Durability testing device for four-channel chassis component of passenger car

Abstract

The invention discloses a passenger car four-channel chassis component durability testing device, and relates to the field of durability testing, the passenger car four-channel chassis component durability testing device comprises a testing frame main body, the bottom end in the testing frame main body is provided with a first tractor, the left side in the testing frame main body is provided with a second tractor, and the right side in the testing frame main body is provided with a third tractor; a control module is installed on the front face of the testing frame body, and the first tractor, the second tractor and the third tractor are used for fixing a tested part and conducting traction to conduct transverse, vertical and torsional fatigue testing. A prediction model is constructed, a test state of a certain period in the future is predicted and obtained based on test data of a current test item, and a mapping relation between an initial period damage state of a test part and a fission and deformation state in the future period is analyzed; therefore, the long-term test performance of the tested object can be quickly obtained through a short-term actual test.

Description

Technical Field

[0001] The invention relates to the technical field of durability testing, in particular to a durability testing device for a four-channel chassis component of a passenger car. Background Art

[0002] With the development of the automobile industry, people's requirements for the safety, reliability and service life of passenger cars are constantly increasing. Modern automobile chassis technology is becoming increasingly complex, involving a variety of materials and components. Automobile manufacturers need to conduct systematic and comprehensive tests on the fatigue, load bearing capacity and damage evolution process of chassis components under various working conditions. During driving, chassis components are simultaneously subjected to multiple composite loads such as lateral, vertical and torsional loads. Durability testing can help manufacturers develop more durable products, thereby reducing resource waste and environmental impact. The rise of big data and machine learning enables the collection and analysis of large amounts of data during the testing process, thereby improving the accuracy and efficiency of testing. Through data analysis, manufacturers can better predict product performance and maintenance needs; Traditional test benches mostly use single-channel or fixed-direction loading, which can only simulate a single vertical or lateral fatigue load. It is difficult to truly reflect the multi-directional composite stresses that chassis components are subjected to in complex road conditions. In actual driving scenarios, the frequency and amplitude of the impact force they bear are not uniform within a short period. Traditional methods predict damage based on empirical formulas or simplified mechanical models, which are difficult to capture complex behaviors such as nonlinear material degradation and microcrack propagation, resulting in large deviations in life prediction. Manual design is required and relies on a fixed sequence, such as static loading followed by dynamic loading, which makes it difficult to quickly respond to test project requirements for new materials or design changes. Summary of the invention

[0003] (I) Technical problems to be solved: In view of the above-mentioned shortcomings of the prior art, the present invention provides a four-channel chassis component durability testing device for passenger cars, which can effectively solve the problems of the prior art.

[0004] (II) Technical solution: To achieve the above objectives, the present invention is implemented through the following technical solutions: The present invention discloses a passenger car four-channel chassis component durability test device, comprising a test frame body, a tractor 1 is installed at the bottom end of the test frame body, a tractor 2 is installed on the left side of the test frame body, a tractor 3 is installed on the right side of the test frame body, and a control module is installed on the front of the test frame body, wherein: The tractor 1, tractor 2 and tractor 3 are used to fix the tested component and to pull it to perform lateral, vertical and torsional fatigue tests; The control module comprises: A test execution unit for registering various test items of a combination, selecting and executing a certain combined test item, and obtaining the test parameters of the component under test in the current cycle; An analysis and prediction unit for performing model analysis on the test parameters obtained from the combined test item and predicting the state data of the damage evolution of the test component in the future cycle; A data generation module for simulating and generating a corresponding virtual test item combination according to the state data of the damage evolution of the test component in the future cycle; An item adjustment module for analyzing the test parameters in the current combined test item according to the difference between the current test data and the predicted data, as well as the virtual test item combination, and calculating the adjusted test parameter combination by using an optimization algorithm as the solution for the next round of testing; An active verification module for generating control instructions for the first tractor, the second tractor, and the third tractor according to the adjusted combined test item solution, obtaining the verified data and comparing it with the predicted data as the basis for updating the prediction model parameters.

