Transfer case vibration simulation analysis optimization method and device

Abstract

The invention discloses a transfer case vibration simulation analysis optimization method and device, and the method comprises the steps: building a transfer case finite element model, simulating each bearing of a transfer case shaft gear, and obtaining a bearing dynamic load; acquiring the inherent frequency and the inherent vibration mode of the first ten orders of the transfer case shell, inputting variable parameters into the finite element model of the shell, and performing simulation solution analysis on the frequency response of the box body through a modal superposition method; establishing parameter sample data points and level numbers on the basis of variable parameters, generating an orthogonal test table, and respectively performing an independent simulation test on each parameter group to obtain an average torque and a vibration acceleration level corresponding to each parameter group; the variable parameters serve as target parameters, a neural network prediction model is established to optimize the multi-target parameters, and an optimal solution set with the maximum average torque and the minimum vibration acceleration level is obtained through iterative operation; optimization is carried out through simulation analysis so as to reduce or avoid vibration, data support is provided for rack construction of a transfer case high-speed verification test, test problems are reduced, and test working efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle simulation analysis, and in particular to a transfer case vibration simulation analysis optimization method and device. Background Art

[0002] With the continuous development of the automobile industry, in order to better improve the safety and comfort of consumers' driving and riding, the requirements for automobile testing are becoming more and more stringent. However, the test is affected by sample processing and manufacturing errors, installation errors, gear load deformation and other reasons. When running at high speed, additional vibration will be generated, which has a great impact on the accuracy of the test.

[0003] For example, when testing a transfer case, power is transmitted through the power motor to cause the gears to mesh and rotate, and then transmitted to the case through the shaft and bearings, which excites the case to vibrate outward, causing the transfer case to deviate from the equilibrium position and rotate at high speed. On the one hand, this vibration is transmitted to the case through the shaft teeth, making the vibration more complex and changeable, causing fatigue damage to the transfer case; on the other hand, when the vibration is close to or even equal to the natural frequency of the transfer case itself, it is easy to cause resonance, causing unpredictable damage to the transfer case structure and making the test site more dangerous, so it should be avoided as much as possible.

[0004] Based on the above theoretical analysis, observation and thinking, it is necessary to simulate and analyze the automobile transfer case for the vibration problem in the transfer case test, and analyze the vibration of the transfer case at high speed, so as to provide technical support for the transfer case test. It is possible to evaluate and analyze the transfer case vibration before the test, determine whether there are problems and risks, and carry out targeted design optimization according to the problems found, so as to speed up problem solving and improve work efficiency. Summary of the invention

[0005] The object of the present invention is to provide a transfer case vibration simulation analysis optimization method and device to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a transfer case vibration simulation analysis optimization method, comprising:

[0007] Establish a transfer case finite element model, which includes a transfer case housing finite element model and a transfer case shaft gear finite element model, simulate the bearings of the transfer case shaft gear, and obtain the bearing dynamic load;

[0008] Obtain the first ten natural frequencies and natural vibration modes of the transfer case housing, input variable parameters into the finite element model of the housing, and simulate and analyze the frequency response of the housing through the modal superposition method. The variable parameters include dynamic excitation, set analysis frequency range, analysis step, and system damping of the transfer case. The dynamic excitation is the support reaction force of the bearing dynamic load.

[0009] Based on the variable parameters, parameter sample data points and level numbers are established to generate an orthogonal test table. Multiple parameter groups are generated according to the parameter sample data points and level numbers of the orthogonal test table. Independent simulation tests are performed on each parameter group to obtain the average torque and vibration acceleration level corresponding to each parameter group.

[0010] Taking the variable parameters as target parameters, a neural network prediction model is established to optimize the multi-objective parameters. Through iterative calculation, the optimal solution set with the maximum average torque and the minimum vibration acceleration level is obtained.

[0011] Furthermore, a finite element model of the transfer case shaft gear is established, and each bearing of the transfer case shaft gear is simulated to obtain the dynamic load of the bearing, which also includes:

[0012] Establish accurate gear profile and obtain transition curve parameters by solving tooth root transition curve parameter equation;

[0013] Define materials and constraints, set contact forces and drives, where drives include speed and torque;

[0014] The second-order tetrahedral mesh unit is used for mesh division, and the bearing and the shaft are connected by tie;

[0015] Each bearing is simulated and the dynamic load of each bearing in the X, Y and Z directions is obtained.

