Electric drive axle gear vibration optimization method and system
By using the combination of EVO-Catboost and NRBO algorithms in the electric drive axle gear, a gear vibration response prediction model is established, and the shape modification parameters are optimized. The vibration noise problem of the electric drive axle is solved, and the efficient and rapid shape modification effect is achieved, which significantly reduces the shape modification effect. The vibration noise of the gear transmission system of the electric drive axle.
Patent Information
- Application Number
- CN202411054974.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-02
- Publication Date
- 2024-11-08
AI Technical Summary
The prior art has problems with long data processing time and low shape correction accuracy in improving the vibration noise of electric drive axle gears in new energy vehicles, and it is difficult to effectively reduce vibration and noise.
The gear vibration response prediction model is established by using the Energy Valley Optimization Algorithm (EVO-Catboost), combined with the NRBO algorithm to find the best, analyze the optimal shape modification parameters through the multi-environment shape modification selection model, reduce the shape modification time, improve the shape modification accuracy, and reduce vibration noise.
Effective shape modification under various working conditions is achieved, the vibration noise of the electric drive axle gear transmission system is significantly reduced, the shape modification accuracy and system stability are improved, and the shape modification time is reduced.
Smart Images

Figure CN118916994A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of gear optimization design, in particular to a vibration optimization method for electric drive axle gears. Background Art
[0002] In order to improve the vibration noise of the electric drive axle gear of new energy vehicles, the vibration noise mainly comes from the transmission error and stress concentration in the gear transmission process. Engineering experience can be used to perform micro-modification of gears to achieve vibration reduction and noise reduction, but there are great limitations. For this reason, model simulation and optimization algorithms are generally used to determine the gear modification parameters. For example, the Chinese patent document "A gear tooth profile modification method based on vibration reliability and genetic algorithm, publication number CN102314534A" includes: establishing a virtual prototype of elastic deformation of gear and modified gear meshing; dynamic simulation of gear transmission error during involute tooth profile and tooth profile modification; analyzing the reliability sensitivity of random parameter meshing transmission error of gear pairs; using genetic algorithm to determine the optimal parameters of gear tooth profile modification; verifying the correctness of parameters; analyzing and calculating the reliability sensitivity of each random modification parameter to the transmission error, and obtaining sensitivity diagrams and scatter diagrams; using genetic algorithms to optimize gear tooth profile modification parameters. The optimization data processing time of general model optimization modification methods is relatively long. Summary of the invention
[0003] The purpose of the present invention is to find an optimization method to reduce vibration and noise in response to the above problems, provide an electric drive axle gear vibration optimization method, establish an effective prediction model to accurately predict the vibration response of the electric drive axle, and use an energy valley optimization algorithm to reduce the shaping time, improve the shaping accuracy, and improve the stability and reliability of the system.
[0004] In order to achieve the above object, the technical solution adopted by the present invention is:
[0005] A method for optimizing vibration of electric drive axle gears, comprising the following contents:
[0006] Step S1, simulation analysis: establishing a parametric simulation model of the electric drive axle gear set, using an involute combined with a tooth direction modification method to obtain a data set of modification parameters of the electric drive axle gear and a peak-to-peak value of the transmission error, a maximum contact stress, and a load distribution coefficient;
[0007] Step S2, establishing a prediction model: establishing a gear vibration response prediction model based on EVO-Catboost, analyzing the mapping relationship between gear modification parameters and peak-to-peak value of transmission error, maximum contact stress and load distribution coefficient; using NRBO algorithm to solve the gear vibration response prediction model, optimizing with the peak-to-peak value of transmission error, maximum contact stress and load distribution coefficient of the electric drive axle gear as optimization targets, and obtaining the optimized modification parameters of the gear;
[0008] Step S3, selection model optimization: establish a multi-environment shaping selection model, design multiple shaping schemes for each optimal shaping parameter, perform shaping parameter sensitivity analysis under different environments, and obtain the optimal shaping parameter combination;
[0009] Step S4, modification effect analysis: importing the optimal modification parameter combination into the gear set parameterized simulation model to perform modification effect analysis.
[0010] Among them, the gear modification parameters include tooth guide crown, tooth guide slope, involute crown and involute slope.
