Search

CN-122021163-A - Simulation design modeling system for core piece of motor performance off-road vehicle

CN122021163ACN 122021163 ACN122021163 ACN 122021163ACN-122021163-A

Abstract

The application discloses a simulation design modeling system and method for a core piece of a motor performance off-road vehicle. The system comprises an excitation spectrum analysis module, a geometric topology mapping module, a self-adaptive hybrid modeling module and a parallel solving module, wherein the excitation spectrum analysis module is used for executing fourth-order Butterworth filtering and two-dimensional Fourier transformation on terrain scanning data to generate a vehicle wheel center input load spectrum under a frequency domain, the geometric topology mapping module is used for deducing modal characteristics of a core piece and calculating a spectrum overlapping coefficient in combination with the load spectrum to generate a space sensitivity field, the self-adaptive hybrid modeling module is used for discretizing the core piece into hybrid topology of a high-order entity unit and a reduced-order superunit in space according to the sensitivity field and realizing interface coupling through a multipoint constraint equation based on inverse equal parameter mapping, and the parallel solving module is used for monitoring the strain energy density in real time in the calculation process to trigger local grid reconstruction. According to the method, the self-adaptive mapping of the frequency domain energy and the grid density is adopted, so that the problem that the simulation calculation efficiency and the simulation calculation precision of the core piece of the off-road vehicle are difficult to consider under the complex working condition is solved.

Inventors

  • WU ZHENWEI

Assignees

  • 浙江立雅机车有限公司

Dates

Publication Date
20260512
Application Date
20260130

Claims (10)

  1. 1. A motor performance off-road vehicle core simulation design modeling system, comprising: The excitation spectrum analysis module is configured to receive three-dimensional terrain scanning data of vehicle running and vehicle running speed parameters, and convert the terrain scanning data into a vehicle wheel center input load spectrum under a frequency domain; The geometric topology mapping module is in communication connection with the excitation spectrum analysis module, is configured for storing a three-dimensional geometric model of the core piece, and generates a spatial sensitivity field describing stress response sensitivity of each region of the core piece based on the matching degree between the input load spectrum of the wheel center of the vehicle and the modal characteristics of the core piece; The self-adaptive hybrid modeling module is in communication connection with the geometric topology mapping module and is configured to discretize the three-dimensional geometric model of the core piece into a heterogeneous finite element grid model according to the space sensitivity field, wherein for the area of the space sensitivity field higher than a preset threshold value, a high-order entity unit is generated, for the area of the space sensitivity field lower than the preset threshold value, a reduced superunit or beam unit is generated, and a multi-point constraint coupling relation between the high-order entity unit and the reduced superunit or beam unit is established; And the parallel solving and reconstructing module is in communication connection with the adaptive hybrid modeling module and is configured for carrying out nonlinear dynamics solving based on the heterogeneous finite element grid model.
  2. 2. The system of claim 1, wherein the geometric topology mapping module comprises a geometric-modal correlation inference unit configured to run a pre-trained three-dimensional convolutional neural network with the voxelized data of the core as input to directly output predicted pre-core N-order modal shape data, wherein the geometric topology mapping module is configured to calculate the spatial sensitivity field in combination with the modal shape data and the vehicle wheel center input load spectrum, and wherein the three-dimensional convolutional neural network is configured to learn a nonlinear mapping relationship between core geometry and modal shape.
  3. 3. The system of claim 1, wherein the adaptive hybrid modeling module comprises an interface coupling controller configured to compute shape function weights using inverse iso-mapping, construct a displacement-coordinated multi-point constraint equation at interface nodes of the higher-order entity units and the reduced superunits or beam units, and force the displacement of higher-order entity unit nodes at the interface to be a shape function weighted sum of corresponding reduced-order unit nodes.
  4. 4. The system of claim 1, wherein the geometric topology mapping module is configured to quantify the degree of matching by calculating a spectral overlap coefficient defined as, for each order mode, calculating an integral value of the vehicle wheel center input load spectrum within a preset bandwidth centered on the order mode frequency, multiplying the integral value by the amplitude of the mode of the order mode, and summing the products for all modes to obtain the spectral overlap coefficient.
  5. 5. The system of claim 1, wherein the excitation spectrum analysis module comprises a pavement filtering subunit configured to apply a fourth order bute Wo Sigao pass filter to filter out low frequency heave data having a wavelength greater than a preset multiple of wheelbase.
  6. 6. The simulation design modeling method for the motor performance off-road vehicle core piece is characterized by comprising the following steps of: receiving three-dimensional terrain scanning data of vehicle running and vehicle running speed parameters, and converting the terrain scanning data into a vehicle wheel center input load spectrum under a frequency domain; Generating a spatial sensitivity field describing stress response sensitivity of each region of the core based on the degree of matching between the vehicle wheel center input load spectrum and modal characteristics of the core; Discretizing a three-dimensional geometric model of the core piece into a heterogeneous finite element grid model according to the space sensitivity field, wherein for the region of the space sensitivity field higher than a preset threshold value, generating a high-order entity unit, for the region of the space sensitivity field lower than the preset threshold value, generating a reduced superunit or beam unit, and establishing a multipoint constraint coupling relation between the high-order entity unit and the reduced superunit or beam unit; And carrying out nonlinear dynamics solving based on the heterogeneous finite element grid model.
  7. 7. The method of claim 6, wherein generating a spatial sensitivity field describing stress response sensitivity of each region of the core based on a degree of matching between the vehicle wheel center input load spectrum and modal features of the core comprises: Utilizing a pre-trained three-dimensional convolutional neural network, taking voxelized data of the core piece as input, and deducing a front N-order mode shape and corresponding frequency of the core piece; Calculating a spectrum overlapping coefficient under each space partition, wherein the spectrum overlapping coefficient represents the energy overlapping degree of the mode shape and the vehicle wheel center input load spectrum on a frequency domain; And performing normalization and threshold segmentation operations on the spectrum overlapping coefficients to generate a binarized spatial sensitivity mask.
  8. 8. The method of claim 6, wherein establishing the multi-point constrained coupling relationship comprises: Identifying interface nodes of the high-order entity units and the reduced superunits or beam units; Based on a reverse iso-mapping algorithm, calculating a shape function value of the high-order entity unit at the interface as a weight coefficient; And constructing a multipoint constraint equation, so that the degree of freedom of the high-order entity unit node is linearly represented by the degree of freedom of the reduced-order unit node and the corresponding weight coefficient.
  9. 9. The method of claim 6, wherein the non-linear dynamics solving based on the heterogeneous finite element mesh model further comprises: Monitoring the plastic strain energy density of each unit in real time; When the plastic strain energy density of any unit exceeds a preset safety threshold value, suspending solving; triggering a local grid splitting instruction on a region where the unit exceeding a preset safety threshold value is located, reconstructing the region into a high-order entity unit, updating a rigidity matrix of a local region with a changed topological structure, assembling the rigidity matrix into a global rigidity matrix, and rolling back to a current time step for recalculation, wherein the preset safety threshold value is set to be 80% of the yield strain energy density of the material.
  10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the method of any of claims 6 to 9.