[0016] Furthermore, the differential equation of motion of the transfer case is also included:

[0017]

[0018] Where [M] is the structural mass matrix; [C] is the damping matrix; [K] is the stiffness coefficient matrix; are the system acceleration vector, velocity vector and displacement vector respectively; F(t) is the dynamic load.

[0019] Furthermore, obtaining the first ten natural frequencies and natural vibration modes of the transfer case housing also includes:

[0020] Tetrahedron elements are used to mesh the finite element model of the transfer case housing.

[0021] Define the shell material and set Young's modulus, Poisson's ratio and density parameters;

[0022] The fixed constraints of the transfer case are set, the bolted connection housing is simulated by rbe2 unit, and the first ten natural frequencies and natural vibration modes of the transfer case housing are solved.

[0023] Furthermore, parameter sample data points and the number of levels are established based on the variable parameters to generate an orthogonal test table, and multiple parameter groups are generated according to the parameter sample data points and the number of levels in the orthogonal test table. An independent simulation test is performed on each parameter group to obtain the average torque and vibration acceleration level corresponding to each parameter group, which also includes:

[0024] Based on the selection requirements of variable parameters of the orthogonal test method, multiple variable parameters are selected to establish parameter sample data points, which are used as collaborative optimization variable parameters for level analysis, and an orthogonal experimental table is generated based on the parameter sample data points and the number of levels;

[0025] An independent simulation test was performed on each parameter group, and the average torque and vibration acceleration level corresponding to each parameter group were obtained.

[0026] Furthermore, the variable parameters are used as target parameters, and a neural network prediction model is established to optimize the multi-target parameters, which also includes:

[0027] Establish BP neural network prediction model;

[0028] The parameter group of the orthogonal test table is used as a data set, and different parameter groups are used as input variables to output the corresponding average torque and vibration acceleration;

[0029] The determination coefficient is set and used as an indicator to evaluate the performance of the BP neural network prediction model, and a mathematical model of the transfer case vibration is established.

[0030] Furthermore, the mathematical model formula of the transfer case vibration is:

[0031]

[0032] Among them, y is the output, namely the average torque and vibration acceleration.

[0033] Furthermore, the BP neural network prediction model is established, which also includes:

[0034] Batch input orthogonal test parameter samples and normalize the input and output quantities;

[0035] Set the initialization parameters of the BP neural network prediction model, including the number of training times, learning accuracy, number of hidden nodes, initial weights, thresholds, and learning rates;

[0036] Input orthogonal experimental sample point samples and calculate the output values ​​of each layer;

[0037] Set the output layer error threshold and calculate the output layer error. If it is not less than the output layer error threshold, adjust the parameter variables and recalculate the output values ​​of each layer. If it is less than the output layer error threshold, output the corresponding average torque and vibration acceleration.

[0038] Furthermore, through iterative calculation, the optimal solution set with the maximum average torque and the minimum vibration acceleration level is obtained, which also includes:

[0039] Based on the multi-objective genetic algorithm, the non-dominated set of the data optimization set is constructed;

[0040] Merge the parent and child data populations so that the next generation has double the selection space to obtain the best Pareto solution set;

[0041] Through Matlab programming, establish the objective function to be optimized, add constraints, set the population size, crossover coefficient, optimal front-end individual coefficient, maximum number of iterations, and constraint tolerance;

[0042] Running the program obtains the optimal solution set with the maximum average torque and the minimum vibration acceleration level.

[0043] On the other hand, a transfer case vibration simulation analysis and optimization device is provided, which uses any of the above transfer case vibration simulation analysis and optimization methods, including:

[0044] A building module for establishing a finite element model of a transfer case;

[0045] The simulation module is used to obtain the first ten natural frequencies and natural vibration modes of the transfer case housing, input variable parameters into the finite element model of the housing, and simulate and analyze the frequency response of the housing through the modal superposition method. The variable parameters include dynamic excitation, set analysis frequency range, analysis step, and system damping of the transfer case. The dynamic excitation is the support reaction force of the dynamic load of the bearing.

[0046] A processing module is used to establish parameter sample data points and level numbers based on variable parameters, generate an orthogonal test table, and generate multiple parameter groups according to the parameter sample data points and level numbers of the orthogonal test table, and perform independent simulation tests on each parameter group to obtain the average torque and vibration acceleration level corresponding to each parameter group;

[0047] The optimization module is used to take the variable parameters as the target parameters, establish a neural network prediction model to optimize the multi-objective parameters, and obtain the optimal solution set with the maximum average torque and the minimum vibration acceleration level through iterative calculation.