[0005] Furthermore, a sub-module is deployed at the lower level of the test execution unit. The sub-module includes: a project management module, a combination selection module, and a data acquisition module. The project management module and the combination selection module are interconnected through a wireless network, and the combination selection module and the data acquisition module are interconnected through a wireless network, where: The project management module is used to record the combinations of the corresponding test items of the first tractor, the second tractor, and the third tractor, and classify the entered combined test items; The combination selection module is used to select the combined test item for the current cycle according to the current test status, component performance, and trend data of historical tests; The data acquisition module is used to perform a complete cycle of durability test on the specified test component according to the selected combined test item, and collect the state of the test equipment and the parameter data of the test component during the test process through sensors.

[0006] Furthermore, the process of the combination selection module for item selection processing is as follows: Step 1: Obtain the estimated performance parameters of the current test component and the trend data of historical tests. The trend data includes the change law of component performance parameters, the damage accumulation rate, and the correlation characteristics of test interruption events in historical tests; Step 2: Dynamically match the component performance parameters in historical tests with the combined test items pre-stored in the project management module, and extract the candidate test item combinations associated with the current test component type, preset durability target, and damage-sensitive direction; Step 3: Preset a scoring model, and perform multi-dimensional scoring on the candidate test item combinations based on the scoring model. The scoring dimensions include the test efficiency weight, the damage feature coverage weight, and the effectiveness feedback weight of historical similar test schemes. Step 4: Screen the candidate combinations with scores higher than the preset threshold, and determine the optimal combined test items for the current cycle through a dynamic programming algorithm according to the real-time test resource occupancy status and the optimization target constraint conditions. Step 5: Send the optimal combined test items to the data acquisition module to trigger the execution of a full-cycle test, and record the currently selected logical parameters as the updated input of the historical trend data.

[0007] Furthermore, sub-modules are deployed under the analysis and prediction unit. The sub-modules include: a damage analysis module, a model construction module, and a damage prediction module. The damage analysis module and the model construction module are interconnected through a wireless network, and the model construction module and the damage prediction module are interconnected through a wireless network. Among them: The damage analysis module is used to perform preliminary analysis on the collected data, identify various damage states that occur in the test components during the loading process, quantitatively record the damage states of each test cycle, establish a corresponding damage state file in combination with historical data, and record the time, location, and damage degree of each detection point. The model construction module is used to construct a prediction model using a machine learning algorithm based on the damage state data obtained in a single-cycle test. The prediction model establishes a mapping relationship between the initial cycle damage state and the fission and deformation states in future cycles, and screens the key feature variables related to damage evolution. The damage prediction module is used to predict the damage states in several future cycles using the constructed prediction model. The model inputs the various test combination data of the current cycle and outputs the state data of the damage evolution of the test components in future cycles.

[0008] Furthermore, in the process of the model construction module screening key feature variables, by receiving the test combination data collected in real time by sensors in the current cycle, the loading force values, displacement amounts, torque parameters, and component deformation feature data of each test component are obtained, the obtained feature data is normalized, and the preset damage feature variables are extracted. The damage feature variables include: stress concentration coefficient, strain gradient value, and material fatigue cumulative factor.

[0009] Furthermore, the prediction model constructed by the model construction module establishes multi-cycle time series associations through an LSTM neural network, outputs damage state prediction data including several future test cycles, and decomposes the prediction data into specific damage types, occurrence locations, damage degrees, and prediction time points. Among them, the damage types include at least two of crack growth rate, plastic deformation amount, and connector loosening gap.

[0010] Furthermore, during the operation of the project adjustment module, a difference quantification model between the measured damage parameters in the current cycle and the predicted data is established, and the deviation characteristic quantity of the test parameters is extracted. Perform multi-objective correlation analysis on the simulation data of the virtual test project combination and the deviation characteristic quantity, and construct optimization constraint conditions including load intensity, action frequency, and test timing sequence. Based on the particle swarm optimization algorithm, perform iterative optimization on the current combined test parameters. With the objective function of minimizing the prediction residual and maximizing the damage coverage rate under the condition of meeting the constraint conditions, solve to obtain the adjusted test parameter combination. Verify the confidence level of the optimized parameter combination with the virtual test results, and screen the solutions that meet the preset reliability threshold as the test instruction set for the next cycle.

