electromagnetic rail launcher;

[0105] In this embodiment of the present invention, finite element simulation analysis was performed using COMSOL Multiphysics software. The geometric parameters constructed using the modeling software were imported into the finite element software, and the composite material properties of the armature and guide rail were defined and assigned. The armature was made of aluminum, the guide rail base layer was copper, and the contact layer was copper-tungsten alloy. Material parameters included, but were not limited to, key physical quantities such as electrical conductivity, thermal conductivity, density, Young's modulus, and Poisson's ratio.

[0106] Step S13: Perform finite element mesh division and optimization, refine the mesh in key areas of the track structure, such as the contact interface and material transition area, and inspect and optimize the mesh quality;

[0107] In this embodiment of the present invention, the geometric model of the electromagnetic rail launcher is meshed using finite element simulation software. Local mesh refinement strategies, such as adding boundary layers, are employed in critical areas of the rail structure, such as the contact interface between the armature and the guide rail and the material transition zone at the interface between different materials, to improve simulation accuracy. Mesh independence is also verified to ensure computational convergence and reliable results.

[0108] Step S14: setting multiple physical fields in the electromagnetic railgun finite element model, including electromagnetic field, temperature field and structural field, and defining the coupling relationship between the physical fields;

[0109] In an embodiment of the present invention, a multi-physics field simulation is performed using the physical field interface provided by COMSOL Multiphysics software. In the electromagnetic field, the current density distribution and the induced magnetic field strength between the track and the armature are obtained by solving the Maxwell equations. In the temperature field, the Joule heat and induction heating generated in the electromagnetic field module are introduced as heat source terms, and the temperature distribution and temperature rise characteristics of each structural area of ​​the track are analyzed by solving the transient heat conduction equation. In the structural field, the temperature rise output by the temperature field and the electromagnetic force in the electromagnetic field are input to calculate the thermal stress distribution and contact stress of the track structure under the action of complex thermoelectric loads.

[0110] Step S15: Establish a parametric interface program, set the contact layer thickness as the main design variable, and realize the automated process of geometric model, mesh, solution and post-processing.

[0111] In this embodiment of the present invention, Python is used for parametric modeling, encapsulating the model's physical field settings, geometric parameters, material definitions, meshing, and other aspects into code. This parameterization significantly improves model simulation efficiency, reduces errors caused by human intervention, and enhances the accuracy of evaluating the impact of contact layer thickness parameters on structural performance.

[0112] In step S14, the electromagnetic field model is constructed. Since the conduction current of the conductor in the electromagnetic emission system is generated by the magnetic field, the displacement current can be ignored. The Maxwell equations based on which the model is constructed are:

[0113]

[0114] Where B is the magnetic induction intensity; E is the electric field intensity; J is the conduction current density; v is the movement speed of the conductor; μ is the magnetic permeability; and t is time.

[0115] As an example of the present invention, refer to Figure 3 As shown, in this example, step S2 includes:

[0116] Step S21: Determine the design variable range of each track section, establish the constraint boundary of the contact layer thickness of each section, and use the experimental design method to generate a test set with different thicknesses;

[0117] In an embodiment of the present invention, the design variable range of the contact layer thickness of each track section is determined based on the actual working conditions and design requirements of the track structure. For example, a central composite design method is adopted to ensure comprehensive and uniform sampling within the design space by rationally arranging design points.

[0118] Step S22: Establishing a batch simulation calculation process, inputting the thickness parameters in the test set into the parametric finite element model established in step S1, and controlling the entire process of geometric parameter updating, mesh regeneration, and multi-physics field solution calculation through an automated script;

[0119] In the embodiment of the present invention, the parametric finite element model established in step S1 is used in combination with an automated script system to sequentially input each set of thickness parameters generated by the central composite design into the model. The system automatically drives the geometric model to update according to the input parameters, and regenerates the mesh according to the mesh division rules, and automatically performs batch simulation calculations.

[0120] Step S23: Based on the calculation results of step S22, extract the multi-physics field response index data in the finite element simulation and build a key performance index data set;

[0121] In an embodiment of the present invention, based on the calculation results of step S22, multi-physical field response index data in the finite element simulation is extracted to construct a key performance indicator data set. The key performance indicators include resistance per unit length, maximum temperature rise, contact surface wear, electromagnetic force distribution peak, surface temperature gradient, maximum stress value, etc.

[0122] Step S24: Using a machine learning method to train the key performance indicator data set to obtain a prediction function, and establish a mapping relationship between the design variables and the multi-physics field key performance indicators.

