Collaborative optimization method for NVH performance structure of automobile tire and tire

Abstract

The invention relates to the technical field of tire structure optimization, in particular to an automobile tire NVH (Noise Vibration and Harshness) performance structure collaborative optimization method and a tire. Performing feature extraction on the data set to obtain a feature parameter matrix representing a transmission relation between structural vibration and noise; inputting the matrix into a coupling model constructed by finite element analysis, statistical energy analysis and neural network prediction, and outputting a prediction result set; processing the result set by adopting a multi-objective optimization algorithm, synchronously adjusting a plurality of design variables such as tread rubber layer thickness distribution, a support rubber stiffness curve, a cap ply winding track and carcass damping distribution, and generating an optimal parameter combination meeting multiple performance constraints; and performing virtual verification on the parameter combination through a digital twin platform, and inputting the verified parameters into an intelligent manufacturing system to complete tire forming. The technical problems of large noise and strong vibration caused by lack of collaborative optimization of tire structure design and NVH (Noise Vibration and Harshness) performance are solved.

Description

Technical Field

[0001] This invention relates to the field of tire structure optimization technology, and in particular to a method for synergistic optimization of NVH performance structure of automobile tires and a tire thereof. Background Technology

[0002] Existing tire NVH performance optimization technologies suffer from the following technical pain points: Firstly, existing methods focus on single optimization techniques, such as adding sound-absorbing materials or adjusting tread pitch, lacking a synergistic design of the overall tire structural parameters, leading to conflicts between material and structural performance. Secondly, in new energy vehicle applications, low engine noise makes tire noise more prominent, but sound-absorbing materials are prone to aging and detachment under high temperature and humidity conditions, resulting in reduced noise reduction. Thirdly, tread pattern optimization to disperse resonance frequencies may weaken wet grip, increasing driving risks. This isolated optimization approach cannot balance NVH performance with key indicators such as rolling resistance and durability, thus causing technical pain points such as noise amplification, increased vibration transmission, and decreased comfort during vehicle operation. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a method for synergistic optimization of NVH performance structure of automobile tires and a tire in general, solving the technical problem of high noise, strong vibration and poor comfort during vehicle operation caused by the lack of synergistic optimization between tire structure design and NVH performance.

[0004] To solve the above-mentioned technical problems, the specific contents of the present invention are as follows: In a first aspect, the present invention provides a method for structural collaborative optimization of NVH performance of automotive tires, comprising: Step 1: Collect multi-physics field data of the tire under driving conditions through the sensing system to generate a multi-physics field dataset. The multi-physics field data includes tread geometric features, structural strain distribution and noise spatial features. Step 2: Extract features from the multiphysics dataset to obtain a feature parameter matrix characterizing the relationship between tire structure vibration and noise transmission; Step 3: Input the feature parameter matrix into the tire structure and NVH performance coupling model. The coupling model is constructed through a hybrid modeling method of finite element analysis, statistical energy analysis and neural network prediction, and outputs a set of prediction results including the impact of structural parameter adjustment on NVH performance. Step 4: The predicted result set is processed using a multi-objective optimization algorithm. Multiple design variables, such as the tread rubber layer thickness distribution, the support rubber stiffness curve, the crown belt winding trajectory, and the tire body damping distribution, are adjusted simultaneously to generate the optimal parameter combination that meets the constraints of NVH performance, rolling resistance performance, and durability. Step 5: Virtually verify the optimal parameter combination through a digital twin platform, and input the verified parameters into the intelligent manufacturing system to complete the tire structure forming.

[0005] Furthermore, in the structural collaborative optimization method for NVH performance of automotive tires described in this invention, step 1 includes: The depth, width, and distribution angle of the tread grooves are obtained using a three-dimensional laser scanning system as geometric feature data. Dynamic strain between the internal structural layers of a tire is collected using a distributed optical fiber sensor network as strain distribution data. The spatial distribution of noise sound pressure is collected using an acoustic camera array arranged in a ring, which serves as noise characteristic data. Simultaneously collect tire inflation pressure, load conditions, and road surface type parameters to generate a set of working condition parameters; The geometric feature data, strain distribution data, noise feature data, and operating condition parameter set are synchronized and aligned in time to generate the multiphysics dataset.

[0006] Furthermore, in the automotive tire NVH performance structure collaborative optimization method of the present invention, step 2 includes: Structural vibration and noise data are extracted from the multiphysics dataset, and the transfer function relationship between the two is analyzed using the transfer path analysis method to generate the analytical results of the transfer function. Based on the multiphysics dataset and transfer function analysis results, machine learning algorithms are used to identify the correlation features between the shape of the tread contact mark and NVH performance, and generate correlation feature identifiers. By combining the multiphysics dataset, transfer function analysis results, and associated feature identifiers, the energy flow distribution characteristics of the coupled vibration between the tire carcass and the tread are extracted, and the structural modal participation factor and acoustic-vibration sensitivity index are calculated, which together constitute the feature parameter matrix.