[0011] Furthermore, during the operation stage of the active verification module, when the deviation between the actual test data and the predicted data exceeds the preset threshold, automatically trigger the retraining process of the prediction model in the model construction module. Adopt the sliding time window algorithm, retain the test data in the recent several cycles as the training sample set, and adjust the neural network hidden layer node number and activation function combination parameters of the prediction model constructed by the model construction module through cross-validation.

[0012] Furthermore, the test execution unit is connected to the repository module through wireless network interaction. The repository module is used to centrally control all the collected data and calculation data, perform classified storage combining cloud and local, and provide an online indexing interface for historical classified data.

[0013] Furthermore, the project adjustment module is connected to the active adjustment module through an electrical medium. The active adjustment module is used to receive the test plan for the next round output by the project adjustment module, provide visual display, and provide user identity verification. Provide the modification permission of the test plan to the verified user. After the modification is completed, feedback the modified data to the project adjustment module for application.

[0014] (III) Beneficial effects: Adopting the technical solution provided by the present invention, compared with the known prior art, it has the following beneficial effects. 1. By constructing a prediction model, based on the test data of the current test project, predict and obtain the test state of a future cycle, analyze the mapping relationship between the initial cycle damage state of the test component and the fission and deformation states of future cycles. Furthermore, the long-term test performance of the object under test can be quickly obtained through short-term actual tests, thereby reducing the test time and ensuring accuracy, and effectively reducing the errors caused by various factors during the actual test process.

[0015] 2. By utilizing an optimization algorithm, the device can adjust the test parameter combination in real time according to the difference between the current test results and the predicted data, so as to adapt to different test requirements. By generating a virtual test project combination, it can simulate before the actual test, reducing the test cost and risk. By combining historical data and real-time test data to establish a damage status file, it can better understand the performance evolution trend of components. By optimizing the test plan, unnecessary repeated tests and resource waste are reduced.

[0016] 3. Through actuators independently controlled in the lateral, vertical, and torsional directions, multi-degree-of-freedom dynamic coupling loading is achieved to simulate the multi-axis linkage effect of the actual forces on the chassis components during vehicle driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is the overall structure diagram of the present invention; Figure 2 It is the overall framework schematic diagram of the present invention; Figure 3 It is the framework schematic diagram of the control module in the present invention.

[0019] The reference numerals in the drawings respectively represent: 1. Main body of the test stand; 2. Tractor 1; 3. Tractor 2; 4. Tractor 3; 5. Control module; 51. Test execution unit; 511. Project management module; 512. Combination selection module; 513. Data acquisition module; 52. Analysis and prediction unit; 521. Damage analysis module; 522. Model construction module; 523. Damage prediction module; 53. Data generation module; 54. Project adjustment module; 55. Active verification module; 56. Active adjustment module; 57. Repository module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0021] The following further describes the present invention with reference to the embodiments.