[0123] In this embodiment of the present invention, a support vector machine (SVM) is used as a machine learning method. Based on the dataset extracted in step S22, a prediction function is trained to construct a mapping relationship between input parameters and multi-physics performance indicators. By selecting an appropriate kernel function and optimizing its parameters, this method ensures that the prediction function has high accuracy and stability across the entire design space. This method exhibits good generalization and strong nonlinear fitting capabilities, making it suitable for high-dimensional small sample problems and capable of effectively capturing the inherent relationships between variables in complex physical systems.

[0124] In step S3, a multi-objective optimization function is defined, and a non-dominated sorting genetic algorithm is used to optimize the contact layer thickness parameters of each stage track to generate a Pareto optimal solution set corresponding to each track structure.

[0125] The multi-objective optimization function is:

[0126] min t F(t)=[f1(t),f2(t),f3(t)] T (2)

[0127] Where f1(t)=R(t) is the resistance per unit length; f2(t)=ΔT(t) is the maximum temperature rise; f3(t)=W(t) is the wear amount.

[0128] The constraints are:

[0129] Thickness constraint: Ensure that the thickness of the contact layer is within the constraint range.

[0130] Performance constraints: The maximum temperature rise shall not exceed the heat resistance temperature of the material, and the armature outlet speed shall not be lower than the design requirement value.

[0131] In step S3, the non-dominated sorting genetic algorithm includes: initializing a population containing multiple thickness parameters, and using the prediction function constructed in step 2 to replace the real simulation model to quickly obtain the performance index corresponding to each individual; performing non-dominated sorting on the individuals in the generated population, and calculating the crowding degree of individuals in each dominated layer; using the tournament method to select individuals, giving priority to individuals with lower non-dominated levels, and if the levels are the same, selecting individuals with a larger crowding distance; performing genetic operations including selection, crossover and mutation operations; merging the parent and child populations, performing non-dominated sorting and crowding degree comparison again, and selecting the next generation population, and continuously iterating.

[0132] As an example of the present invention, refer to Figure 4 As shown, in this example, step S4 includes:

[0133] Step S41: Select a representative solution from the solution set of the multi-objective optimization of the initial stage, and extract the corresponding state output data as the initial condition of the acceleration stage simulation;

[0134] Step S42: Based on the initial conditions delivered in step S41, perform multi-physics field coupling simulation of the acceleration section, and then execute step S2 to construct a data set and establish an acceleration section prediction function model;

[0135] Step S43: Execute the multi-objective optimization of step S3 based on the prediction function model of S42, use the multi-objective optimization algorithm to generate the Pareto solution set of the acceleration segment, select a representative solution from the solution set, and extract the corresponding state output data as the initial condition of the launch segment simulation;

[0136] Step S44: Based on the initial conditions passed in step S43, perform multi-physics field coupling simulation of the launch segment, execute step S2, construct the data set, and establish the launch segment prediction function model; execute the multi-objective optimization of step S3 to generate the Pareto solution set of the launch segment.

[0137] Step S45: Establish a combination model, systematically combine the geometric parameters, material properties and thickness distribution of the starting segment, acceleration segment and launch segment to form a complete full-track thickness combination solution, perform full-track multi-physics field simulation, and calculate and obtain the performance output of the full-track thickness combination.

[0138] In step S4, the representative solution can be determined according to different strategies, such as a weighted balance point of resistance, temperature rise and wear in the Pareto front, an extreme optimal point or a boundary solution with less congestion.

[0139] In step S5, the decision-making method adopts the entropy weight TOPSIS method, which is used to perform comprehensive evaluation and decision analysis on the full track combination solution in step S4 to select the ideal solution. The steps of this method are as follows:

[0140] (1) Constructing the original evaluation matrix

[0141] The performance index data of all track thickness combinations are obtained from step S4 to form an original evaluation matrix, which can be expressed as:

[0142]

[0143] Where: X is the original evaluation matrix; x ij represents the value of the i-th optimization solution on the j-th performance index; m is the number of optimization solutions; n is the number of performance indicators contained in each solution.

[0144] (2) Normalization of the range method

[0145] Since the performance indicators are all inverse indicators, the range method is used to normalize the original evaluation matrix. The range method formula can be expressed as:

[0146]

[0147] Where: is the maximum value of the jth index; is the minimum value of the jth index; r ij ∈[0,1] is an element in the standardized decision matrix.

[0148] (3) Determine indicator weights using entropy weight method

[0149] The entropy weight method indicator weight is expressed as:

[0150]

[0151] Where: w j is the weight of the j-th indicator; e j is the entropy value of the j-th indicator; is the weight of the j-th indicator.