[0007] Furthermore, in the automotive tire NVH performance structure collaborative optimization method of the present invention, step 3 includes: Call the finite element analysis tool, input the material properties and structural parameters of the crown reinforcement and tread compound, simulate their interaction, and output the finite element analysis results; Call the statistical energy analysis tool, input the material properties of the high-damping rubber layer, calculate the frequency-dependent vibration energy dissipation characteristics, and output the statistical energy analysis results; Prepare historical test datasets, train neural networks to establish a nonlinear mapping relationship between structural parameters and NVH performance, and generate neural network models; By integrating the finite element analysis results, statistical energy analysis results, and neural network model, a coupled model of tire structure and NVH performance is constructed.

[0008] Furthermore, in the automotive tire NVH performance structure collaborative optimization method of the present invention, step 4 includes: Using the predicted result set as sample data, a response surface model between design variables and NVH performance indicators is established; Under constraints of NVH performance, rolling resistance performance and durability, a Pareto front search is performed on the response surface model to obtain the Pareto solution set; Sensitivity analysis is performed on the Pareto solution set to verify the stability of the solution, and the optimal parameter combination is selected from it.

[0009] Furthermore, in the automotive tire NVH performance structure collaborative optimization method of the present invention, step 5 includes: Based on the optimal parameter combination, a virtual tire prototype is constructed in the digital twin platform, and the prototype is connected to the hardware-in-the-loop simulation system. Drive the hardware-in-the-loop simulation system to simulate the NVH performance of a virtual tire prototype under extreme working conditions and generate virtual test results. The virtual test results are compared with the real vehicle test data, and the tire structure and NVH performance coupling model is modified based on the comparison results to form a closed-loop optimization.

[0010] Furthermore, in the automotive tire NVH performance structure collaborative optimization method of the present invention, step 2 includes: The strain and stress response of the support adhesive were analyzed from a multiphysics dataset to identify the influence of the nonlinear stiffness characteristics of the support adhesive on vibration attenuation. Based on the identified influence patterns, stiffness gradient distribution mapping parameters are quantized and generated, and these parameters are incorporated into the feature parameter matrix.

[0011] Furthermore, in the automotive tire NVH performance structure collaborative optimization method of the present invention, step 4 includes: The crown band winding angle and winding tension parameters are used as coupled design variables for synchronous optimization, so that the optimized crown band winding trajectory and the tire body damping distribution work together to suppress vibration.

[0012] Furthermore, in the structural collaborative optimization method for NVH performance of automotive tires described in this invention, step 4 further includes: Based on the tread ground pressure distribution obtained from finite element analysis, the thickness gradient of the tread rubber layer is adjusted. During the optimization process, the changes in thickness gradient are correlated with the changes in the stiffness curve of the support adhesive, and synchronous optimization is performed in a coupled manner.

[0013] Secondly, the present invention provides a car tire, wherein the structural parameters of the tire are generated by the car tire NVH performance structure collaborative optimization method, and the structural parameters include the optimized tread rubber layer thickness distribution, support rubber stiffness curve, crown belt layer winding trajectory and tire body damping distribution.

[0014] Beneficial effects of this invention; This invention constructs a feature parameter matrix that comprehensively reflects the dynamic characteristics of tires through multi-physics data acquisition and feature extraction. It establishes a coupled tire structure and NVH performance model using a hybrid modeling approach combining finite element analysis, statistical energy analysis, and neural network prediction. This model accurately predicts the nonlinear impact of structural parameter adjustments on NVH performance. Through a multi-objective optimization algorithm, it simultaneously adjusts multiple design variables, including tread rubber layer thickness distribution, support rubber stiffness curve, crown belt winding trajectory, and tire body damping distribution. Under constraints of NVH performance, rolling resistance, and durability, it generates an optimal parameter combination, enabling synergistic effects among structural parameters. For example, the nonlinear characteristics of support rubber stiffness and crown belt winding trajectory jointly suppress vibration, and the tread thickness gradient and ground pressure distribution are mutually adapted, thus solving the performance imbalance problem caused by parameter conflicts in existing single optimization methods. Finally, through virtual verification and closed-loop optimization on a digital twin platform, the verified parameters are input into an intelligent manufacturing system to achieve precise tire structure forming. This significantly improves tire noise suppression, vibration attenuation, and comfort during driving, while also considering rolling resistance and durability requirements. Attached Figure Description

[0015] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on the drawings without creative effort.

[0016] Figure 1 This is a flowchart of the structural collaborative optimization method for NVH performance of automobile tires according to the present invention. Detailed Implementation

[0017] To make the technical solution of the present invention clearer, the present invention will be clearly and completely described below with reference to specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. The present invention provided by various embodiments will be described in detail below with reference to the accompanying drawings. To better understand the purpose of the present invention, the present invention will be described in further detail below.

[0018] Firstly, please refer to Figure 1 The present invention provides a method for structural synergistic optimization of NVH performance of automobile tires, comprising: Step 1: Collect multi-physics field data of the tire under driving conditions through the sensing system to generate a multi-physics field dataset. The multi-physics field data includes tread geometric features, structural strain distribution and noise spatial features. Step 2: Extract features from the multiphysics dataset to obtain a feature parameter matrix characterizing the relationship between tire structure vibration and noise transmission; Step 3: Input the feature parameter matrix into the tire structure and NVH performance coupling model. The coupling model is constructed through a hybrid modeling method of finite element analysis, statistical energy analysis and neural network prediction, and outputs a set of prediction results including the impact of structural parameter adjustment on NVH performance. Step 4: The predicted result set is processed using a multi-objective optimization algorithm. Multiple design variables, such as the tread rubber layer thickness distribution, the support rubber stiffness curve, the crown belt winding trajectory, and the tire body damping distribution, are adjusted simultaneously to generate the optimal parameter combination that meets the constraints of NVH performance, rolling resistance performance, and durability. Step 5: Virtually verify the optimal parameter combination through a digital twin platform, and input the verified parameters into the intelligent manufacturing system to complete the tire structure forming.