[0048] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention provides a high-speed vibration simulation analysis method for an automobile transfer case, which relates to the field of automobile vibration simulation modeling and analysis. Mainly based on the vibration problem of the transfer case housing, the finite element model of the transfer case is simplified and modeled; starting from the modal analysis method, by obtaining the key parameters such as the Young's modulus and Poisson's ratio of the transfer case housing, shaft gear and other mechanisms, the inherent modal analysis and vibration analysis of the transfer case are performed using the simulation analysis method, and based on this, the transfer case shaft gear is used as the dynamic load excitation, and the inherent modal analysis and vibration response analysis of the transfer case are performed to obtain the overall natural frequency and vibration type analysis of the transfer case. Then, the key influencing parameters are selected, the orthogonal test is designed, and the orthogonal sample points are learned and predicted based on the BP neural network algorithm to obtain the mathematical model of the key influencing parameters and the system vibration; on the basis of the model, the multi-objective genetic algorithm is used to find the global optimal solution, so as to reduce or avoid the generation of vibration, provide data support for the bench construction of the transfer case high-speed verification test, reduce test problems, and improve test work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A flow chart of a vibration simulation analysis and optimization method for a transfer case in an embodiment of the present invention;

[0050] Figure 2 It is a connection block diagram of a transfer case vibration simulation analysis and optimization device in an embodiment of the present invention;

[0051] Figure 3 It is a logic block diagram of the BP neural network method in an embodiment of the present invention;

[0052] Figure 4 It is a logic block diagram of a multi-objective genetic algorithm in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] In the description of the present invention, unless otherwise specified, "plurality" means two or more than two; the orientations or positional relationships indicated by the terms "upper", "lower", "left", "right", "inner", "outer", "front end", "rear end", "head", "tail", etc. are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0055] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0056] Please refer to the attached drawings of the specification, the present invention provides a technical solution: a transfer case vibration simulation analysis optimization method, comprising the following steps:

[0057] S102, establishing a transfer case finite element model, the transfer case finite element model including a transfer case housing finite element model and a transfer case shaft gear finite element model, simulating each bearing of the transfer case shaft gear to obtain a bearing dynamic load;

[0058] It can be understood that, without reducing the system simulation accuracy and maintaining the rotational vibration characteristics of the transfer case, the finite element modeling of the transfer case is simplified as much as possible to conform to the theoretical model; the simplified finite element model of the transfer case is obtained, and the simplified finite element model of the transfer case includes a finite element model of the transfer case housing and a finite element model of the transfer case shaft gear.

[0059] Specifically, an accurate gear profile is established, and transition curve parameters are obtained by solving the tooth root transition curve parameter equation;

[0060] Define materials and constraints, set contact forces and drives, where drives include speed and torque;

[0061] The second-order tetrahedral mesh unit is used for mesh division, and the bearing and the shaft are connected by tie;

[0062] Each bearing is simulated and the dynamic load of each bearing in the X, Y and Z directions is obtained.

[0063] Among them, material definition, constraint definition, contact force setting, drive setting, including speed and torque, etc., mesh division, using second-order tetrahedral mesh units, the bearing and the shaft are connected by ties, and the slight chamfers of the structure are ignored. Each bearing is simulated to obtain the dynamic load of each bearing in the X, Y, and Z directions.

[0064] Specifically, it also includes the differential equation of motion of the transfer case:

[0065]

[0066] Where [M] is the structural mass matrix; [C] is the damping matrix; [K] is the stiffness coefficient matrix; are the system acceleration vector, velocity vector and displacement vector respectively; F(t) is the dynamic load.

[0067] S104, obtaining the first ten natural frequencies and natural vibration modes of the transfer case housing, inputting variable parameters into the finite element model of the housing, and performing simulation and analysis on the frequency response of the housing by means of a modal superposition method, wherein the variable parameters include dynamic excitation, a set analysis frequency range, an analysis step, and system damping of the transfer case, wherein the dynamic excitation is the support reaction force of the dynamic load of the bearing;

[0068] Specifically, tetrahedral elements are used to mesh the finite element model of the transfer case housing;

[0069] Define the shell material and set Young's modulus, Poisson's ratio and density parameters;

[0070] The fixed constraints of the transfer case are set, the bolted connection housing is simulated by rbe2 unit, and the first ten natural frequencies and natural vibration modes of the transfer case housing are solved.

