Vehicle modal identification measuring point setting method and device based on particle swarm algorithm

Abstract

The application discloses a vehicle modal identification measuring point setting method and device based on a particle swarm algorithm, and the method comprises the following steps: constructing a vehicle finite element model based on the size parameters of the vehicle, the material parameters of each vehicle body element and the setting positions of each vehicle equipment; performing eigenvalue solving on the vehicle finite element model to obtain first modal parameters; performing a modal test based on the vehicle entity and preset sensor setting position information to obtain second modal parameters; when the difference between the first natural frequency and the second natural frequency is greater than a preset frequency difference, performing particle swarm optimization iteration solving based on the first modal parameters to determine optimal sensor setting position information; the application can globally optimize and select the sensor measuring points under the condition that the number of sensors is limited, reduces the correlation between different modes, improves the independence, accuracy of vehicle modal identification and the scientificity of measuring point arrangement.

Description

Technical Field

[0001] This invention relates to the field of structural dynamics testing technology, and in particular to a method and apparatus for setting up vehicle modal identification measurement points based on particle swarm optimization algorithm. Background Technology

[0002] In the study of rail vehicle structural dynamics, modal parameters, as key indicators characterizing the dynamic properties of the vehicle body, directly affect the optimization of vehicle structural design, operational status assessment, and subsequent maintenance decisions. During modal testing, how to rationally arrange measuring points under limited sensor quantity to ensure comprehensive and accurate identification of vehicle body modal parameters has become a pressing technical problem in this field.

[0003] To address the aforementioned technical challenges, existing modal testing for rail vehicles typically employs a method of uniformly distributing sensors across typical cross-sections and key locations of the vehicle body to acquire vibration response data. To improve the rationality of the measurement point layout, various optimization methods have been proposed, including optimization criteria based on information matrices, modal strain energy, controllability and observability, model reduction, and modal guarantee principles. These methods are combined with traditional algorithms such as sequence selection and effective independence methods to optimize the measurement point layout design.

[0004] However, existing technologies still have certain shortcomings. On the one hand, uniform or empirically based point placement methods cannot fully reflect the differences in the contribution of different measurement points to modal identification, easily leading to measurement point redundancy or omission of key modal information. On the other hand, modal confidence indices are mostly used in the experimental result analysis stage, with less application in the pre-experimental measurement point optimization design, resulting in a lack of specificity in measurement point optimization. In addition, some optimization methods mainly focus on theoretical analysis or simulation verification, lacking systematic real-vehicle test verification for complex rail vehicle structures. Under conditions of multi-mode coexistence and noise interference, the stability and engineering applicability of their optimization results still need further improvement.

[0005] This section is intended to provide background or context for the embodiments of the invention set forth in the claims. The description herein is not an admission that it is prior art simply because it is included in this section. Summary of the Invention

[0006] This invention provides a method for setting measurement points for vehicle modality recognition based on particle swarm optimization, which can globally optimize the selection of sensor measurement points under the condition of limited number of sensors, reduce the correlation between different modalities, and improve the independence, accuracy and scientific nature of vehicle modality recognition and measurement point layout.

[0007] The method for setting measurement points for vehicle modal recognition based on particle swarm optimization includes: A finite element model of the vehicle is constructed based on the vehicle's size parameters, the material parameters of each vehicle body component, and the location of each in-vehicle device. The eigenvalues ​​of the vehicle finite element model are solved to obtain the first modal parameters; the first modal parameters include: the first mode shape vector of each mode and its corresponding first natural frequency; Modal tests are conducted based on the vehicle entity and preset sensor location information to obtain second modal parameters; the second modal parameters include: the second modal shape vector of each mode and its corresponding second natural frequency; When the difference between the first natural frequency and the second natural frequency is greater than a preset frequency difference, particle swarm optimization iterative solution is performed based on the first modal parameters to determine the optimal sensor setting position information.

[0008] In some embodiments, constructing a vehicle finite element model based on the vehicle's size parameters, the material parameters of each vehicle body component, and the location of each in-vehicle device includes: A first geometric model of the vehicle entity is established based on the dimensional parameters and the material parameters of each vehicle body component. According to the preset mesh density and element type, the vehicle body components in the first geometric model are meshed to obtain multiple shell element nodes; Each in-vehicle device is represented as a mass point, and each mass point is set at the corresponding position in the first geometric model according to the setting position of the corresponding in-vehicle device. The mass points are connected to the shell element nodes in the first geometric model by multi-point constraint elements (MPC) to obtain the vehicle finite element model.

[0009] In some embodiments, the modal testing based on the vehicle entity and preset sensor location information to obtain the second modal parameters includes: A second geometric model of the vehicle entity is established based on the vehicle entity and a preset set of sensor location information. Modal analysis is performed based on the second geometric model to obtain the second modal parameters.

[0010] In some embodiments, establishing a second geometric model of the vehicle entity based on the vehicle entity and a preset set of sensor location information includes: According to the sensor location information set, multiple sensors and vibration excitation devices are installed in the vehicle body; The second geometric model is created based on the vehicle entity, and test nodes corresponding to each sensor are set on each cross section of the second geometric model.

[0011] In some embodiments, the modal analysis based on the second geometric model to obtain the second modal parameters includes: Associate the acquisition channels of each of the aforementioned sensors and the excitation device with the corresponding test nodes; The excitation signal and the acceleration signal of each test node are acquired through the acquisition channel; The second mode parameters are determined by calling the least squares complex exponential function based on the acceleration signal and the excitation signal.

