A vehicle side slope estimation closed-loop test method based on a rotary table IMU simulation and HIL system

Abstract

The application discloses a kind of vehicle side slope estimation closed-loop test method based on rotary table IMU simulation and HIL system, it is related to intelligent vehicle test verification technical field, comprising the following steps: constructing vehicle side slope estimation closed-loop test bench, configuration HIL real-time simulation environment, high-precision attitude rotary table and whole vehicle area control hardware interaction environment, establish whole vehicle dynamics model, road side slope scene library, road adhesion model and vehicle driving condition model by HIL real-time simulation environment, generate fixed side slope, dynamic gradual side slope, jolt coupling side slope and steering coupling side slope corresponding side slope true value data.The application is fused by rotary table IMU physical attitude simulation and HIL real-time simulation, realize the high-fidelity closed-loop test of vehicle side slope estimation, support full-condition, multiple version algorithm and cross-domain control verification, and have automation test, fault injection, data quantification evaluation capacity, improve algorithm verification efficiency, accuracy and reliability.

Description

Technical Field

[0001] This invention relates to the field of intelligent vehicle testing and verification technology, specifically to a closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system. Background Technology

[0002] With the rapid development of intelligent vehicle domain control architecture and advanced intelligent driving technologies, the importance of vehicle attitude perception and intelligent chassis control technologies is constantly increasing. Among these, vehicle lateral slope estimation, as a crucial foundational perception technology for intelligent chassis attitude control, air suspension leveling, rear-wheel steering coordination, intelligent driving lane keeping, and rollover risk protection, directly impacts vehicle driving safety and ride comfort due to its estimation accuracy, adaptability to various operating conditions, and cross-domain collaborative capabilities. To ensure that the lateral slope estimation algorithm meets the requirements of mass production applications, it needs to undergo high-precision, repeatable, and quantifiable testing and verification covering various road environments, vehicle motion states, and abnormal operating conditions. Currently, vehicle lateral slope estimation algorithm testing primarily relies on real-vehicle road testing, tilt platform testing, and hardware-in-the-loop (HIL) simulation testing. Among them, real vehicle road testing can obtain real vehicle operation data, but the testing process is greatly affected by road environment, weather conditions and vehicle condition, and has problems such as uncontrollable test scenarios, poor repeatability, difficulty in reproducing extreme dangerous conditions, long test cycle and high cost; traditional tilt test platforms mainly achieve slope simulation by adjusting vehicle posture through mechanical structure, but usually can only achieve static side slope simulation at a fixed angle, and cannot simulate dynamic posture changes caused by factors such as bumps, steering, acceleration and deceleration during vehicle driving; existing general HIL test systems mainly generate virtual IMU data through software models to simulate vehicle posture sensor output, but lack the real physical posture excitation process, and cannot fully reflect the characteristics of noise, temperature drift, vibration interference and installation errors in actual vehicle-mounted IMU sensors.

[0003] Existing technical solutions still have the following shortcomings: First, in traditional pure software HIL testing methods, IMU data is mainly generated by vehicle model calculations, lacking the response characteristics of real sensors during physical attitude changes. This leads to a deviation between the laboratory testing environment and the actual vehicle operating environment, easily resulting in the problem that the algorithm passes laboratory verification but fails in real-vehicle applications. Second, existing tilt platforms mostly use static angle adjustment methods, which cannot realize dynamic side slope changes and the coupled simulation of side slope with vehicle steering, bumps, acceleration, deceleration, and other motion states. They cannot effectively verify the decoupling ability of the side slope estimation algorithm on the combined effects of vehicle roll and road side slope. Third, existing testing platforms are usually verified for a single algorithm version or a single sensor configuration, and cannot simultaneously support basic side slope estimation algorithms without intelligent tires and multi-source information side slope estimation algorithms that integrate intelligent tire data, resulting in insufficient versatility of the testing platform. At the same time, existing testing methods mainly focus on the accuracy of side slope estimation results, lacking closed-loop linkage verification with chassis domain controllers, intelligent driving domain controllers, and body domain controllers. This makes it impossible to verify the actual impact of side slope estimation results on functions such as air suspension leveling, rear wheel steering compensation, intelligent driving decision-making, and vehicle comfort control. In addition, existing testing systems lack the ability to automatically inject faults and quantitatively evaluate them for extreme conditions such as large side slopes, low-adhesion road surfaces, sensor malfunctions, and signal failures. The testing process relies on manual operation and cannot achieve automatic traversal of all working conditions, automatic comparison of test data, and standardized evaluation of algorithm robustness. This makes it difficult to meet the high-efficiency and high-reliability verification requirements in the mass production development of intelligent vehicle chassis algorithms.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system, so as to solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system, comprising the following steps: A closed-loop test bench for vehicle side slope estimation was constructed, and a HIL real-time simulation environment, a high-precision attitude rotary table, and a vehicle domain control hardware interaction environment were configured. The vehicle dynamics model, road side slope scenario library, road surface adhesion model, and vehicle driving condition model were established through the HIL real-time simulation environment to generate true side slope data corresponding to fixed side slope, dynamic gradual side slope, bumpy coupled side slope, and steering coupled side slope. Dynamic attitude adjustment is achieved by controlling a high-precision attitude rotary table based on true side slope data. The vehicle's IMU sensor is installed on the high-precision attitude rotary table, which simulates the vehicle's side slope driving attitude change. Raw IMU data containing noise and hysteresis characteristics is collected and output to the vehicle side slope estimation algorithm for side slope estimation. Based on the raw IMU data and the output of the slope estimation algorithm, the slope estimation closed-loop verification was carried out through the vehicle domain control hardware interaction environment. The chassis control response, intelligent driving decision response and vehicle comfort control response were verified respectively. The data fusion effect of different slope estimation algorithms was verified by switching between test modes without intelligent tires and test modes with intelligent tires. The closed-loop test process for vehicle side slope estimation is automated with traversal of operating conditions and fault injection testing. Test data is collected according to preset side slope angle, vehicle speed, road surface adhesion status and vehicle motion status. Error calculation and robustness evaluation are performed on the side slope estimation results under IMU anomaly, signal drift, signal disconnection, intelligent tire data anomaly and suspension signal anomaly.

