An automobile acoustic package technical scheme data batch generation method and system

Abstract

The application discloses a kind of automobile acoustic package technical scheme data batch generation method and system, it is related to automobile NVH engineering technical field, it solves the technical problem that existing technology relies on manual operation inefficiency and error is prone to occur and data enhancement method based on linear interpolation or combination is difficult to generate real, effective training sample.The present application comprises the following steps: S1: constructing automobile acoustic package performance hierarchical decomposition tree, and defining parameter space;S2: based on optimal Latin hypercube algorithm, generate discrete design sample set in the parameter space;S3: parse XML format acoustic package structure file, automatically reconstruct parameterized model, obtain updated acoustic package structure file;S4: based on statistical energy analysis method simulation solving kernel, the updated acoustic package structure file is carried out batch processing calculation to obtain acoustic response result;S5: processing the acoustic response result, constructs simulation dataset, and the present application has the advantages such as batch efficient generation design scheme.

Description

Technical Field

[0001] This invention belongs to the field of automotive NVH engineering technology, specifically relating to a method and system for batch generation of automotive acoustic package technical data. Background Technology

[0002] With rising living standards and diversified user needs, automotive NVH performance has become a core indicator for measuring vehicle quality. In-vehicle noise control is particularly crucial, and acoustic packages are a key passive means of suppressing airborne sound transmission, having a decisive impact on in-vehicle acoustic quality. In the development of automotive acoustic packages, VA One simulation software based on statistical energy analysis has been widely used for performance prediction and solution verification. In recent years, data-driven technologies have provided new paths for acoustic prediction, but model training relies on large-scale, high-quality samples. While existing simulation data collection is more cost-effective than experimental testing, it relies on manually modifying parameters, submitting calculations, and recording results, which is inefficient and error-prone, resulting in datasets that fail to meet training requirements in terms of scale, quality, and consistency.

[0003] To address the efficiency issues in the modeling and simulation stages of VA One, existing patents have explored related research. For example, patent CN113553734B proposes a method for importing the overall digital model through a preprocessing module, determining the acoustic cavity structure partitioning data, simplifying the model, segmenting components, and meshing before exporting it to VA One to build an SEA model. This effectively improves the accuracy and efficiency of single or small-scale modeling. Patent CN108170970A imports a 3D automotive model into Hypermesh for geometry cleanup, subsystem partitioning, mesh generation, and attribute establishment, then exports it to Nastran files and imports it into different windows of VA One for modeling before copying and connecting. These methods all focus on the preprocessing modeling stage of SEA models, aiming to solve the problems of modeling efficiency and accuracy.

[0004] However, existing patents have the following shortcomings: First, they lack a fully automated process from design sample generation to simulation result storage, making it impossible to efficiently produce the large-scale, reliable samples required for data-driven models. Second, they rely excessively on manual operation of specific software, lacking the versatility for batch reproduction and flexible expansion. Third, traditional linear data augmentation methods cannot effectively generate acoustic package simulation samples with strong nonlinear characteristics, making it difficult to expand high-quality training datasets through post-processing. Furthermore, given the urgent need for large-scale samples in data-driven models, commonly used data augmentation methods in fields such as image processing (e.g., Mixup, Cutmix) are essentially based on linear interpolation or linear combination to generate new samples. However, the overall sound insulation performance of automotive acoustic packages exhibits significant strong nonlinearity, high dimensionality, and multi-scale coupling characteristics with design parameters (e.g., material type, thickness, coverage). Therefore, traditional linear data augmentation methods are not applicable in this field. Thus, a method for efficiently generating automotive acoustic package simulation data in batches is urgently needed to address these issues. Summary of the Invention

[0005] This invention discloses a method and system for batch generation of data for automotive acoustic package technology, which aims to solve the technical problems of existing technologies that rely on manual operation, are inefficient and prone to errors, and that data augmentation methods based on linear interpolation or combination are difficult to generate real and effective training samples.

[0006] To solve the aforementioned technical problems, the present invention adopts the following technical solution:

[0007] A method for batch generating data for an automotive acoustic package technology solution includes the following steps:

[0008] S1: Construct a hierarchical decomposition tree of the automotive acoustic package performance and define the parameter space;

[0009] S2: Based on the optimal Latin hypercube algorithm, a discrete design sample set is generated in the parameter space;

[0010] S3: Parse the XML format acoustic package structure file, automatically reconstruct the parametric model based on the design sample set, and obtain the updated acoustic package structure file;

[0011] S4: The simulation solution kernel based on statistical energy analysis performs batch processing calculations on the updated acoustic packet structure file to obtain the acoustic response results;

[0012] S5: Process the acoustic response results and construct a simulation dataset.

[0013] In this invention, steps S1-S5 establish an automated batch generation process for automotive acoustic package simulation data. Specifically, in S1, the acoustic performance of the entire vehicle is decoupled and decomposed step by step along the airborne sound transmission path until quantifiable design parameters are reached, and a parameter space is defined to provide standardized input for subsequent sampling. In S2, sampling is performed within the parameter space to generate a comprehensive design sample set. In S3, the XML-formatted acoustic package structure file is automatically parsed and its parameters are overwritten to reconstruct the parameterized model. In S4, batch processing calculations are performed based on statistical energy analysis to obtain the acoustic response results corresponding to each design scheme. In S5, the acoustic response results are processed to construct a simulation dataset. S1 to S5 form a complete automated data production process from parameter space definition, experimental design sampling, automatic model reconstruction, batch processing calculation to dataset generation, achieving efficient batch generation of acoustic package simulation data. This avoids the problems of low efficiency and inability to meet the training requirements of data-driven models in traditional manual operations.

[0014] Furthermore, S1 includes the following steps:

[0015] S11: Based on the airborne sound transmission path of the whole vehicle, the overall vehicle performance of the acoustic package is decoupled into the system-level sound insulation performance of multiple independent acoustic subsystems, and decomposed into the component sound insulation performance of sheet metal parts, guard plates, sound insulation pads and glass. The sound insulation performance of each component is decomposed into acoustic package design parameters. The overall vehicle performance of the acoustic package includes the sound transfer function and the amount of vehicle noise reduction.

[0016] S12: Construct a fourth-order mapping structure from the overall vehicle performance of the acoustic package to the design parameters of the acoustic package. The fourth-order mapping structure includes, in sequence, the vehicle level, the system level, the component level, and the design parameter level.

[0017] S13: Define the parameter space, determine the value range and variable type of the acoustic package design parameters, the acoustic package design parameters include at least the sheet metal area and thickness distribution; the protective plate area and thickness distribution; the sound insulation pad coverage, material type and thickness distribution; the glass area, material and thickness; the variable types include continuous variables and discrete variables, the continuous variables include at least the sheet metal thickness and the sound insulation pad coverage, and the discrete variables include at least the sound insulation pad material type and the glass material.