[0012] In some embodiments, the step of determining the preferred sensor placement location information by performing particle swarm optimization iterative solution based on the first modal parameters includes: Based on the current position vector of each particle in the particle swarm and the first mode shape vector of each mode, the MAC matrix of each particle is constructed. The fitness value of each particle is determined based on the MAC matrix; The optimal position of each particle is updated based on its fitness value. Based on the updated individual optimal positions of each particle, the group optimal position of the particle swarm is determined; Based on the group's preferred position, position update function, velocity update function, current position vector and current velocity vector of each particle, and the updated individual preferred position, determine the updated position vector and velocity vector of each particle; The above steps are executed iteratively. When the preset termination condition is met, the combination of measuring points corresponding to the preferred position of the group is obtained.

[0013] In some embodiments, constructing the MAC matrix for each particle based on its current position vector and the first mode shape vector of each mode in the particle swarm includes: Based on the current position vector of each particle, determine the combination of measurement points for each particle and its corresponding finite element degrees of freedom; Based on the combination of measurement points of each particle and its corresponding finite element degrees of freedom, the first mode shape vector of each particle's mode is extracted to obtain the first mode shape sub-vector of each particle's mode. Based on the first mode shape subvectors of each particle's first mode, determine the MAC matrix between any two first mode shape subvectors of each particle's first mode.

[0014] This invention also provides a vehicle modality recognition measurement point setting device based on particle swarm optimization algorithm, which is used to globally optimize the selection of sensor measurement points under the condition of limited number of sensors, reduce the correlation between different modalities, and improve the independence, accuracy and scientific nature of vehicle modality recognition and measurement point layout.

[0015] The vehicle modality recognition measurement point setting device based on particle swarm optimization algorithm includes: The finite element model building module is used to build a finite element model of the vehicle based on the vehicle's size parameters, the material parameters of each vehicle body component, and the location of each in-vehicle device. The first modal parameter acquisition module is used to solve the eigenvalues ​​of the vehicle finite element model to obtain the first modal parameters; the first modal parameters include: the first mode shape vector of each mode and its corresponding first natural frequency; The reference modal test module is used to conduct modal tests based on the vehicle entity and preset sensor setting position information to obtain second modal parameters; the second modal parameters include: the second modal shape vector of each mode and its corresponding second natural frequency; The measurement point optimization triggering and solving module is used to determine the optimal sensor setting position information by performing particle swarm optimization iterative solution based on the first modal parameters when the difference between the first natural frequency and the second natural frequency is greater than the preset frequency difference.

[0016] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described method for setting vehicle modal recognition measurement points based on particle swarm optimization algorithm.

[0017] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for setting vehicle modal recognition measurement points based on particle swarm optimization.

[0018] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described method for setting vehicle modal recognition measurement points based on particle swarm optimization algorithm.

[0019] The vehicle modal recognition measurement point setting method and apparatus based on particle swarm optimization (PSO) provided in this invention combines vehicle finite element modal analysis with actual vehicle modal testing. It optimizes measurement points while ensuring consistency between the finite element model and the actual vehicle's dynamic characteristics, thus eliminating reliance on manual experience or uniform distribution for sensor measurement point placement. When the difference between the natural frequencies calculated by the finite element model and the measured natural frequencies identified by the modal tests exceeds a preset threshold, a particle swarm optimization iterative process based on the first modal parameter is introduced to perform a global search of sensor measurement point combinations. This automatically determines a more favorable measurement point arrangement scheme for modal differentiation even with a limited number of sensors. This method fully utilizes the overall structural modal characteristics reflected by the finite element model, transforming the sensor measurement point selection problem into an optimization process based on modal parameter evaluation. This effectively reduces the correlation between different modes and improves the stability and consistency of modal recognition results. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is a flowchart illustrating a method for setting measurement points for vehicle modal recognition based on particle swarm optimization in one embodiment of the present invention. Figure 2 This is a schematic diagram of the structure of a vehicle finite element model in another embodiment of the present invention; Figure 3 This is a flowchart illustrating the method for setting measurement points for vehicle modal recognition based on particle swarm optimization in an embodiment of the present invention. Figure 4 This is a flowchart illustrating the method for setting measurement points for vehicle modal recognition based on particle swarm optimization algorithm in another embodiment of the present invention. Figure 5 This is a flowchart illustrating the method for setting measurement points for vehicle modal recognition based on particle swarm optimization algorithm in another embodiment of the present invention. Figure 6 This is a schematic diagram of the sensor placement location in an embodiment of the present invention; Figure 7 This is a flowchart illustrating the method for setting measurement points for vehicle modal recognition based on particle swarm optimization algorithm in another embodiment of the present invention. Figure 8 This is a flowchart illustrating the method for setting measurement points for vehicle modal recognition based on particle swarm optimization algorithm in another embodiment of the present invention. Figure 9This is a flowchart illustrating the method for setting measurement points for vehicle modal recognition based on particle swarm optimization algorithm in another embodiment of the present invention. Figure 10 This is a schematic diagram of the vehicle modality recognition measurement point setting device based on particle swarm optimization algorithm in an embodiment of the present invention; Figure 11 This is a schematic diagram of the physical structure of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments and their descriptions are used to explain the present invention, but are not intended to limit the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be arbitrarily combined with each other. The acquisition, storage, use, and processing of data in the technical solutions of this application all comply with relevant laws and regulations. The user information in the embodiments of this application is obtained through legal and compliant means, and the acquisition, storage, use, and processing of user information have been authorized and agreed upon by the customer.

[0022] To facilitate understanding of the technical solution provided in this application, the relevant content of the technical solution in this application will be explained below.

[0023] To overcome the problem that existing vehicle modal testing relies mainly on manual experience or uniform arrangement for sensor placement, making it difficult to simultaneously ensure the integrity and accuracy of modal recognition under limited sensor quantity, this application provides a vehicle modal recognition sensor placement method based on particle swarm optimization (PSO) algorithm. This method constructs a vehicle finite element model considering the structural characteristics of the vehicle body and the mass distribution of in-vehicle equipment to obtain the vehicle's first modal parameters. Combined with second modal parameters obtained from modal testing of the vehicle entity, the modal consistency of the finite element model is judged. Based on meeting preset frequency consistency requirements, a particle swarm optimization algorithm is introduced to model the sensor placement problem as a discrete combinatorial optimization problem. A global search and iterative update of candidate sensor combination is performed, automatically determining the optimal sensor placement location information that minimizes the correlation between different modes. Through this technical solution, this application realizes a shift from experience-driven to model- and data-driven sensor placement in vehicle modal testing, improving the modal discrimination capability and the rationality and stability of sensor placement during vehicle modal recognition.