[0007] The HIL real-time simulation environment includes a real-time processor, a vehicle dynamics model, a road side slope scenario library, a road surface adhesion model, and a vehicle driving condition model. The vehicle dynamics model is used to simulate the motion response of the vehicle under different speeds, steering states, and road conditions. The road side slope scenario library is used to configure test scenarios for fixed side slopes, dynamic gradual side slopes, bumpy coupled side slopes, and steering coupled side slopes. The road surface adhesion model is used to simulate different road adhesion states. The vehicle driving condition model is used to generate vehicle operating state data and output the corresponding side slope ground truth data as a benchmark for evaluating the accuracy of the side slope estimation algorithm.

[0008] A high-precision attitude rotary table is fixedly mounted on the test base. The actual vehicle-produced IMU sensor is rigidly mounted on the table surface of the high-precision attitude rotary table. The high-precision attitude rotary table supports single-axis lateral rotation and continuous angle adjustment, with an angle control accuracy of no more than 0.05°. The high-precision attitude rotary table interacts with the HIL real-time simulation environment in real time through a communication interface. It receives dynamic angle control commands based on the true side slope data and drives the actual vehicle-produced IMU sensor to generate corresponding attitude changes, so as to output raw IMU data containing sensor noise characteristics and response hysteresis characteristics.

[0009] The vehicle domain control hardware interaction environment includes a chassis domain controller, an intelligent driving domain controller, a body domain controller, an intelligent tire simulator, an air suspension centralized air supply module, and a suspension height signal simulation module. Among them, the chassis domain controller is used to execute the side slope estimation algorithm and output chassis control commands, the intelligent driving domain controller is used to execute driving decision control based on the side slope state, the body domain controller is used to execute posture comfort control, and the air suspension centralized air supply module and the suspension height signal simulation module are used to simulate the vehicle suspension control process, forming a hardware closed-loop verification environment for vehicle side slope estimation and vehicle control feedback.

[0010] In the non-intelligent tire test mode, the intelligent tire simulation signal input is turned off. The vehicle speed signal, steering signal and road condition signal are output through the HIL real-time simulation environment. The IMU dynamic attitude data is output through the high-precision attitude rotary table. The vehicle suspension height difference signal is output through the suspension height signal simulation module. The side slope estimation algorithm is based on the IMU data and the suspension height difference signal to estimate the side slope, which is used to verify the accuracy and stability of the side slope estimation algorithm based on the IMU and suspension geometry decoupling method.

[0011] In the intelligent tire test mode, the intelligent tire simulator is activated, outputting dynamic tire lateral force, ground pressure, and tread deformation data. The intelligent tire simulation data is then fused with IMU raw data, suspension height data, and visual simulation data. The performance of the multi-domain fusion side slope estimation algorithm is verified through multi-source information fusion. The intelligent tire simulator adjusts the output data according to the vehicle's operating status and the road side slope status to simulate the changes in tire force under different road conditions.

[0012] The closed-loop verification of slope estimation includes verification at the perception layer, chassis control layer, intelligent driving decision layer, and vehicle comfort layer. Specifically, the perception layer calculates static accuracy, dynamic following accuracy, steady-state error, and transient response error by comparing the output value of the slope estimation algorithm with the true slope value output by the HIL real-time simulation environment. The chassis control layer verifies the response of the air suspension to independent left and right leveling, centralized high and low pressure air supply switching, and rear-wheel reverse steering control triggered by the slope estimation results. The intelligent driving decision layer verifies lane keeping compensation, slope speed limiting, and rollover risk assessment logic. The vehicle comfort layer verifies seat posture adjustment and low-pressure air source adaptation control logic.