[0011] The optimization solution processing flow of the vibration response prediction model of step S2 is as follows: step a1, initializing the prediction model parameters, setting iterations, learning rate, depth and L2 regularization; step a2, evaluating the fitness value of the candidate particles; step a3, determining the enrichment and binding of the particles; step a4, determining the enrichment level and stability level of the particles; step a5, finding the particles with the optimal stability level; step a6, judging whether the optimal parameters are found, if so, outputting the optimal parameters and ending the program, otherwise returning to execute step a2.
[0012] The optimization processing flow of the selection model of step S3: step b1, establish the optimization target as the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient under different working conditions; step b2, design i kinds of shaping schemes, i is the number of optimization targets; step b3, perform sensitivity analysis of shaping parameters under acceleration, deceleration and constant speed environment; step b4, set the optimization target weight.
[0013] As mentioned above, data sets are collected through modeling and simulation, and the gear vibration response prediction model is established using EVO-Catboost and optimized through NRBO. Then, the sensitivity analysis of the modification parameters under different environments is carried out through the multi-environment modification selection model, and finally the optimal modification parameter combination is obtained.
[0014] Based on the above scheme, in the improved scheme, the optimization method further includes the following contents: Step S5, bench verification: verifying the shaping effect of the optimal shaping parameter combination through the electric drive axle bench test. In this way, the shaping effect can be verified through the bench test.
[0015] Due to the adoption of the above technical solution, the present invention has the following beneficial effects:
[0016] The present invention collects data sets through modeling and simulation, uses EVO-Catboost to establish a gear vibration response prediction model and performs optimization through NRBO, and then uses a multi-environment modification selection model to perform modification parameter sensitivity analysis under different environments, and finally obtains the optimal modification parameter combination, ensuring that the modification scheme can achieve good results under various working conditions, reducing the modification time, improving the modification accuracy, and realizing effective and rapid optimization. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] Figure 1 It is a step diagram of Example 1 of the optimization method of the present invention.
[0018] Figure 2 It is a step diagram of Example 2 of the optimization method of the present invention.
[0019] Figure 3 It is a processing flow chart of Example 2 of the optimization method of the present invention.
[0020] Figure 4 It is a flow chart of the EVO optimized Catboost parameter processing of the optimization method example 2 of the present invention. DETAILED DESCRIPTION
[0021] The specific implementation of the invention is further described below with reference to the accompanying drawings.
[0022] Example 1
[0023] like Figure 1-Figure 2 As shown, as mentioned above, the basic example of the optimization method described in the present application includes the step S1-step S4 scheme, and an improved example includes the step S1-step S5 scheme. The following only takes the step S1-step S5 scheme as an example for specific description. For other example technical feature combinations, please refer to the above.
[0024] See also Figure 2-Figure 4 , a vibration optimization method for an electric drive axle gear of this embodiment 1 comprises the following steps:
[0025] Step S1, simulation analysis:
[0026] First, the simulation analysis and contact spot verification were carried out to establish a parameterized model of the electric drive axle gear set, and the effectiveness of the model was verified through gear contact spot simulation and experiments. The transmission process of the electric drive axle was analyzed by simulation, and the contact spot verification test was carried out. The verification results showed that the test results were basically consistent with the simulation results, proving the accuracy of the simulation model. This provides a reliable model foundation for subsequent optimization.
[0027] Then, the gear micro-modification is optimized by using the involute combined with tooth direction modification method to obtain the data sets of modification parameters of the electric drive axle gear set and the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient.
[0028] According to GB / Z6413.1-2003, the involute modification formula is as follows:
[0029]
[0030] In the formula, K A is the service factor; K mp is the branching coefficient; F t is the tangential force (unit N); b is the tooth width (unit mm); α t is the end pressure angle; c γ is the meshing stiffness.
[0031] The tooth modification formula is as follows:
[0032] C a =0.25b×10 -3 +0.5f g
[0033] f g =A(0.1b+10)
[0034] Where, b is the tooth width (unit: mm); f g is the helix error (unit: μm); A is the gear processing accuracy coefficient.
[0035] Based on the simulation model, the micro-modification of the gear is optimized and an effective improvement plan is proposed, thereby reducing the vibration and noise of the electric drive axle and improving the performance of the gear transmission system.