[0024] This invention provides a method for setting measurement points for vehicle modality recognition based on particle swarm optimization algorithm, such as... Figure 1 As shown, the method for setting vehicle modal recognition measurement points includes steps 101 to 104.

[0025] Step 101: Construct a finite element model of the vehicle based on its size parameters, material parameters of each vehicle body component, and the location of each in-vehicle device.

[0026] Step 102: Solve for the eigenvalues ​​of the vehicle finite element model to obtain the first modal parameters. The first modal parameters include: the first mode shape vector of each mode and its corresponding first natural frequency.

[0027] Step 103: Conduct modal testing based on the vehicle entity and preset sensor position information to obtain the second modal parameters. The second modal parameters include: the second modal shape vectors of each mode and their corresponding second natural frequencies.

[0028] Step 104: When the difference between the first natural frequency and the second natural frequency is greater than the preset frequency difference, perform particle swarm optimization iterative solution based on the first modal parameters to determine the preferred sensor setting position information.

[0029] According to the above embodiments, by combining vehicle finite element modal analysis with actual vehicle modal testing, measurement point optimization is carried out while ensuring consistency between the finite element model and the actual vehicle's dynamic characteristics. This allows sensor measurement point placement to no longer rely solely on manual experience or uniform distribution. When the difference between the natural frequencies calculated by the finite element model and the measured natural frequencies identified by the modal tests exceeds a preset threshold, a particle swarm optimization iterative process based on the first modal parameter is introduced to perform a global search of the sensor measurement point combination. This enables the automatic determination of a measurement point placement scheme that is more conducive to modal differentiation under the condition of limited sensor quantity. This method fully utilizes the overall structural modal characteristics reflected by the finite element model, transforming the sensor measurement point selection problem into an optimization process based on modal parameter evaluation. This effectively reduces the correlation between different modes and improves the stability and consistency of modal identification results.

[0030] In some embodiments, such as Figure 2 As shown, step 101 includes steps 201 to 204.

[0031] Step 201: Establish the first geometric model of the vehicle entity based on the dimensional parameters and the material parameters of each vehicle body component.

[0032] Step 202: According to the preset mesh density and element type, mesh the vehicle body components in the first geometric model to obtain multiple shell element nodes.

[0033] Step 203: Equip each in-vehicle device with a mass point, and set each mass point in the corresponding position of the first geometric model according to the setting position of the corresponding in-vehicle device.

[0034] Step 204: Connect the mass points to the shell element nodes in the first geometric model using multi-point constraint elements (MPC) to obtain the vehicle finite element model.

[0035] In this embodiment of the invention, firstly, based on the actual geometric dimensions of the rail vehicle, the material parameters of the vehicle body components, and the setting of the auxiliary equipment inside the vehicle body, a three-dimensional geometric model consistent with the vehicle's physical structure and mass distribution characteristics is established using finite element analysis software, and then a vehicle finite element model is constructed on this basis.

[0036] Specifically, a complete three-dimensional geometric model of the vehicle body is established based on the design drawings and material parameters. The vehicle body is a typical large thin-walled shell structure, mainly composed of thin-walled components such as the floor, side walls, roof, end walls, and driver's cab. Considering these structural characteristics, four-node shell elements are preferentially used to discretize and model each thin-walled component during the finite element modeling process. This effectively characterizes the bending stiffness and in-plane stiffness characteristics of the vehicle body structure, thereby meeting the modeling requirements for the dynamic characteristics analysis of the rail vehicle body. For areas with complex local geometric transitions or irregular structures, triangular shell elements are used for mesh generation to improve the overall mesh quality and ensure the numerical stability of the finite element calculation.

[0037] During the mesh generation process, based on computational accuracy and efficiency, the modal calculation results under different mesh size conditions were compared and analyzed. Finally, the mesh size was controlled within the range of 20–30 mm, so that the obtained vehicle finite element model could accurately reflect the main modal characteristics of the vehicle body at a reasonable computational cost.

[0038] Furthermore, to simulate the impact of numerous auxiliary devices on the overall vehicle mass distribution and dynamic characteristics under actual maintenance conditions, a lumped mass modeling method is introduced into the finite element model. These auxiliary devices include onboard electrical equipment, seats, and interior trim. Equivalent lumped mass points are set at corresponding locations on the vehicle body, ensuring that the total added mass is consistent with the mass level of the actual vehicle under maintenance conditions. This total added mass is approximately 4.5 t, but this invention is not limited to this.

[0039] Specifically, based on the installation location and mass information of the vehicle's auxiliary equipment, the auxiliary equipment distributed within the same area is equivalently represented as several concentrated mass points. In the vehicle finite element model, structural nodes or node sets corresponding to the installation locations of the auxiliary equipment are selected, and the concentrated mass points are set at these corresponding locations. Subsequently, the concentrated mass points are connected to multiple shell element nodes of the vehicle body structure through multi-point constraint elements, allowing the concentrated mass points to participate in the transfer of inertial loads without introducing additional structural stiffness, thereby simulating the impact of the auxiliary equipment on the overall vehicle mass distribution and dynamic characteristics.

[0040] After completing the above finite element modeling steps, the result is as follows: Figure 3 The vehicle finite element model shown.