[0152] (4) Constructing a weighted normalization matrix

[0153] Use weight w j Weighting the standardized decision matrix gives the weighted decision matrix:

[0154] v ij =w j ×r ij (6)

[0155] Where: v ij is an element in the weighted decision matrix.

[0156] (5) Determine the positive ideal solution and the negative ideal solution

[0157]

[0158] Where: A + represents the positive ideal solution set; A - represents the set of positive ideal solutions; represents a positive ideal solution; represents a negative ideal solution.

[0159] (6) Calculate the distance from the ideal solution

[0160] The Euclidean distance between the i-th optimization solution and the positive ideal solution is:

[0161]

[0162] The Euclidean distance between the i-th optimization solution and the negative ideal solution is:

[0163]

[0164] (7) Calculate proximity

[0165] The closeness coefficient of the i-th optimization solution:

[0166]

[0167] Among them C i ∈[0,1],C i The larger the value, the closer the optimization solution is to the ideal solution, and the performance is the best. Select C i The largest solution is regarded as the optimal combination.

[0168] Example 2

[0169] Figure 5 The figure shows a module diagram of a track design system for an electromagnetic track launcher. Figure 5 As shown, a track design system for an electromagnetic track launch device includes a model building module, a data building module, an optimization solution module and a decision evaluation module.

[0170] A model construction module is used to construct a finite element model of an electromagnetic track with a parameterized structure. The track is divided into three parts: a starting section, an acceleration section, and a launch section according to the armature motion process. A double-layer structure consisting of a conductive base layer and a functional contact layer is adopted, and the thickness of the contact layer is set as an adjustable design variable. The model construction includes determining basic geometric parameters, defining the material properties of the track composite structure, performing finite element meshing and optimization, defining multi-physics field coupling, and establishing a parameterized interface program.

[0171] The data acquisition module is used to generate experimental sets with different contact layer thicknesses using experimental design methods, run finite element simulation models to establish a batch simulation calculation process, extract multi-physics field response indicator data under various design variable combinations, and use machine learning methods to train key performance indicator data sets to obtain prediction functions, thereby forming a mapping relationship between track structure thickness design and performance indicators;

[0172] The optimization solution module defines the multi-objective optimization function and constraints based on the data set and prediction function built by the data construction module. It uses the non-dominated sorting genetic algorithm to optimize the contact layer thickness parameters of each stage of the track and generate the Pareto optimal solution set corresponding to each section of the track structure.

[0173] The decision-making evaluation module is used to perform normalized analysis and ranking of the full track performance indicators of each segment combination solution. It uses the entropy weight TOPSIS decision-making method to perform multi-objective compromise evaluation on all candidate solutions, and finally selects the optimal three-segment track configuration combination and outputs the corresponding performance indicator report and structure distribution diagram.

[0174] In summary, the electromagnetic rail launcher track design system method and system disclosed in the embodiments of the present invention have at least the following beneficial effects:

[0175] 1. The design concept of segmented optimization is adopted to divide the electromagnetic track into three functional segments. Different designs are carried out according to the characteristic requirements of different stages. The different physical characteristics of the track in the starting, acceleration and launch stages are fully considered, so that the performance of each segment can be adjusted according to its specific working conditions. Compared with the traditional integrated design, this invention achieves the best matching of each track segment under different load conditions, thereby effectively extending the service life of the track and reducing the occurrence of damage such as ablation and wear.

[0176] 2. During the simulation process, an inter-segment state transfer mechanism is established, and the boundary conditions at the end of the current stage are used as the initial boundary conditions for the next stage for simulation. Compared with traditional methods, this reduces the simulation workload and ensures the continuity and coordination between different sections of the track.

[0177] 3. This invention combines genetic algorithms and machine learning models in the multi-objective optimization process of electromagnetic tracks, effectively improving optimization efficiency and accuracy. Firstly, a non-dominated sorting genetic algorithm is used to achieve decoupled optimization of multiple track performance indicators, ensuring that the optimization results possess good multi-objective balance. Secondly, a support vector machine machine learning method is introduced to train and model finite element simulation data, establishing a nonlinear mapping relationship between design parameters and key response variables, significantly reducing the computational cost of repeated simulations.