[0013] The automated test cycle includes automatically executing tests according to preset side slope angles, vehicle speeds, road surface adhesion states, and vehicle motion states. The side slope angle ranges from 0° to 15°, and the test conditions include different gradient side slopes, high and low vehicle speeds, different road surfaces with different adhesion, and coupled conditions of turning, straight driving, and bumpy driving. During the test, the HIL real-time simulation environment, high-precision attitude rotary table, and vehicle domain control hardware interaction environment are automatically controlled to operate in coordination, realizing continuous testing of the side slope estimation algorithm under multiple conditions.

[0014] The fault injection test includes injecting noise into the IMU sensor data, simulating IMU signal drift and signal disconnection, as well as simulating smart tire data failure and suspension signal abnormalities. Under fault conditions, the output of the side slope estimation algorithm is compared with the true side slope value output by the HIL real-time simulation environment to verify the data fault tolerance, adaptive weighted degradation capability, and estimation stability of the side slope estimation algorithm under multi-source signal abnormalities.

[0015] Test data acquisition includes synchronously acquiring high-precision attitude table attitude data, IMU raw data, slope estimation algorithm output data, and execution feedback data from each domain controller, and achieving multi-source data synchronization through a unified timestamp; the slope estimation error is automatically calculated based on the acquired data, and accuracy error curves, working condition pass rates, and fault robustness evaluation reports are generated for slope estimation algorithm parameter calibration, performance optimization, and mass production testing and acceptance.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention constructs a test architecture that combines a real-time HIL simulation environment with physical rotary table IMU attitude simulation. It abandons the traditional method of generating virtual IMU data purely through software, instead utilizing a high-precision physical rotary table to directly drive mass-produced vehicle IMU sensors to generate realistic attitude responses. This ensures that the IMU output data includes attitude changes, sensor noise, and response hysteresis characteristics from actual operation. By using physical attitude excitation, the consistency between the laboratory test environment and the real vehicle road operating environment is improved, effectively reducing the deviation between simulation test results and real-vehicle test results, and enhancing the authenticity and reliability of the side slope estimation algorithm verification results.

[0017] This invention utilizes the HIL real-time simulation environment, a road slope scenario library, and automated test scripts to achieve comprehensive verification of static slopes, dynamic gradual slopes, bumpy coupled slopes, steering coupled slopes, and boundary fault conditions. During testing, it can simulate different slope angles, vehicle speeds, road surface adhesion states, and vehicle motion states, solving the problems of traditional test benches' limited testing scenarios, lack of dynamic conditions, and difficulty in reproducing extreme road environments. This allows the slope estimation algorithm to be verified under various conditions closely resembling real-world vehicle operating environments, improving the algorithm's testing coverage and meeting the needs of mass-production-level algorithm development and verification.

[0018] This invention enables a single test bench to be compatible with different vehicle configurations and different side slope estimation algorithms by setting up a test mode without intelligent tires and a test mode with intelligent tires. In the test mode without intelligent tires, the accuracy and stability of the basic side slope estimation algorithm can be verified based on IMU data and suspension signals. In the test mode with intelligent tires, data such as intelligent tire lateral force, ground pressure, and tread deformation can be introduced to verify the performance of the multi-source information fusion side slope estimation algorithm. Through the dual-mode compatible design, it can adapt to the development needs of high and low configuration vehicle platforms, reduce the cost of repeatedly building test equipment for different algorithm versions, and improve the versatility and utilization efficiency of the test platform.

[0019] This invention constructs a full-link closed-loop verification system consisting of attitude simulation, slope perception, chassis control, intelligent driving decision-making, and vehicle comfort control. This system not only evaluates the accuracy of the slope estimation algorithm output but also further verifies the practical application effects of the slope estimation results in different control domains of the vehicle. Through data interaction between the chassis domain controller, intelligent driving domain controller, and vehicle domain controller, the impact of the slope estimation results on functions such as air suspension leveling, centralized air supply control, rear-wheel steering compensation, lane keeping compensation, slope speed limiting, rollover risk assessment, and seat posture adjustment can be verified. This achieves complete testing from perception algorithm verification to cross-domain collaborative control verification of the entire vehicle, meeting the vehicle-level verification requirements under the intelligent vehicle domain control architecture.

[0020] This invention utilizes a built-in, fully automated test script to achieve automatic configuration of test scenarios, automatic traversal of test conditions, automatic data acquisition, and automatic analysis of test results. During testing, it can automatically execute tests based on preset conditions, combining different side slope angles, vehicle speeds, road adhesion states, and vehicle motion states, while simultaneously acquiring true side slope data, algorithm estimation data, and control feedback data from each domain. By automatically calculating error indicators, generating accuracy reports, and outputting robustness evaluation reports, it reduces manual operation and data processing, improves test execution efficiency, shortens the development and verification cycle of the side slope estimation algorithm, and meets the rapid iterative development needs of intelligent vehicle algorithms.