[0041] After the finite element model is constructed, modal analysis is further performed on the vehicle finite element model using finite element analysis software (such as the combined platform of HyperMesh and OptiStruct) to solve for the first 20 natural frequencies of the vehicle body and their corresponding mode shape vectors. The obtained mode shapes include typical vehicle body modal forms such as the first-order rhomboid mode, vertical bending mode, torsional mode, and lateral bending mode.

[0042] For example, the influence of damping factors was not considered in the above modal analysis process. This is because the material damping of the rail vehicle body structure is relatively small, and its impact on the identification results of natural frequencies and mode shapes is limited. In analyses aimed at obtaining modal parameters and optimizing measurement points, it can be ignored, thus treating the modal analysis as a linear eigenvalue problem. The process of establishing the above vehicle finite element model, including geometric model processing, mesh generation, material property assignment, and the setting of lumped mass and constraint relationships, is all completed in the finite element analysis software.

[0043] According to the above embodiments, by systematically modeling the vehicle entity in finite element analysis software, the structural characteristics of the vehicle body and the mass distribution of the in-vehicle equipment are uniformly incorporated into the finite element model construction process, enabling the resulting vehicle finite element model to realistically reflect the structural characteristics and dynamic behavior of the actual vehicle in its prepared state. By meshing the thin-walled components of the vehicle body according to a preset mesh density and element type, both the accuracy of the calculation of the main modal characteristics and computational efficiency are ensured. Furthermore, by equating the in-vehicle equipment with concentrated mass points and setting them in corresponding positions, the influence of auxiliary equipment on the overall vehicle mass distribution is reasonably reflected. Further, by using multi-point constraint elements to connect the concentrated mass points with multiple shell element nodes, the inertial load can be reasonably transferred without introducing additional local stiffness, thereby avoiding model stiffness distortion.

[0044] In some embodiments, such as Figure 4 As shown, step 103 includes steps 401 to 402.

[0045] Step 401: Based on the vehicle entity and the preset set of sensor position information, establish a second geometric model of the vehicle entity.

[0046] Step 402: Perform modal analysis based on the second geometric model to obtain the second modal parameters.

[0047] According to the above embodiments, by introducing a preset set of sensor location information onto the vehicle entity, a second geometric model corresponding to the actual modal test setup is established. This allows the spatial distribution of sensor measurement points during the modal test to be described and managed in a structured form, ensuring a clear and unambiguous correspondence between test data and the spatial positions of the vehicle entity. Modal analysis based on the second geometric model facilitates accurate correlation between measured acceleration response data and corresponding test nodes, improving the standardization and consistency of the modal parameter identification process. The second modal parameters obtained in this way can realistically reflect the dynamic characteristics of the vehicle under actual test conditions, providing reliable data support for subsequent finite element model verification and measurement point optimization. This helps improve the engineering applicability and reliability of the overall modal analysis and measurement point setup process.

[0048] In some embodiments, such as Figure 5 As shown, step 401 includes steps 501 to 502.

[0049] Step 501: Install multiple sensors and vibration excitation devices in the vehicle body according to the sensor location information set. Among them, the sensor is an acceleration sensor.

[0050] Step 502: Create a second geometric model based on the vehicle entity, and set the test nodes corresponding to each sensor on each section of the second geometric model.

[0051] In this embodiment of the invention, a benchmark modal test is conducted on a physical rail vehicle to obtain measured modal data of the vehicle, which serves as a reference benchmark for subsequent measurement point optimization. Modal testing equipment is installed on the physical vehicle in a static state. The modal testing equipment includes: a vibration excitation device, an acceleration sensor, and a data acquisition module, wherein the data acquisition module integrates data analysis software for modal analysis.

[0052] The vibration excitation device includes two electrically powered vibrators with an operating frequency range of 5 Hz–40 Hz, positioned at the front and middle of the vehicle body underframe respectively, to apply controllable excitation to the vehicle structure. The accelerometer system comprises 40 triaxial accelerometers with a sensitivity of 100 mV / g and an operating frequency range of 0.5 Hz–2000 Hz, used to acquire vibration response signals of the vehicle body in different directions. The data acquisition module employs a 24-bit high-resolution data acquisition instrument with a sampling frequency set to 2048 Hz. Data analysis software, such as LMS Test.Lab or ME'scope, is used to calculate the frequency response function and identify modal parameters.

[0053] During the baseline modal testing phase, triaxial accelerometers are uniformly positioned at typical cross-sectional locations on the vehicle body, including several sections at the front, middle, and rear. At each cross-section, 8–10 sensor measurement points are uniformly arranged along the transverse, vertical, and longitudinal directions, forming a total of 40 measurement points to cover the vehicle body's vibration response in the primary directions. This sensor arrangement serves as the initial sensor location information set for the baseline modal testing phase (e.g., ...). Figure 6 As shown in the figure, although the sensor setup will be further adjusted in the subsequent measurement point optimization process, it can obtain relatively comprehensive vibration response data of the main areas of the vehicle body in the initial stage.

[0054] For example, based on relevant industry standards (such as TB / T 3115-2005 "Test Methods for Dynamic Performance of Locomotives and Rolling Stock"), the excitation frequency range of the benchmark modal test is selected as 5 Hz to 40 Hz to cover the main low-order modal characteristics of rail vehicles.

[0055] In the LMS Test.Lab software, a test geometry model is created that matches the sensor placement locations on the actual vehicle, based on the vehicle's physical structure. This model is typically a simplified 3D wireframe or dot-line model, used to represent the vehicle's spatial structural relationships. Multiple test nodes are then created at the corresponding spatial locations within this test geometry model, according to the sensor placement scheme described above. Each test node has a unique node number and spatial coordinates.

[0056] According to the above embodiments, by setting the accelerometer and vibration excitation device in the vehicle body according to a preset set of sensor location information, and constructing a second geometric model corresponding to the actual modal test setup, the physical installation position of each sensor during the modal test can be clearly mapped in the second geometric model. By setting test nodes corresponding one-to-one with each sensor at each cross-sectional position of the second geometric model, it is beneficial to accurately correlate the measured vibration response data with the specific spatial position of the vehicle body, avoiding ambiguity between data and structural position, thereby improving the standardization and consistency of test data management and modal parameter identification processes.