[0071] Among them, the three-dimensional model is imported into the finite element software for calculation, the shell material is defined, and key parameters such as Young's modulus, Poisson's ratio, and density are set; meshing, the introduced transfer case shell model is meshed using tetrahedron units. A better mesh can make the model conform to the actual situation, and is conducive to solving the time required and improving the design efficiency; the fixed constraints of the transfer case are set, and the bolted shell is simulated by the rbe2 unit to solve the first 10 natural frequencies and natural vibration modes of the transfer case shell; the modal superposition method is used to solve and analyze the frequency response of the box, and the dynamic support reaction force of the bearing is used as the dynamic load amplitude for simulation, the three-dimensional model of the shell is imported, the mesh is divided, and it is added as a dynamic excitation, and the analysis frequency range and analysis step are set, and the system damping is set.

[0072] S106, establishing parameter sample data points and the number of levels based on the variable parameters, generating an orthogonal test table, and generating multiple parameter groups according to the parameter sample data points and the number of levels in the orthogonal test table, performing independent simulation tests on each parameter group, and obtaining the average torque and vibration acceleration level corresponding to each parameter group;

[0073] Specifically, based on the requirements of the orthogonal experimental method for the selection of variable parameters, multiple variable parameters are selected to establish parameter sample data points, which are used as collaborative optimization variable parameters for level analysis. An orthogonal experimental table is generated based on the parameter sample data points and the number of levels. Independent simulation tests are performed on each parameter group to obtain the average torque and vibration acceleration level corresponding to each parameter group.

[0074] Among them, based on the orthogonal test method's requirements for variable parameter selection, multiple reliable parameters are selected as collaborative optimization variables, and orthogonal test sample points are constructed. In order to save time cost and reduce the number of test samples, representative and evenly distributed parameters are selected as collaborative optimization variables for horizontal analysis. Based on the orthogonal test method, parameter sample data points and level numbers are designed, and an orthogonal test table is established. According to the designed orthogonal test table, independent simulation tests are performed to obtain the average torque and vibration acceleration level corresponding to each set of optimization parameters.

[0075] S108. Using the variable parameters as target parameters, a neural network prediction model is established to optimize the multi-target parameters, and an optimal solution set with the maximum average torque and the minimum vibration acceleration level is obtained through iterative calculations.

[0076] Specifically, a BP neural network prediction model is established;

[0077] The parameter group of the orthogonal test table is used as a data set, and different parameter groups are used as input variables to output the corresponding average torque and vibration acceleration;

[0078] The determination coefficient is set and used as an indicator to evaluate the performance of the BP neural network prediction model, and a mathematical model of the transfer case vibration is established.

[0079] Among them, Figure 3 As shown in the figure, the BP neural network method with high fitting accuracy and strong stability is used to obtain the function expression between random parameters and dynamic response. The orthogonal test sample points are used as data sets, different parameters are used as input variables, and the average torque and vibration acceleration are output. The determination coefficient is used as an indicator to evaluate the performance of the model. The range of the determination coefficient R is in [0,1]. The closer it is to 1, the better the model fits. Therefore, the mathematical model of transfer case vibration is established.

[0080] Specifically, the mathematical model formula of the transfer case vibration is:

[0081]

[0082] Among them, y is the output, namely the average torque and vibration acceleration.

[0083] Specifically, establishing a BP neural network prediction model also includes:

[0084] Batch input orthogonal test parameter samples and normalize the input and output quantities;

[0085] Set the initialization parameters of the BP neural network prediction model, including the number of training times, learning accuracy, number of hidden nodes, initial weights, thresholds, and learning rates;

[0086] Input orthogonal experimental sample point samples and calculate the output values ​​of each layer;

[0087] Set the output layer error threshold and calculate the output layer error. If it is not less than the output layer error threshold, adjust the parameter variables and recalculate the output values ​​of each layer. If it is less than the output layer error threshold, output the corresponding average torque and vibration acceleration.