[0019] In step 1, a multiphysics dataset is generated by collecting multiphysics data of the tire during driving using a sensing system. Specifically, a 3D laser scanning system scans the rotating tire tread to obtain microscopic geometric features such as the depth, width, and distribution angle of the tread grooves, forming geometric feature data. A distributed fiber optic sensor network is embedded between the structural layers inside the tire carcass to monitor the dynamic strain distribution in areas such as the tire sidewall and tire crown in real time during vehicle operation, generating strain distribution data. An acoustic camera array is arranged in a ring around the tire to collect the noise sound pressure signal and its spatial distribution characteristics generated when the tire contacts the road surface, obtaining noise feature data. During the test, tire inflation pressure, vertical load, and road surface roughness are recorded simultaneously to generate a working condition parameter set. Finally, the geometric feature data, strain distribution data, noise feature data, and working condition parameter set are synchronized in time to eliminate data deviations caused by differences in the response time of the acquisition devices, generating a complete multiphysics dataset.

[0020] Step 2 involves feature extraction from the multiphysics dataset to obtain a feature parameter matrix characterizing the relationship between tire structural vibration and noise transmission. First, structural vibration and noise signals are separated from the multiphysics dataset. A transmission path analysis method is used to analyze the vibration-noise transfer function relationship from the tire excitation point to the vehicle's response point, generating the analytical results of the transfer function. Next, based on the multiphysics dataset and the analytical results of the transfer function, a convolutional neural network algorithm is used to identify the implicit correlation between the shape change of the tire tread contact patch and the sound pressure level of specific frequency noise inside the vehicle, generating associated feature identifiers. Finally, combining the multiphysics dataset, the analytical results of the transfer function, and the associated feature identifiers, the vibration energy flow distribution characteristics of the tire carcass and tread coupling region are extracted, and the structural modal participation factor and acoustic-vibration sensitivity index are calculated. All these feature parameters together constitute the feature parameter matrix.

[0021] Step 3 involves inputting the feature parameter matrix into the tire structure and NVH performance coupling model. Model construction requires using a finite element analysis (FEM) tool, inputting material properties and structural parameters such as the cord angle and density of the crown reinforcement and the Shore hardness of the tread compound, simulating the interaction mechanism between the belt layer winding tension and the rubber flow, and outputting the FEM results. Simultaneously, a statistical energy analysis (SAE) tool is used, inputting the loss factor parameters of the high-damping rubber layer, calculating its vibration energy dissipation characteristics in the 50-500Hz frequency range, and outputting the SAE results. Historical test datasets including different tire specifications are prepared, and a deep neural network is trained to establish a nonlinear mapping relationship between tread thickness changes and second-order in-vehicle noise, generating a neural network model. Finally, the FEM results, SAE results, and neural network model are integrated to construct a tire structure and NVH performance coupling model capable of accurately predicting the impact of structural parameter adjustments on NVH performance.

[0022] Step 4 employs a multi-objective optimization algorithm to process the prediction result set. Using the prediction result set as sample data, a response surface model is established to connect design variables such as tread rubber layer thickness distribution and support rubber stiffness curve with performance indicators such as overall in-vehicle noise sound pressure level and rolling resistance coefficient. Under multi-objective constraints of NVH performance, rolling resistance performance, and durability, a multi-island genetic algorithm is used to search the response surface model to obtain the Pareto solution set. Sobol sensitivity analysis is performed on the Pareto solution set to verify the stability of the impact of each design variable on performance, and the optimal parameter combination with the best overall performance is selected from the solution set.

[0023] Step 5 involves virtual verification of the optimal parameter combination using a digital twin platform. Based on the optimal parameter combination, a virtual tire prototype with detailed material properties is constructed in the digital twin platform and connected to a hardware-in-the-loop simulation system. The hardware-in-the-loop simulation system is then driven to simulate the NVH performance of the virtual tire prototype under extreme low-adhesion road conditions, generating virtual test results including vibration acceleration spectra. These virtual test results are compared with test data collected from a real vehicle at the test track. When the error exceeds a threshold, the contact algorithm parameters in the tire structure and NVH performance coupling model are corrected, forming a closed-loop optimization. Finally, the verified parameter set is input into a six-axis winding machine in the intelligent manufacturing system to complete the precise molding of the tire structure.

[0024] A 3D laser scanning system performs non-contact scanning of the tire tread, capturing the microscopic geometric contours of the tread grooves to form geometric feature data. A distributed fiber optic sensor network is embedded between the tire's internal layers to monitor the dynamic strain changes in various areas of the tire body in real time during driving, outputting strain distribution data. An acoustic camera array is arranged around the tire to collect the sound pressure signal and its spatial distribution of tire-road contact noise, obtaining noise feature data. The testing system simultaneously records operating parameters such as tire inflation pressure, axial load, and road surface type, generating a set of operating parameters. All collected data are aligned using a timestamp synchronization algorithm to eliminate timing errors and are finally fused into a multiphysics dataset.

