Method, system and equipment for extracting wheel center force for whole vehicle road noise simulation

Abstract

The invention provides a wheel center force extraction method, system and device for whole vehicle road noise simulation, and the method comprises the steps: building a vibration transmission relation between a wheel center excitation point and a steering knuckle response region based on a whole vehicle NVH finite element model, and determining frequency domain transmission data; collecting dynamic response signals of a plurality of steering knuckle response areas and road noise synchronization signals of a preset reference position in a time-sharing manner when the vehicle runs on an actual road surface, and performing phase alignment on frequency domain responses in different time periods according to the road noise synchronization signals to obtain a target frequency domain response matrix; determining a frequency domain excitation vector of the wheel center excitation point according to the frequency domain transmission data and the target frequency domain response matrix; and combining the frequency domain transmission data to reconstruct a theoretical response of a steering knuckle response region, and generating a target wheel center force. According to the method, the wheel center force suitable for high-frequency-band whole vehicle road noise simulation can be rapidly and accurately extracted with low cost on the premise of not depending on a high-cost sensor and a special test bench, so that optimization design of NVH (Noise Vibration and Harshness) performance of a new energy vehicle is supported.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of vehicle noise, vibration and harshness (NVH) analysis, and in particular to a wheel hub force extraction method, system and device for vehicle road noise simulation. BACKGROUND

[0002] With the popularization of new energy vehicles, the masking effect of noise and vibration generated by traditional internal combustion engines disappears, and the problem of structure-borne road noise caused by road roughness is increasingly prominent. Wheel hub force, as a key input parameter for vehicle road noise simulation, faces multiple technical challenges. Existing six-component force sensors are costly and difficult to apply on a large scale, and their upper frequency response limit is usually lower than 80 Hz, which cannot meet the demand for road noise analysis above 300 Hz. The method of bench testing vibration transfer function has a complicated process, a long cycle and strong dependence on equipment. Full-vehicle multi-point synchronous acquisition requires up to 16 acceleration sensors, which is complex in wiring and difficult in data management, and lacks an effective error monitoring mechanism to ensure data quality. In addition, the lack of a unified phase reference in the time-sharing acquisition process easily leads to distortion in data splicing in different time periods, affecting the accuracy of wheel hub force inversion. SUMMARY

[0003] The purpose of the present application is to provide a wheel hub force extraction method, system and device for vehicle road noise simulation, which can quickly, accurately and cost-effectively extract wheel hub force suitable for high-frequency vehicle road noise simulation without relying on high-cost sensors and special test benches, to support the optimization design of new energy vehicle NVH performance.

[0004] In a first aspect, the present application provides a wheel hub force extraction method for vehicle road noise simulation, comprising: Based on the vehicle NVH finite element model, a vibration transfer relationship between the wheel hub excitation point and the knuckle response area is constructed, and frequency domain transfer data is determined based on the vibration transfer relationship; Time-sharing acquisition of dynamic response signals of multiple knuckle response areas and road noise synchronous signals of a preset reference position when the vehicle is running on an actual road, and phase alignment of frequency domain responses in different time periods according to the road noise synchronous signals to obtain a target frequency domain response matrix; According to the frequency domain transfer data and the target frequency domain response matrix, a frequency domain excitation vector of the wheel hub excitation point is determined; Based on the frequency domain excitation vector and the frequency domain transfer data, the theoretical response of the knuckle response area is reconstructed, and the target wheel hub force for vehicle road noise simulation analysis is generated based on the theoretical response.

[0005] In an optional embodiment, time-sharing acquisition of dynamic response signals of multiple knuckle response areas and road noise synchronous signals of a preset reference position when the vehicle is running on an actual road comprises: collecting dynamic response signals of response regions of the plurality of steering knuckles in response to the vehicle driving on the actual road surface in a time-sharing manner based on a plurality of sensing units arranged on structures where the steering knuckles are located, wherein the spatial distribution of the sensing units satisfies a non-coplanar constraint condition; collecting the road noise synchronous signals based on a common sensing unit arranged at a preset reference position selected inside the vehicle.

[0006] In an optional implementation, the dynamic response signals of the response regions of the plurality of steering knuckles in response to the vehicle driving on the actual road surface are collected in a time-sharing manner, including: dividing all the wheels into a plurality of test groups, and sequentially obtaining dynamic response signals of the response regions of the steering knuckles corresponding to each test group; synchronously collecting the dynamic response signals of the response regions of the steering knuckles included in the same test group, and realizing data correlation between different test groups through the common sensing unit.

[0007] In an optional implementation, the target frequency domain response matrix is obtained by phase alignment of frequency domain responses of different time periods according to the road noise synchronous signals, including: establishing a mapping relationship between excitation points and response points in the simulation model and data channels in actual measurement; unifying data of different sources to the same identification system by using the mapping relationship; performing Fourier transform on the dynamic response signals collected by each test group and the corresponding road noise synchronous signals, respectively, to obtain frequency domain response subsets of each test group and frequency domain synchronous reference data; calculating cross-group phase deviations based on the phase relationship between the frequency domain synchronous reference data of each test group, and performing phase compensation processing on the corresponding frequency domain response subsets; combining the frequency domain response subsets after phase compensation to obtain the target frequency domain response matrix.

[0008] In an optional implementation, the frequency domain excitation vector of the wheel hub excitation point is determined according to the frequency domain transfer data and the target frequency domain response matrix, including: constructing an excitation-response equation set according to the frequency domain transfer data and the target frequency domain response matrix; performing pseudo-inverse operation on the coefficient matrix in the excitation-response equation set to obtain the frequency domain excitation vector.

[0009] In an optional implementation, the theoretical response of the response region of the steering knuckle is reconstructed based on the frequency domain excitation vector and the frequency domain transfer data, and the target wheel hub force for vehicle road noise simulation analysis is generated based on the theoretical response and the measured response, including: inputting the frequency domain excitation vector into a forward propagation model represented by the vibration transfer relationship; computing expected dynamic responses of each knuckle response area under the same working condition, and generating a theoretical response in a corresponding frequency domain; comparing the theoretical response with the measured response for consistency, and generating a data reliability evaluation result of each knuckle response area; screening a target frequency domain excitation component according to the data reliability evaluation result, and determining a target wheel center force of each wheel center position for vehicle road noise simulation analysis based on the target frequency domain excitation vector.

[0010] In an optional implementation, after the target wheel center force for vehicle road noise simulation analysis is generated based on the theoretical response, the method further includes: encapsulating the target wheel center force into an excitation input file in a standardized format according to a predefined data interface specification; loading the excitation input file into a vehicle road noise simulation environment automatically to determine a boundary condition for vibration noise performance prediction.

[0011] In a second aspect, the present application provides a wheel center force extraction system for vehicle road noise simulation, comprising: a construction unit configured to construct a vibration transmission relationship between a wheel center excitation point and a knuckle response area based on a vehicle NVH finite element model, and determine frequency domain transmission data based on the vibration transmission relationship; a signal processing unit configured to collect dynamic response signals of multiple knuckle response areas and a road noise synchronous signal of a preset reference position when a vehicle is running on an actual road in different time periods, and perform phase alignment on frequency domain responses in different time periods to obtain a target frequency domain response matrix according to the road noise synchronous signal; a determination unit configured to determine a frequency domain excitation vector of the wheel center excitation point according to the frequency domain transmission data and the target frequency domain response matrix; a generation unit configured to reconstruct a theoretical response of the knuckle response area based on the frequency domain excitation vector and the frequency domain transmission data, and generate a target wheel center force for vehicle road noise simulation analysis based on the theoretical response and a measured response.