[0088] Among them, BP neural network is a back propagation algorithm with good nonlinear mapping ability. It adjusts weights through negative feedback of errors until the error between network output and expected output reaches the required performance index. It is a network composed of three parts: input layer, hidden layer and output layer. The obtained orthogonal test sample points are used as data sets, different parameters are used as input vectors, and the output layer refers to vibration acceleration and average torque; its workflow is to first process parameters, normalize the parameters, and then set the number of training times, learning accuracy, number of hidden nodes, initial tertiary value, threshold, learning rate, etc. required by the algorithm, and then input orthogonal test sample points for calculation, and finally calculate the error. If the error does not meet the requirements, continue.

[0089] Specifically, through iterative calculation, the optimal solution set with the maximum average torque and the minimum vibration acceleration level is obtained, which also includes:

[0090] Based on the multi-objective genetic algorithm, the non-dominated set of the data optimization set is constructed;

[0091] Merge the parent and child data populations so that the next generation has double the selection space to obtain the best Pareto solution set;

[0092] Through Matlab programming, establish the objective function to be optimized, add constraints, set the population size, crossover coefficient, optimal front-end individual coefficient, maximum number of iterations, and constraint tolerance;

[0093] Running the program obtains the optimal solution set with the maximum average torque and the minimum vibration acceleration level.

[0094] Among them, based on the multi-objective parameter optimization of the neural network model: as shown in the attached Figure 4 , through the multi-objective genetic algorithm (NSGA-Ⅱ), the influence of parameter changes on torque and vibration is collaboratively considered, the non-dominated set of the data optimization set is constructed, and the parent data and child data populations are merged so that the next generation has double the selection space, thereby obtaining relatively excellent data individuals, and thus obtaining the best Pareto solution set. Through Matlab programming, the objective function to be optimized is established, constraints are added, the population size, crossover coefficient, optimal front-end individual coefficient, maximum number of iterations, and constraint tolerance are set. Running the program obtains the optimal solution set with the maximum average torque and the minimum vibration acceleration level; Figure 4 As shown in the figure, its workflow is as follows: first initialize the population and randomly generate a group of individuals (parameters), each of which represents a potential solution; then calculate the fitness and use a multi-objective function to calculate the fitness value of the individual; then perform Pareto front sorting, sort the individuals according to the fitness value and divide them into different Pareto fronts; then perform crowding evaluation, and in each Pareto front, calculate the distance between individuals to maintain the diversity of the population; then perform selection, crossover, and mutation, select a group of individuals from the population, and then generate new individuals through crossover and mutation operations, and then update the population, replace the original individuals with the new individuals, and form the next generation of population; finally, judge the termination condition, and stop the algorithm when the predetermined number of iterations is reached or a satisfactory solution is obtained.

[0095] The simulation analysis method of the transfer case under high-speed operation uses the transfer case shaft teeth as the external excitation source to realize the vibration simulation analysis caused by the transfer case operation, determine the weak position of the vibration response, and use the orthogonal experimental data to establish the BP neural network prediction model. Then, the parameters are optimized through the neural network model to provide technical support for the implementation and optimization of the test plan.

[0096] On the other hand, a transfer case vibration simulation analysis and optimization device is provided, comprising:

[0097] A construction module 10 is used to establish a finite element model of a transfer case;

[0098] Among them, the precise gear profile is established, and the transition curve parameters are obtained by solving the tooth root transition curve parameter equation;

[0099] Define materials and constraints, set contact forces and drives, where drives include speed and torque;

[0100] The second-order tetrahedral mesh unit is used for mesh division, and the bearing and the shaft are connected by tie;

[0101] Each bearing is simulated and the dynamic load of each bearing in the X, Y and Z directions is obtained.

[0102] The simulation module 20 is used to obtain the first ten natural frequencies and natural vibration modes of the transfer case housing, input variable parameters into the finite element model of the housing, and simulate and solve the frequency response of the housing by the modal superposition method, wherein the variable parameters include dynamic excitation, a set analysis frequency range, an analysis step, and the system damping of the transfer case, wherein the dynamic excitation is the support reaction force of the dynamic load of the bearing;

[0103] Among them, tetrahedral units are used to divide the mesh of the finite element model of the transfer case housing; the shell material is defined, and Young's modulus, Poisson's ratio and density parameters are set; the fixed constraints of the transfer case are set, and the bolted shell is simulated by rbe2 units to solve the first ten natural frequencies and natural vibration modes of the transfer case housing.