[0025] Structural vibration acceleration signals and noise spectrum data were separated from a multiphysics dataset. The vibration transfer function from the tire excitation point to the vehicle seat rail was calculated using transfer path analysis, generating analytical results. Clustering algorithms from machine learning were used to analyze the correlation between tire tread contact patch images and NVH performance parameters, generating associated feature identifiers. Vibration energy flow analysis was combined to extract the energy distribution characteristics of the tire carcass-tread interface, and modal participation factors and acoustic-vibration sensitivity indices were calculated. These parameters together form a feature parameter matrix.

[0026] The finite element analysis (FEM) tool takes into account the cord spacing of the tire crown reinforcement and the dynamic modulus parameters of the tread compound to simulate the stress coupling effect between the belt layer and the rubber compound, and outputs the FEM results. The statistical energy analysis (SAE) tool takes into account the loss factor data of the high-damping rubber to calculate the attenuation characteristics of vibration energy over a wide frequency band, and outputs the SAE results. The neural network model is trained using a historical tire test database to learn the nonlinear relationship between structural parameters and in-vehicle noise levels, generating a neural network model. Finally, a model fusion algorithm integrates the three types of analysis results to establish a coupled model.

[0027] Using the predicted result set as training samples, a response surface model of tread rubber layer thickness and noise index is constructed. A multi-objective genetic algorithm is employed to search for the Pareto optimal solution set under constraints of NVH performance, rolling resistance, and durability. Analysis of variance is used to evaluate the sensitivity of each design variable in the solution set, and the parameter combination with optimal stability is selected.

[0028] A virtual tire prototype is built in a twin platform based on the optimal parameter combination. The prototype includes material constitutive relations and boundary conditions. The hardware-in-the-loop simulation system drives the prototype to simulate vibration response under extreme conditions and generate virtual test results. The virtual results are compared with real vehicle test data using a differential algorithm. When the deviation exceeds a threshold, the contact stiffness parameters of the coupled model are corrected in reverse to achieve closed-loop optimization.

[0029] Strain-stress curves of the support adhesive region were extracted from a multiphysics dataset to identify the attenuation law of vibration transmission due to its nonlinear stiffness characteristics. The stiffness gradient distribution was quantized into mapping parameters through parametric modeling and incorporated into the characteristic parameter matrix.

[0030] By synchronously iterating with winding angle and tension as correlated variables, the optimized winding trajectory and the tire damping distribution form a phase cancellation effect, thereby enhancing vibration suppression.

[0031] Based on the ground pressure contour map obtained from finite element analysis, the thickness gradient of different regions of the tire tread is adjusted. The optimization algorithm correlates the thickness variation with the stiffness curve of the support rubber, achieving coupled adjustment of structural parameters.

[0032] Secondly, the present invention provides a car tire, wherein the structural parameters of the tire are generated by the car tire NVH performance structure collaborative optimization method, and the structural parameters include the optimized tread rubber layer thickness distribution, support rubber stiffness curve, crown belt layer winding trajectory and tire body damping distribution.

[0033] The tire tread rubber layer thickness distribution, support rubber stiffness curve, crown belt layer winding trajectory, and tire body damping distribution are all generated by the automotive tire NVH performance structure collaborative optimization method. The synergistic effect of each parameter enables the tire to have optimized NVH performance, rolling resistance performance, and durability.

[0034] The technical solution of the present invention is described in detail through the following embodiments. In the application scenario of new energy vehicles, in order to address the problems of easy aging of sound-absorbing materials and the impact of tread pattern adjustment on grip in existing tire optimization methods, this solution adopts a multi-physics data fusion and collaborative optimization strategy.

[0035] First, a 3D laser scanning system is used to perform high-speed scanning of the tire tread during operation, acquiring data on the depth distribution of the tread grooves. A distributed fiber optic sensor network is embedded between the tire carcass layers to collect real-time data on the dynamic strain distribution of the tire as it alternates between asphalt and concrete surfaces. An acoustic camera array, arranged in a ring, collects the noise spectrum characteristics generated by the tire when traveling over rough surfaces. The testing system simultaneously records the tire's operating parameters under different loads and inflation pressures, and all data is integrated into a multiphysics dataset using a timestamp synchronization algorithm.

[0036] When extracting features from a multiphysics dataset, the structural vibration signal is first separated from the strain data. Then, combined with the noise spectrum, a transmission path analysis method is used to calculate the vibration transfer function from the tire contact point to the vehicle floor. Machine learning algorithms are used to analyze the correlation between the shape change of the tire contact mark and the mid-frequency noise inside the vehicle. Simultaneously, the strain-stress curve of the support rubber region is extracted to identify its nonlinear stiffness characteristics. Finally, by integrating the vibration energy flow distribution characteristics and the acoustic sensitivity index, a feature parameter matrix including stiffness gradient distribution mapping parameters is formed.