[0022] ① Embodiment 1: The four-channel chassis component durability test device for a passenger vehicle in this embodiment is as shown Figures 1 - 3 in the figure, and includes a test rack main body 1. A first traction device 2 is installed at the bottom end inside the test rack main body 1. A second traction device 3 is installed on the left side inside the test rack main body 1. A third traction device 4 is installed on the right side inside the test rack main body 1. A control module 5 is installed on the front of the test rack main body 1, where: The first traction device 2, the second traction device 3, and the third traction device 4 are used to fix the components to be tested and perform lateral, vertical, and torsional fatigue tests; The control module 5 includes: a test execution unit 51, which is used to register various combined test items, select and execute a certain combined test item, and obtain the test parameters of the component to be tested in the current cycle. Sub-modules are deployed under the test execution unit 51, and the sub-modules include: a project management module 511, a combination selection module 512, and a data acquisition module 513. The project management module 511 and the combination selection module 512 are interconnected through a wireless network. The combination selection module 512 and the data acquisition module 513 are interconnected through a wireless network, where: The project management module 511 is used to record the combinations of the first traction device 2, the second traction device 3, and the third traction device 4 corresponding to various test items, and classify the entered combined test items; The combination selection module 512 is used to perform project selection processing according to the estimated performance of the current test component and the trend data of its historical tests, and obtain the combined test item in the current cycle; The data acquisition module 513 is used to perform a complete cycle of durability test on the specified test component according to the selected combined test item, and collect the state data of the test equipment and the parameter data of the test component during the test process through sensors; An analysis and prediction unit 52 is used to perform model analysis on the test parameters obtained from the combined test item, and predict the state data of the damage evolution of the test component in the future cycle. Sub-modules are deployed under the analysis and prediction unit 52, and the sub-modules include: a damage analysis module 521, a model construction module 522, and a damage prediction module 523. The damage analysis module 521 and the model construction module 522 are interconnected through a wireless network. The model construction module 522 and the damage prediction module 523 are interconnected through a wireless network, where: The damage analysis module 521 is used to perform preliminary analysis on the collected data, identify various damage states that occur in the test component during the loading process, quantitatively record the damage states in each test cycle, establish a corresponding damage state file in combination with historical data, and record the time, location, and damage degree of each detection point; The model construction module 522 is configured to construct a prediction model based on the damage state profiles obtained in the single-cycle test using machine learning algorithms. The prediction model establishes a mapping relationship between the initial cycle damage state and the future cycle fission and deformation states, and screens key feature variables. During the process of screening key feature variables, by receiving the test combination data collected in real time by sensors within the current cycle, the loading force values, displacement amounts, torque parameters, and component deformation feature data of each test component are obtained, the obtained feature data is normalized, and preset damage feature variables are extracted. The damage feature variables include: stress concentration coefficient, strain gradient value, and material fatigue cumulative factor; The prediction model establishes multi-cycle time series associations through an LSTM neural network and outputs damage state prediction data for several future test cycles. The prediction data is decomposed into specific damage types, occurrence locations, damage degrees, and prediction time points, where the damage types include at least two of crack growth rate, plastic deformation amount, and connector looseness gap; The damage prediction module 523 is configured to obtain the prediction model constructed by the model construction module 522, input the key feature variables of various test data in the current cycle into the prediction model, and the prediction model outputs the state data of the damage evolution of the test component in the future cycle; The data generation module 53 is configured to simulate and generate a corresponding virtual test project combination according to the state data of the damage evolution of the test component in the future cycle; The project adjustment module 54 is configured to analyze each test parameter in the current combined test project according to the difference between the current test data and the prediction data, and the virtual test project combination, and calculate the adjusted test parameter combination using an optimization algorithm as the plan for the next round of test. The project adjustment module 54 is electrically connected to the active adjustment module 56. The active adjustment module 56 is configured to receive the plan for the next round of test output by the project adjustment module 54, provide visual display, and provide user identity verification, provide the modification permission of the test plan to the user who passes the verification, and after the modification is completed, feedback the modified data to the project adjustment module 54 for application; A user identity verification mechanism is designed to ensure that only verified users can modify the test plan, improving the security of the system; The active verification module 55 is configured to generate control instructions for the tractor 1, tractor 2, and tractor 3 according to the adjusted combined test project plan, obtain the verified data and compare it with the prediction data as the basis for updating the prediction model parameters.

[0023] As a preferred implementation manner in this embodiment, such as Figure 3As shown, the test execution unit 51 is interactively connected to the repository module 57 via a wireless network. The repository module 57 is used to control all collected data and calculated data, perform classified storage combining cloud and local, and provide an online indexing interface for historical classified data. Through the combination of cloud and local, classified storage and online indexing of data are realized, which facilitates the review of historical data at any time and promotes data sharing and collaborative work.

[0024] Compared with existing technologies, it can simultaneously conduct lateral, vertical and torsional fatigue tests, and can more comprehensively test the durability of chassis components, thereby better reflecting the performance under actual use conditions. Through the design of combined test items and sub-modules, it can intelligently record, classify and process test data. This design makes the test process more efficient, and can analyze historical data and optimize subsequent test strategies; The multi-period prediction model built using the LSTM neural network can predict the future damage evolution state of components in real time, helping engineers identify potential risks in advance and reduce the probability of failure. By analyzing the differences between current test data and predicted data and using optimization algorithms to adjust the test parameter combination, the test plan can be continuously improved to improve the accuracy and effectiveness of the test. Based on the current test data and prediction results, virtual test items can be generated to optimize the test process, improve test efficiency, and reduce resource waste.