[0012] In a third aspect, the present application provides an electronic device comprising a processor and a memory, wherein the memory stores computer executable instructions capable of being executed by the processor, and the processor executes the computer executable instructions to implement the wheel center force extraction method for vehicle road noise simulation according to any one of the preceding embodiments.

[0013] In a fourth aspect, the present application provides a computer readable storage medium storing computer executable instructions, which, when invoked and executed by a processor, cause the processor to implement the wheel hub force extraction method for vehicle road noise simulation according to any one of the preceding embodiments.

[0014] The wheel hub force extraction method, system and device for vehicle road noise simulation provided by the present application avoid relying on measured transfer functions by constructing the vibration transmission relationship between the wheel hub excitation point and the knuckle response area, solve the problems of complicated bench test process and long cycle; based on finite element simulation to obtain frequency domain transfer data, break through the limitation of low upper limit of frequency of six-component force sensor, support road noise analysis of high frequency band above 300Hz. The dynamic response signal is collected in time sharing mode, and the road noise synchronization signal of the preset reference position is introduced for phase alignment, which effectively reduces the number of sensors and solves the problems of complex traditional full-channel synchronous wiring and high cost. Cross-period data fusion is realized through phase compensation, which improves the accuracy of data splicing under time-sharing collection. The wheel hub excitation vector is inverted combined with the target frequency domain response matrix and the transfer data, and the theoretical response is reconstructed for error evaluation, realizing the identification and quality control of abnormal data, and improving the reliability and engineering practicability of wheel hub force extraction. BRIEF DESCRIPTION OF DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the specific embodiments or prior art of the present application, the drawings needed in the description of the specific embodiments or prior art will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0016] Figure 1 A flowchart of a wheel hub force extraction method for vehicle road noise simulation provided by an embodiment of the present application; Figure 2 A schematic diagram of acceleration sensor arrangement on a knuckle provided by an embodiment of the present application; Figure 3 A wheel hub force error evaluation schematic diagram provided by an embodiment of the present application; Figure 4 A mapping representation diagram of analysis and test provided by an embodiment of the present application; Figure 5 A schematic diagram of vibration transfer function analysis result file content provided by an embodiment of the present application; Figure 6 A schematic diagram of acceleration response test result data file content provided by an embodiment of the present application; Figure 7A flow chart of a specific wheel hub force extraction method for vehicle road noise simulation provided by the embodiment of the present application is provided. Figure 8 A structural diagram of a wheel hub force extraction system for vehicle road noise simulation provided by the embodiment of the present application is provided. Figure 9 A structural diagram of an electronic device provided by the embodiment of the present application is provided. DETAILED DESCRIPTION

[0017] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations.

[0018] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without making creative efforts fall within the scope of the present application.

[0019] It should be noted that: similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0020] The embodiment of the present application provides a wheel hub force extraction method for vehicle road noise simulation, as shown in the figure, which mainly comprises the following steps: Figure 1 S110, based on a vehicle NVH finite element model, a vibration transmission relationship between a wheel hub excitation point and a steering knuckle response area is constructed, and frequency domain transmission data is determined based on the vibration transmission relationship. S110, based on a vehicle NVH finite element model, a vibration transmission relationship between a wheel hub excitation point and a steering knuckle response area is constructed, and frequency domain transmission data is determined based on the vibration transmission relationship.

[0021] The above-mentioned vehicle NVH finite element model refers to a structural finite element model containing a vehicle body, a subframe, a suspension system and connecting parts, that is, a vehicle NVH finite element model without tires, which is used to simulate the vibration response behavior of the vehicle under dynamic excitation; the wheel hub excitation point is a virtual node at the assembly center of the tire and the wheel, which is usually located at the end of the main shaft or the suspension hard point position; the steering knuckle response area refers to the spatial distribution area where a plurality of acceleration sensors are installed on the steering knuckle body, and the physical position has a non-coplanar characteristic to capture multi-dimensional vibration modes. The vibration transmission relationship specifically represents the frequency response function (FRF) from the wheel hub to each sensor arrangement point under the action of unit excitation force, that is, the complex ratio of acceleration / force, which constitutes the input-output mapping of the linear system in the frequency domain.

[0022] In the specific implementation, a complete vehicle NVH finite element model is established in a pre-processing software such as HyperMesh, and flexible tire modeling is excluded to simplify the calculation. A unit force (1N) excitation is respectively applied to the wheel center position corresponding to each wheel in X, Y and Z directions, and a frequency response analysis is performed through an Optistruct or Nastran solver, and the frequency range is set to 0-400 Hz with a resolution of not less than 1 Hz. The acceleration response results of the preset sensor positions on the steering knuckle under each excitation condition are extracted, as shown in FIG. 21, which is located near the steering rod mounting hole, 22 is located at the front end ear of the steering knuckle, 23 is located at the brake disc mounting bolt position, and 24 is located at the upper end groove of the steering knuckle. The acceleration response results are organized in matrix form according to the excitation source and the response channel, and output as a PCH format file to form the above frequency domain transfer data as the basic transfer path information for subsequent inversion calculation. Figure 2

[0023] In S120, dynamic response signals of multiple steering knuckle response regions and synchronous road noise signals of a preset reference position are collected at different times when the vehicle is driving on an actual road, and the frequency domain responses of different time periods are phase-aligned according to the synchronous road noise signals to obtain a target frequency domain response matrix.

[0024] The time-sharing collection in the embodiment refers to dividing the four wheels into several test groups (for example, one group for the front wheels and one group for the rear wheels, or testing each tire separately), and sequentially completing the real vehicle road test of each group at different time periods; the dynamic response signal is a time domain vibration signal collected by a three-axis acceleration sensor, and in actual application, the sampling frequency can be set to not less than 1024 Hz, so as to ensure covering the target frequency band. The above preset reference position is a fixed measuring point on the rigid structure in the vehicle, such as a cross beam or a seat mounting point, which is used to deploy a common acceleration sensor to obtain a synchronous reference signal. The phase alignment is to use the cross spectrum phase difference of the reference signal between different test periods to compensate the frequency domain of each group of response data, and eliminate the phase mismatch caused by time misalignment.

[0025] In the specific implementation, four three-axis acceleration sensors can be installed on the steering knuckles of the current test group, and a fifth (or ninth) sensor can be fixed at the reference position in the vehicle to collect a synchronous signal. The vehicle passes through the same rough road at the same speed, and records the acceleration time domain data of all channels and saves them as UFF format files. Fourier transform is performed on each group of data to obtain a frequency domain response subset, and a frequency domain phase sequence of the reference channel is extracted. Taking a group as a reference, the phase deviation of the remaining groups at the same frequency is calculated, and the phase compensation of the response data of the group is performed. Finally, all the corrected response subsets are spliced in the order of wheel positions to form a complete target frequency domain response matrix.

[0026] ​S130, determining the frequency domain excitation vector of the wheel center excitation point according to the frequency domain transfer data and the target frequency domain response matrix.

[0027] The frequency domain transfer data is used to characterize the system forward transfer characteristics, and the target frequency domain response matrix is the measured response set after phase alignment. The frequency domain excitation vector is used to characterize the six-direction or three-direction excitation force spectrum of each wheel center position in the frequency domain.