[0018] After adopting this technical solution, it should be noted that there is a multi-layered coupling relationship between the overall vehicle performance and the acoustic package design parameters. The overall vehicle performance is jointly determined by the system-level sound insulation performance of multiple subsystems, and the sound insulation performance of each subsystem depends on the intrinsic sound insulation performance of components such as sheet metal parts, protective panels, sound insulation pads, and glass. Ultimately, the performance of these components is driven by specific acoustic package design parameters, such as material type, thickness, coverage, and area. In S12, a four-order mapping structure from overall vehicle performance to acoustic package design parameters is constructed, namely, vehicle level, system level, component level, and design parameter level. This decouples and decomposes the overall vehicle acoustic performance indicators step by step down along the airborne sound transmission path until reaching the quantifiable and adjustable design parameter level, thereby transforming the overall vehicle performance target into specific engineering operable variables. Based on this, S13 further distinguishes between continuous variables (such as sheet metal thickness and sound insulation pad coverage) and discrete variables (such as sound insulation pad material type and glass material), configuring their value ranges or candidate sets respectively. Through the above hierarchical decomposition and parameter space definition, it is ensured that the values ​​of each subsequent design parameter have clear physical meaning and engineering feasibility.

[0019] Furthermore, S2 includes the following steps:

[0020] S21: Determine the range of acoustic package design parameters, construct a parameter sampling space for mixed variable types, configure upper and lower limit ranges for continuous variables, establish a mapping dictionary between material types and integer index values ​​for discrete variables, and convert non-numerical material options into integer codes;

[0021] S22: Perform optimal Latin hypercube sampling for multidimensional space filling within the parameter sampling space, perform hierarchical random sampling on the variable intervals of each dimension, and generate an original sampling matrix covering the high-dimensional parameter space.

[0022] S23: Perform reverse parsing on the original sampling matrix, reverse map the integer codes back to the corresponding material types, serialize all parameter combinations into a structured data table, and output a standardized acoustic package design sample set according to the preset sampling scale.

[0023] After adopting this technical solution, it should be noted that the acoustic package design parameter space has the characteristics of mixed variable types, including both continuous physical quantities, such as thickness and coverage, and discrete material options, such as PET cotton and PU foam. The optimal Latin hypercube sampling algorithm requires a fully numerical parameter space as input. In S21, a mapping dictionary between material type and integer index values ​​is established for discrete variables, such as mapping "PET cotton" to 1 and "PU foam" to 2. Non-numerical material options are converted into an integer encoding space, thereby constructing a fully digital parameter sampling space. The optimal Latin hypercube algorithm can perform high-dimensional space filling and stratified random sampling within a unified numerical framework. In step S22, the optimal Latin hypercube algorithm performs stratified sampling on the variable intervals of each dimension, ensuring that the final generated sampling points are uniformly distributed in the high-dimensional space. This avoids the sampling points clustering in local areas of the parameter space while ignoring other areas, thus guaranteeing sufficient coverage of the entire parameter space by the sampling results. In step S23, the integer codes are restored to the corresponding material type names through a mapping dictionary, while retaining the physical quantity values ​​of continuous variables. All parameters are combined and organized into a structured data table for output. Through the above steps, the computability of the sampling algorithm within a unified numerical framework is guaranteed.

[0024] Furthermore, S3 includes the following steps:

[0025] S31: Perform tree structure parsing on the acoustic package structure file, traverse the node levels of the acoustic package structure file, identify and map the names of sub-elements containing acoustic package design parameters, excitation conditions, acoustic constraints and solver settings and their corresponding attribute contents, and construct a parameter node index table;

[0026] S32: Match each acoustic package design parameter in the acoustic package design sample set with the parameter node index table to locate the corresponding sub-element attribute in the acoustic package structure file;

[0027] S33: Overwrite the values ​​of the acoustic package design parameters into the mapped sub-element attributes, thereby automatically completing the reconstruction of the parametric model and generating an updated acoustic package structure file corresponding to the acoustic package design sample set.

[0028] After adopting this technical solution, it should be noted that in S31, by traversing the nodes at each level, it identifies which node attributes correspond to design parameters, which correspond to excitation conditions, which correspond to acoustic constraints, and which correspond to solver settings, and establishes a mapping relationship of parameter name → node path → attribute name, i.e., a parameter node index table. This index table allows the system to quickly locate the corresponding writing position in the XML file simply by using the parameter name. S32 matches each dimension parameter in the design sample set with the index table to determine its target node and attribute in the XML file; S33 writes the specific value of each design parameter into the located attribute, completing the update of the XML file. Through the above index table construction and parameter overwriting mechanism, the automatic conversion from the design sample set to the parameterized model file is realized, enabling each set of parameter combinations in the design sample set to quickly generate a corresponding independent model file.

[0029] Furthermore, in S33, by establishing a mapping relationship between the design parameters of each acoustic package in the design sample set and the parameter node index table, based on this mapping relationship, the values ​​of the design parameters of each acoustic package are batch overwritten into the corresponding node attributes in the acoustic package structure file, thereby generating multiple updated acoustic package structure files corresponding to the design sample set.

[0030] After adopting this technical solution, it should be noted that the design samples usually contain multiple different combinations of design parameters, where each set of parameters represents an independent acoustic package design scheme. Here, based on the established mapping relationship, an overwrite operation is performed on each set of parameter combinations in the design sample set. Each overwrite is based on the same original XML template file, replacing the values ​​of the corresponding node attributes with the specific parameter values ​​of the scheme, thereby generating a unique updated XML file for that scheme. Through batch overwriting, the automatic and efficient conversion from the design sample set to a large number of parametric model files is achieved.

[0031] Furthermore, in S4, the simulation solver kernel supports importing acoustic package structure parameters in XML file format and performing acoustic response calculations.

[0032] Furthermore, S4 includes the following steps:

[0033] S41: Based on acoustic simulation software, create a blank database file, import the updated acoustic package structure file into the blank database file, and generate an executable calculation file that corresponds one-to-one with each updated acoustic package structure file. The calculation file contains complete model parameters and simulation parameters, and each calculation file corresponds to a design scheme.

[0034] S42: Call the acoustic simulation software kernel, submit the calculation task to the calculation file, and obtain the in-vehicle acoustic cavity noise response corresponding to each design scheme;

[0035] S43: Export the noise response results of the in-vehicle acoustic cavity and store them in a structured format in a preset HDF5 data file.