[0037] When constructing the coupled model of tire structure and NVH performance, the finite element analysis module inputs the cord angle parameters of the crown reinforcement to simulate the stress distribution of the belt layer and tread rubber. The statistical energy analysis module calculates the energy dissipation efficiency of high-damping rubber in a specific frequency band, while the neural network module establishes a mapping relationship between thickness parameters and noise through training with historical data. The results of these three types of analysis are integrated into a coupled model through a model fusion algorithm, which can predict the combined impact of structural adjustments on NVH performance.

[0038] In the multi-objective optimization process, the prediction result set output by the coupled model is used as a sample to establish a response surface model including variables such as tread thickness distribution and support rubber stiffness curve. A genetic algorithm is employed for multi-objective optimization under constraints of NVH performance, rolling resistance, and durability, simultaneously adjusting the crown belt winding angle and tension parameters to achieve a synergistic vibration suppression effect with the tire carcass damping distribution. Simultaneously, the thickness gradient is optimized based on the ground pressure contour map obtained from finite element analysis, thus establishing a coupling relationship between the thickness distribution and the support rubber stiffness curve.

[0039] In the digital twin verification phase, a virtual tire prototype is built based on optimized parameters, and the vibration response under wet road conditions is simulated using a hardware-in-the-loop simulation system. The simulation data is compared with real-vehicle test results, and when the error exceeds a set threshold, the contact algorithm parameters in the coupled model are automatically corrected, forming a closed-loop optimization. Finally, the verified parameter set is input into a six-axis winding machine to manufacture a finished tire with collaboratively optimized structural parameters.

[0040] The structural parameters of the finished tire, such as the thickness distribution of the tread rubber layer and the stiffness curve of the support rubber, are all generated through the above-mentioned collaborative optimization method. The parameters are intrinsically related, which enables the tire to significantly improve noise and vibration performance while maintaining wet grip. It is particularly suitable for new energy vehicle application scenarios with strict NVH performance requirements.

[0041] This invention addresses the problems of high noise, intense vibration, and poor comfort caused by the lack of coordinated optimization between tire structure design and NVH performance by constructing a complete technical chain based on multi-physics data fusion, structure-NVH performance coupled modeling, multi-objective collaborative optimization, and digital twin verification. Existing optimization methods often adjust sound-absorbing materials or tread pitch individually, making it difficult to balance NVH performance with key indicators such as rolling resistance and durability. This invention employs a collaborative optimization strategy. First, it collects tire tread geometry, structural strain distribution, and noise spatial characteristics during tire operation using 3D laser scanning, distributed fiber optic sensors, and an acoustic camera array, forming a multi-physics dataset that provides a comprehensive data foundation for optimization. Next, it uses transfer path analysis and machine learning algorithms to extract feature parameter matrices from the dataset, accurately characterizing the relationship between tire structure vibration and noise transmission, overcoming the limitations of existing methods that rely on empirical formulas. Then, it establishes a coupled model integrating finite element analysis, statistical energy analysis, and neural networks to simulate the nonlinear impact of structural parameter adjustments on NVH performance, avoiding prediction biases from single models. In the multi-objective optimization stage, multiple design variables, including tread rubber layer thickness distribution, support rubber stiffness curve, crown belt winding trajectory, and tire carcass damping distribution, are adjusted simultaneously. Optimal parameter combinations satisfying multiple performance constraints are generated through response surface methodology and Pareto front search, enabling synergistic effects among structural parameters. For example, the nonlinear characteristics of support rubber stiffness and crown belt winding trajectory jointly suppress vibration, and the tread thickness gradient and ground pressure distribution are mutually adapted, thus resolving parameter conflicts in existing methods. Finally, the optimization results are virtually verified using a digital twin platform, and the model is corrected by comparing hardware-in-the-loop simulation with real-vehicle test data, forming a closed-loop optimization that ensures an overall improvement in NVH performance, rolling resistance, and durability during tire structure forming. This method, through data-driven and multidisciplinary coupled optimization, achieves intrinsic synergy among tire structural parameters, fundamentally solving the performance imbalance problem caused by isolated optimization.

[0042] This invention relates to a method for co-optimizing the NVH performance of automobile tires and a tire. Addressing the technical problem that tire noise becomes more prominent due to reduced engine noise in new energy vehicle applications, and that existing optimization methods focus on single approaches such as adding sound-absorbing materials or adjusting tread pitch, lacking overall structural parameter co-design leading to material-performance conflicts, this method achieves co-design of tire structural parameters through multi-physics data fusion, structure-NVH performance coupled modeling, multi-objective co-optimization, and digital twin verification.

[0043] During the data acquisition phase, a 3D laser scanning system is used to perform non-contact scanning of the tire tread while it is in motion, acquiring the depth, width, and distribution angle of the tread grooves as geometric feature data. A distributed fiber optic sensor network is embedded between the internal structural layers of the tire to collect the dynamic strain distribution of the sidewall and crown areas in real time as strain distribution data. An acoustic camera array is arranged in a ring to collect the spatial distribution of noise sound pressure generated when the tire contacts the road surface as noise feature data. Tire inflation pressure, load conditions, and road type parameters are recorded simultaneously to generate a set of operating parameters. All acquired data are processed using a time synchronization alignment algorithm to eliminate timing deviations and generate a multiphysics dataset.