[0028] In an embodiment, the frequency domain transfer matrix output by S110 and the target response matrix generated by S120 are introduced into a Python calculation program, and an excitation-response equation is constructed at each frequency point. Since the frequency domain transfer matrix is a thin matrix and may be ill-conditioned, the least squares pseudo-inverse method is used for solution. By traversing the entire frequency range, the excitation component of each wheel center is calculated at each frequency point, and finally a complete frequency domain excitation vector is synthesized as the preliminary identification result of the wheel center force.

[0029] S140, reconstructing the theoretical response of the knuckle response area based on the frequency domain excitation vector and the frequency domain transfer data, and generating a target wheel center force for vehicle road noise simulation analysis based on the theoretical response.

[0030] The reconstructed theoretical response refers to re-entering the inverted excitation vector into the forward transfer model to calculate the response value that should be generated, which is used for comparison and verification with the measured value. The target wheel center force is high-credibility excitation data after error evaluation and screening, which can be used for subsequent vehicle road noise simulation boundary condition loading.

[0031] In specific implementation, the frequency domain excitation vector obtained as described above can be substituted into the forward propagation model to calculate the theoretical acceleration response of each knuckle response point. The calculated value is compared with the measured response in S120 channel by channel, and the decibel error of the amplitude ratio is calculated, as shown in Figure 3 the wheel center force error evaluation diagram. In the diagram, the X-axis is the channel name of each acceleration sensor in different directions, and the Y-axis is the error size G_b between the calculated value and the test value. If the error of two directions of one sensor exceeds 0.4, it is determined that the channel data is abnormal, triggering a data replacement or retest prompt, and the data of another sensor needs to be replaced to re-calculate the wheel center force. If it still cannot meet the requirements, the data needs to be retested and collected.

[0032] Through the closed-loop verification mechanism, reliable excitation components are screened out to generate the final target wheel center force, and a standard format excitation card file is automatically generated according to the mapping table information (see Figure 4 ), which is used for vehicle road noise simulation analysis and calling.

[0033] In summary, by constructing the vibration transmission relationship from the wheel center to the knuckle, combining finite element simulation with real-time measurement data, using reference signals to align the phase of frequency domain responses, the number of sensors is effectively reduced and the test process is simplified. Based on the inverse matrix method to solve the wheel center force, and through the error evaluation of the reconstructed response, the calculation accuracy and reliability are improved. The wheel center force extraction for vehicle road noise simulation is fast, low cost and high precision, which is significantly better than the traditional bench test and multi-sensor synchronous acquisition method.

[0034] For ease of understanding, the wheel center force extraction method for vehicle road noise simulation provided by the embodiments of the present application is described in detail below.

[0035] In an optional implementation, the above-mentioned time-sharing acquisition of dynamic response signals of multiple knuckle response regions when the vehicle is running on an actual road and the road noise synchronous signal of the preset reference position can include the following steps 1-1 to 1-2: Step 1-1, based on the multiple sensing units arranged on the structure of each knuckle of the vehicle, time-sharing acquisition of dynamic response signals of multiple knuckle response regions when the vehicle is running on an actual road, wherein the spatial distribution of the sensing units satisfies the non-coplanar constraint condition.

[0036] The above-mentioned multiple sensing units refer to three-axis acceleration sensors used to collect vibration acceleration signals. Each sensor has three independent measurement channels in orthogonal directions (X, Y, Z), which correspond to the longitudinal direction (forward direction), lateral direction (left-right direction) and vertical direction (up-down direction) in the vehicle coordinate system. Through three-axis measurement, the full-degree-of-freedom dynamic response of the knuckle region under complex road excitation can be completely captured.

[0037] The dynamic response signal refers to the acceleration time-domain signal caused by road roughness excitation during vehicle running, which is transmitted to the wheel center through the tire and further conducted to the knuckle structure. This signal contains rich frequency components, especially covering the medium and high frequency bands (usually 300-400 Hz and above) concerned in vehicle road noise analysis, which is the key input data for subsequent inversion of wheel center force.

[0038] The spatial distribution satisfying the non-coplanar constraint condition means that the installation positions of multiple sensing units (preferably 4 or 5) corresponding to the same wheel are not located in the same geometric plane in three-dimensional space. This arrangement principle aims to improve the independence of vibration modal identification and the stability of matrix solution, avoiding rank deficiency of the transfer function matrix due to sensor layout degradation, thereby improving the accuracy of wheel center force inversion.

[0039] In practical application, at least four triaxial acceleration sensors are arranged on each wheel, on its corresponding knuckle and its direct connection structure. Preferably, four sensors are configured for each of the front left, front right, rear left and rear right wheels, totaling 16 sensors. However, in actual testing, in order to reduce equipment cost and wiring complexity, a grouping test strategy can be used: only one or a pair of wheels (such as a front wheel group or a single wheel) is installed with all sensors at a certain time, and the remaining wheels are not installed or only synchronous reference sensors are retained.

[0040] During the acquisition process, the vehicle can drive at a constant speed (such as 60 km / h or 80 km / h) through typical rough road surfaces (such as Belgian roads, gravel roads, asphalt random road surfaces, etc.), and record the acceleration time domain signals output by each sensor, with a sampling frequency not less than 2048 Hz to fully cover the target frequency band (≥400 Hz) and meet the Nyquist sampling theorem. The collected data is stored in a standard universal file format (UFF, Universal File Format), and each channel contains channel identification, direction information, unit (m / s 2 ), sampling time interval and complete time domain waveform data (see Figure 6 Examples).

[0041] Step 1-2, the common sensor unit deployed based on the preset reference position selected inside the vehicle acquires the road noise synchronization signal.

[0042] The preset reference position selected inside the vehicle refers to a position on the main body structure of the vehicle body that has high stiffness and is far from local vibration sources, and is used to deploy a common acceleration sensor as a global synchronization reference. This preset reference position can be arranged inside the vehicle, such as a seat rail fixed point. In addition, it can also be arranged at the central channel reinforcement, the lower node of the A-pillar, etc. These positions exhibit overall response in the vehicle NVH behavior and are less affected by local suspension vibration, making them suitable as time synchronization and phase alignment references for cross-group testing.

[0043] The common sensor unit refers to a triaxial acceleration sensor, which has a different function from the local response sensor on the knuckle, and is mainly used to acquire the road noise synchronization signal throughout the entire test process. This signal refers to the overall vibration acceleration signal of the vehicle body transmitted through the chassis caused by road excitation when the vehicle is driving on the same actual road. Since this signal is derived from the same external excitation source (i.e. road profile), even if different wheels are tested in different groups at different times, as long as the vehicle passes through the same road segment at a similar speed, the signal should have high similarity in the time domain. Therefore, it can be used as a time alignment anchor point between multiple batches of test data.

[0044] Before each test, a triaxial acceleration sensor is firmly installed at the above-mentioned preset reference position (for example, the left rail mounting bolt of the front seat) and is always connected throughout the entire group test process. The sensor uses the same data acquisition system as all the steering knuckle sensors to ensure uniform timestamps.

[0045] During the group test (for example, the front wheels are tested first, and the rear wheels are tested later), the triaxial acceleration signals of the common sensor are synchronously collected for each test group. In subsequent data processing, the common signal is used as a reference to correct the time synchronization of the steering knuckle response data collected at different time periods through cross-correlation analysis or phase alignment algorithm, eliminating the misalignment caused by start-up delay, GPS clock drift, and other factors.

[0046] The design of the common sensor unit significantly reduces the dependence on a 16-channel synchronous acquisition system, allowing only 5 sensors (4 for the current wheels + 1 for the vehicle) to complete the four-wheel hub force extraction, greatly reducing the test cost and operation difficulty while ensuring data consistency and inversion accuracy.