[0025] ② Embodiment 2: In other aspects, this embodiment also provides another optimization mechanism based on Embodiment 1, specifically a process of project selection and processing, specifically: Step 1: Obtain the estimated performance parameters of the current test component and the trend data of historical tests. The trend data includes the change rules of component performance parameters in historical tests, the damage accumulation rate, and the correlation characteristics of test interruption events; Step 2: Dynamically match component performance parameters in historical tests with combined test items pre-stored in the item management module 511, and extract candidate test item combinations associated with the current test component type, preset durability target, and damage-sensitive direction; Step 3: Preset a scoring model and perform multi-dimensional scoring on the candidate test item combination based on the scoring model. The scoring dimensions include the test efficiency weight, the damage feature coverage weight, and the effectiveness feedback weight of historical similar test solutions; Step 4: Filter candidate combinations with scores higher than the preset threshold, and determine the optimal combination test items for the current cycle through a dynamic programming algorithm based on the real-time test resource occupancy status and optimization target constraints; Step 5: Send the optimal combination test items to the data acquisition module 513 to trigger the execution of the full cycle test, and record the currently selected logic parameters as the update input of the historical trend data.

[0026] Compared with the prior art, by obtaining the estimated performance parameters of the current test component and the historical test trend data, this mechanism can perform dynamic matching, which is more scientific and accurate than static or experience-based matching methods, and can quickly adapt to different test conditions and component characteristics; Introducing a scoring model to perform multi-dimensional scoring on the candidate test item combinations can comprehensively consider multiple influencing factors, making the item selection more comprehensive and objective, reducing the errors caused by subjective judgment. Among the selected candidate combinations, the occupied state of real-time test resources and the optimization target constraint conditions can be considered, and the combination is optimized through the dynamic programming algorithm, ensuring that the best test effect can be achieved and the utilization efficiency of resources can be improved under limited resources; Recording the currently selected logical parameters as the updated input of the historical trend data helps to optimize the subsequent test plan, continuously improve the intelligent level and adaptive ability of the test. Through the effective screening and optimization of test items, invalid tests and resource waste can be effectively reduced, thereby shortening the overall test cycle time and reducing the test cost.

[0027] ③ Embodiment 3: This embodiment provides a personalized adjustment strategy for the original test items, including the following processes: Establish a difference quantification model between the measured damage parameters and the predicted data in the current cycle, and extract the deviation characteristic quantities of the test parameters; Perform multi-objective correlation analysis on the simulation data of the virtual test item combination and the deviation characteristic quantities, and construct optimization constraint conditions including load intensity, action frequency, and test time sequence; Based on the particle swarm optimization algorithm, iteratively optimize the current combined test parameters. With the goal of minimizing the prediction residual and maximizing the damage coverage rate under the constraint conditions, solve to obtain the adjusted test parameter combination; Verify the confidence level of the optimized parameter combination with the virtual test results, and screen the solutions that meet the preset reliability threshold as the test instruction set for the next cycle.

[0028] Compared with the prior art, by establishing a difference quantification model between the measured damage parameters and the predicted data in the current cycle, the difference between the actual test results and the prediction model can be accurately identified and quantified, which is more accurate than simple statistical analysis, helping to deeply understand the deficiencies of the test performance. Performing correlation analysis on the simulation data of the virtual test item combination and the deviation characteristic quantities can more comprehensively consider the interaction between different factors, helping to identify the factors that have the greatest impact on the test results, thus providing a more targeted basis for subsequent optimization; Construct optimization constraints including load intensity, action frequency, and test timing sequence, making the optimization process more flexible and precise, thus allowing adjustments in more complex test environments, ensuring the practical operability of the selected solution, emphasizing personalized adjustment of the original test project, and being able to flexibly adjust test parameters according to actual test conditions and requirements. This enables the test process to better adapt to specific component characteristics and usage scenarios.