[0044] In the feature extraction stage, structural vibration signals and noise spectrum data are separated from the multiphysics dataset. The transfer path analysis method is used to analyze the vibration-noise transfer function relationship from the tire excitation point to the vehicle's response point, generating the analytical results of the transfer function. Based on the multiphysics dataset and the analytical results of the transfer function, a clustering algorithm in machine learning is used to identify the correlation features between the shape change of the tire tread contact patch and the sound pressure level of specific frequency noise inside the vehicle, generating associated feature identifiers. Combining the multiphysics dataset, the analytical results of the transfer function, and the associated feature identifiers, the vibration energy flow distribution characteristics of the tire carcass and tread coupling region are extracted, and the structural modal participation factor and acoustic-vibration sensitivity index are calculated, collectively forming the feature parameter matrix. Specifically, the strain and stress response of the support rubber are analyzed from the multiphysics dataset to identify the influence of the nonlinear characteristics of the support rubber stiffness on vibration attenuation, quantifying and generating stiffness gradient distribution mapping parameters and incorporating them into the feature parameter matrix.

[0045] In the coupled modeling phase, finite element analysis (FEM) tools are used, inputting material properties and structural parameters such as the cord angle and density of the tire crown reinforcement and the Shore hardness of the tread rubber compound. The interaction mechanism between the winding tension of the belt layer and the flow of the rubber compound is simulated, and the FEM results are output. Statistical energy analysis (SAE) tools are then used, inputting the loss factor parameters of the high-damping rubber layer, calculating the frequency-dependent vibration energy dissipation characteristics, and outputting SAE results. Historical test datasets are prepared, and a deep neural network is trained to establish a nonlinear mapping relationship between structural parameters such as tread thickness and in-vehicle noise level, generating a neural network model. Integrating the FEM results, SAE results, and the neural network model, a coupled model of tire structure and NVH performance is constructed using a model fusion algorithm. This model can predict the combined impact of structural parameter adjustments on NVH performance.

[0046] In the multi-objective optimization stage, the predicted result set output by the coupled model is used as sample data to establish a response surface model between design variables such as tread rubber layer thickness distribution, support rubber stiffness curve, crown belt winding trajectory, and carcass damping distribution, and NVH performance, rolling resistance, and durability indicators. Under the multi-objective constraints of NVH performance, rolling resistance, and durability, a multi-island genetic algorithm is used to perform Pareto front search on the response surface model to obtain the Pareto solution set. Sobol sensitivity analysis is performed on the Pareto solution set to verify the stability of the influence of each design variable on performance, and the optimal parameter combination is selected. During the optimization process, multiple design variables are adjusted simultaneously, and the crown belt winding angle and winding tension parameters are used as coupled design variables for simultaneous optimization, so that the optimized crown belt winding trajectory and carcass damping distribution work together to suppress vibration. Based on the tread ground pressure distribution obtained from finite element analysis, the tread rubber layer thickness gradient is adjusted, and the thickness gradient change is correlated with the support rubber stiffness curve change during the optimization process for simultaneous optimization in a coupled manner.

[0047] In the virtual verification phase, based on the optimal parameter combination, a virtual tire prototype, including material constitutive relations and boundary conditions, is constructed in a digital twin platform and connected to a hardware-in-the-loop simulation system. The hardware-in-the-loop simulation system simulates the NVH performance of the virtual tire prototype under conditions such as extreme low-adhesion road surfaces, generating virtual test results. These virtual test results are compared with real-vehicle test data using a differential algorithm. When the error exceeds a set threshold, the contact stiffness parameters in the tire structure and NVH performance coupling model are corrected, forming a closed-loop optimization. Finally, the verified parameter set is input into a six-axis winding machine in the intelligent manufacturing system to complete the precise molding of the tire structure.

[0048] The automotive tires generated by the above method have structural parameters including optimized tread rubber layer thickness distribution, support rubber stiffness curve, crown belt layer winding trajectory, and tire body damping distribution. These parameters are intrinsically correlated through synergistic optimization, which significantly improves the noise and vibration performance of the tires in new energy vehicle applications, while taking into account rolling resistance performance and durability requirements.