[0057] In some embodiments, such as Figure 7 As shown, step 402 includes steps 701 to 703.

[0058] Step 701: Associate the acquisition channels of each sensor and excitation device with the corresponding test node.

[0059] Step 702: Acquire acceleration signals and excitation signals of each test node through the acquisition channel.

[0060] Step 703: Based on the acceleration signal and excitation signal, call the least squares complex exponential function to determine the second mode parameters.

[0061] In this embodiment of the invention, a corresponding measurement degree of freedom (DOF) is defined for each test node in the LMS Test.Lab software. This measurement degree of freedom includes the X, Y, and Z directions. A triaxial accelerometer corresponds to the X, Y, and Z measurement degrees of freedom for the same test node and is used to acquire the vibration response of the test node in different directions.

[0062] In the LMS Test.Lab software, each acquisition channel is associated with its corresponding test node and its measurement degrees of freedom through software configuration, thereby achieving a one-to-one correspondence between the acquisition channel and the spatial position and measurement direction of the test node. Specifically, each accelerometer corresponds to one acquisition channel, and the acquired acceleration signals are automatically mapped to the measurement degrees of freedom of the test node in the software.

[0063] The acquisition channels include an excitation force channel and a response channel. The excitation force channel is the acquisition channel corresponding to the excitation device, used to acquire the excitation signal applied to the vehicle structure. The response channel is the acquisition channel corresponding to the acceleration sensor, used to acquire the acceleration signal of the vehicle structure at each test node.

[0064] In the LMS Test.Lab software, modal identification analysis was performed on the acquired acceleration signals. Before entering the modal identification process, the acceleration signals underwent data preprocessing, including denoising, filtering, and normalization, to improve signal quality and reduce the impact of noise interference. Subsequently, the frequency response function (FRF) was calculated based on the preprocessed acceleration signals, and modal parameter identification methods were called for analysis to extract the natural frequencies of the vehicle structure and their corresponding mode shape vectors. Modal parameter identification methods included: multi-reference point, multi-output subspace identification methods or the least squares complex exponential method.

[0065] According to the above embodiments, by explicitly associating the acquisition channels of the sensors and excitation devices with the test nodes, the acquired acceleration signals and excitation signals can correspond one-to-one with the specific spatial locations of the vehicle structure, thereby ensuring the consistency and traceability of the test data in the spatial dimension. Synchronously acquiring the acceleration signals and corresponding excitation signals of each test node through the acquisition channels provides a data foundation for constructing a complete and reliable input-output relationship. Based on this, the least squares complex exponential method is used to identify modal parameters of the acceleration signals. This enables the stable extraction of second modal parameters such as the natural frequencies and mode shape vectors of the vehicle structure under multi-measurement point and multi-modal conditions, improving the accuracy and robustness of the modal identification results. The obtained second modal parameters can serve as an important basis for finite element model verification and measurement point optimization, helping to improve the engineering applicability and reliability of the overall modal analysis and measurement point setting process.

[0066] In some embodiments, typical modes (such as the first-order rhomboid mode and the vertical bending mode) corresponding to each order are selected from the modal parameters calculated from the vehicle finite element model and the modal parameters extracted through modal tests. Each modal parameter includes the mode shape vector of the corresponding order and its corresponding natural frequency. For the selected modes of the same order, the difference between the natural frequency output by the finite element model and the natural frequency extracted from the modal tests is calculated, and the frequency difference ratio is determined based on the difference.

[0067] For example, an acceptable threshold for the frequency difference ratio is set to 15%. When the frequency difference ratio is greater than 15%, the modal prediction accuracy of the current vehicle finite element model is determined to be unsatisfactory, and the relevant parameters of the vehicle finite element model are adjusted or optimized to update its output modal parameters. When the frequency difference ratio is less than or equal to 15%, the modal prediction result of the current vehicle finite element model is determined to meet the consistency requirements, and the modal parameters output by the current vehicle finite element model are accepted for subsequent measurement point optimization calculations based on the modal parameters.

[0068] In some embodiments, such as Figure 8 As shown, step 104 includes steps 801 to 806.

[0069] Step 801: Construct the MAC matrix for each particle based on its current position vector and the first mode shape vector of each mode in the particle swarm.

[0070] Step 802: Determine the fitness value of each particle based on the MAC matrix.

[0071] Step 803: Update the individual preferred position of each particle based on its fitness value.

[0072] Step 804: Determine the group optimal position of the particle swarm based on the updated individual optimal positions of each particle.

[0073] Step 805: Based on the group's preferred position, position update function, velocity update function, current position vector and current velocity vector of each particle, and the updated individual preferred position, determine the updated position vector and velocity vector of each particle.

[0074] Step 806: Iteratively execute the above steps. When the preset termination condition is met, obtain the combination of measuring points corresponding to the preferred position of the group.

[0075] According to the above embodiments, by modeling the sensor measurement point selection process as a global optimization problem based on particle swarm optimization, and utilizing different sensor setup schemes represented by each particle, and using the MAC matrix constructed with the modality guarantee criterion as the evaluation basis, a quantitative evaluation of the merits of measurement point combinations is achieved. By continuously updating the individual and collective optimal positions of particles during the iteration process, the particle swarm can be guided to gradually converge to a measurement point arrangement scheme that is more conducive to modality differentiation within the global search space. This method avoids the problem of missing important modal information caused by traditional reliance on manual experience or local search strategies, effectively reduces the correlation between different modes, improves the stability and accuracy of modality recognition, and achieves automatic optimization of sensor measurement point arrangement under the condition of limited number of sensors, demonstrating good engineering applicability and result reliability.