[0029] Working principle: During the operation of this device, the tested component is jointly fixed by the traction device one 2 at the bottom inside the main body 1 of the test stand, the traction device two 3 on the left side, and the traction device three 4 on the right side to conduct horizontal or vertical pulling, compression, and torsional fatigue tests. The control module 5 is responsible for registering combined test items, the project management module 511 records the combination of each test item, the combination selection module 512 selects the combined test items for the current cycle, and sends them to the data acquisition module 513 to execute a complete test cycle, ensuring comprehensive and reliable data acquisition. Next, the analysis and prediction unit 52 analyzes the test data, uses machine learning algorithms to establish a prediction model to identify the damage state and predict changes in future cycles, screens out key feature variables. The project adjustment module 54 adjusts the test parameters through an optimization algorithm based on the difference between the real-time test results and the prediction data, forms the next round of test plan, and combines it with virtual test items to verify the effectiveness of the plan. The active verification module 55 starts the retraining process when the deviation between the actual test data and the prediction data exceeds the threshold to ensure the accuracy of the model. At the same time, the repository module 57 conducts data aggregation and classified storage, providing convenient online indexing. The entire work process realizes efficient interaction and data sharing among them through a wireless network. This device can comprehensively evaluate the durability of components by combining horizontal, vertical, and torsional fatigue tests. At the same time, using machine learning algorithms and optimization techniques, the device can analyze test data in real time, predict future damage states, and dynamically adjust test parameters to ensure the effectiveness and accuracy of the test plan. In addition, the modular design and wireless network connection improve the flexibility and data management efficiency of the system, making the test process more intelligent and automated.

[0030] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A durability test device for four-channel chassis components of a passenger car, characterized in that, It includes a test rack body, where a first tractor is installed at the bottom inside the test rack body, a second tractor is installed on the left side inside the test rack body, a third tractor is installed on the right side inside the test rack body, and a control module is installed on the front of the test rack body, where: The first tractor, the second tractor and the third tractor are used to fix the tested component and traction for horizontal, vertical and torsional fatigue tests; The control module includes: A test execution unit, which is used to register various test items of the combination type, select and execute a certain combined test item, and obtain the test parameters of the tested component in the current cycle; An analysis and prediction unit, which is used to perform model analysis on the test parameters obtained from the combined test item and predict the state data of the damage evolution of the test component in the future cycle; A data generation module, which is used to simulate and generate a corresponding virtual test item combination according to the state data of the damage evolution of the test component in the future cycle; A project adjustment module, which is used to analyze the test parameters in the current combined test item according to the difference between the current test data and the predicted data, as well as the virtual test item combination, and use an optimization algorithm to calculate the adjusted test parameter combination as the plan for the next round of tests; An active verification module, which is used to generate control instructions for the first tractor, the second tractor and the third tractor according to the adjusted combined test item plan, obtain the verified data and compare it with the predicted data as the basis for updating the prediction model parameters.

2. The durability test device for four-channel chassis components of a passenger car according to claim 1, characterized in that Sub-modules are deployed at the lower level of the test execution unit. The sub-modules include: a project management module, a combination selection module and a data acquisition module. The project management module and the combination selection module are interconnected through a wireless network, and the combination selection module and the data acquisition module are interconnected through a wireless network, where: The project management module is used to record the combination of the first tractor, the second tractor and the third tractor corresponding to each test item and classify the entered combined test items; The combination selection module is used to perform project selection processing according to the estimated performance of the current test component and the trend data of its historical tests, and obtain the combined test item in the current cycle; The data acquisition module is used to perform a complete cycle of durability test on the specified test component according to the selected combined test item, and collect the state of the test equipment and the parameter data of the test component during the test process through sensors.

3. The durability test device for the four-channel chassis components of a passenger car according to claim 2, characterized in that, The process of the combination selection module performing project selection processing is as follows: Step 1: Obtain the estimated performance parameters of the current test component and the trend data of its historical tests. The trend data includes the change law of the component performance parameters, the damage accumulation rate and the correlation characteristics of the test interruption events in the historical tests; Step 2: Dynamically match the component performance parameters in the historical tests with the combined test items pre-stored in the project management module, and extract the candidate test item combinations associated with the type of the current test component, the preset durability target and the damage-sensitive direction; Step 3: Preset a scoring model, and perform multi-dimensional scoring on the candidate test item combinations based on the scoring model. The scoring dimensions include the test efficiency weight, the damage feature coverage weight and the effectiveness feedback weight of the historical similar test scheme; Step 4: Screen the candidate combinations with scores higher than the preset threshold, and determine the optimal combined test items for the current cycle through the dynamic programming algorithm according to the real-time test resource occupancy status and the optimization target constraint conditions; Step 5: Send the optimal combined test items to the data acquisition module to trigger the execution of the full-cycle test, and at the same time record the currently selected logical parameters as the updated input of the historical trend data.