[0036] After adopting this technical solution, it should be noted that the external program can send instructions to the acoustic simulation software to control it to perform a series of operations such as creating files, importing models, submitting calculations, and exporting results. S41 calls the acoustic simulation software, first creating a new blank database file in the software environment, and then importing the updated acoustic package structure file generated in S33 into this blank database. During the import process, the software automatically parses all model information in the XML file, generating an executable calculation file that can be directly used for subsequent calculations. This executable calculation file contains all the information required for a complete model. In S42, the simulation kernel is based on statistical energy analysis. By solving the energy balance equations between each subsystem, it obtains the sound pressure response results of the in-vehicle acoustic cavity under a given noise source excitation. S43 exports the calculated sound pressure response results from the software environment and stores them in HDF5 format. HDF5 is a file format designed for large-scale scientific data. It uses a hierarchical grouping structure to organize data, supports storing multiple sets of response results from different schemes in the same file, and supports quick reading of data for specific schemes by group index, making it suitable as a storage medium for large amounts of simulation result data. The aforementioned acoustic simulation software and batch processing workflow enabled the automated submission and result aggregation of multi-scheme simulation calculation tasks.

[0037] Furthermore, S5 includes the following steps:

[0038] S51: The HDF5 data file is parsed using a script to extract the in-vehicle acoustic cavity noise response results for each design scheme, and the results are converted from sound pressure physical quantities into sound pressure level characteristics.

[0039] S52: Calculate the sound transfer function and the reduction in overall vehicle noise based on the noise response results;

[0040] S53: Associate and map each design scheme with its corresponding acoustic transfer function and the reduction in vehicle noise, and write it into a pre-set data sample table in a structured manner to complete the batch collection of the vehicle acoustic performance of different design schemes and obtain the simulation dataset.

[0041] After adopting this technical solution, it should be noted that the HDF5 data file stored in S43 contains the in-vehicle acoustic cavity noise response results corresponding to each design scheme, in physical form as sound pressure (unit Pa). However, in engineering practice, acoustic performance indicators are usually characterized in the form of sound pressure level (unit dB) and sound transfer function, which are more consistent with the human ear's perception of sound intensity; the sound transfer function characterizes the sound energy transfer characteristics from the noise source to the receiving position in the vehicle. Therefore, S51 first reads the sound pressure response data corresponding to each scheme in the HDF5 file through the data parsing program and converts it into sound pressure level form according to the sound pressure level conversion formula. S52 calculates the sound transfer function of each transmission path and the overall vehicle noise reduction based on the converted sound pressure level data. S53 associates the design parameter combination of each design scheme with the calculated sound transfer function and the overall vehicle noise reduction one-to-one, forming a structured data pair of design parameters → acoustic performance, and writes it into a preset data sample table.

[0042] Furthermore, the following operations are performed in the preset script sequence: reading the design sample set, overwriting the numerical values ​​of the acoustic package design parameters, creating a blank database and importing the updated acoustic package structure file, submitting the calculation task and storing the calculation results.

[0043] A system for batch generating data for automotive acoustic package technology solutions includes:

[0044] The parameter space definition module is used to construct the performance hierarchy decomposition tree of the acoustic package and define the parameter space, generating parameter space data that includes the value range and variable types of the acoustic package design parameters.

[0045] The experimental design sampling module is used to communicate with the parameter space definition module, receive the parameter space data, and perform sampling based on the optimal Latin hypercube algorithm to generate a design sample set.

[0046] The XML parsing and generation module is used to parse the acoustic package structure file and, based on the design sample set, batch overwrite parameters to generate multiple updated acoustic package structure files corresponding to each design scheme.

[0047] The simulation batch processing module is used to communicate with the XML parsing and generation module, and call the acoustic simulation software kernel to sequentially import the updated acoustic package structure file and perform batch processing calculations, outputting an HDF5 file containing acoustic response results;

[0048] The whole vehicle sound insulation performance calculation module is used to communicate with the simulation batch processing module, parse the HDF5 file, convert the sound pressure physical quantity into a sound pressure level characterization, and calculate the sound transfer function and the whole vehicle noise reduction.

[0049] The dataset construction module is used to communicate with the vehicle sound insulation performance calculation module and the experimental design sampling module, and to associate and map the acoustic package design parameters with the corresponding acoustic performance indicators to construct a structured simulation dataset.

[0050] The central scheduling module communicates with the parameter space definition module, the experimental design sampling module, the XML parsing and generation module, the simulation batch processing module, the whole vehicle sound insulation performance calculation module, and the dataset construction module, and performs timing and data interaction.

[0051] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0052] 1. The present invention provides a method and system for batch generation of automotive acoustic package technical solution data. By constructing a hierarchical decomposition tree, optimal Latin hypercube sampling, XML parsing and parameter overwriting, batch processing calculation and post-processing dataset generation, the traditionally manual process is completed automatically by the system, and the design scheme is generated in batches and efficiently, which meets the needs of data-driven models for large-scale training samples.

[0053] 2. The present invention provides a method and system for batch generation of data for automotive acoustic package technology. By using the optimal Latin hypercube algorithm to perform hierarchical sampling in a high-dimensional mixed variable parameter space, the sampling points are uniformly covered throughout the parameter space, avoiding the problem of uneven parameter space coverage caused by sampling clustering, and providing a comprehensive and diverse sample basis for subsequent model training.

[0054] 3. The present invention provides a method and system for batch generation of automotive acoustic package technical solution data. Through the construction of parameter node index table and batch overwrite mechanism, it realizes the automatic conversion from design sample set to a large number of parameterized model files, avoids parameter input errors caused by negligence when manually modifying the model, and ensures the high quality of the generated dataset. Attached Figure Description

[0055] The present invention will be described by way of example and with reference to the accompanying drawings, wherein:

[0056] Figure 1 This is an overall flowchart of the present invention;

[0057] Figure 2 This is a flowchart illustrating the specific process of this invention;

[0058] Figure 3 This is a schematic diagram of the hierarchical decomposition structure of the present invention;

[0059] Figure 4 This is a schematic diagram illustrating the generation of design parameter schemes based on optimal Latin hypercube sampling in this invention;

[0060] Figure 5This is a flowchart illustrating the process of parsing the acoustic package structure file and automatically reconstructing the parameterized model according to the present invention.

[0061] Figure 6 This is a schematic diagram of the process of obtaining acoustic response results by simulation batch processing calculation based on the kernel call of acoustic simulation software according to the present invention;

[0062] Figure 7 This is a schematic diagram of the process for analyzing and processing acoustic response results in this invention;

[0063] Figure 8 This is a schematic diagram of the acoustic transfer function and the vehicle noise reduction dataset generated by the present invention.