[0076] In some embodiments, such as Figure 9 As shown, step 901 includes steps 901 to 903.

[0077] Step 901: Determine the combination of measurement points for each particle and its corresponding finite element degrees of freedom based on the current position vector of each particle.

[0078] Step 902: Based on the combination of measurement points of each particle and its corresponding finite element degrees of freedom, extract the first mode shape vector of each particle's mode to obtain the first mode shape sub-vector of each particle's mode.

[0079] Step 903: Based on the first mode shape subvectors of each particle's first mode, determine the MAC matrix between any two first mode shape subvectors of each particle's first mode.

[0080] In this embodiment of the invention, the position vector of the p-th particle at the t-th iteration is represented as:

[0081] in, This represents the total number of candidate sensor measurement points. It is a binary variable; when When, it indicates that the k-th candidate sensor measurement point is selected. When the k-th candidate sensor measurement point is not selected, it indicates that the k-th candidate sensor measurement point is not selected. The candidate sensor measurement points are represented in the form of node number and measurement direction, for example (node ​​12, X direction), where each measurement direction corresponds to a measurement degree of freedom.

[0082] After obtaining the modal parameters of each order output from the vehicle finite element model, the modal parameters include: the mode shape vectors of each mode and their corresponding natural frequencies. This is based on the position vector of the p-th particle. The corresponding sensor position information set and its corresponding measurement degrees of freedom are used to extract components from the full-degree-of-freedom mode shape vectors of each order mode, obtaining the mode shape sub-vectors corresponding to the candidate sensor measurement points and their measurement degrees of freedom. Norm normalization is then performed on the mode shape sub-vectors to obtain the set of normalized mode shape sub-vectors corresponding to each order for the p-th particle. Where N is the modal order involved in the optimization.

[0083] Based on the set of normalized mode shape sub-vectors corresponding to the p-th particle. Construct a modal assurance criterion matrix between any two modes. The modal assurance criterion (MAC) matrix is ​​used to measure the similarity between different mode shape vectors, and its calculation method is shown in formula (1):

[0084] in, Let i be the sub-vector of the i-th mode shape of the p-th particle. Let be the subvector of the j-th mode shape of the p-th particle. The MAC value ranges from... When the MAC value is close to 1, it indicates that there is a high correlation between the corresponding mode shapes; when the MAC value is close to 0, it indicates that there is good independence between the corresponding mode shapes.

[0085] According to the above embodiments, in the process of optimizing sensor measurement points, minimizing the off-diagonal elements of the MAC matrix is ​​used as the optimization objective, thereby reducing the correlation between different modes and improving the independence and accuracy of mode recognition.

[0086] To avoid missing important modes that might result from relying solely on manual experience to set sensor measurement points, this invention further introduces Particle Swarm Optimization (PSO) algorithm to perform global optimization of the sensor measurement points. In PSO, each particle represents a sensor setup scheme, and its position vector characterizes the set of sensor position information.

[0087] In some embodiments, steps 702 to 706 specifically include the following calculation process.

[0088] Based on the aforementioned modal guarantee criterion matrix, a fitness function for the particle swarm optimization is constructed to evaluate the merits of different sensor configuration schemes. The fitness function is:

[0089] in, This represents the sub-vector of the i-th mode shape corresponding to the p-th particle. This represents the sub-vector of the j-th mode shape corresponding to the p-th particle. The fitness function is optimized by minimizing the correlation between different modes, thereby improving the independence between modes. Fitness Function Used to determine the position vector of the p-th particle at the t-th iteration. The fitness value of the particle is obtained by evaluating the corresponding set of sensor location information.

[0090] The fitness value of the p-th particle in the t-th iteration If the current position vector of the particle is less than the particle's historical best fitness value, then... Updated to the individual optimal position vector of the particle. The individual optimal position vector of all particles. In the process, the optimal position vector of the individual with the smallest fitness value is selected as the population optimal position vector of the particle swarm. .in, This represents the component of the individual's optimal position vector in the dimension corresponding to the k-th candidate sensor measurement point. The component of the optimal position vector of the group in the dimension corresponding to the k-th candidate sensor measurement point.

[0091] Based on this, the individual optimal position vector of the p-th particle Components in the k-th dimension Optimal position vector of the group Components in the k-th dimension The component of the particle's position vector in the k-th dimension at the t-th iteration. and the corresponding velocity vector Substituting into the velocity update formula, we update the velocity of the p-th particle in the k-th dimension to obtain the updated velocity components. .

[0092]

[0093] in, This is the inertia weighting coefficient. This represents the velocity vector of the p-th particle in the dimension corresponding to the k-th candidate sensor measurement point. Let represent the component of the optimal position of the p-th particle in the k-th dimension. This represents the position component of the p-th particle at the k-th candidate sensor measurement point during the t-th iteration. For individual learning factors, As a group learning factor, and It is a random number. This represents the components of the optimal position of the group in k dimensions.

[0094] The component of the position vector in the k-th dimension at the t-th iteration. and the updated velocity vector Substituting into the position update formula, we obtain the updated position vector. .

[0095]

[0096] in, This represents the position component of the p-th particle at the k-th candidate sensor measurement point during the t-th iteration. This represents the updated velocity component of the p-th particle at the dimension corresponding to the k-th candidate sensor measurement point.

[0097] During the iteration process, when the preset maximum number of iterations or the convergence threshold condition of the fitness function is met, the particle swarm iteration calculation is stopped, and the set of sensor position information corresponding to the optimal position G of the swarm is output as the optimal sensor setting scheme.