4. The durability test device for a four-channel chassis component of a passenger vehicle according to claim 1, characterized in that, Sub-modules are deployed under the analysis and prediction unit. The sub-modules include: a damage analysis module, a model construction module, and a damage prediction module. The damage analysis module is connected to the model construction module through wireless network interaction, and the model construction module is connected to the damage prediction module through wireless network interaction, where: The damage analysis module is used to perform preliminary analysis on the collected data, identify various damage states that occur in the test components during the loading process, quantitatively record the damage states of each test cycle, establish a corresponding damage state file in combination with historical data, and record the time, location, and damage degree of each detection point; The model construction module is used to construct a prediction model using machine learning algorithms based on the damage state file obtained in the single-cycle test. The prediction model establishes a mapping relationship between the initial cycle damage state and the fission and deformation states of future cycles, and screens key feature variables; The damage prediction module is used to obtain the prediction model constructed by the model construction module, input the key feature variables of various test data in the current cycle into the prediction model, and the prediction model outputs the state data of the damage evolution of the test components in future cycles.

5. The durability test device for the four-channel chassis components of a passenger car according to claim 4, characterized in that, During the process of the model construction module screening key feature variables, by receiving the test combination data collected in real time by the sensor in the current cycle, obtain the loading force value, displacement, torque parameter, and component deformation characteristic data of each test component, normalize the obtained feature data, and extract the preset damage feature variables, The damage feature variables include: stress concentration coefficient, strain gradient value, and material fatigue cumulative factor.

6. The durability test device for the four-channel chassis components of a passenger car according to claim 4, characterized in that, The prediction model constructed by the model construction module establishes multi-cycle time series associations through the LSTM neural network, and outputs the damage state prediction data including several future test cycles. The prediction data is decomposed into specific damage types, occurrence locations, damage degrees, and prediction time points, where the damage types include at least two of crack growth rate, plastic deformation amount, and connector loosening gap.

7. The durability test device for the four-channel chassis components of a passenger car according to claim 1, characterized in that, During the operation of the project adjustment module, a difference quantification model between the measured damage parameters and the prediction data in the current cycle is established, and the deviation feature quantity of the test parameters is extracted; Perform multi-objective correlation analysis on the simulation data of the virtual test item combination and the deviation feature quantity, and construct optimization constraint conditions including load intensity, action frequency, and test timing; Based on the particle swarm optimization algorithm, iteratively optimize the current combined test parameters, and use minimizing the prediction residual and maximizing the damage coverage rate as the objective function under the condition of meeting the constraint conditions to solve and obtain the adjusted test parameter combination; Verify the confidence level of the optimized parameter combination and the virtual test results, and screen the solutions that meet the preset reliability threshold as the test instruction set for the next cycle.

8. The durability test device for the four-channel chassis components of a passenger car according to claim 1, characterized in that During the operation phase, when the deviation between the actual test data and the predicted data exceeds the preset threshold, the active verification module automatically triggers the retraining process of the prediction model in the model construction module. Using the sliding time window algorithm, it retains the test data in the recent several cycles as the training sample set, and adjusts the number of neural network hidden layer nodes and the activation function combination parameters of the prediction model constructed by the model construction module through cross-validation.

9. The durability test device for a four-channel chassis component of a passenger car according to claim 1, characterized in that, The test execution unit is connected to the repository module through wireless network interaction. The repository module is used to centrally control all the collected data and calculation data, perform classified storage combining cloud and local, and provide an online indexing interface for historical classified data of each category.

10. The durability test device for the four-channel chassis components of a passenger car according to claim 1, characterized in that The project adjustment module is connected to the active adjustment module through an electrical medium. The active adjustment module is used to receive the next round of test plan output by the project adjustment module, provide visual display, and provide user authentication. It provides the modification permission of the test plan to the users who pass the authentication. After the modification is completed, it feeds back the modified data to the project adjustment module for application.

Images