[0064] Figure 9 This is a comparison chart of the efficiency of the present invention based on the solution of this application and the traditional manual method. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the embodiments and accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and marked in the accompanying drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0066] In the description of the embodiments of this application, it should be noted that the terms "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of the invention is in use. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. In addition, the terms "first," "second," and "third," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0067] The following is combined Figures 1-8 The present invention will be described in detail below.

[0068] A method for batch data generation of automotive acoustic package technology solutions, such as Figure 1 , Figure 2 As shown, it includes the following steps:

[0069] S1: Construct a hierarchical decomposition tree of the automotive acoustic package performance and define the parameter space;

[0070] S2: Based on the optimal Latin hypercube algorithm, a discrete design sample set is generated in the parameter space;

[0071] S3: Parse the XML format acoustic package structure file, automatically reconstruct the parametric model based on the design sample set, and obtain the updated acoustic package structure file;

[0072] S4: The simulation solution kernel based on statistical energy analysis performs batch processing calculations on the updated acoustic packet structure file to obtain the acoustic response results;

[0073] S5: Process the acoustic response results and construct a simulation dataset.

[0074] In this embodiment, steps S1-S5 establish an automated batch generation process for automotive acoustic package simulation data. Specifically, in S1, the acoustic performance of the entire vehicle is decoupled and decomposed step by step along the airborne sound transmission path until the quantifiable design parameters are defined, and a parameter space is defined to provide standardized input for subsequent sampling. In S2, sampling is performed within the parameter space to generate a comprehensive design sample set. In S3, the XML-formatted acoustic package structure file is automatically parsed and its parameters are overwritten to reconstruct the parameterized model. In S4, batch processing calculations are performed based on statistical energy analysis to obtain the acoustic response results corresponding to each design scheme. In S5, the acoustic response results are processed to construct the simulation dataset. S1 to S5 form a complete automated data production process from parameter space definition, experimental design sampling, automatic model reconstruction, batch processing calculation to dataset generation, achieving efficient batch generation of acoustic package simulation data. This avoids the problems of low efficiency and inability to meet the training requirements of data-driven models in traditional manual operations.

[0075] In this embodiment, S1 includes the following steps:

[0076] S11: Based on the airborne sound transmission path of the whole vehicle, the overall vehicle performance of the acoustic package is decoupled into the system-level sound insulation performance of multiple independent acoustic subsystems, and decomposed into the component sound insulation performance of sheet metal parts, guard plates, sound insulation pads and glass. The sound insulation performance of each component is decomposed into acoustic package design parameters. The overall vehicle performance of the acoustic package includes the sound transfer function and the amount of vehicle noise reduction.

[0077] In S11, the core logic of transmission path analysis is to decompose the noise ultimately heard inside the vehicle into a superposition of contributions from different sources through different paths, thereby helping engineers accurately locate and solve noise problems. In this embodiment, using the target sound pressure level inside the vehicle as the final constraint, the vehicle performance requirements are allocated progressively and in reverse to system-level performance, component-level performance, and design parameters along the sound propagation path from the source to the receiver. Specifically, using the target sound pressure level inside the vehicle as the top-level constraint, the sound energy transmission from major noise sources such as the engine and tires to the vehicle through air and structural paths is first identified and quantified. Based on this, according to the location of each acoustic component inside the vehicle, such as... Figure 3 As shown, the vehicle acoustic package is divided into nine independent subsystems: front bulkhead system, floor system, rear wheel hub system, roof system, side panel system, front door system, rear door system, tailgate system, and rear bulkhead system. The overall vehicle-level performance indicators of the acoustic package are decomposed into system-level sound insulation performance. The key components carrying sound insulation functions within each subsystem are analyzed, and these system-level sound insulation performances are further decomposed into sound insulation performance targets for components such as sheet metal parts, protective panels, sound insulation pads, and glass. Finally, the component-level performance requirements are further transformed into quantifiable acoustic package design parameters such as materials, area, and thickness, thus establishing a complete causal chain from overall vehicle targets to design parameters.

[0078] S12: Construct a fourth-order mapping structure from the overall vehicle performance of the acoustic package to the design parameters of the acoustic package. The fourth-order mapping structure includes, in sequence, the vehicle level, the system level, the component level, and the design parameter level.

[0079] like Figure 3 As shown, in S12, the fourth-order mapping structure includes, in sequence, the vehicle level characterizing the overall vehicle performance, the system level characterizing the sound insulation performance of each subsystem, the component level characterizing the intrinsic sound insulation performance of the components, and the acoustic package design parameter level characterizing the changes in driving sound insulation performance. The tree structure does not contain weight coefficients, and the levels are naturally related through physical transmission paths.

[0080] S13: Define the parameter space, determine the value range and variable types of the acoustic package design parameters. The acoustic package design parameters include at least the sheet metal area and thickness distribution; the guard plate area and thickness distribution; the sound insulation pad coverage, material type, and thickness distribution; and the glass area, material, and thickness. The variable types include continuous and discrete variables. Continuous variables include at least the sheet metal thickness and sound insulation pad coverage, while discrete variables include at least the sound insulation pad material type and the glass material. There is a multi-layered coupling relationship between the overall vehicle performance and the acoustic package design parameters. The overall vehicle performance is jointly determined by the system-level sound insulation performance of multiple subsystems, while the sound insulation performance of each subsystem depends on the intrinsic sound insulation performance of components such as sheet metal, guard plates, sound insulation pads, and glass. Ultimately, the performance of these components is driven by specific acoustic package design parameters, such as material type, thickness, coverage, and area. S12 constructs a four-order mapping structure from vehicle performance to acoustic package design parameters: vehicle level, system level, component level, and design parameter level. This decouples and decomposes the vehicle's acoustic performance indicators step-by-step along the airborne sound transmission path until reaching the quantifiable and adjustable design parameter level, thus transforming vehicle performance targets into specific, engineering-operable variables. Building on this, S13 further distinguishes between continuous variables (such as sheet metal thickness and sound insulation pad coverage) and discrete variables (such as sound insulation pad material type and glass material), configuring their respective value ranges or candidate sets. Through this hierarchical decomposition and parameter space definition, it is ensured that the values ​​of each subsequent design parameter have clear physical meaning and engineering feasibility.

[0081] In this embodiment, S2 includes the following steps:

[0082] S21: As Figure 4 As shown, the range of acoustic package design parameters is determined, a parameter sampling space of mixed variable type is constructed, and upper and lower limit ranges of values ​​are configured for continuous variables. For example, in this embodiment, the thickness range of the front sound insulation pad is limited to 0mm to 30mm, and the coverage range is 0.80 to 0.95. For discrete variables, a mapping dictionary between material type and integer index value is established, and non-numerical material options are converted into integer codes. For example, in this embodiment, the material type can be PET cotton, EVA, and PU foam, which are mapped to integers 1, 2, and 3, respectively.