[0098] For example, the parameters of the particle swarm optimization algorithm are set as follows: number of particles is 50, maximum number of iterations is 200, and inertia weight is... The initial value was set to 0.8, and gradually decreased to 0.4 during the iteration process; this is the individual learning factor. Set to 1.5, group learning factor The value is set to 1.5. In each iteration, each particle represents a sensor configuration scheme, and its position vector uses a 0 / 1 encoding to indicate whether the corresponding candidate node is selected as a sensor measurement point. Through iterative updates of the particle velocity and position vectors, the particle swarm gradually converges towards the solution that minimizes the modality guarantee criterion value. After optimization, approximately 36 sensor measurement points are finally obtained, a reduction of about 10% compared to the initial 40 measurement points, while maintaining the integrity of the modality recognition results.

[0099] According to the above embodiments, 36 sensors were arranged on the vehicle body according to the optimized sensor setup scheme, and modal tests were repeated under the same test conditions and parameter settings as the baseline modal test. Test results show that the optimized sensor arrangement scheme can completely identify the first 15 main modes of the vehicle, with corresponding modal frequency errors all less than 2%. Simultaneously, the average value of the off-diagonal elements in the modal guarantee criterion matrix decreased from 0.25 before optimization to 0.12, indicating a significant reduction in the correlation between different modes and an effective improvement in modal independence. Furthermore, the reduction in the number of sensors reduces the complexity of test wiring, correspondingly reduces installation and debugging workload, and shortens test preparation time by approximately 15%. Therefore, the above method effectively reduces the number of sensor points while ensuring modal recognition accuracy and modal independence, helping to reduce modal test costs and improve test implementation efficiency, and has good engineering application value.

[0100] This application provides a vehicle modal recognition measurement point setting device based on particle swarm optimization (PSO) algorithm, applied to the aforementioned vehicle modal recognition measurement point setting method based on PSO algorithm. This vehicle modal recognition measurement point setting device based on PSO algorithm is based on the same inventive concept as the vehicle modal recognition measurement point setting method based on PSO algorithm in one embodiment of this application, and the principle of solving the problem is similar. Therefore, the implementation of the vehicle modal recognition measurement point setting device based on PSO algorithm is the same as the implementation of the vehicle modal recognition measurement point setting method based on PSO algorithm in one embodiment of this application, and repeated details will not be described again. As used below, the terms "unit" or "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0101] like Figure 10 As shown, the vehicle modality recognition measurement point setting device based on particle swarm optimization algorithm includes: Finite element model construction module 1001 is used to construct a vehicle finite element model based on the vehicle's size parameters, the material parameters of each vehicle body component, and the location of each in-vehicle device. The first modal parameter acquisition module 1002 is used to solve the eigenvalues ​​of the vehicle finite element model to obtain the first modal parameters; the first modal parameters include: the first mode shape vector of each mode and its corresponding first natural frequency; The reference modal test module 1003 is used to conduct modal tests based on the vehicle entity and preset sensor setting position information to obtain second modal parameters; the second modal parameters include: the second modal shape vector of each order mode and its corresponding second natural frequency; The measurement point optimization triggering and solving module 1004 is used to determine the optimal sensor setting position information by performing particle swarm optimization iterative solution based on the first modal parameters when the difference between the first natural frequency and the second natural frequency is greater than the preset frequency difference.

[0102] In some embodiments, the finite element model construction module 1001 includes: The first geometric model establishment submodule is used to establish the first geometric model of the vehicle entity based on the dimensional parameters and the material parameters of each vehicle body component. The mesh generation submodule is used to mesh each vehicle body component in the first geometric model according to a preset mesh density and element type, thereby obtaining multiple shell element nodes. The centralized mass modeling submodule is used to represent each in-vehicle device as a mass point, and to set each mass point at the corresponding position of the first geometric model according to the setting position of the corresponding in-vehicle device. The multi-point constraint connection submodule is used to connect the mass points to each shell element node in the first geometric model through multi-point constraint elements (MPC) to obtain the vehicle finite element model.

[0103] In some embodiments, the reference modal testing module 1003 includes: The second geometric model establishment submodule is used to establish a second geometric model of the vehicle entity based on the vehicle entity and a preset set of sensor position information. The experimental modal analysis submodule is used to perform modal analysis based on the second geometric model to obtain the second modal parameters.

[0104] In some embodiments, the second geometric model building submodule includes: A sensor and vibration device arrangement unit is used to arrange multiple sensors and vibration devices in the vehicle body according to the sensor position information set; The test node setting unit is used to create the second geometric model based on the vehicle entity, and set the test nodes corresponding to each sensor on each section of the second geometric model.

[0105] In some embodiments, the test modal analysis submodule includes: The channel association unit is used to associate the acquisition channels of each of the sensors and the excitation device with the corresponding test nodes; The signal acquisition unit is used to acquire the excitation signal and the acceleration signal of each of the test nodes through the acquisition channel; The modal parameter identification unit is used to determine the second modal parameters by calling the least squares complex exponential function based on the acceleration signal and the excitation signal.

[0106] In some embodiments, the measurement point optimization triggering and solving module 1004 includes: The MAC matrix construction submodule is used to construct the MAC matrix of each particle based on the current position vector of each particle in the particle swarm and the first mode shape vector of each mode. The fitness calculation submodule is used to determine the fitness value of each particle based on the MAC matrix; The Individual Preferred Position Update Submodule is used to update the individual preferred position of each particle based on its fitness value. The group optimal position determination submodule is used to determine the group optimal position of the particle swarm based on the updated individual optimal positions of each particle. The position and velocity update submodule is used to determine the updated position vector and velocity vector of each particle based on the group's preferred position, position update function, velocity update function, the current position vector and current velocity vector of each particle, and the updated individual preferred position. The iteration termination and output submodule is used to iteratively execute the above steps. When the preset termination condition is met, the combination of measurement points corresponding to the preferred position of the group is obtained.