[0036] Step S2: Establish a prediction model:
[0037] In order to reduce the shaping time and improve the shaping accuracy, a gear micro-shaping parameter-vibration response prediction model was established based on EVO-Catboost. EVO (Energy valley optimizer) is an energy valley optimization algorithm, and Catboost (categorical boosting) is a machine learning algorithm. Its model based on the gradient boosting tree (GBM) combines the optimization of classification and regression tasks; a mathematical model called "objective function" is used, which can learn model parameters based on the input characteristics and output labels in the data set; the input of the prediction model is set as the shaping parameter, and the output is set as the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient, and the mapping relationship between them is explored, that is, the mapping relationship between the gear shaping parameters and the peak-to-peak value of the gear transmission error, the maximum contact stress and the load distribution coefficient is explored. Taking the peak-to-peak value, the maximum contact stress and the load distribution coefficient of the electric drive axle gear transmission error as the optimization target, the NRBO (Newton-Raphson-based optimizer) algorithm is introduced to optimize the micro-shaping parameters of the gear.
[0038] The optimization solution processing flow of the vibration response prediction model of step S2 is as follows: step a1, initializing the prediction model parameters, setting iterations, learning rate, depth and L2 regularization; step a2, evaluating the fitness value of the candidate particles; step a3, determining the enrichment and binding of the particles; step a4, determining the enrichment level and stability level of the particles; step a5, finding the particles with the optimal stability level; step a6, judging whether the optimal parameters are found, if so, outputting the optimal parameters and ending the program, otherwise returning to execute step a2.
[0039] Step S3, selection model optimization: To ensure that the shaping scheme can achieve good results under various working conditions, sensitivity analysis is performed to search for the shaping parameters with the strongest correlation, and a multi-environment shaping selection model is established to obtain the optimal shaping parameter combination.
[0040] The optimization processing flow of the selection model of step S3: step b1, establish the optimization target as the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient under different working conditions; step b2, design i kinds of modification schemes, i is the number of optimization targets; step b3, perform sensitivity analysis of the modification parameters under acceleration, deceleration and constant speed environments; step b4, set the optimization target weight: based on the sensitivity analysis, the peak-to-peak value of the transmission error under different working conditions has a greater impact on the vibration and noise of the gear transmission, while the load distribution coefficient and the maximum contact stress have a smaller impact. Therefore, the weight of the peak-to-peak value of the transmission error is set to a, and the weight of the load distribution coefficient is set to a / 2, where a≥0.5.
[0041] Step S4, modification effect analysis: importing the optimal modification parameter combination into the gear set parameterized simulation model to perform modification effect analysis.
[0042] Step S5, bench verification: verify the shaping effect of the optimal shaping parameter combination through the electric drive axle bench test. In this way, the shaping effect can be verified through the bench test. The electric drive axle bench test results show that the peak value of the vibration noise of the electric drive axle is reduced by 4% after the shaping, which effectively reduces the vibration noise of the electric drive axle gear transmission system.
[0043] As mentioned above, the optimization method collects data sets through modeling and simulation, uses EVO-Catboost to establish a gear vibration response prediction model and optimizes it through NRBO, and then uses a multi-environment modification selection model to perform sensitivity analysis of modification parameters under different environments, and finally obtains the optimal modification parameter combination.
[0044] This optimization method combines machine learning with optimization algorithms, proposes a gear modification method based on EVO-Catboost, and combines it with the optimization algorithm (NRBO) to achieve effective and rapid optimization. Compared with traditional theoretical models and dynamic models, the combination of machine learning and optimization algorithms can more efficiently predict and optimize the performance of gear transmission systems.
[0045] The electric drive axle vibration response prediction model proposed by this optimization method has practical engineering significance and can provide a reference for the optimization scenarios of various gear transmission systems. The application of machine learning in engineering practice solves practical problems and demonstrates its wide application potential.
[0046] In summary, this optimization method solves the vibration and noise problem of the electric drive axle. Through the comprehensive application of simulation analysis, gear micro-modification, prediction model and optimization algorithm, it achieves significant optimization effect and has high innovation and engineering application value.
[0047] Example 2
[0048] This embodiment 2 is based on the optimization method of the aforementioned embodiment 1 to form an optimization system, which is briefly described here. For other examples of technical feature combinations and other unexplained details, please refer to the aforementioned embodiment 1.