[0083] S22: Perform optimal Latin hypercube sampling for multidimensional space filling within the parameter sampling space, perform hierarchical random sampling on the variable intervals of each dimension, and generate an original sampling matrix covering the high-dimensional parameter space.

[0084] In S22, the value range of each design parameter is divided into several equally probable intervals. A sample point is randomly selected in each interval, and stratified random sampling is performed on the variable intervals of each dimension to generate an original sampling matrix covering the high-dimensional parameter space. In this embodiment, a single-system change scheme (adjusting only the design parameters related to one system), a multi-system change scheme (adjusting the design parameters related to multiple systems simultaneously), and a full-system change scheme (adjusting the design parameters related to all nine systems simultaneously) can be generated. Secondly, in this embodiment, the changes in the design parameters of each system are controlled by a preset selection parameter. The length of the selection is a vector equal to the total number of systems, i.e., 9 systems. Each element in the list controls whether the design parameters of all components in a system change. For example, if only one element in the selection is 1 and the other elements are 0, it corresponds to a single-system change scheme; if multiple elements are 1 and some elements are 0, it corresponds to a multi-system change scheme; if all elements are 1, it corresponds to a full-system change scheme. The system executes the corresponding sampling strategy in the parameter sampling space according to the preset selection parameter to generate an original sampling matrix covering the high-dimensional parameter space.

[0085] In this embodiment, taking the front sound insulation pad as an example, the original sampling matrix corresponding to some samples in the following original sampling matrix is ​​as follows: each row represents one sample, the first column corresponds to the material type code, the second column corresponds to the coverage rate, and the third to tenth columns correspond to the distribution ratio of eight thickness levels from 5mm to 40mm:

[0086]

[0087] As can be seen from the matrix, the first column consists of integer codes, reflecting the mapping result of discrete variables to integer codes in S21; the second column is the coverage value, which is distributed in the range of [0.85, 0.95]; the sum of each row in the third to tenth columns is 1, representing the distribution ratio of each thickness level in the component area, reflecting the distribution relationship of the thickness distribution parameter among different thickness levels.

[0088] S23: Perform reverse parsing on the original sampling matrix, mapping the integer codes back to the corresponding material types (e.g., mapping the integer 1 back to PET cotton), and serialize all parameter combinations into a structured data table. Output a standardized acoustic package design sample set according to a preset sampling scale. In this embodiment, the preset sampling scale is 2000 samples.

[0089] In this embodiment, step S21 establishes a mapping dictionary between material types and integer index values ​​for discrete variables, such as mapping "PET cotton" to 1, transforming non-numerical material options into an integer coding space. This constructs a fully digital parameter sampling space, enabling the optimal Latin hypercube algorithm to perform high-dimensional space filling and stratified random sampling within a unified numerical framework. In step S22, the optimal Latin hypercube algorithm performs stratified sampling on the variable intervals of each dimension, ensuring that the final generated sampling points are evenly distributed in the high-dimensional space. This avoids sampling points clustering in local areas of the parameter space while ignoring other areas, thus guaranteeing sufficient coverage of the entire parameter space by the sampling results. In step S23, the integer codes are restored to their corresponding material type names through the mapping dictionary, while retaining the physical quantity values ​​of continuous variables. All parameters are combined and organized into a structured data table for output. These steps ensure the computability of the sampling algorithm within a unified numerical framework.

[0090] In this embodiment, as Figure 5 As shown, S3 includes the following steps:

[0091] S31: Perform tree structure parsing on the acoustic package structure file, traverse the node levels of the acoustic package structure file, identify and map the names of sub-elements containing acoustic package design parameters, excitation conditions, acoustic constraints, and solver settings, and their corresponding attribute contents, and construct a parameter node index table. This index table records the precise path of each modifiable parameter in the XML tree structure. It can be understood as writing the path of the element corresponding to the design parameter of each component in the XML in the index table. Subsequent programs modify the element attributes at the corresponding positions in the XML according to the index table. For example, the path of the coverage parameter of the sound insulation pad is " / Faces / FlatPlateFace / PositiveFace".

[0092] S32: Match each acoustic package design parameter in the acoustic package design sample set with the parameter node index table to locate the corresponding sub-element attribute in the acoustic package structure file;

[0093] S33: The numerical values ​​of the acoustic package design parameters are overwritten into the mapped sub-element attributes, thereby automatically completing the reconstruction of the parametric model and generating an updated acoustic package structure file corresponding to the acoustic package design sample set. In this embodiment, the specific overwriting process is as follows: Using a scripting language (such as Python), the acoustic package structure file XML before the update is parsed, and its node hierarchy is traversed to locate the target sub-element; the name and attribute identifier of the sub-element are obtained and compared with the parameter node index table; when the name and attribute identifier meet the mapping conditions with the design parameters, the corresponding target design parameter values ​​are extracted from the design sample set, and the attributes of the sub-element are automatically overwritten. If the current node does not meet the mapping conditions, continue comparing the next node until a matching node is found or all nodes in the current XML file have been traversed. For example, if the path corresponding to the sound insulation pad coverage is " / Faces / FlatPlateFace / PositiveFace", but there are several sound insulation pad coverage rates under this path, to modify the sound insulation rate of the front sound insulation pad, it is necessary to determine whether the name and attribute representation of a certain element under this path maps to the front sound insulation pad. If so, the overwrite is performed; otherwise, continue comparing the next node until a matching node is found or the element name and attributes corresponding to the front sound insulation pad are found after traversing all elements under the entire path. When all nodes of the acoustic package structure file XML before the update have been traversed, the parameter overwrite corresponding to the current design scheme is completed, and the updated acoustic package structure file XML is saved. Subsequently, it is determined whether there are any unprocessed design schemes in the design sample set. If so, the next design scheme is obtained, the previous acoustic package structure file XML is reloaded as a template, and the above node traversal and parameter overriding steps are repeated. If not, the parametric model reconstruction of all schemes is completed, resulting in multiple updated acoustic package structure files XML corresponding one-to-one with the design sample set. Through the above batch overriding steps, the acoustic package structure files XML corresponding to different samples are automatically generated, realizing the automated reconstruction of the automotive acoustic package parametric model. This process can be completed without opening the simulation software interface and is not limited by the simulation software version.