[0049] Embodiment 1 of this invention: Specific implementation of a tire NVH performance structure-based collaborative optimization method under typical urban road conditions for new energy vehicles. A three-dimensional laser scanning system collects depth data of the tire tread grooves with an accuracy of 0.1 mm, ranging from 1.5 mm to 2.5 mm in depth and 3 mm to 5 mm in width, with a distribution angle of 30 degrees to 60 degrees, forming geometric feature data. A distributed fiber optic sensor network monitors the dynamic strain between tire layers in real time at a sampling frequency of 1000 Hz, achieving a strain resolution of 5 micro-strains, generating strain distribution data. An acoustic camera array is arranged in a ring around the tire at a spacing of 0.5 meters, collecting the spatial distribution of noise sound pressure, covering a frequency range of 50 Hz to 500 Hz, with a sound pressure level measurement accuracy of ±1 dB, serving as noise feature data. Simultaneously, tire inflation pressure (250 kPa), load (500 N), and asphalt pavement type parameters are recorded, generating a set of working condition parameters. All data are aligned using a timestamp synchronization algorithm with a delay of less than 1 millisecond, forming a multiphysics dataset. In the feature extraction stage, vibration acceleration signals and noise spectra are separated from the dataset. A transfer path analysis method is used to calculate the transfer function from the tire tread contact point to the vehicle seat rail, with a frequency resolution of 2 Hz, generating analytical results for the transfer function. Machine learning clustering algorithms are used to analyze the tire tread contact imprint image, identifying correlation features with the 125 Hz noise sound pressure level inside the vehicle, generating associated feature identifiers. Combined with vibration energy flow analysis, the energy distribution in the tire carcass and tread coupling region is extracted, and the structural modal participation factor (CMDP) is calculated to be greater than 0.8 and the acoustic-vibration sensitivity index (SHR) to be less than 0.5, forming a feature parameter matrix. In the coupling modeling stage, the finite element analysis tool is used with inputs of a 90-degree cord angle for the tire crown reinforcement, a density of 1500 cords / meter, and a Shore hardness of 65 for the tread rubber, simulating the interaction between the belt layer and the rubber compound, and outputting a stress distribution cloud map. The statistical energy analysis tool is used with an input high-damping rubber loss factor of 0.3 to calculate energy dissipation in the 50 Hz to 500 Hz frequency band, outputting attenuation characteristic curves. The neural network model was trained using a historical dataset containing 1000 tire specifications, establishing a mapping between tread thickness and noise, with 50 hidden layer nodes, and outputting a set of prediction results. Multi-objective optimization used the prediction set as samples to construct a response surface model, employing a multi-island genetic algorithm with a population size of 100. Under NVH performance constraints of in-vehicle noise below 65 dB, rolling resistance coefficient less than 6.5 N / kN, and 100,000 durability test cycles, a Pareto front search was performed. After obtaining the solution set, the optimal parameter combination was selected through Sobol sensitivity analysis, such as a tread rubber layer thickness gradient of 0.5 mm to 1.5 mm, a nonlinear exponent of 1.2 for the support rubber stiffness curve, a crown belt winding angle of 15 degrees to 20 degrees, and a tire body damping distribution loss factor of 0.4.In the virtual verification phase, a virtual prototype of the tire is constructed on the digital twin platform. The Mooney-Rivlin model is used for the material constitutive relation. The hardware-in-the-loop simulation system is connected to simulate the working condition of a friction coefficient of 0.3 on a wet and slippery road surface, and a vibration acceleration spectrum is generated. The error rate threshold between the spectrum and the actual vehicle test data is 5%. The contact stiffness parameters of the coupled model are corrected, and the optimized parameters are input into a six-axis winding device to complete the tire forming.

[0050] Embodiment 2 of this invention: For new energy vehicles operating under extreme conditions of high-speed driving and high temperature and humidity, this method emphasizes durability and NVH balance. During data acquisition, a 3D laser scanning system scans the tread grooves at a tire speed of 100 km / h, with a depth of 1.8 mm to 2.8 mm, a width of 4 mm to 6 mm, and a distribution angle of 40 degrees to 70 degrees. A distributed fiber optic sensor network monitors tire strain within a temperature range of -10°C to 50°C and a humidity range of 30% to 90%, compensating for temperature drift with strain data. An acoustic camera array collects spatial noise characteristics, focusing on the 200 Hz to 400 Hz frequency band, with a sound pressure level dynamic range of 30 dB to 90 dB. Operating parameters include an inflation pressure of 280 kPa, a load of 600 N, and a concrete road surface, generating a multiphysics dataset. During feature extraction, the phase difference of the vibration transfer function analyzed by the transmission path analysis was less than 10 degrees. Machine learning identified the correlation between grounding marks and NVH. The strain threshold of the nonlinear characteristics of the support rubber stiffness was 0.1, and the slope of the quantized stiffness gradient distribution mapping parameters was 0.8. Coupled model integrated finite element analysis simulated the modulus change of the crown reinforcement and rubber at high temperature. Statistical energy analysis calculated the loss factor of high-damping rubber at 40 degrees Celsius to be 0.25, and the neural network predicted the nonlinear mapping error to be less than 3%. In multi-objective optimization, the response surface model included interaction terms. Pareto search, under the constraints of NVH performance, rolling resistance coefficient of 6.0 N / kN, and durability of 150,000 cycles, simultaneously optimized the tread thickness distribution gradient from 0.8 mm to 2.0 mm, the stress at the inflection point of the support rubber stiffness curve of 1.5 MPa, and the crown layer winding tension from 200 N to 300 N, in synergy with the tire body damping distribution to suppress vibration energy transmission. Virtual verification uses a digital twin platform to simulate extreme high-temperature conditions of 60 degrees Celsius. Hardware-in-the-loop simulation generates a noise spectrum, which is compared with real vehicle data to correct the model. Optimized parameters are input into the intelligent manufacturing system, resulting in tires that reduce noise by 3 decibels and reduce vibration at high speeds.