[0104] The processing module 30 is used to establish parameter sample data points and the number of levels based on the variable parameters, generate an orthogonal test table, and generate multiple parameter groups according to the parameter sample data points and the number of levels in the orthogonal test table, and perform independent simulation tests on each parameter group to obtain the average torque and vibration acceleration level corresponding to each parameter group;

[0105] Among them, based on the requirements of the orthogonal experimental method for the selection of variable parameters, multiple variable parameters are selected to establish parameter sample data points, which are used as collaborative optimization variable parameters for level analysis, and an orthogonal experimental table is generated based on the parameter sample data points and the number of levels; independent simulation tests are performed on each parameter group, and the average torque and vibration acceleration level corresponding to each parameter group are obtained.

[0106] The optimization module 40 is used to use the variable parameters as target parameters, establish a neural network prediction model to optimize the multi-target parameters, and obtain the optimal solution set with the maximum average torque and the minimum vibration acceleration level through iterative calculation.

[0107] Among them, a BP neural network prediction model is established; the parameter group of the orthogonal test table is used as the data set, and different parameter groups are used as input variables to output the corresponding average torque and vibration acceleration; the determination coefficient is set, and the determination coefficient is used as an indicator to evaluate the performance of the BP neural network prediction model, and a mathematical model of transfer case vibration is established.

[0108] Among them, a BP neural network prediction model is established, orthogonal test parameter samples are input in batches, and the input and output quantities are normalized; the initialization parameters of the BP neural network prediction model are set, and the initialization parameters include the number of training times, learning accuracy, number of hidden nodes, initial weights, thresholds, and learning rates; orthogonal test sample point samples are input, and the output values ​​of each layer are calculated; the output layer error threshold is set, and the output layer error is calculated. If it is not less than the output layer error threshold, the parameter variables are adjusted to recalculate the output values ​​of each layer. If it is less than the output layer error threshold, the corresponding average torque and vibration acceleration are output;

[0109] Among them, based on the multi-objective genetic algorithm, a non-dominated set of the data optimization set is constructed; the parent data and child data populations are merged so that the next generation has double the selection space to obtain the best Pareto solution set; through Matlab programming, the objective function to be optimized is established, constraints are added, the population size, crossover coefficient, optimal front-end individual coefficient, maximum number of iterations, and constraint tolerance are set; and the program is run to obtain the optimal solution set with the maximum average torque and the minimum vibration acceleration level.

[0110] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A transfer case vibration simulation analysis and optimization method, characterized in that: include: Establish a transfer case finite element model, which includes a transfer case housing finite element model and a transfer case shaft gear finite element model, simulate the bearings of the transfer case shaft gear, and obtain the bearing dynamic load; Obtain the first ten natural frequencies and natural vibration modes of the transfer case housing, input variable parameters into the finite element model of the housing, and simulate and analyze the frequency response of the housing through the modal superposition method. The variable parameters include dynamic excitation, set analysis frequency range, analysis step, and system damping of the transfer case. The dynamic excitation is the support reaction force of the bearing dynamic load. Based on the variable parameters, parameter sample data points and level numbers are established to generate an orthogonal test table. Multiple parameter groups are generated according to the parameter sample data points and level numbers of the orthogonal test table. Independent simulation tests are performed on each parameter group to obtain the average torque and vibration acceleration level corresponding to each parameter group. Taking the variable parameters as target parameters, a neural network prediction model is established to optimize the multi-objective parameters. Through iterative calculation, the optimal solution set with the maximum average torque and the minimum vibration acceleration level is obtained.

2. A transfer case vibration simulation analysis and optimization method according to claim 1, characterized in that: Establish a finite element model of the transfer case shaft gear, simulate the bearings of the transfer case shaft gear, and obtain the dynamic load of the bearings, including: Establish accurate gear profile and obtain transition curve parameters by solving tooth root transition curve parameter equation; Define materials and constraints, set contact forces and drives, where drives include speed and torque; The second-order tetrahedral mesh unit is used for mesh division, and the bearing and the shaft are connected by tie; Each bearing is simulated and the dynamic load of each bearing in the X, Y and Z directions is obtained.

3. A transfer case vibration simulation analysis and optimization method according to claim 2, characterized in that: The differential equations of motion for the transfer case are also included: Where [M] is the structural mass matrix; [C] is the damping matrix; [K] is the stiffness coefficient matrix; are the system acceleration vector, velocity vector and displacement vector respectively; F(t) is the dynamic load.