[0049] An electric drive axle gear vibration optimization system of this embodiment 2 includes the following contents:
[0050] Simulation analysis module: used to establish a parametric simulation model of the electric drive axle gear set, using the involute combined with the tooth direction modification method to obtain the corresponding electric drive axle gear modification parameters and the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient data set;
[0051] Establishing prediction model module: used to establish gear vibration response prediction model based on EVO-Catboost, mapping gear modification parameters to transmission error peak-to-peak value, maximum contact stress and load distribution coefficient; using NRBO algorithm to solve gear vibration response prediction model, optimizing with peak-to-peak value of transmission error, maximum contact stress and load distribution coefficient of electric drive axle gear as optimization target, and obtaining optimized modification parameters of gear;
[0052] Selection model optimization module: used to establish a multi-environment modification selection model, design a variety of modification schemes for each optimization modification parameter, conduct sensitivity analysis of modification parameters under different environments, and obtain the optimal modification parameter combination;
[0053] Modification effect analysis module: used to import the optimal modification parameter combination into the gear set parametric simulation model for modification effect analysis.
[0054] Among them, the gear modification parameters include tooth guide crown, tooth guide slope, involute crown and involute slope.
[0055] The optimization solution processing flow of the vibration response prediction model of the prediction model optimization module is as follows: step a1, initializing the prediction model parameters, setting iterations, learning rate, depth and L2 regularization; step a2, evaluating the fitness value of the candidate particles; step a3, determining the enrichment and binding of the particles; step a4, determining the enrichment level and stability level of the particles; step a5, finding particles with the optimal stability level; step a6, judging whether the optimal parameters are found, if so, outputting the optimal parameters and ending the program, otherwise returning to step a2.
[0056] The optimization processing flow of the selection model of the selection model optimization module: step b1, establish the optimization target as the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient under different working conditions; step b2, design i kinds of shaping schemes, i is the number of optimization targets; step b3, conduct sensitivity analysis of shaping parameters under acceleration, deceleration and constant speed environment; step b4, set the optimization target weight.
[0057] As mentioned above, data sets are collected through modeling and simulation, and the gear vibration response prediction model is established using EVO-Catboost and optimized through NRBO. Then, the sensitivity analysis of the modification parameters under different environments is carried out through the multi-environment modification selection model, and finally the optimal modification parameter combination is obtained.
[0058] Based on the above example, in an improved example, the optimization system also includes the following contents: a bench verification module: used to verify the shaping effect of the optimal shaping parameter combination through an electric drive axle bench test.
[0059] It should be pointed out that the examples of the above-mentioned embodiments can be preferably combined with one or more of them according to actual needs, and multiple examples use a set of illustrations of combined technical features, which will not be described one by one here.
[0060] The above descriptions are detailed descriptions and illustrations of the preferred embodiments of the present invention, but these descriptions are not intended to limit the scope of protection required by the present invention. All equivalent changes or modified modifications achieved under the technical teachings suggested by the present invention should fall within the scope of patent protection covered by the present invention.
Claims
1. A method for optimizing the vibration of electric drive axle gears, characterized in that: Includes the following: Step S1, simulation analysis: establishing a parametric simulation model of the electric drive axle gear set, using an involute combined with a tooth direction modification method to obtain a data set of modification parameters of the electric drive axle gear and a peak-to-peak value of the transmission error, a maximum contact stress, and a load distribution coefficient; Step S2, establishing a prediction model: establishing a gear vibration response prediction model based on EVO-Catboost, analyzing the mapping relationship between gear modification parameters and peak-to-peak value of transmission error, maximum contact stress and load distribution coefficient; using NRBO algorithm to solve the gear vibration response prediction model, optimizing with the peak-to-peak value of transmission error, maximum contact stress and load distribution coefficient of the electric drive axle gear as optimization targets, and obtaining the optimized modification parameters of the gear; Step S3, selection model optimization: establish a multi-environment shaping selection model, design multiple shaping schemes for each optimal shaping parameter, perform shaping parameter sensitivity analysis under different environments, and obtain the optimal shaping parameter combination; Step S4, modification effect analysis: importing the optimal modification parameter combination into the gear set parameterized simulation model to perform modification effect analysis.
2. The electric drive axle gear vibration optimization method according to claim 1, characterized in that: The gear modification parameters include tooth guide crowning, tooth guide slope, involute crowning and involute slope; Among them, the calculation formula for involute modification is as follows: In the formula, K A is the service factor; K mp is the branching coefficient; F t is the tangential force; b is the tooth width; α t is the end pressure angle; c γ is the meshing stiffness; The tooth modification calculation formula is as follows: C a =0.25b×10 -3 +0.5f g f g =A(0.1b+10) Where b is the tooth width; f g is the helix error; A is the gear processing accuracy coefficient.