[0107] In some embodiments, the MAC matrix construction submodule includes: The measurement point combination determination unit is used to determine the measurement point combination of each particle and its corresponding finite element degrees of freedom based on the current position vector of each particle. The modal shape extraction unit is used to extract the first modal shape vector of each particle's first mode based on the combination of measurement points of each particle and its corresponding finite element degrees of freedom, so as to obtain the first modal shape sub-vector of each particle's first mode. The MAC matrix generation unit is used to determine the MAC matrix between the first mode shape vectors of any two modes of each particle based on the first mode shape vectors of each particle's first mode shape vectors.

[0108] Figure 11 This is a schematic diagram of the physical structure of a computer device provided in an embodiment of the present invention, such as... Figure 11 As shown, the computer device includes: a processor 1101, a memory 1102, and a bus 1103.

[0109] The processor 1101 and the memory 1102 communicate with each other via the bus 1103.

[0110] The processor 1101 is used to call program instructions in the memory 1102 to execute the methods provided in the above-described method embodiments.

[0111] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for setting vehicle modal recognition measurement points based on particle swarm optimization.

[0112] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-described method for setting vehicle modal recognition measurement points based on particle swarm optimization algorithm.

[0113] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0114] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0115] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0116] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0117] In the description of this specification, the references to terms such as "an embodiment," "a specific embodiment," "some embodiments," "for example," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0118] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for setting measurement points for vehicle modal recognition based on particle swarm optimization algorithm, characterized in that, include: A finite element model of the vehicle is constructed based on the vehicle's size parameters, the material parameters of each vehicle body component, and the location of each in-vehicle device. The first modal parameters are obtained by solving the eigenvalues ​​of the finite element model of the vehicle. The first modal parameters include: the first mode shape vector of each mode and its corresponding first natural frequency; Modal tests are conducted based on the vehicle entity and preset sensor location information to obtain second modal parameters; the second modal parameters include: the second modal shape vector of each mode and its corresponding second natural frequency; When the difference between the first natural frequency and the second natural frequency is greater than a preset frequency difference, particle swarm optimization iterative solution is performed based on the first modal parameters to determine the optimal sensor setting position information.

2. The method according to claim 1, characterized in that, The finite element model of the vehicle is constructed based on the vehicle's size parameters, the material parameters of each vehicle body component, and the location of each in-vehicle device, including: A first geometric model of the vehicle entity is established based on the dimensional parameters and the material parameters of each vehicle body component. According to the preset mesh density and element type, the vehicle body components in the first geometric model are meshed to obtain multiple shell element nodes; Each in-vehicle device is represented as a mass point, and each mass point is set at the corresponding position in the first geometric model according to the setting position of the corresponding in-vehicle device. The mass points are connected to the shell element nodes in the first geometric model by multi-point constraint elements (MPC) to obtain the vehicle finite element model.

3. The method according to claim 1, characterized in that, The modal test based on the vehicle entity and preset sensor location information is used to obtain the second modal parameters, including: A second geometric model of the vehicle entity is established based on the vehicle entity and a preset set of sensor location information. Modal analysis is performed based on the second geometric model to obtain the second modal parameters.

4. The method according to claim 3, characterized in that, The process of establishing a second geometric model of the vehicle entity based on the vehicle entity and a preset set of sensor location information includes: According to the sensor location information set, multiple sensors and vibration excitation devices are installed in the vehicle body; The second geometric model is created based on the vehicle entity, and test nodes corresponding to each sensor are set on each cross section of the second geometric model.

5. The method according to claim 4, characterized in that, The modal analysis based on the second geometric model to obtain the second modal parameters includes: Associate the acquisition channels of each of the aforementioned sensors and the excitation device with the corresponding test nodes; The excitation signal and the acceleration signal of each test node are acquired through the acquisition channel; The second mode parameters are determined by calling the least squares complex exponential function based on the acceleration signal and the excitation signal.

6. The method according to claim 1, characterized in that, The step of performing particle swarm optimization iterative solution based on the first modal parameters to determine the optimal sensor placement information includes: Based on the current position vector of each particle in the particle swarm and the first mode shape vector of each mode, the MAC matrix of each particle is constructed. The fitness value of each particle is determined based on the MAC matrix; The optimal position of each particle is updated based on its fitness value. Based on the updated individual optimal positions of each particle, the group optimal position of the particle swarm is determined; Based on the group's preferred position, position update function, velocity update function, current position vector and current velocity vector of each particle, and the updated individual preferred position, determine the updated position vector and velocity vector of each particle; The above steps are executed iteratively. When the preset termination condition is met, the combination of measuring points corresponding to the preferred position of the group is obtained.

7. The method according to claim 6, characterized in that, The construction of the MAC matrix for each particle based on its current position vector and the first mode shape vector of each mode in the particle swarm includes: Based on the current position vector of each particle, determine the combination of measurement points for each particle and its corresponding finite element degrees of freedom; Based on the combination of measurement points of each particle and its corresponding finite element degrees of freedom, the first mode shape vector of each particle's mode is extracted to obtain the first mode shape sub-vector of each particle's mode. Based on the first mode shape subvectors of each particle's first mode, determine the MAC matrix between any two first mode shape subvectors of each particle's first mode.

8. A vehicle modal recognition measurement point setting device based on particle swarm optimization algorithm, characterized in that, include: The finite element model building module is used to build a finite element model of the vehicle based on the vehicle's size parameters, the material parameters of each vehicle body component, and the location of each in-vehicle device. The first modal parameter acquisition module is used to solve the eigenvalues ​​of the vehicle finite element model to obtain the first modal parameters; The first modal parameters include: the first mode shape vector of each mode and its corresponding first natural frequency; The reference modal test module is used to conduct modal tests based on the vehicle entity and preset sensor setting position information to obtain second modal parameters; the second modal parameters include: the second modal shape vector of each mode and its corresponding second natural frequency; The measurement point optimization triggering and solving module is used to determine the optimal sensor setting position information by performing particle swarm optimization iterative solution based on the first modal parameters when the difference between the first natural frequency and the second natural frequency is greater than the preset frequency difference.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

11. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 7.

Images