[0047] Further, the dynamic response signals of multiple steering knuckle response regions when the vehicle is driving on actual road surfaces can include the following steps 2-1 to 2-2: Step 2-1, divide all the wheels into several test groups, and sequentially obtain the dynamic response signals of the steering knuckle response regions corresponding to each test group.

[0048] The above division into several test groups means that according to the actual test resources (such as the number of sensors, the number of acquisition channels, labor hours, etc.), the originally required 16 triaxial acceleration signals (4 wheels x 4 sensors per wheel x 3 directions) are simplified into a combination measurement strategy of time division and regional division. This division aims to reduce the number of sensors required for a single test, thereby reducing equipment costs and on-site wiring complexity.

[0049] Specifically, the "test group" can adopt two typical division modes: Double-group division mode: divide the four wheels into two test groups, "front wheel group" (including front left and front right) and "rear wheel group" (including rear left and rear right), according to the front and rear axles. During each test, 4 triaxial acceleration sensors are arranged on each wheel in the current group, a total of 8 sensors; plus the in-vehicle common synchronization sensor, a total of 9 sensors can complete the entire vehicle test.

[0050] Four groups division mode: divide four wheels into four independent test groups in single wheel unit. Test dynamic response of steering knuckle area corresponding to one wheel at a time, i.e. use 4 sensors to collect data around one wheel at a time. Cooperate with in-vehicle shared sensors, the whole test process can be completed with only 5 physical sensors (4 for current wheel + 1 in-vehicle shared).

[0051] In specific implementation, first determine the test route and standard driving conditions (such as uniform speed of 60 km / h through Belgium road). Then perform the collection task of each test group in the preset order: if two groups are divided, first install the sensors on the steering knuckles of the front wheels (8 sensors in total), run a complete road test, then remove the front wheel sensors and move them to the rear wheels, run the same road test again, keep the in-vehicle shared sensors in continuous working state during each test to ensure consistent time reference; if four groups are divided, test each tire one by one: for example, first test the front left wheel, arrange 4 sensors on its corresponding steering knuckle, and do not install on the rest; after completing one pass, change to the front right wheel, and so on.

[0052] To ensure the comparability of data between groups, all tests are completed under the same environmental conditions as far as possible (such as road surface condition, air temperature, tire pressure, vehicle speed control accuracy, etc.), and GPS time stamp or external trigger signal is recorded during the test process to facilitate data alignment in later stage.

[0053] Step 2-2, in the same test group, synchronously collect the dynamic response signals of the steering knuckle response area included in the test group, and realize data correlation between different test groups through the shared sensor unit.

[0054] The same test group refers to a group of steering knuckle response areas participating in data collection at the same time in one test run. For example, the "front wheel group" in two-group division includes the front left and front right steering knuckle areas; in four-group division, "single wheel" is a minimum test unit, which includes 4 sensor arrangement points corresponding to the wheel.

[0055] In specific implementation, all sensors are connected to the same multi-channel dynamic signal acquisition instrument (such as LMS SCADAS, Siemens SCOUT or NI DAQ system), and are triggered and started to collect by the main control module. The collection software sets uniform engineering name, channel label (according to Figure 6 format), direction definition (X / Y / Z), unit (m / s 2 ) and file output format (UFF). After starting collection, the system automatically records the complete time domain waveform of each channel, and embeds accurate timestamp information.

[0056] The data association between different test groups through the shared sensing unit refers to using the shared triaxial accelerometer deployed in the preset reference position in the vehicle in steps 1-2 as the time and excitation consistency benchmark for cross-group testing, and establishing the mapping relationship between the data of each batch.

[0057] Because the data from different test groups were collected at different time periods, even though the vehicles were traveling on the same road surface, there may still be slight speed fluctuations, path deviations, or starting delays, resulting in phase differences in the overall vibration signal. Therefore, test data from different groups cannot be directly spliced ​​together and used.

[0058] In addition, in frequency domain modeling, if a test group only contains some wheel center excitations (such as only testing the front left wheel), then only the corresponding block is retained in the transfer function matrix, and the rest are set to zero or interpolated. The coherence of the test group with other groups is verified by using a shared reference signal (generally, the coherence coefficient is required to be >0.8).

[0059] This data association mechanism effectively solves the spatiotemporal mismatch problem caused by time-sharing testing, enabling the reconstruction of the complete four-wheel center force frequency response characteristics even with very few sensors (such as 5), demonstrating the outstanding advantages of this invention in data fusion capability and engineering practicality.

[0060] This method rationally divides the test units and makes full use of the synchronous anchoring effect of the shared sensing units, which greatly saves test resources and ensures the accuracy and reliability of the wheel center force inversion results, providing an efficient and feasible technical path for road noise simulation of new energy vehicles.

[0061] Furthermore, the process of obtaining the target frequency domain response matrix by phase alignment of the frequency domain responses at different time periods based on the road noise synchronization signal may include the following steps 3-1 to 3-5: Step 3-1: Establish the mapping relationship between the excitation points and response points in the simulation model and the measured data channels.

[0062] In the simulation model, the excitation point refers to the key node location used to apply unit force to calculate the vibration transmission path in the whole vehicle NVH finite element analysis model, specifically corresponding to the virtual wheel center (or main shaft center point) of each wheel. The response point refers to the location used to output the acceleration response in the finite element model, specifically corresponding to the physical location on the actual vehicle where a triaxial acceleration sensor is installed, including the front lug of the steering knuckle, near the tie rod mounting hole, at the brake disc bolt, and the upper groove, etc. (see [reference]). Figure 2 ).

[0063] The measured data channel refers to the unique identification information of the acceleration signal collected and recorded by the sensor in the actual test process, which contains three parts: channel name (Channel Label), direction (Direction) and unit (m / s 2 ). For example, "FL_Knuckle_Top_Z" represents the Z-direction (vertical) acceleration signal of the front left knuckle top sensor.

[0064] The mapping relationship refers to establishing a one-to-one correspondence table of the above three types of information: simulation excitation point (wheel center node ID), simulation response point (knuckle node ID) and measured response channel (test label), to ensure that the simulation results and measured data are completely matched in spatial position and physical meaning.

[0065] In specific implementation, a structured analysis and test mapping table (Mapping Table) can be constructed, and the mapping table is shown in Figure 4 , which represents the intention of analysis and test mapping, wherein ExcitationPointID represents the wheel center node ID number, KnuckleResponsePointID represents the knuckle sensor node ID number in the whole vehicle NVH finite element model, MappingKnuckleLabels represents the channel name of the knuckle sensor response data in the test data, and ReferenceKnuckleLabelsDir represents the channel name and direction of the in-vehicle sensor response data.

[0066] The role of the mapping table is: 1) to ensure that the acceleration sensor is consistent in position in the whole vehicle NVH finite element model and the actual vehicle, which is set through the wheel center node ID number in the ExcitationPointID column; 2) the grouping of the wheel center and the sensor on each tire is corresponding, one wheel center has four or five sensors respectively, which is set through the ExcitationPointID column and the KnuckleResponsePointID column; 3) to mark the correspondence of the same response point in the vibration transfer function finite element analysis result file (see Figure 5 ) and the acceleration response test result data file (see Figure 6 ), which is set through the KnuckleResponsePointID column and the MappingKnuckleLabels column; 4) to mark the data channel name and direction for synchronization, which is set through the ReferenceKnuckleLabelsDir column.