[0094] In step S33, a mapping relationship is established between the design parameters of each acoustic package in the design sample set and the parameter node index table. Based on this mapping relationship, the values ​​of the design parameters of each acoustic package are batch overwritten into the corresponding node attributes in the acoustic package structure file, generating multiple updated acoustic package structure files corresponding to the design sample set. Design samples typically contain multiple different combinations of design parameters, each representing an independent acoustic package design scheme. Here, based on the established mapping relationship, an overwrite operation is performed on each parameter combination in the design sample set. Each overwrite is based on the same original XML template file, replacing the values ​​of the corresponding node attributes with the specific parameter values ​​of that scheme, thereby generating an updated XML file unique to that scheme. Through batch overwriting, an automatic and efficient conversion from the design sample set to a large number of parameterized model files is achieved.

[0095] In this embodiment, in S4, the simulation solver kernel supports importing acoustic package structure parameters in XML file format and performing acoustic response calculations.

[0096] In this embodiment, as Figure 6 As shown, S4 includes the following steps:

[0097] S41: Based on acoustic simulation software, create a blank database file, import the updated acoustic package structure file into the blank database file, and generate an executable calculation file that corresponds one-to-one with each updated acoustic package structure file. The calculation file contains complete model parameters and simulation parameters, and each calculation file corresponds to a design scheme.

[0098] In S41, the blank database file refers to the initial file that has not been modeled and parameter set. In this embodiment, it is in .va1 format. Different acoustic package schemes are identified and calculated sequentially using a preset numbering rule, such as "car_1", "car_2"... "car_N". A serial calculation processing method is adopted to avoid competition for system resources when multiple calculation tasks run at the same time.

[0099] S42: Call the acoustic simulation software kernel, submit the calculation task to the calculation file, and obtain the in-vehicle acoustic cavity noise response corresponding to each design scheme;

[0100] S43: Export the noise response results of the in-vehicle acoustic cavity and store them in a structured format in a preset HDF5 data file.

[0101] In this embodiment, an external program can send instructions to the acoustic simulation software to control it to perform a series of operations such as creating files, importing models, submitting calculations, and exporting results. S41 calls the acoustic simulation software, first creating a new blank database file in the software environment, and then importing the updated acoustic package structure file generated in S33 into this blank database. During the import process, the software automatically parses all model information in the XML file, generating an executable calculation file that can be directly used for subsequent calculations. This executable calculation file contains all the information required for a complete model. In S42, the simulation kernel is based on statistical energy analysis. By solving the energy balance equations between subsystems, it obtains the sound pressure response results of the in-vehicle acoustic cavity under a given noise source excitation. The basic idea of ​​statistical energy analysis is to decompose a complex system into several coupled subsystems and establish an energy balance equation for each subsystem. The energy received by a subsystem is equal to the sum of its own energy loss and the energy transferred to other subsystems. For two coupled subsystems, their power flow equation can be expressed as:

[0102]

[0103] In the formula, The net power transferred from subsystem 1 to subsystem 2 is expressed in watts (W). To analyze the center frequency within the frequency band, the unit is rad / s. The coupling loss factor between subsystem 1 and subsystem 2. The total vibrational energy of subsystem 1 within the analysis frequency band. This represents the total vibrational energy of subsystem 2 within the analysis frequency band.

[0104] When the system contains N subsystems, the above equations are expanded into a set of energy balance equations in matrix form:

[0105]

[0106] In the formula, Representation Subsystem Damping loss factor , For subsystem With subsystem Coupling loss factor between , To analyze the center frequency (rad / s) within the frequency band. Let i be the total vibrational energy of subsystem i within the analysis frequency band. For subsystem The input power; that is, in this embodiment, the input power for each subsystem. Equal to the energy dissipation of the subsystem itself In addition to the net power it transfers to all other subsystems .

[0107] The simulation kernel in S42 is based on statistical energy analysis. By solving the energy balance equations between each subsystem, it obtains the sound pressure response results of the in-vehicle acoustic cavity under a given noise source excitation. S43 exports the calculated sound pressure response results from the software environment and stores them in HDF5 format. HDF5 is a file format designed for large-scale scientific data. It uses a hierarchical grouping structure to organize data, supports storing multiple sets of response results from different schemes in the same file, and supports quick retrieval of data for specific schemes by group index, making it suitable as a storage medium for large amounts of simulation results. Through the aforementioned acoustic simulation software and batch processing workflow, the automated submission and result aggregation of multi-scheme simulation calculation tasks are achieved.

[0108] In this embodiment, as Figure 7 As shown, S5 includes the following steps:

[0109] S51: The HDF5 data file is parsed using a script to extract the in-vehicle acoustic cavity noise response results for each design scheme, and the results are converted from sound pressure physical quantities into sound pressure level characteristics. In this embodiment, the conversion formula is as follows:

[0110]

[0111] In the formula, Sound pressure level (dB) The measured sound pressure (Pa) is given. Pa is the reference sound pressure level.

[0112] S52: Calculate the sound transfer function and the overall vehicle noise reduction based on the noise response results. In this embodiment, the calculation formulas for the sound transfer function and the overall vehicle noise reduction are as follows:

[0113]

[0114]

[0115]

[0116]

[0117] In the formula, For cabin acoustic transfer function; The tire sound transfer function; For the acoustic transfer function of the aft cabin; To reduce overall vehicle noise; Sound pressure level of cabin noise sources, measured in dB; This refers to the sound pressure level of the tire noise source, measured in dB. The sound pressure level of the noise source in the rear cabin is expressed in dB. The sound pressure level of an outdoor noise source with reverberation is expressed in dB. The noise sound pressure level at the left front of the head is expressed in dB. The noise sound pressure level at the right rear head position is expressed in dB. Furthermore, in this embodiment, a standard sedan consists of three independent "compartments": an engine compartment, a passenger compartment, and a luggage compartment. The engine compartment represents the front noise source of the vehicle, specifically the engine compartment, while the rear engine compartment represents the luggage compartment. It characterizes the contribution of engine compartment noise to vehicle interior noise. Under the same excitation, the smaller the value, the better the sound insulation performance of the whole vehicle acoustic package for engine compartment noise. The smaller the value, which characterizes the contribution of tire noise to in-vehicle noise under the same excitation, the better the sound insulation performance of the whole vehicle acoustic package for tire noise. The value of the rear engine compartment noise characterizes the contribution of the rear engine compartment noise to the vehicle interior noise. Under the same excitation, the smaller the value, the better the sound insulation performance of the whole vehicle acoustic package for the rear engine compartment noise. These four indicators characterize the reduction level of in-vehicle noise relative to external noise; a higher value indicates better overall sound insulation performance of the vehicle's acoustic package against external noise. All four indicators are used to measure the overall sound insulation performance of the vehicle's acoustic package.