[0021] This invention outputs the true value of the standard slope through a real-time simulation environment using HIL (High-Intensity Logic) and simultaneously records rotary table attitude data, IMU raw data, algorithm-output slope data, and multi-domain control execution data using a data acquisition unit, achieving data correlation and result traceability throughout the testing process. By comparing and analyzing the true value data with the algorithm-estimated data, the slope estimation error can be accurately quantified, obtaining algorithm performance indicators under different working conditions. Simultaneously, the test results provide standardized data for algorithm parameter adjustment, accuracy calibration, fault optimization, and mass production acceptance, making the slope estimation algorithm testing process measurable, evaluable, reproducible, and traceable, thus improving the quality of algorithm development and verification. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0023] Figure 1 This is a flowchart of a closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to the present invention. Detailed Implementation

[0024] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0025] This invention provides, for example Figure 1 The following is a closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system. The specific steps are as follows: The vehicle side slope estimation closed-loop test bench adopts a three-layer integrated test architecture. The overall architecture consists of an upper HIL real-time simulation layer, a middle rotary table IMU physical attitude simulation layer, and a lower vehicle domain control hardware closed-loop layer. The three layers form a complete virtual-real combined test link through real-time data interaction. The chassis domain controller is used as the core control object, and it also links the intelligent driving domain controller, body domain controller, and intelligent tire simulation unit to realize a full-dimensional closed-loop test from road environment simulation, vehicle attitude reproduction, side slope estimation perception verification to vehicle control response evaluation.

[0026] The upper-layer HIL real-time simulation layer is used to build the real-time simulation foundation for vehicle operating environment and side slope test scenarios. By incorporating a vehicle dynamics model, a road side slope scenario library, a road surface adhesion model, and a vehicle driving condition model, it digitally simulates the road conditions, vehicle motion states, and attitude changes during actual vehicle operation. The vehicle dynamics model describes the vehicle's response characteristics under different road conditions, speeds, and motion states. The road side slope scenario library stores different types of side slope road conditions, including fixed side slopes, dynamically changing side slopes, bumpy coupled side slopes, and steering coupled side slopes. The road surface adhesion model simulates the interaction between the vehicle tires and the road under different adhesion conditions, while the vehicle driving condition model simulates the dynamic changes of the vehicle during straight-line driving, turning, acceleration / deceleration, and complex motion states.

[0027] The HIL real-time simulation layer generates simulation data for different slope angles and vehicle operating states based on preset test requirements, and outputs standard true slope angles as benchmark data for evaluating the accuracy of slope estimation algorithms. During testing, the HIL real-time simulation layer can adjust the road slope variation pattern according to different test requirements, enabling the repeated reproduction of static slope, dynamic slope, and complex coupled slope scenarios, avoiding the problems of uncontrollable environment and difficulty in repeating test conditions in traditional road testing. Simultaneously, the HIL real-time simulation layer maintains real-time communication with hardware control devices, sending vehicle operating status, road environment information, and true slope data to the corresponding control unit, providing real-time input for subsequent physical attitude simulation and algorithm verification.

[0028] The intermediate rotary table IMU physical attitude simulation layer is used to physically reproduce the vehicle's side slope attitude. By setting up a high-precision attitude rotary table and mounting the actual vehicle's IMU sensors on the rotary table surface, the IMU sensors can directly sense real mechanical attitude changes. The high-precision attitude rotary table supports single-axis lateral rotation, with continuously adjustable rotation angles and an angle control accuracy of ≤0.05°. It can dynamically respond to the side slope conditions output by the HIL real-time simulation layer, thereby simulating the actual attitude changes of the vehicle during its journey on a side slope road.

[0029] Unlike traditional methods that directly generate IMU data through software models, this attitude simulation method directly applies mechanical motion to the real IMU sensor. This ensures that the IMU output data includes the actual noise characteristics, response hysteresis characteristics, and attitude changes generated during sensor operation. By using physical attitude excitation, the output state of the IMU sensor during vehicle side slope driving can be more realistically reproduced, improving the consistency between the side slope estimation algorithm test environment and the actual vehicle application environment.

[0030] The high-precision attitude rotary table can adjust the attitude angle in real time according to different road side slope scenarios. For example, it can maintain a specified tilt angle in fixed side slope tests, continuously adjust the rotation angle according to the side slope change trend in dynamic gradual side slope tests, and perform dynamic attitude response according to the vehicle's motion state in bumpy coupled side slope and steering coupled side slope tests. The raw IMU data output by the rotary table can be used to verify the adaptability of the side slope estimation algorithm to dynamic attitude changes, sensor errors, and complex motion disturbance conditions.