Description

Simulation design modeling system for core piece of motor performance off-road vehicle Technical Field The application relates to the technical field of intersection of Computer Aided Engineering (CAE) and vehicle engineering, in particular to a simulation design modeling system for a core piece of a motor performance off-road vehicle. Background During development of automotive off-road vehicles, the core load bearing components of the suspension control arms, knuckles, etc. need to maintain structural integrity under extremely complex road surface excitation. The prior art is faced with a fundamental technical contradiction between computational efficiency and simulation fidelity when performing simulation modeling on the components. In order to capture the millisecond impact response, the traditional full-size Finite Element Analysis (FEA) has to perform high-density meshing on the whole component, so that the number of degrees of freedom is increased sharply, the calculation time is often up to hours or even days, and the iterative design speed is severely restricted. On the other hand, the multi-body dynamics (MBD) model or the excessively simplified beam unit model adopted for efficiency pursuit have high calculation speed, but the fatigue life of the stress concentration point cannot be accurately predicted due to neglecting the local flexible deformation and the contact nonlinearity of the component, so that the design scheme has potential safety hazards. The modeling method of the 'one-cut' type ignores the non-uniformity of the road surface excitation energy distribution in the frequency domain and the space domain, so that the calculation force is wasted in a non-critical area, and the precision of the critical area is not guaranteed. Therefore, a need exists for a simulation modeling system that is capable of adaptively adjusting the topology of a model based on external excitation characteristics. Disclosure of Invention In a first aspect, the application provides a simulation design modeling system for a core piece of a motor performance off-road vehicle, which comprises an excitation spectrum analysis module, a geometric topology mapping module, an adaptive hybrid modeling module and a parallel solving and reconstructing module. The excitation spectrum analysis module is configured to receive three-dimensional terrain scanning data of vehicle running and vehicle running speed parameters, and convert the terrain scanning data into a vehicle wheel center input load spectrum in a frequency domain. The geometric topology mapping module is in communication connection with the excitation spectrum analysis module, is configured for storing a three-dimensional geometric model of the core piece, and generates a spatial sensitivity field describing stress response sensitivity of each region of the core piece based on the matching degree between the input load spectrum of the wheel center of the vehicle and the modal characteristics of the core piece. The self-adaptive hybrid modeling module is in communication connection with the geometric topology mapping module and is configured to discretize a three-dimensional geometric model of the core piece into a heterogeneous finite element grid model according to the space sensitivity field, wherein a high-order entity unit is generated for a region with the space sensitivity field higher than a preset threshold value, a reduced superunit or beam unit is generated for a region with the space sensitivity field lower than the preset threshold value, and a multipoint constraint coupling relation between the high-order entity unit and the reduced superunit or beam unit is established. The parallel solving and reconstructing module is in communication connection with the adaptive hybrid modeling module and is configured for nonlinear dynamics solving based on the heterogeneous finite element mesh model. Optionally, the geometric topology mapping module comprises a geometric-modal correlation inference unit configured to run a pre-trained three-dimensional convolutional neural network, take the voxelized data of the core as input, directly output predicted front-N-order modal shape data of the core, calculate the spatial sensitivity field by combining the modal shape data and the vehicle wheel center input load spectrum, and learn a nonlinear mapping relation between geometric features of the core and the modal shape. Optionally, the adaptive hybrid modeling module includes an interface coupling controller configured to calculate shape function weights using inverse iso-mapping, construct a displacement-coordinated multi-point constraint equation at interface nodes of the higher-order entity units and the reduced superunits or beam units, and force the displacements of the higher-order entity unit nodes at the interface to be shape function weighted sums of the corresponding reduced-order unit nodes. Optionally, the geometric topology mapping module is configured t