[0178] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for designing a track of an electromagnetic track launcher, characterized in that: The following steps are involved: Step S1: Construct a finite element model of an electromagnetic track with a parameterized structure. The track is divided into a starting section, an acceleration section, and a launch section according to the armature motion process. A double-layer structure consisting of a conductive base layer and a functional contact layer is adopted, and the thickness of the functional contact layer is set as an adjustable design variable. Step S2: Using an experimental design method to generate a test set with different functional contact layer thicknesses; running the electromagnetic track finite element model, establishing a batch simulation calculation process based on the design variables, extracting multi-physics field response indicator data under each design variable, and constructing a key performance indicator data set; using a machine learning method to train the key performance indicator data set to obtain a prediction function, forming a mapping relationship between the track composite structure thickness design and the key performance indicators; Step S3: Based on the key performance indicator dataset constructed in step S2, a multi-objective optimization function and constraints are defined; a non-dominated sorting genetic algorithm is used to optimize the thickness parameters of the functional contact layer of each track segment, and a Pareto optimal solution set corresponding to the composite structure of each track segment is generated; The multi-objective function is defined as follows: min t F(t)=[f1(t),f2(t),f3(t)] T (2) Where f1(t) = R(t) is the resistance per unit length; f2(t) = ΔT(t) is the maximum temperature rise; f3(t) = W(t) is the wear amount; The constraints are: (1) Thickness constraint: Ensure that the thickness of the functional contact layer is within the constraint range; (2) Performance constraints: The maximum temperature rise of the track does not exceed the heat resistance temperature of the material, and the armature outlet speed is not lower than the design requirement value; Step S4: Using the representative solution in the Pareto optimal solution set of the starting segment as the input of the electromagnetic track finite element model, the corresponding state output is obtained and used as the initial condition for the acceleration segment simulation. This state output is then transferred backwards to build an inter-segment state transfer mechanism, realizing the linkage of performance states between different segments and obtaining the performance output of the entire track combination. Step S5: Perform normalized analysis and ranking of the full track performance indicators of each combined solution. Use a decision-making method based on the target weight priority to perform a multi-objective compromise evaluation on all candidate solutions. Finally, select the optimal three-segment track configuration combination and output the corresponding performance indicator report and structure distribution diagram.

2. The electromagnetic rail launcher track design method and system according to claim 1, characterized in that: Step S1 is specifically as follows: Step S11: determining the geometric parameters of the electromagnetic track finite element model, dividing the track into a starting section, an acceleration section, and a launch section, and determining the geometric dimensions of the starting section, the acceleration section, and the launch section respectively; Step S12: defining the material properties of the rail composite structure in the finite element simulation software, and setting the material properties of the armature guide rail portion of the electromagnetic rail launcher; Step S13: performing finite element mesh division and optimization, performing mesh refinement processing on key areas of the track composite structure, and inspecting and optimizing the mesh quality; wherein the key areas include contact interfaces and material transition areas; Step S14: setting multiple physical fields in the electromagnetic track finite element model, including electromagnetic field, temperature field and structural field, and defining the coupling relationship between the physical fields; Step S15: Establish a parameterized interface program, set the thickness of the functional contact layer as the main design variable, and realize the automated process of geometric parameters, meshing, solving and post-processing.

3. The electromagnetic rail launcher track design method and system according to claim 1, characterized in that: Step S2 is specifically as follows: Step S21: Determine the design variable range of each track section, establish the constraint boundary of the functional contact layer thickness of each section, and use the experimental design method to generate a test set with different functional contact layer thicknesses; Step S22: Establishing a batch simulation calculation process, inputting the functional contact layer thickness in each test set into the electromagnetic track finite element model established in step S1, and controlling the entire process of geometric parameter update, mesh regeneration, and multi-physics field solution calculation of the electromagnetic track finite element model through an automated script; Step S23: Based on the calculation results of step S22, extract the multi-physics field response index data in the finite element simulation and build a key performance index data set; Step S24: Using a machine learning method to train the key performance indicator data set to obtain a prediction function, and establish a mapping relationship between the design variables and the multi-physics field key performance indicators.