[0118] S53: Associate and map each design scheme with its corresponding acoustic transfer function and overall vehicle noise reduction, and structure this association into a preset data sample table. This completes the batch collection of the overall vehicle acoustic performance for different design schemes, resulting in the simulation dataset. In this embodiment, one data sample table corresponds to a set of acoustic package design parameter schemes and their corresponding acoustic transfer functions and overall vehicle noise reduction. Each column of the data sample table corresponds to a frequency point, and each row corresponds to an overall vehicle performance level (e.g., ...). , , , ).

[0119] In this embodiment, the HDF5 data file stored in S43 contains the in-vehicle acoustic cavity noise response results corresponding to each design scheme, which are in the physical form of sound pressure (unit Pa). However, in engineering practice, acoustic performance indicators are usually characterized in the form of sound pressure level (unit dB) and sound transfer function, which are more consistent with the human ear's perception of sound intensity. The sound transfer function characterizes the sound energy transfer characteristics from the noise source to the receiving position in the vehicle. Therefore, S51 first reads the sound pressure response data corresponding to each scheme in the HDF5 file through the data parsing program and converts it into sound pressure level form according to the sound pressure level conversion formula. S52 calculates the sound transfer function of each transmission path and the overall vehicle noise reduction based on the converted sound pressure level data. S53 associates the design parameter combination of each design scheme with the calculated sound transfer function and the overall vehicle noise reduction one-to-one, forming a structured data pair of design parameters → acoustic performance, and writes it into a preset data sample table.

[0120] An exception handling logic is integrated into the result parsing process, using a try-except exception capture statement. When any solution's result file contains an exception or the simulation diverges, the exception for that solution is captured and skipped, without affecting the batch extraction of result files for other solutions. All erroneous solution numbers are output uniformly after batch processing is completed.

[0121] In this embodiment, the reading of the design sample set, the parameter overwriting of the acoustic package structure file, the creation and import of the database, the submission of the simulation task, and the storage of the calculation results are executed sequentially through the preset script. S1-S5 are executed sequentially through the preset automated control script, which includes a parameter space definition module, an experimental design sampling module, an XML parsing and generation module, a simulation batch processing module, a whole vehicle sound insulation performance calculation module, and a central scheduling module. It completes the generation of the acoustic package technical solution design sample set, the parameter overwriting of the acoustic package structure file XML, the submission of the calculation task, and the post-processing and structured storage of the calculation results without human intervention.

[0122] like Figure 8The figure illustrates the relationship between the acoustic transfer function and the vehicle noise reduction curves corresponding to 2000 different acoustic package technical solutions. It shows that while the curves for different solutions exhibit a consistent trend, they differ in specific values, reflecting the impact of different design parameter combinations on acoustic performance. This figure visually demonstrates the large scale and diversity of the dataset generated by this invention. Furthermore, the figure shows that the nonlinear response curve generated through physical simulation in this embodiment exhibits a clear non-monotonic, nonlinear fluctuation pattern, fully preserving the nonlinear performance characteristics of the automotive acoustic package system and overcoming the technical deficiency of traditional linear data augmentation methods in accurately characterizing the nonlinear response relationship of acoustic packages. Generating simulation samples within the full design parameter space through optimal Latin hypercube sampling effectively covers the boundary region of the design space, obtaining a complete nonlinear response surface, which can be directly used for constructing large-scale, high-quality training samples for AI agent models.

[0123] Example 2

[0124] A system for batch generating data for automotive acoustic package technology solutions includes:

[0125] The parameter space definition module is used to construct the performance hierarchy decomposition tree of the acoustic package and define the parameter space, generating parameter space data that includes the value range and variable types of the acoustic package design parameters.

[0126] The experimental design sampling module is used to communicate with the parameter space definition module, receive the parameter space data, and perform sampling based on the optimal Latin hypercube algorithm to generate a design sample set.

[0127] The XML parsing and generation module is used to parse the acoustic package structure file and, based on the design sample set, batch overwrite parameters to generate multiple updated acoustic package structure files corresponding to each design scheme.

[0128] The simulation batch processing module is used to communicate with the XML parsing and generation module, and call the acoustic simulation software kernel to sequentially import the updated acoustic package structure file and perform batch processing calculations, outputting an HDF5 file containing acoustic response results;

[0129] The whole vehicle sound insulation performance calculation module is used to communicate with the simulation batch processing module, parse the HDF5 file, convert the sound pressure physical quantity into a sound pressure level characterization, and calculate the sound transfer function and the whole vehicle noise reduction.

[0130] The dataset construction module is used to communicate with the vehicle sound insulation performance calculation module and the experimental design sampling module, and to associate and map the acoustic package design parameters with the corresponding acoustic performance indicators to construct a structured simulation dataset.

[0131] The central scheduling module communicates with the parameter space definition module, the experimental design sampling module, the XML parsing and generation module, the simulation batch processing module, the whole vehicle sound insulation performance calculation module, and the dataset construction module, and performs timing and data interaction.

[0132] Comparative Example 1

[0133] In this comparative example, taking the collection of simulation data for a single vehicle model with identical design samples as an example, we compare the use of script-based batch processing with manual operation. Specifically, taking the simulation data collection task of 2000 design samples for a single vehicle model as an example, we obtain the following results: Figure 9 The efficiency comparison chart between the proposed method and the traditional manual method is shown in the figure. The proposed method, through a scripted batch processing mechanism, reduces the total processing time from approximately 218 hours to approximately 28 hours and 40 minutes using the traditional manual method, representing an efficiency improvement of approximately 7.6 times. Specifically, in the design sample set generation stage, the proposed method takes 3 hours and 30 minutes, only about 1 / 7 of the 24 hours and 24 minutes required by manual methods; in the parametric model reconstruction stage, the proposed method takes 8 hours and 54 minutes, significantly better than the 252 hours and 42 minutes required by manual modifications; in the simulation solution stage, the proposed method takes 29 hours and 42 minutes, better than the 38 hours and 6 minutes required by manual submission; and in the acoustic response result processing stage, the proposed method only requires 30 minutes, while manual recording and organization takes 16 hours and 42 minutes. The efficiency improvement becomes more significant with the increase in the number of solutions, while simultaneously eliminating human errors during manual data entry.

[0134] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for batch generating data for an automotive acoustic package technical solution, characterized in that, Includes the following steps: S1: Construct a hierarchical decomposition tree of the automotive acoustic package performance and define the parameter space; S2: Based on the optimal Latin hypercube algorithm, a discrete design sample set is generated in the parameter space; S3: Parse the XML format acoustic package structure file, automatically reconstruct the parametric model based on the design sample set, and obtain the updated acoustic package structure file; S4: The simulation solution kernel based on statistical energy analysis performs batch processing calculations on the updated acoustic packet structure file to obtain the acoustic response results; S5: Process the acoustic response results and construct a simulation dataset.