4. The transfer case vibration simulation analysis and optimization method according to claim 1, characterized in that: Obtain the first ten natural frequencies and natural vibration modes of the transfer case housing, including: Tetrahedron elements are used to mesh the finite element model of the transfer case housing. Define the shell material and set Young's modulus, Poisson's ratio and density parameters; The fixed constraints of the transfer case are set, the bolted connection housing is simulated by rbe2 unit, and the first ten natural frequencies and natural vibration modes of the transfer case housing are solved.

5. The transfer case vibration simulation analysis and optimization method according to claim 1, characterized in that: Based on the variable parameters, parameter sample data points and level numbers are established to generate an orthogonal test table. Multiple parameter groups are generated according to the parameter sample data points and level numbers of the orthogonal test table. Independent simulation tests are performed on each parameter group to obtain the average torque and vibration acceleration level corresponding to each parameter group, including: Based on the selection requirements of variable parameters of the orthogonal test method, multiple variable parameters are selected to establish parameter sample data points, which are used as collaborative optimization variable parameters for level analysis, and an orthogonal experimental table is generated based on the parameter sample data points and the number of levels; An independent simulation test was performed on each parameter group, and the average torque and vibration acceleration level corresponding to each parameter group were obtained.

6. The transfer case vibration simulation analysis and optimization method according to claim 1, characterized in that: Taking variable parameters as target parameters, establishing a neural network prediction model to optimize multi-objective parameters also includes: Establish BP neural network prediction model; The parameter group of the orthogonal test table is used as a data set, and different parameter groups are used as input variables to output the corresponding average torque and vibration acceleration; The determination coefficient is set and used as an indicator to evaluate the performance of the BP neural network prediction model, and a mathematical model of the transfer case vibration is established.

7. The transfer case vibration simulation analysis and optimization method according to claim 6, characterized in that: Mathematical model formula of transfer case vibration: Among them, y is the output, namely the average torque and vibration acceleration.

8. The transfer case vibration simulation analysis and optimization method according to claim 6, characterized in that: Establishing the BP neural network prediction model also includes: Batch input orthogonal test parameter samples and normalize the input and output quantities; Set the initialization parameters of the BP neural network prediction model, including the number of training times, learning accuracy, number of hidden nodes, initial weights, thresholds, and learning rates; Input orthogonal experimental sample point samples and calculate the output values ​​of each layer; Set the output layer error threshold and calculate the output layer error. If it is not less than the output layer error threshold, adjust the parameter variables and recalculate the output values ​​of each layer. If it is less than the output layer error threshold, output the corresponding average torque and vibration acceleration.

9. The transfer case vibration simulation analysis and optimization method according to claim 1, characterized in that: Through iterative calculation, the optimal solution set with the maximum average torque and the minimum vibration acceleration level is obtained, which also includes: Based on the multi-objective genetic algorithm, the non-dominated set of the data optimization set is constructed; Merge the parent and child data populations so that the next generation has double the selection space to obtain the best Pareto solution set; Through Matlab programming, establish the objective function to be optimized, add constraints, set the population size, crossover coefficient, optimal front-end individual coefficient, maximum number of iterations, and constraint tolerance; Running the program obtains the optimal solution set with the maximum average torque and the minimum vibration acceleration level.

10. A transfer case vibration simulation analysis and optimization device, characterized in that: include: A building module for establishing a finite element model of a transfer case; The simulation module is used to obtain the first ten natural frequencies and natural vibration modes of the transfer case housing, input variable parameters into the finite element model of the housing, and simulate and analyze the frequency response of the housing through the modal superposition method. The variable parameters include dynamic excitation, set analysis frequency range, analysis step, and system damping of the transfer case. The dynamic excitation is the support reaction force of the dynamic load of the bearing. A processing module is used to establish parameter sample data points and level numbers based on variable parameters, generate an orthogonal test table, and generate multiple parameter groups according to the parameter sample data points and level numbers of the orthogonal test table, and perform independent simulation tests on each parameter group to obtain the average torque and vibration acceleration level corresponding to each parameter group; The optimization module is used to take the variable parameters as the target parameters, establish a neural network prediction model to optimize the multi-objective parameters, and obtain the optimal solution set with the maximum average torque and the minimum vibration acceleration level through iterative calculation.

Images