[0031] The lower-level vehicle domain control hardware closed-loop layer is used to realize the real hardware interaction between various vehicle control units during testing. This includes the chassis domain controller, intelligent driving domain controller, body domain controller, intelligent tire simulator, air suspension centralized air supply module, and suspension height signal simulation module. The chassis domain controller, as the core control unit, receives IMU sensor output data and vehicle status information provided by the HIL real-time simulation environment, runs the side slope estimation algorithm, and outputs the corresponding control results. Simultaneously, it interacts with the intelligent driving domain controller, body domain controller, and other execution units through hardware connections to verify the impact of side slope estimation results on the vehicle control function.

[0032] The chassis domain controller executes corresponding chassis control strategies based on the slope estimation results, such as independent left and right leveling control of the air suspension, high and low pressure switching control of centralized air supply, and rear-wheel counter-steering control. The intelligent driving domain controller executes intelligent driving control logic such as lane keeping compensation, slope speed limiting, and rollover risk assessment based on the slope status information. The body domain controller executes seat posture adjustment and comfort control strategies based on changes in vehicle posture. Through data interaction between multiple domain controllers, the practical application effect of the slope estimation algorithm output results in different functional domains of the vehicle can be verified.

[0033] Meanwhile, the intelligent tire simulator can output dynamic tire lateral force, ground contact pressure, and tread deformation data according to test requirements, providing data input for the vehicle side slope estimation algorithm equipped with intelligent tires. By fusing information between the intelligent tire simulator and IMU sensors, suspension height signals, and other simulation data, the stability and accuracy of the multi-source data fusion side slope estimation algorithm under complex road conditions can be verified.

[0034] The centralized air supply module and suspension height signal simulation module for air suspension are used to simulate the actual operating state of a vehicle's air suspension system. This provides corresponding signal inputs for the slope estimation algorithm based on the IMU and suspension geometry, and simultaneously verifies the feedback effect of the slope estimation results on the air suspension control strategy. Through the vehicle domain control hardware closed-loop layer, each control unit participates in testing using a hardware environment close to that of a mass-produced vehicle. This allows the testing process to not only verify the accuracy of slope estimation but also the logical correctness and response effect when the slope perception results drive the vehicle control function.

[0035] Through the coordinated operation of the upper-layer HIL real-time simulation layer, the middle-layer rotary table IMU physical attitude simulation layer, and the lower-layer vehicle domain control hardware closed-loop layer, a complete closed-loop verification architecture is formed, encompassing virtual road environment generation, physical attitude excitation, sensor data acquisition, side slope algorithm calculation, multi-domain control feedback, and test result evaluation. This architecture supports the reproduction of multiple types of side slope scenarios, testing of various vehicle configurations, and verification of multi-domain control functions, providing a highly realistic, highly repeatable, and mass-production-grade testing and verification environment for vehicle side slope estimation algorithms.

[0036] The test modes without smart tires and with smart tires are used to verify the compatibility of the vehicle slope estimation algorithm with different vehicle configurations, different sensor combinations, and different data fusion methods. By switching between the two test modes, the slope estimation performance testing requirements of low-configuration vehicles and high-configuration smart vehicle platforms can be met.

[0037] In the test mode without intelligent tires, the testing process mainly focused on verifying the slope estimation algorithm based on basic sensor information. The HIL real-time simulation environment outputs vehicle speed signals, steering signals, and road condition signals according to preset vehicle operating conditions and road slope scenarios, providing the slope estimation algorithm with input information on vehicle motion state and road environment. Specifically, the vehicle speed signal simulates the attitude changes of the vehicle under different operating speed conditions, the steering signal simulates the lateral motion state of the vehicle during steering, and the road condition signal describes the road environment and corresponding slope changes during vehicle operation.

[0038] In this test mode, the high-precision attitude rotary table dynamically adjusts the vehicle's attitude based on the side slope conditions generated by the HIL real-time simulation environment. Dynamic attitude data generated during vehicle attitude changes is collected by IMU sensors mounted on the rotary table. Since the IMU sensors are directly affected by the rotary table's mechanical attitude changes, the output data includes noise characteristics, attitude response characteristics, and dynamic hysteresis characteristics generated during actual sensor operation, improving the consistency between the test data and the actual vehicle's operating state. Simultaneously, the suspension simulation module outputs corresponding suspension height difference signals based on the vehicle's side slope condition and body attitude changes, simulating the lateral height changes of the actual vehicle's air suspension system under different side slope road conditions.

[0039] The performance of a slope estimation algorithm based on the geometric decoupling of the IMU and suspension was verified by comprehensively inputting vehicle speed signals, steering signals, road condition signals, IMU dynamic attitude data, and suspension height difference signals. During testing, the slope estimation algorithm obtains the vehicle's current attitude change based on IMU attitude information and analyzes the relationship between vehicle roll and road slope using the suspension height difference signal. The algorithm calculates the true road slope angle through geometric decoupling. The test platform evaluates the slope estimation accuracy, dynamic following ability, and stability under different vehicle motion states by comparing the standard true slope values ​​output from the HIL real-time simulation environment with the algorithm's estimation results.