Claims

1. A method for structural collaborative optimization of NVH performance of a vehicle tire, characterized in that, The method comprises the following steps: Step 1: Collecting multi-physical field data of the tire in the driving state through a sensing system to generate a multi-physical field data set, wherein the multi-physical field data comprises tread geometric characteristics, structural strain distribution and noise spatial characteristics; Step 2: Extracting features from the multi-physical field data set to obtain a feature parameter matrix representing the relationship between tire structure vibration and noise transmission; Step 3: Inputting the feature parameter matrix into a tire structure and NVH performance coupling model, wherein the coupling model is constructed by a hybrid modeling method of finite element analysis, statistical energy analysis and neural network prediction, and outputting a prediction result set comprising the influence of structural parameter adjustment on NVH performance; Step 4: Processing the prediction result set by using a multi-objective optimization algorithm to simultaneously adjust the thickness distribution of the tread rubber layer, the support rubber stiffness curve, the crown belt layer winding trajectory and the tire body damping distribution of multiple design variables to generate an optimal parameter combination meeting the constraints of NVH performance, rolling resistance performance and durability; Step 5: Virtually verifying the optimal parameter combination through a digital twin platform, and inputting the verified parameters into an intelligent manufacturing system to complete the tire structure molding.

2. The method for collaborative optimization of automotive tire NVH performance structure according to claim 1, characterized in that, The step 1 comprises: Obtaining the depth, width and distribution angle of the tread pattern groove through a three-dimensional laser scanning system as geometric feature data; Collecting the dynamic strain between the internal structural layers of the tire through a distributed optical fiber sensor network as strain distribution data; Collecting the spatial distribution of noise pressure through an acoustic camera array arranged in a ring shape as noise characteristic data; Synchronously collecting tire inflation pressure, load conditions and road type working condition parameters to generate a working condition parameter set; Synchronously aligning the geometric feature data, strain distribution data, noise characteristic data and working condition parameter set in time to generate the multi-physical field data set.

3. The method for collaborative optimization of automotive tire NVH performance structure according to claim 2, characterized in that, The step 2 comprises: Extracting structural vibration and noise data from the multi-physical field data set, analyzing the transfer function relationship between the two by using a transfer path analysis method to generate a transfer function analysis result; Identifying the correlation features between the tread footprint shape and the NVH performance based on the multi-physical field data set and the transfer function analysis result by using a machine learning algorithm to generate a correlation feature identifier; Extracting the coupling vibration energy flow distribution characteristics of the tire body and the tread in combination with the multi-physical field data set, the transfer function analysis result and the correlation feature identifier, and calculating the structural modal participation factor and the acoustic vibration sensitivity index to jointly constitute the feature parameter matrix.

4. The method for collaborative optimization of automotive tire NVH performance structure according to claim 1, characterized in that, The step 3 comprises: Calling a finite element analysis tool, inputting the material properties and structural parameters of the crown reinforcement and the tread rubber, simulating the interaction between the two, and outputting the finite element analysis result; Calling a statistical energy analysis tool, inputting the material properties of the high-damping rubber layer, calculating the frequency-dependent vibration energy dissipation characteristics, and outputting the statistical energy analysis result; Preparing a historical test data set, training a neural network to establish a nonlinear mapping relationship between the structural parameters and the NVH performance, and generating a neural network model; Integrating the finite element analysis result, the statistical energy analysis result and the neural network model to construct the tire structure and NVH performance coupling model.

5. The method for synergistic optimization of automotive tire NVH performance structure according to claim 4, characterized in that, The step 4 comprises: The response surface model between the design variables and the NVH performance indicators is established by taking the prediction result set as sample data; Pareto frontier search is performed on the response surface model under the constraints of NVH performance, rolling resistance performance and durability to obtain a Pareto solution set; Sensitivity analysis is performed on the Pareto solution set to verify the stability of the solution, and the optimal parameter combination is selected from the Pareto solution set.

6. The method for synergistic optimization of automotive tire NVH performance structure according to claim 5, characterized in that, The step 5 includes: Based on the optimal parameter combination, a virtual prototype of the tire is constructed in the digital twin platform, and the prototype is connected to the hardware-in-the-loop simulation system; The hardware-in-the-loop simulation system is driven to simulate the NVH performance of the virtual prototype of the tire under extreme conditions, and virtual test results are generated; The virtual test results are compared with the real vehicle test data, and the tire structure and NVH performance coupling model is corrected based on the comparison results to form a closed-loop optimization.

7. The method for collaborative optimization of automotive tire NVH performance structure according to claim 1, characterized in that, The step 2 includes: The strain and stress response of the support rubber is analyzed from the multi-physical field data set to identify the influence law of the nonlinear characteristics of the support rubber stiffness on vibration attenuation; Based on the identified influence law, the stiffness gradient distribution mapping parameters are quantitatively generated and included in the feature parameter matrix.

8. The method for synergistic optimization of automotive tire NVH performance structure according to claim 7, characterized in that, The step 4 includes: The crown belt layer winding angle and winding tension parameters are used as coupled design variables for synchronous optimization, so that the optimized crown belt layer winding trajectory and carcass damping distribution work together to suppress vibration.

9. The method for synergistic optimization of automotive tire NVH performance structure according to claim 8, characterized in that, The step 4 also includes: Based on the tread ground pressure distribution obtained by finite element analysis, the thickness gradient of the tread rubber layer is adjusted; During the optimization process, the thickness gradient change is associated with the support rubber stiffness curve change for synchronous optimization in a coupled manner.

10. A vehicle tire characterized by comprising: The structure parameters of the tire are generated by the tire NVH performance structure collaborative optimization method of any one of claims 1 and 9, and the structure parameters include the optimized tread rubber layer thickness distribution, support rubber stiffness curve, crown belt layer winding trajectory and carcass damping distribution.

Images