4. The electromagnetic rail launcher track design method and system according to claim 1, characterized in that: Step S4 is specifically as follows: Step S41: Select a representative solution from the Pareto optimal solution set of the initial stage, and extract the corresponding state output data as the initial condition of the acceleration stage simulation; Step S42: Based on the initial conditions delivered in step S41, perform a multi-physics field coupling simulation of the acceleration section, then perform the key performance indicator data set construction in step S2 to establish an acceleration section prediction function model; Step S43: Execute the multi-objective optimization of step S3 based on the acceleration segment prediction function model of S42, use the multi-objective optimization algorithm to generate the Pareto optimal solution set of the acceleration segment, select a representative solution from the Pareto optimal solution set, and extract the corresponding state output data as the initial conditions of the launch segment simulation; Step S44: Based on the initial conditions passed in step S43, perform multi-physics field coupling simulation of the launch segment, execute the key performance indicator data set construction of step S2, and establish the launch segment prediction function model; execute the multi-objective optimization of step S3 to generate the Pareto solution set of the launch segment; Step S45: Establish a combination model, systematically combine the geometric parameters, material properties and thickness distribution of the starting segment, acceleration segment and launch segment to form a complete full-track thickness combination solution, perform full-track multi-physics field simulation, and calculate and obtain the performance index data of the full-track thickness combination.

5. The electromagnetic rail launcher track design method and system according to claim 2, characterized in that: In step S12, the material properties include electrical, thermal and mechanical parameters.

6. The electromagnetic rail launcher track design method and system according to claim 3, characterized in that: In step S21, the experimental design method adopts the central composite design method.

7. The electromagnetic rail launcher track design method and system according to claim 3, characterized in that: In step S24, the machine learning method adopts a support vector machine method.

8. The electromagnetic rail launcher track design method and system according to claim 1, characterized in that: The decision-making method in step S5 adopts the entropy weight TOPSIS method, which includes: (1) Construct the original evaluation matrix; The performance index data of all track thickness combinations are obtained from step S4 to form an original evaluation matrix, which is expressed as: Where: X is the original evaluation matrix; x ij represents the value of the i-th optimization solution on the j-th performance index; m is the number of optimization solutions; n is the number of performance indicators contained in each solution; (2) Normalization by range method; Since the performance indicators are all inverse indicators, the range method is used to normalize the original evaluation matrix. The range method is expressed as: Where: is the maximum value of the jth index; is the minimum value of the jth index; r ij ∈[0,1] is an element in the standardized decision matrix; (3) Entropy weight method to determine indicator weights; The entropy weight method indicator weight is expressed as: Where: w j is the weight of the j-th indicator; e j is the entropy value of the j-th indicator; is the proportion of the jth indicator; (4) Constructing a weighted normalization matrix; Use weight w j Weighting the standardized decision matrix gives the weighted decision matrix: v ij =w j ×r ij (6) Where: v ij is the element in the weighted decision matrix; (5) Determine the positive ideal solution and the negative ideal solution; Where: A + represents the positive ideal solution set; A - represents the set of positive ideal solutions; represents a positive ideal solution; represents a negative ideal solution; (6) Calculate the distance to the ideal solution; The Euclidean distance between the i-th optimization solution and the positive ideal solution is: The Euclidean distance between the i-th optimization solution and the negative ideal solution is: (7) Calculate proximity; The closeness coefficient of the i-th optimization solution: Among them C i ∈[0,1],C i The larger the value, the closer the optimization solution is to the ideal solution, and the performance is the best. Select C i The largest solution is regarded as the optimal combination.

9. A track design system for an electromagnetic track launcher, characterized in that: An electromagnetic rail launcher track design system that executes the electromagnetic rail launcher track design method according to claims 1-8 comprises: A model construction module is used to construct a finite element model of an electromagnetic track with a parameterized structure. The track is divided into three parts: a starting section, an acceleration section, and a launch section according to the armature motion process. A double-layer structure consisting of a conductive base layer and a functional contact layer is adopted, and the thickness of the contact layer is set as an adjustable design variable. The model construction includes determining geometric parameters, defining the material properties of the track composite structure, performing finite element meshing and optimization, defining multi-physics field coupling, and establishing a parameterized interface program. The data acquisition module is used to generate a test set with different contact layer thicknesses using an experimental design method, run the electromagnetic track finite element model to establish a batch simulation calculation process, extract multi-physics field response indicator data under various design variable combinations, and use machine learning methods to train the key performance indicator data set to obtain a prediction function, thereby forming a mapping relationship between the track structure thickness design and performance indicators; The optimization solution module defines the multi-objective optimization function and constraints based on the data set and prediction function built by the data construction module. It uses the non-dominated sorting genetic algorithm to optimize the contact layer thickness parameters of each stage of the track and generate the Pareto optimal solution set corresponding to each section of the track structure. The decision-making evaluation module is used to perform normalized analysis and ranking of the full track performance indicators of each segment combination solution. It uses the entropy weight TOPSIS decision-making method to perform multi-objective compromise evaluation on all candidate solutions, and finally selects the optimal three-segment track configuration combination and outputs the corresponding performance indicator report and structure distribution diagram.

Images