2. The method for batch data generation of an automotive acoustic package technical solution according to claim 1, characterized in that, S1 includes the following steps: S11: Based on the airborne sound transmission path of the whole vehicle, the overall vehicle performance of the acoustic package is decoupled into the system-level sound insulation performance of multiple independent acoustic subsystems, and decomposed into the component sound insulation performance of sheet metal parts, guard plates, sound insulation pads and glass. The sound insulation performance of each component is decomposed into acoustic package design parameters. The overall vehicle performance of the acoustic package includes the sound transfer function and the amount of vehicle noise reduction. S12: Construct a fourth-order mapping structure from the overall vehicle performance of the acoustic package to the design parameters of the acoustic package. The fourth-order mapping structure includes, in sequence, the vehicle level, the system level, the component level, and the design parameter level. S13: Define the parameter space, determine the value range and variable type of the acoustic package design parameters, the acoustic package design parameters include at least the sheet metal area and thickness distribution; the protective plate area and thickness distribution; the sound insulation pad coverage, material type and thickness distribution; the glass area, material and thickness; the variable types include continuous variables and discrete variables, the continuous variables include at least the sheet metal thickness and the sound insulation pad coverage, and the discrete variables include at least the sound insulation pad material type and the glass material.

3. The method for batch data generation of an automotive acoustic package technical solution according to claim 2, characterized in that, S2 includes the following steps: S21: Determine the range of acoustic package design parameters, construct a parameter sampling space for mixed variable types, configure upper and lower limit ranges for continuous variables, establish a mapping dictionary between material types and integer index values ​​for discrete variables, and convert non-numerical material options into integer codes; S22: Perform optimal Latin hypercube sampling for multidimensional space filling within the parameter sampling space, perform hierarchical random sampling on the variable intervals of each dimension, and generate an original sampling matrix covering the high-dimensional parameter space. S23: Perform reverse parsing on the original sampling matrix, reverse map the integer codes back to the corresponding material types, serialize all parameter combinations into a structured data table, and output a standardized acoustic package design sample set according to the preset sampling scale.

4. The method for batch data generation of an automotive acoustic package technical solution according to claim 3, characterized in that, S3 includes the following steps: S31: Perform tree structure parsing on the acoustic package structure file, traverse the node levels of the acoustic package structure file, identify and map the names of sub-elements containing acoustic package design parameters, excitation conditions, acoustic constraints and solver settings and their corresponding attribute contents, and construct a parameter node index table; S32: Match each acoustic package design parameter in the acoustic package design sample set with the parameter node index table to locate the corresponding sub-element attribute in the acoustic package structure file; S33: Overwrite the values ​​of the acoustic package design parameters into the mapped sub-element attributes, thereby automatically completing the reconstruction of the parametric model and generating an updated acoustic package structure file corresponding to the acoustic package design sample set.

5. The method for batch data generation of an automotive acoustic package technical solution according to claim 4, characterized in that, In S33, by establishing a mapping relationship between the design parameters of each acoustic package in the design sample set and the parameter node index table, based on this mapping relationship, the values ​​of the design parameters of each acoustic package are batch overwritten into the corresponding node attributes in the acoustic package structure file, thereby generating multiple updated acoustic package structure files corresponding to the design sample set.

6. A method for batch generating data of an automotive acoustic package technical solution according to any one of claims 1-5, characterized in that, In S4, the simulation solver kernel supports importing acoustic package structure parameters in XML file format and performing acoustic response calculations.

7. A method for batch generating data of an automotive acoustic package technical solution according to any one of claims 1-5, characterized in that, S4 includes the following steps: S41: Based on acoustic simulation software, create a blank database file, import the updated acoustic package structure file into the blank database file, and generate an executable calculation file that corresponds one-to-one with each updated acoustic package structure file. The calculation file contains complete model parameters and simulation parameters, and each calculation file corresponds to a design scheme. S42: Call the acoustic simulation software kernel, submit the calculation task to the calculation file, and obtain the in-vehicle acoustic cavity noise response corresponding to each design scheme; S43: Export the noise response results of the in-vehicle acoustic cavity and store them in a structured format in a preset HDF5 data file.

8. The method for batch data generation of an automotive acoustic package technical solution according to claim 7, characterized in that, S5 includes the following steps: S51: The HDF5 data file is parsed using a script to extract the in-vehicle acoustic cavity noise response results for each design scheme, and the results are converted from sound pressure physical quantities into sound pressure level characteristics. S52: Calculate the sound transfer function and the reduction in overall vehicle noise based on the noise response results; S53: Associate and map each design scheme with its corresponding acoustic transfer function and the reduction in vehicle noise, and write it into a pre-set data sample table in a structured manner to complete the batch collection of the vehicle acoustic performance of different design schemes and obtain the simulation dataset.

9. The method for batch generating data of an automotive acoustic package technical solution according to claim 8, characterized in that, The following operations are performed in the preset script sequence: reading the design sample set, overwriting the numerical values ​​of the acoustic package design parameters, creating a blank database and importing the updated acoustic package structure file, submitting the calculation task and storing the calculation results.

10. A system for batch generating data of automotive acoustic package technology solutions, used to implement the method described in any one of claims 1-9, characterized in that, include: The parameter space definition module is used to construct the performance hierarchy decomposition tree of the acoustic package and define the parameter space, generating parameter space data that includes the value range and variable types of the acoustic package design parameters. The experimental design sampling module is used to communicate with the parameter space definition module, receive the parameter space data, and perform sampling based on the optimal Latin hypercube algorithm to generate a design sample set. The XML parsing and generation module is used to parse the acoustic package structure file and, based on the design sample set, batch overwrite parameters to generate multiple updated acoustic package structure files corresponding to each design scheme. The simulation batch processing module is used to communicate with the XML parsing and generation module, and call the acoustic simulation software kernel to sequentially import the updated acoustic package structure file and perform batch processing calculations, outputting an HDF5 file containing acoustic response results; The whole vehicle sound insulation performance calculation module is used to communicate with the simulation batch processing module, parse the HDF5 file, convert the sound pressure physical quantity into a sound pressure level characterization, and calculate the sound transfer function and the whole vehicle noise reduction. The dataset construction module is used to communicate with the vehicle sound insulation performance calculation module and the experimental design sampling module, and to associate and map the acoustic package design parameters with the corresponding acoustic performance indicators to construct a structured simulation dataset. The central scheduling module communicates with the parameter space definition module, the experimental design sampling module, the XML parsing and generation module, the simulation batch processing module, the whole vehicle sound insulation performance calculation module, and the dataset construction module, and performs timing and data interaction.

Images