[0067] Step 3-2, unify the data of different sources to the same identification system by using the mapping relationship.

[0068] The data of different sources mainly include two types: one is the vibration transfer function simulation data from the finite element solver (such as Nastran or Optistruct), and the identification system thereof is based on the node ID and the working condition naming rules of the finite element model (such as GRID ID + DOF in the PCH file); the other is the acceleration response measured data generated from the test system (such as LMS Test.Lab), and the identification system thereof depends on the channel name, DAQ number and direction code (UFF file format, see Figure 6 ). Since the original naming rules of the two are inconsistent, they cannot directly participate in matrix operation, and therefore, they must be unified to the same logical identification system through the mapping relationship established in step 3-1.

[0069] Step 3-3, Fourier transform is respectively performed on the dynamic response signals collected by each test group and the corresponding road noise synchronous signals to obtain the frequency domain response subsets of each test group and the frequency domain synchronous reference data.

[0070] In specific implementation, the collected data of each test group can be first preprocessed to remove the direct current offset, apply the Hann window to suppress leakage, and be divided into several equal-length time periods (such as 4 seconds each) for averaging processing; FFT operation is respectively performed on each three-axis acceleration channel; then the auto-power spectral density (Auto-PSD) and cross-power spectral density (Cross-PSD) matrices of each channel are calculated to constitute the frequency domain response subset of the test group; FFT is synchronously performed on the signals of the three directions of the shared sensor to obtain the corresponding frequency domain synchronous reference data of the group, which is used for subsequent cross-group phase alignment.

[0071] Step 3-4, based on the phase relationship between the frequency domain synchronous reference data of each test group, the cross-group phase deviation is calculated, and the corresponding frequency domain response subset is phase compensated.

[0072] The above cross-group phase deviation refers to the phase difference between the shared reference signals of each group under the same road excitation due to the different running times of different test groups. The deviation mainly comes from the slight delay of vehicle start, speed fluctuation or GPS clock error, and if not corrected, it will seriously affect the solving accuracy of the wheel hub moment matrix.

[0073] In specific implementation, the frequency spectrum of the shared reference signal of the first test group (such as the front left wheel) can be first selected as the reference; the cross spectrum of the shared reference signal of each of the remaining groups and the reference signal is calculated; the phase angle of the cross spectrum is extracted, and each element in the corresponding frequency domain response subset of the group is phase compensated to obtain the frequency domain response subset after phase correction.

[0074] Step 3-5, the frequency domain response subsets after phase compensation are combined to obtain the target frequency domain response matrix.

[0075] The phase-compensated frequency-domain response subsets refer to the local response matrices of each test group that have completed phase alignment, and each subset only covers the response channels corresponding to the measured wheel center.

[0076] Steps 3-1 to 3-5 effectively solve the time-space separation problem caused by grouping tests by establishing an accurate mapping, a unified identification system, and introducing a common reference signal phase correction mechanism, significantly improve the test flexibility and engineering practicability, and at the same time guarantee the accuracy and robustness of the wheel center force inversion results, fully embodying the high degree of technical innovation of the application in the field of intelligent testing and simulation collaboration.

[0077] Further, the determination of the frequency-domain excitation vector of the wheel center excitation point according to the frequency-domain transfer data and the target frequency-domain response matrix can include the following steps 4-1 to 4-2 in specific implementation: Step 4-1, constructing an excitation-response equation set according to the frequency-domain transfer data and the target frequency-domain response matrix.

[0078] The frequency-domain transfer data refers to the vibration transfer function matrix from the virtual wheel center excitation point to the knuckle acceleration sensor arrangement position obtained by simulation of the whole vehicle NVH finite element model, and the target frequency-domain response matrix refers to the actual measured acceleration response frequency-domain data obtained after mapping, unification, phase compensation and combination in step 3-5, each row of which corresponds to the power spectral density or average acceleration spectrum value of a sensor in a certain direction. The excitation-response equation set refers to the matrix equation formed by substituting the above two groups of data into the frequency-domain dynamics model of the linear system, which constitutes the core mathematical model of the wheel center force inversion.

[0079] In specific implementation, the program automatically reads the PCH file generated by Nastran or Optistruct, parses the target frequency-domain response matrix at each frequency point, matches the simulation nodes and measured data according to the channel name according to the mapping table as shown in Figure 4 , and ensures that the data are completely aligned in the spatial dimension; for each analysis frequency (such as scanning point by point from 1 Hz to 400 Hz), the corresponding complex linear equation set is constructed, and the equation set is stored in the memory in the form of a standard matrix for calling by the pseudo-inverse operation in the next step.

[0080] Step 4-2 performs pseudo-inverse operation on the coefficient matrix in the excitation-response equation set to obtain the frequency-domain excitation vector.

[0081] At each frequency point, pseudo-inverse operation is performed on the current coefficient matrix, and the pseudo-inverse matrix is multiplied by the target response vector of the current frequency point to obtain the wheel center force vector (i.e. the frequency domain excitation vector) at the frequency. The entire target frequency band (such as 1-400 Hz) is sequentially traversed to obtain the force components (FX, FY, FZ, RX, RY, RZ) of each wheel in six directions, and then the high-frequency wheel center excitation force is obtained.

[0082] Further, the above reconstructs the theoretical response of the knuckle response area based on the frequency domain excitation vector and the frequency domain transfer data, generates the target wheel center force for the whole vehicle road noise simulation analysis based on the theoretical response and the measured response, and in specific implementation, can include the following steps 5-1 to 5-4: Step 5-1, input the frequency domain excitation vector into the forward propagation model represented by the vibration transfer relationship.

[0083] The frequency domain excitation vector refers to the input force estimation value set of the four wheel centers obtained by pseudo-inverse operation in step 4-2, which includes the complex force components (unit: N) of each wheel in X (longitudinal), Y (lateral), Z (vertical), RX (rotation around X axis), RY (rotation around Y axis), and RZ (rotation around Z axis) six directions, and the frequency range covers 0-400 Hz.

[0084] The forward propagation model represented by the vibration transfer relationship refers to a linear time-invariant system constructed based on the whole vehicle NVH finite element analysis, which is used to represent the dynamic response path between the wheel center excitation point and the knuckle sensor arrangement position.

[0085] Step 5-2, calculate the expected dynamic response of each knuckle response area under the same working condition, and generate the theoretical response in the corresponding frequency domain.

[0086] The expected dynamic response of each knuckle response area under the same working condition refers to the acceleration response signal at each sensor installation position predicted by the forward propagation model under the premise of inputting the wheel center force obtained by inversion, also known as "theoretical response" or "reconstructed response". Generating the theoretical response in the corresponding frequency domain refers to organizing the data calculated in step 5-1 into a standard frequency domain data structure, and clearly labeling the information such as channel name, direction, unit, etc. so as to directly compare with the measured response.

[0087] Step 5-3, compare the consistency of the theoretical response and the measured response, and generate the data reliability evaluation result of each knuckle response area.

[0088] The deviation between the two is quantified by comparing the generated forward prediction response (i.e. theoretical response) with the actual target frequency domain response matrix obtained by acquisition and processing, in terms of consistency (i.e. calculating the relative error (decibel difference) of the two in the frequency domain).

[0089] The data reliability evaluation result refers to the channel-level error score, abnormality label and recommended processing opinion, which is used for subsequent fine screening of wheel forces.

[0090] Step 5-4: Screening target frequency domain excitation components according to the data reliability evaluation result, and determining target wheel forces for vehicle road noise simulation analysis based on the target frequency domain excitation vector.