3. The electric drive axle gear vibration optimization method according to claim 1, characterized in that: The optimization solution processing flow of the vibration response prediction model of step S2 is as follows: step a1, initializing the prediction model parameters, setting iterations, learning rate, depth and L2 regularization; step a2, evaluating the fitness value of the candidate particles; step a3, determining the enrichment and binding of the particles; step a4, determining the enrichment level and stability level of the particles; step a5, finding particles with the optimal stability level; step a6, judging whether the optimal parameters are found, if so, outputting the optimal parameters and terminating this procedure, otherwise returning to execute step a2.
4. The electric drive axle gear vibration optimization method according to claim 1, characterized in that: The optimization processing flow of the selection model of step S3 is as follows: step b1, establishing the optimization target as the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient under different working conditions; step b2, designing i kinds of shaping schemes, i is the number of optimization targets; step b3, performing sensitivity analysis of shaping parameters under acceleration, deceleration and constant speed environments; step b4, setting the optimization target weights, the weight of the peak-to-peak value of the transmission error is set to a, and the weight of the load distribution coefficient is set to a / 2, where a≥0.
5.
5. The electric drive axle gear vibration optimization method according to claim 1, characterized in that: Also included are the following: Step S5, bench verification: verifying the shaping effect of the optimal shaping parameter combination through an electric drive axle bench test.
6. An electric drive axle gear vibration optimization system, characterized in that: Includes the following: Simulation analysis module: used to establish a parametric simulation model of the electric drive axle gear, using the involute combined with the tooth direction modification method to obtain the corresponding electric drive axle gear modification parameters and the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient data set; Establishing prediction model module: It is used to establish a gear vibration response prediction model based on EVO-Catboost, analyze the mapping relationship between gear modification parameters and peak-to-peak value of transmission error, maximum contact stress and load distribution coefficient; use NRBO algorithm to solve the gear vibration response prediction model, and optimize the peak-to-peak value of transmission error, maximum contact stress and load distribution coefficient of electric drive axle gear as optimization targets to obtain the optimal modification parameters of the gear; Selection model optimization module: used to establish a multi-environment modification selection model, design a variety of modification schemes for each optimization modification parameter, conduct sensitivity analysis of modification parameters under different environments, and obtain the optimal modification parameter combination; Modification effect analysis module: used to import the optimal modification parameter combination into the gear set parametric simulation model for modification effect analysis.
7. The electric drive axle gear vibration optimization system according to claim 6, characterized in that: The gear modification parameters include tooth guide crowning, tooth guide slope, involute crowning and involute slope; Among them, the calculation formula for involute modification is as follows: In the formula, K A is the service factor; K mp is the branching coefficient; F t is the tangential force; b is the tooth width; α t is the end pressure angle; c γ is the meshing stiffness; The tooth modification calculation formula is as follows: C a =0.25b×10 -3 +0.5f g f g =A(0.1b+10) Where b is the tooth width; f g is the helix error; A is the gear processing accuracy coefficient.
8. The electric drive axle gear vibration optimization system according to claim 6, characterized in that: The optimization solution processing flow of the vibration response prediction model of the prediction model optimization module is as follows: step a1, initializing the prediction model parameters, setting iterations, learning rate, depth and L2 regularization; step a2, evaluating the fitness value of the candidate particles; step a3, determining the enrichment and binding of the particles; step a4, determining the enrichment level and stability level of the particles; step a5, finding particles with the optimal stability level; step a6, judging whether the optimal parameters are found, if so, outputting the optimal parameters and ending the program, otherwise returning to execute step a2.
9. The electric drive axle gear vibration optimization system according to claim 6, characterized in that: The optimization processing flow of the selection model of the selection model optimization module includes: step b1, establishing the optimization target as the peak-to-peak value of the transmission error, the maximum contact stress and the load distribution coefficient under different working conditions; step b2, designing i kinds of shaping schemes, i is the number of optimization targets; step b3, conducting sensitivity analysis of shaping parameters under acceleration, deceleration and constant speed environments; step b4, setting the optimization target weights, the weight of the peak-to-peak value of the transmission error is set to a, and the weight of the load distribution coefficient is set to a / 2, where a≥0.
5.
10. The electric drive axle gear vibration optimization system according to claim 6, characterized in that: It also includes the following contents: Bench verification module: used to verify the shaping effect of the optimal shaping parameter combination through electric drive axle bench experiments.
Citation Information
Patent Citations
Gear profile modification method based on vibration reliability and genetic algorithm
CN102314534A