[0040] For different side slope conditions, the test mode without smart tires can be configured with test scenarios such as fixed side slope, dynamically changing side slope, bump-coupled side slope, and steering-coupled side slope. Under fixed side slope conditions, the algorithm's basic estimation capability for different slope angles can be verified; under dynamically changing side slope conditions, the algorithm's tracking response capability for continuously changing side slopes can be verified; under bump-coupled side slope conditions, the algorithm's anti-interference capability for attitude changes caused by road disturbances can be verified; under steering-coupled side slope conditions, the algorithm's decoupling capability for the combined effects of vehicle lateral movement and road side slope can be verified. Through the above test methods, the accuracy and stability of the basic side slope estimation algorithm can be verified without smart tire data input.

[0041] In the smart tire testing mode, the testing platform further introduces a smart tire simulator to verify the high-precision side slope estimation algorithm that integrates multi-source vehicle state information. After starting the smart tire simulator, based on the vehicle operating status and road condition information generated by the HIL real-time simulation environment, it outputs dynamic tire lateral force, ground pressure, and tread deformation data to simulate the actual perception information of the smart tire under different side slope road conditions and different vehicle motion states.

[0042] The data output from the intelligent tire simulator is fused with IMU dynamic attitude data, suspension height difference signals, and visual simulation data to provide complete data input for the multi-source information fusion slope estimation algorithm. Specifically, IMU data reflects overall vehicle attitude changes, suspension data reflects vehicle geometry changes, intelligent tire data provides information on tire contact status, lateral forces, and road conditions, and visual simulation data supplements road environment perception information. This multi-source data fusion further enhances the slope estimation process's adaptability to complex road environments and extreme vehicle conditions.

[0043] In this test mode, a multi-domain fusion UKF high-precision slope estimation algorithm is used to fuse the input data. By comprehensively analyzing data from different sources, high-precision estimation of the vehicle's slope state is achieved. During the test, the road slope angle, vehicle speed, road surface adhesion state, and vehicle motion state are adjusted to simulate the actual vehicle operation in a complex road environment, verifying the estimation performance of the multi-domain fusion UKF high-precision slope estimation algorithm under different working conditions.

[0044] The intelligent tire testing mode can cover more complex vehicle operating environments, such as low-adhesion roads, steep side slopes, and side slope estimation verification under dynamic vehicle motion conditions. Under low-adhesion road conditions, the algorithm's adaptability to tire state changes is verified by outputting tire lateral force and ground pressure variation data through the intelligent tire simulator. Under steep side slope conditions, the algorithm's estimation stability for extreme attitude environments is verified by fusing IMU, suspension, and intelligent tire data. Under dynamic motion conditions, the algorithm's ability to handle complex coupled conditions is verified through joint simulation of vehicle motion state and road side slope changes.

[0045] Through a compatible design with both intelligent tire-less and intelligent tire-with testing modes, the same testing platform can cover different vehicle configurations and different slope estimation algorithms. The intelligent tire-less testing mode primarily focuses on verifying basic slope estimation algorithms based on IMU and suspension information, emphasizing the evaluation of the algorithm's basic accuracy and stability. The intelligent tire-with testing mode primarily focuses on verifying high-precision slope estimation algorithms that integrate intelligent tire data, emphasizing the evaluation of estimation performance under multi-source information fusion conditions. The two testing modes can be switched according to vehicle configuration requirements, achieving complete coverage from basic slope perception verification to multi-domain fusion high-precision estimation verification. This compatible testing mechanism effectively improves the versatility of the slope estimation algorithm testing platform, meeting the verification needs of different vehicle platforms, sensor configurations, and algorithm versions.

[0046] The perception layer, chassis control layer, intelligent driving decision layer, and vehicle comfort layer constitute a multi-level verification system in the side slope estimation closed-loop testing process. By comprehensively evaluating the side slope estimation results, vehicle control response, and the overall vehicle function linkage effect, a complete closed-loop test is achieved, from basic perception algorithm verification to vehicle function application verification. This verification logic uses the side slope estimation results as the core input and, through data interaction between different control domains, comprehensively evaluates the accuracy, stability, and functional adaptability of the side slope estimation algorithm in real-world vehicle application environments.

[0047] During the verification process at the perception layer, the accuracy and dynamic response capability of the slope estimation algorithm's output are primarily evaluated. During testing, standard slope ground truth data is generated using the HIL real-time simulation environment as a benchmark for the slope estimation results. Simultaneously, the real-time slope estimation values ​​output by the algorithm are collected, and the algorithm's estimation results are compared and analyzed with the HIL ground truth values ​​to calculate the error variation between the two. Through data comparison, the algorithm's static accuracy, dynamic tracking accuracy, steady-state error, and transient response error under different test conditions can be quantitatively evaluated.