[0091] Screening target frequency domain excitation components is to identify and eliminate wheel force components (such as a certain wheel Z-direction force distorted due to the failure of the top sensor) greatly affected by abnormal sensors according to the reliability evaluation result, and to retain excitation components corresponding to high-quality data. Determining target wheel forces for vehicle road noise simulation analysis is to encapsulate the screened excitation vector into a standardized input file for NVH simulation.

[0092] Further, after generating the target wheel forces for vehicle road noise simulation analysis based on the theoretical response, the following steps 6-1 to 6-2 are further included: Step 6-1: Encapsulating the target wheel forces into an excitation input file in a standardized format according to a pre-defined data interface specification.

[0093] The pre-defined data interface specification refers to a set of unified data organization rules and technical standards formulated for seamless transmission of wheel force data between different software platforms. The specification covers key elements such as file format, field naming, coordinate system definition, unit system, frequency accuracy, complex number representation form and metadata information, ensuring that the output results have good compatibility and traceability. Among them, the file format such as PCH format (Nastran Punch File), UFF format, and optional CSV / TXT text format for debugging and manual review.

[0094] The target wheel forces generated by the present application are used as the subsequent vehicle road noise simulation structural excitation source, which are applied to the wheel hub connection nodes of the body finite element model, constituting the key boundary conditions for vibration and noise performance prediction. Wheel forces are not the final noise results, but input conditions for driving simulation.

[0095] Step 6-2: Automatically loading the excitation input file into the vehicle road noise simulation environment to determine the boundary conditions for vibration and noise performance prediction.

[0096] The whole vehicle road noise simulation environment can include Optistruct or Nastran, and the excitation file is imported into the simulation environment without manual intervention through automatic loading, that is, through scripting or API interface, and is bound to the correct wheel hub node and degree of freedom to complete the boundary condition setting.

[0097] Specifically, after solving the wheel hub excitation vector and verifying the data consistency, the determined target wheel hub force is expressed in the frequency domain, including the six force components (three translational direction forces and three rotational direction moments) at each wheel hub position, and a corresponding loading card is generated according to a preset data interface format. The loading card is imported into the whole vehicle NVH finite element model or hybrid test simulation environment as an external excitation source applied to the wheel hub connection node. Since the excitation reflects the real vibration energy transmitted to the vehicle body through the suspension system under actual road conditions, it constitutes the core boundary condition for structure-borne noise analysis in simulation.

[0098] This way defines a standardized data interface, encapsulates high-reliability wheel hub forces, and automatically loads them into mainstream simulation environments, so that the extracted wheel hub forces can be used for simulation input. This not only greatly shortens the NVH development cycle, but also improves the accuracy and consistency of road noise prediction.

[0099] Figure 7 A flowchart of a specific whole vehicle road noise simulation wheel hub force extraction method is shown, including the following steps: S1, a whole vehicle NVH analysis model is established, and the vibration transfer function of each tire wheel hub to the steering knuckle is calculated.

[0100] The corresponding vibration transfer function calculation formula is: = , where is the transfer function between k points and j points, is the excitation force at k points, is the acceleration response at j points. The acceleration response of j points (i.e. the acceleration sensor) is output after loading a unit force at k points (i.e. the wheel hub) on the whole vehicle NVH finite element analysis model, and the results are shown in Figure 5 The vibration transfer function analysis result file content diagram in is shown in the figure, and the content format is PCH, which is automatically generated by the finite element software Nastran or Optistruct. Each tire needs to be calculated, and at least 4 sensors need to be arranged for each tire; if the number of sensors is sufficient, it can be considered to increase to 5 to facilitate subsequent data screening. The new sensor position needs to be located on the steering knuckle or the same part near the wheel hub, and at the same time, it must be ensured that the positions of all sensors cannot be located in the same plane.

[0101] S2, whole vehicle road noise test is carried out to obtain acceleration response time domain data of the knuckle.

[0102] Acceleration response time domain data is shown in Figure 6 The content of the acceleration response test result data file is shown in the figure. The content of the test result data file is in the UFF (Universal File Format), which is used in the field of structural dynamic test and vibration, acoustic analysis, and is automatically generated by the test software LMS Test Lab. Figure 6 In the figure, 58 represents the data type, which is a dynamic characteristic parameter such as a transfer function, FL1 represents the sensor channel name, 2 represents the channel ID number, -3 represents the response direction, 22528 represents the number of response data, 3.04129e-04 represents the starting sampling time, 4.88281e-04 represents the sampling time interval, m / s^2 represents the response unit, and the subsequent detailed test response values are shown.

[0103] S3, video conversion and synchronization processing are performed on the knuckle time domain test data corresponding to each tire, and inverse matrix calculation is performed to output acceleration frequency domain data.

[0104] The obtained acceleration time domain data is converted into power spectral density frequency domain data through Fourier transform and power spectral density calculation formula, and then the corresponding relationship in the mapping table of analysis and test and the transfer function result obtained through simulation calculation at S1 are used to calculate the wheel center force according to Figure 4 The mapping of analysis and test and the calculation formula of the wheel center force are assembled into a matrix to calculate the wheel center force.

[0105] The calculation formula of the wheel center force in the frequency domain is as follows:

[0106] wherein, represents the wheel center force matrix, represents the power spectral density matrix of the knuckle sensor acceleration test data, represents the vibration transfer function matrix of the excitation force and acceleration response between the wheel center and the knuckle sensor, wherein the superscript + represents pseudo-inverse, and the superscript H represents Hermitian transpose.

[0107] In actual test, in view of 16 acceleration sensors, we often divide four tires into two groups (front and rear) or four groups (front left, front right, rear left and rear right) to detect acceleration response. When grouping into two groups, 9 sensors are needed (9 = 2 x 4 + 1, 2 represents that sensors are arranged on two tires at the same time, 4 represents that each tire corresponds to 4 acceleration sensors, and 1 represents that 1 acceleration sensor is used for in-vehicle test data synchronization), and four groups only need 5 (5 = 1 x 4 + 1, 1 represents that sensors are arranged on one tire, 4 represents that each tire corresponds to 4 acceleration sensors, and 1 represents that 1 acceleration sensor is used for in-vehicle test data synchronization). At the same time, in order to ensure the accuracy of synchronous acquisition, an acceleration sensor is also needed to be reserved in the vehicle (such as the seat mounting point on the cross beam) as a reference data source.

[0108] If the acceleration response is collected in the grouping mode, the corresponding matrix needs to be reconstructed Matrix, see the following formula:

[0109] Among them, the data of the diagonal position , , , does not need to be synchronized, and the data of the non-diagonal position is synchronized by the following formula:

[0110] Where subscript i and j represent the data of different directions of the acceleration sensor on the knuckle, and subscript m and n represent the data of different directions of the in-vehicle sensor in different groups.

[0111] S4, judge whether the error analysis and evaluation is qualified, if qualified, turn to S5, if not qualified, replace the abnormal sensor according to the error evaluation result, retest and collect data and turn to S2.

[0112] Error analysis and evaluation can be performed by generating the wheel center force result, which can be calculated reversely by the following formula :

[0113] And the acceleration response obtained by test is divided by the wheel center force result, and then decibel dB processing is performed, see Figure 3The error evaluation of the middle knuckle force is shown in the schematic diagram. The X axis is the channel name of the different directions of each acceleration sensor, and the Y axis is the error size between the calculated value and the test value. If the error of two directions of a sensor exceeds 0.4, the sensor data is abnormal, and the data of another sensor needs to be replaced to recalculate the knuckle force. If it still cannot meet the requirements, the data collection needs to be retested.