[0048] The static accuracy evaluation primarily verifies the ability of the slope estimation algorithm to identify different slope angles when the vehicle is on a fixed slope. By setting different slope angle conditions, the algorithm's output results are compared with the standard slope angle to analyze the algorithm's basic estimation deviation. The dynamic following accuracy evaluation mainly targets complex conditions such as dynamically changing slopes, bumpy coupled slopes, and steering coupled slopes. By observing the real-time tracking of the algorithm's output slope angle as it changes with the road slope, the algorithm's response capability to dynamically changing environments is evaluated.

[0049] Steady-state error evaluation is mainly used to analyze the fluctuation of slope estimation results under stable driving conditions, and to determine the stability of the algorithm in continuous slope road environments. Transient response error evaluation mainly focuses on the response speed and error changes of the algorithm output during vehicle entry into and exit from the slope, as well as during rapid changes in road conditions, and is used to evaluate the algorithm's adaptability to sudden changes in conditions. Through the above-mentioned perception layer verification method, the performance of the slope estimation algorithm under different road environments, different vehicle motion states, and different slope change conditions can be comprehensively analyzed, providing data basis for algorithm parameter optimization and accuracy calibration.

[0050] The chassis control layer verification is primarily used to evaluate the impact of slope estimation results on the vehicle's chassis control functions. By monitoring the chassis control process triggered by the slope estimation value, the accuracy of the control response and the correctness of the control logic are verified. During the test, the slope estimation algorithm outputs the vehicle's current road slope state. The chassis domain controller executes the corresponding control strategy based on this slope state and collects and analyzes the execution results of the air suspension system, centralized air supply system, and rear-wheel steering system.

[0051] For the air suspension control function, the testing focused on monitoring the effect of independent left and right air spring leveling control. Under different side slope angles, the suspension heights on the left and right sides of the vehicle may differ. The chassis domain controller controls the left and right air springs to adjust independently based on the side slope estimation results to improve the vehicle's attitude. By collecting data on suspension height changes, adjustment response time, and target height tracking, the accuracy of the side slope estimation results in supporting the air suspension leveling control was verified.

[0052] For the centralized air supply control function, the high and low pressure air source switching logic of the air suspension system was monitored during the test, and the air source adjustment strategy was judged to meet the control requirements based on different side slope conditions. By analyzing the consistency between control commands and actual execution results, the logical correctness of the side slope estimation results participating in chassis control was verified.

[0053] For the rear-wheel steering control function, the test monitored the rear-wheel counter-steering control response under changes in side slope conditions, and analyzed whether the changes in rear-wheel steering angle met the control strategy requirements. By comparing the side slope estimation results, control commands, and execution feedback data, the supporting effect of side slope perception information on vehicle attitude control and driving stability was verified.

[0054] The intelligent driving decision-making layer verification is mainly used to evaluate the application effect of side slope information in intelligent driving functions. By simulating road environments with different side slope angles, it verifies the effectiveness of the intelligent driving system in making vehicle control decisions based on the side slope state. During the test, the HIL real-time simulation environment is configured with different side slope conditions to provide vehicle state information and road environment information to the intelligent driving domain controller. At the same time, the side slope estimation results serve as an important input for intelligent driving decisions.

[0055] For the lane-keeping compensation function, the driving conditions of the vehicle under different side slopes were simulated to verify whether the intelligent driving system can compensate for changes in the vehicle's lateral control strategy according to changes in the side slope, thereby improving the vehicle's lane-keeping stability under side slope conditions. The testing process focused on analyzing the correlation between changes in side slope estimation and lane-keeping compensation control, and evaluating whether the control strategy response conformed to the preset logic.

[0056] For the slope-limiting function, different slope angles were set to verify whether the intelligent driving system could adjust the vehicle speed limit strategy according to the road slope conditions. Under larger slope conditions, the system made a safety judgment based on the slope estimation results and took corresponding speed limit control measures. The testing verified the triggering conditions and execution effect of the speed limit strategy.

[0057] For the rollover risk assessment function, the effectiveness of the intelligent driving system in identifying vehicle risks based on slope estimation results was verified by simulating a road environment with a large side slope. During the test, the correlation between changes in side slope angle, risk assessment results, and control response was analyzed to evaluate the supporting capability of side slope estimation data for intelligent driving safety decision-making.

[0058] The vehicle comfort layer verification is primarily used to evaluate the impact of side slope estimation results on vehicle comfort control functions. By simulating a large side slope condition, it verifies the linkage control logic of the vehicle domain controller in adjusting comfort based on changes in vehicle posture. During the test, the side slope estimation results are input into the vehicle control function, and the vehicle controller performs seat posture adjustment and low-pressure air supply adaptation control based on the current vehicle posture.