[0114] S5, error evaluation is qualified, and the calculation card containing the knuckle force is output for road noise simulation analysis.

[0115] After the error evaluation is passed, the system can generate a vehicle road noise analysis card containing the calculation results of the rear knuckle force according to the input mapping table information, including the knuckle node ID number and the response point ID number. The mapping table is shown in Figure 4 The mapping table for analysis and testing is shown in the schematic diagram. ExcitationPointID represents the knuckle node ID number, KnuckleResponsePointID represents the steering knuckle sensor node ID number in the vehicle NVH finite element model, MappingKnuckleLabels represents the channel name of the steering knuckle sensor response data in the test data, and ReferenceKnuckleLabelsDir represents the channel name and direction of the vehicle sensor response data.

[0116] Through the selection of the mapping table, it can be ensured that the acceleration sensor is consistent in the position of the vehicle NVH finite element model and the actual vehicle, which is set through the ExcitationPointID column; it can be ensured that the grouping of the knuckle and the sensor on each tire is corresponding, one knuckle is respectively four or five sensors, which is set through the ExcitationPointID column and the KnuckleResponsePointID column; in addition, it can also be noted that the same response point is respectively in the vibration transfer function finite element analysis result file (see Figure 5 ) and the acceleration response test result data file (see Figure 6 ), which is set through the KnuckleResponsePointID column and the MappingKnuckleLabels column; and the data channel name and direction for synchronization are noted, which is set through the ReferenceKnuckleLabelsDir column.

[0117] In summary, by establishing a mapping relationship between the simulation transfer function and the measured response, combined with grouping testing and common synchronous sensors, the number of required acceleration sensors is significantly reduced (as few as 5), and the testing cost and complexity are reduced. By using the frequency domain inversion algorithm and pseudo-inverse solution, the wheel core force in the high frequency band of 300-400 Hz is accurately extracted, breaking through the frequency limitation of traditional sensors. The closed-loop error evaluation and data reliability screening mechanism is introduced to ensure the accuracy of the results. Finally, the standardized excitation file is automatically generated and connected to the mainstream simulation environment, realizing the full-process automation from testing to road noise simulation, and greatly improving the efficiency and engineering practicability of new energy vehicle NVH development.

[0118] Based on the above method embodiments, the embodiments of the present application also provide a wheel core force extraction system for vehicle road noise simulation, as shown in Figure 8 The device mainly includes the following parts: The construction unit 810 is configured to construct a vibration transmission relationship between a wheel core excitation point and a knuckle response area based on a vehicle NVH finite element model, and determine frequency domain transmission data based on the vibration transmission relationship; The signal processing unit 820 is configured to collect dynamic response signals of multiple knuckle response areas and a road noise synchronous signal of a preset reference position when the vehicle is running on an actual road in a time-sharing manner, and perform phase alignment on the frequency domain responses of different time periods according to the road noise synchronous signal to obtain a target frequency domain response matrix; The determination unit 830 is configured to determine a frequency domain excitation vector of the wheel core excitation point according to the frequency domain transmission data and the target frequency domain response matrix; The generation unit 840 is configured to reconstruct a theoretical response of the knuckle response area based on the frequency domain excitation vector and the frequency domain transmission data, and generate a target wheel core force for vehicle road noise simulation analysis based on the theoretical response and the measured response.

[0119] In a feasible implementation, the signal processing unit 820 is further configured to: Collect the dynamic response signals of the multiple knuckle response areas when the vehicle is running on the actual road in a time-sharing manner based on multiple sensing units arranged on the structures where the vehicle knuckles are located, wherein the spatial distribution of the sensing units satisfies a non-coplanar constraint condition; Collect the road noise synchronous signal based on a common sensing unit arranged at a preset reference position selected inside the vehicle.

[0120] In a feasible implementation, the signal processing unit 820 is further configured to: Divide all the wheels into a plurality of test groups, and sequentially obtain the dynamic response signals of the knuckle response areas corresponding to each test group; In the same test group, the dynamic response signals of the knuckle response areas included in the test group are synchronously collected, and the data correlation is realized through the common sensing unit between different test groups.

[0121] In an implementable embodiment, the signal processing unit 820 is further configured to: establish a mapping relationship between the excitation points and response points in the simulation model and the measured data channels; unify the data of different sources to the same identification system by using the mapping relationship; perform Fourier transform on the dynamic response signals and the corresponding road noise synchronous signals collected by each test group, respectively, to obtain the frequency domain response subsets and the frequency domain synchronous reference data of each test group; calculate the cross-group phase deviation based on the phase relationship between the frequency domain synchronous reference data of each test group, and perform phase compensation processing on the corresponding frequency domain response subsets; combine the frequency domain response subsets after phase compensation to obtain the target frequency domain response matrix.

[0122] In an implementable embodiment, the determining unit 830 is further configured to: construct an excitation-response equation set according to the frequency domain transfer data and the target frequency domain response matrix; perform pseudo-inverse operation on the coefficient matrix in the excitation-response equation set to obtain the frequency domain excitation vector.

[0123] In an implementable embodiment, the generating unit 840 is further configured to: input the frequency domain excitation vector into a forward propagation model represented by the vibration transfer relationship; calculate the expected dynamic response of each knuckle response area under the same working condition and generate the theoretical response in the corresponding frequency domain; perform consistency comparison on the theoretical response and the measured response to generate a data reliability evaluation result of each knuckle response area; filter target frequency domain excitation components according to the data reliability evaluation result, and determine target wheel center forces of each wheel center position for vehicle road noise simulation analysis based on the target frequency domain excitation vector.

[0124] In an implementable embodiment, the system further comprises a simulation module configured to: encapsulate the target wheel center forces into an excitation input file in a standardized format according to a pre-defined data interface specification; automatically load the excitation input file into a vehicle road noise simulation environment to determine the boundary conditions for vibration and noise performance prediction.

[0125] The vehicle road noise simulation wheel core force extraction system provided by the embodiments of the present application has the same implementation principle and technical effects as the foregoing method embodiments. For brevity, the embodiments of the vehicle road noise simulation wheel core force extraction system are not mentioned in the foregoing method embodiments, and the corresponding content can be referred to in the foregoing method embodiments.

[0126] The embodiments of the present application also provide an electronic device, as shown in the accompanying drawings, which is a structural schematic diagram of the electronic device. The electronic device 100 includes a processor 91 and a memory 90. The memory 90 stores computer executable instructions capable of being executed by the processor 91. The processor 91 executes the computer executable instructions to implement any one of the foregoing vehicle road noise simulation wheel core force extraction methods. Figure 9

[0127] In the embodiment shown in the accompanying drawings, the electronic device further includes a bus 92 and a communication interface 93. The processor 91, the communication interface 93 and the memory 90 are connected through the bus 92. Figure 9

[0128] The memory 90 can include a high-speed random access memory (RAM) and can also include a non-volatile memory, for example, at least one disk memory. The communication connection between the system network element and at least one other network element is implemented through at least one communication interface 93 (which can be wired or wireless), and the Internet, a wide area network, a local area network, a metropolitan area network, etc. can be used. The bus 92 can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 92 can be divided into an address bus, a data bus, a control bus, etc. For brevity, Figure 9 In the accompanying drawings, only one bidirectional arrow is used to represent the bus, but it does not mean that there is only one bus or only one type of bus.

[0129] ​​The processor 91 can be an integrated circuit chip with signal processing capability. In the implementation process, the steps of the above method can be completed by the integrated logic circuit of hardware in the processor 91 or the instruction in the form of software. The processor 91 described above can be a general processor, including a central processing unit (CPU), a network processor (NP), etc.; can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as a hardware decoding processor for execution, or a combination of hardware and software modules in the decoding processor for execution. The software module can be located in a random access memory, a flash memory, a read only memory, a programmable read only memory or an electrically erasable programmable memory, a register, or other mature storage media in the art. The storage medium is located in the memory, and the processor 91 reads the information in the memory, and combines the hardware to complete the steps of the whole vehicle road noise simulation wheel core force extraction method described in the foregoing embodiments.

[0130] The embodiment of the present application also provides a computer readable storage medium, which stores computer executable instructions. When the computer executable instructions are called and executed by a processor, the computer executable instructions cause the processor to implement the whole vehicle road noise simulation wheel core force extraction method described above. For specific implementation, refer to the foregoing method embodiments, which will not be described here.

[0131] The computer program product of the whole vehicle road noise simulation wheel core force extraction method, system and device provided by the embodiment of the present application includes a computer readable storage medium storing program codes. The instructions included in the program codes can be used to execute the method described in the foregoing method embodiments. For specific implementation, refer to the method embodiments, which will not be described here.

[0132] Unless otherwise specifically stated, the relative steps, numerical expressions and numerical values of the components and steps set forth in these embodiments do not limit the scope of the present application.

[0133] If the functions are realized in the form of software function units and sold or used as independent products, they can be stored in a nonvolatile computer readable storage medium executable by a processor. Based on this understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0134] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for extracting wheel center force for whole-vehicle road noise simulation, characterized in that, include: Based on the whole vehicle NVH finite element model, the vibration transmission relationship between the wheel center excitation point and the steering knuckle response area is constructed, and the frequency domain transmission data is determined based on the vibration transmission relationship. The system collects dynamic response signals of multiple steering knuckle response areas and road noise synchronization signals at preset reference positions when the vehicle is driving on the actual road surface in a time-division manner. The system then performs phase alignment of the frequency domain response at different time periods based on the road noise synchronization signals to obtain the target frequency domain response matrix. The frequency domain excitation vector of the wheel center excitation point is determined based on the frequency domain transfer data and the target frequency domain response matrix. The theoretical response of the steering knuckle response region is reconstructed based on the frequency domain excitation vector and the frequency domain transfer data, and a target wheel center force for whole vehicle road noise simulation analysis is generated based on the theoretical response.

2. The method for extracting wheel center force for vehicle road noise simulation according to claim 1, characterized in that, The system collects dynamic response signals from multiple steering knuckle response areas and road noise synchronization signals from preset reference positions in a time-sharing manner when the vehicle is traveling on the actual road surface. Based on multiple sensing units arranged on the structure where each steering knuckle of the vehicle is located, dynamic response signals of multiple steering knuckle response areas are collected in time-division when the vehicle is driving on the actual road surface. The spatial distribution of the sensing units satisfies the non-coplanar constraint condition. The road noise synchronization signal is collected by a shared sensing unit deployed at a preset reference position selected inside the vehicle.

3. The method for extracting wheel center force for vehicle road noise simulation according to claim 1, characterized in that, The system collects dynamic response signals from multiple steering knuckle response areas in a time-sharing manner when the vehicle is traveling on the actual road surface, including: All wheels were divided into several test groups, and the dynamic response signals of the steering knuckle response area corresponding to each test group were obtained in sequence. Within the same test group, dynamic response signals of the steering knuckle response area included in the test group are collected synchronously, and data association between different test groups is achieved by sharing a sensing unit.

4. The method for extracting wheel center force for vehicle road noise simulation according to claim 1, characterized in that, The target frequency domain response matrix is ​​obtained by phase alignment of the frequency domain response at different time periods based on the road noise synchronization signal, including: Establish the mapping relationship between the excitation points and response points in the simulation model and the measured data channels; The mapping relationship is used to unify data from different sources under the same identification system; Fourier transforms were performed on the dynamic response signal and the corresponding road noise synchronization signal collected for each test group to obtain the frequency domain response subset and frequency domain synchronization reference data for each test group. Based on the phase relationship between the frequency domain synchronization reference data of each test group, the cross-group phase deviation is calculated, and phase compensation processing is performed on the corresponding frequency domain response subset. The target frequency domain response matrix is ​​obtained by combining the subsets of frequency domain responses after phase compensation.

5. The method for extracting wheel center force for vehicle road noise simulation according to claim 1, characterized in that, Based on the frequency domain transfer data and the target frequency domain response matrix, the frequency domain excitation vector of the wheel center excitation point is determined, including: A set of excitation-response equations is constructed based on the frequency domain transfer data and the target frequency domain response matrix; The frequency domain excitation vector is obtained by performing a pseudo-inverse operation on the coefficient matrix in the excitation-response equation system.

6. The method for extracting wheel center force for vehicle road noise simulation according to claim 1, characterized in that, Based on the frequency domain excitation vector and the frequency domain transfer data, the theoretical response of the steering knuckle response region is reconstructed. Based on the theoretical response and the measured response, a target wheel center force for whole-vehicle road noise simulation analysis is generated, including: The frequency domain excitation vector is input into the forward propagation model characterized by the vibration transmission relationship; Calculate the expected dynamic response of each steering knuckle response region under the same working conditions, and generate the theoretical response in the corresponding frequency domain; The consistency between the theoretical response and the measured response is compared to generate the data credibility evaluation results for each steering knuckle response region. Based on the data reliability assessment results, target frequency domain excitation components are selected, and target wheel center forces for vehicle road noise simulation analysis are determined at each wheel center position based on the target frequency domain excitation vector.

7. The method for extracting wheel center force for vehicle road noise simulation according to claim 1 or 6, characterized in that, After generating the target wheel center force for whole-vehicle road noise simulation analysis based on the theoretical response, the following steps are also included: Based on a predefined data interface specification, the target wheel center force is encapsulated into a standardized excitation input file. The excitation input file is automatically loaded into the vehicle road noise simulation environment to determine the boundary conditions for vibration and noise performance prediction.

8. A system for extracting wheel center force for simulating road noise in a vehicle, characterized in that, include: The construction unit is used to construct the vibration transmission relationship between the wheel center excitation point and the steering knuckle response area based on the whole vehicle NVH finite element model, and to determine the frequency domain transmission data based on the vibration transmission relationship; The signal processing unit is used to collect dynamic response signals of multiple steering knuckle response areas and road noise synchronization signals at preset reference positions when the vehicle is driving on the actual road surface in a time-division manner, and to perform phase alignment of the frequency domain response of different time periods according to the road noise synchronization signal to obtain the target frequency domain response matrix. The determining unit is used to determine the frequency domain excitation vector of the wheel center excitation point based on the frequency domain transmission data and the target frequency domain response matrix; The generation unit is used to reconstruct the theoretical response of the steering knuckle response region based on the frequency domain excitation vector and the frequency domain transfer data, and to generate a target wheel center force for whole vehicle road noise simulation analysis based on the theoretical response and the measured response.

9. An electronic device, characterized in that, The method includes a processor and a memory, the memory storing computer-executable instructions that can be executed by the processor, the processor executing the computer-executable instructions to implement the wheel center force extraction method for whole vehicle road noise simulation according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the wheel center force extraction method for whole vehicle road noise simulation as described in any one of claims 1 to 7.

Images