[0059] Regarding the seat posture adjustment function, the system simulates the posture changes of a vehicle on a steep side slope to verify whether the seat system can adjust its posture accordingly to improve passenger comfort. During testing, estimated side slope values, seat adjustment commands, and actual execution status were collected. The correlation between these three parameters was analyzed to verify whether the control logic meets the design requirements.

[0060] For the low-pressure air source adaptation function, the coordination relationship between the vehicle comfort system and vehicle attitude control was verified by detecting the low-pressure air source control state under large side slope conditions. The adaptability of the vehicle comfort control strategy was evaluated by analyzing the correlation between changes in side slope estimation results and low-pressure air source adjustment.

[0061] By verifying the perception layer, chassis control layer, intelligent driving decision layer, and vehicle comfort layer, a multi-level closed-loop verification system is formed, covering algorithm accuracy evaluation, chassis execution control, intelligent driving decision-making, and occupant comfort adjustment. This verification method can not only evaluate the computational accuracy and stability of the slope estimation algorithm itself, but also further verify the practical application effect of the slope estimation results in different functional domains of the vehicle, realizing the complete verification of slope perception data from input, processing to vehicle control feedback.

[0062] The automated testing and fault injection function is used to automatically verify the vehicle slope estimation algorithm under all operating conditions and abnormal states. Through automatic execution of preset test scenarios, fault state simulation, test data acquisition, and evaluation result generation, it enables quantitative analysis of the slope estimation algorithm's performance and robustness evaluation. This function reduces the work of parameter adjustment, operating condition switching, and data processing during manual testing, improving the consistency, repeatability, and efficiency of the testing process.

[0063] The testing platform has a built-in automated test script for all operating conditions, which can automatically execute test tasks under different road environments, vehicle states, and side slope changes according to preset test conditions. The automated test script can automatically traverse side slope conditions with gradients from 0° to 15°. By setting different side slope angles, it simulates the vehicle's operating state in environments with gentle roads, small-angle side slopes, and large side slopes, and verifies the estimation accuracy of the side slope estimation algorithm under different slope conditions.

[0064] During the side slope angle variation test, the test script can automatically adjust the road side slope parameters in the HIL real-time simulation environment according to the set side slope gradient, and synchronously control the attitude rotation table to output the corresponding IMU dynamic attitude changes, so that the vehicle side slope estimation algorithm can receive sensor input data that matches the actual road conditions. By continuously traversing different side slope angles, the algorithm's adaptability to different slope conditions can be comprehensively evaluated, and the error variation between the algorithm output and the true side slope value can be analyzed.

[0065] Meanwhile, the automated test scripts support testing under different vehicle speed conditions, including high and low speed operating states. By adjusting the vehicle speed parameters in the HIL real-time simulation environment, the attitude changes of vehicles traversing side slope roads under different speed conditions are simulated, verifying the dynamic response capability and stability of the side slope estimation algorithm under different movement speeds. Since changes in vehicle speed affect the vehicle attitude change characteristics and sensor output states, the algorithm's adaptability to different vehicle operating states can be evaluated by traversing high and low speed conditions.

[0066] For different road surface adhesion conditions, the automated test script can call the road surface adhesion model to switch test environments and simulate the vehicle's side slope driving under different road adhesion states. By changing the road adhesion conditions, the impact of changes in tire-road interaction on the vehicle's attitude state and side slope estimation results can be analyzed, verifying the stability of the algorithm under different road surface environments.

[0067] Furthermore, the automated test scripts support testing under steering, straight-line driving, and bumpy coupled conditions. Under the straight-line coupled side-slope condition, the test vehicle travels along the side-slope road to verify the algorithm's ability to recognize the basic side-slope state. Under the steering coupled side-slope condition, by simulating the lateral movement of the vehicle during steering, the algorithm's ability to decouple the vehicle's roll posture from changes in the road side-slope is verified. Under the bumpy coupled side-slope condition, by simulating vehicle posture disturbances caused by uneven road surfaces, the algorithm's ability to suppress dynamic disturbances is verified. Through automatic traversal of these various coupled conditions, the complex side-slope environments that may occur during actual vehicle operation can be covered, improving the algorithm's verification coverage.

[0068] The testing platform simultaneously supports the injection of multiple fault scenarios to verify the reliability and redundancy degradation capability of the slope estimation algorithm under abnormal conditions. The fault injection function can simulate various fault states such as IMU noise injection, signal drift, signal disconnection, smart tire data failure, and suspension signal anomalies, enabling the testing environment to reproduce sensor and data anomalies that may occur during actual vehicle operation.

[0069] For IMU noise injection testing, simulated noise was added to the raw IMU data to verify the anti-interference capability of the slope estimation algorithm under conditions of sensor output interference. During the test, the degree of noise influence was adjusted, and the changes in the algorithm's output slope estimate were analyzed to evaluate the algorithm's adaptability to abnormal fluctuations in IMU data.

[0070] To address IMU signal drift faults, the algorithm's ability to handle slowly changing sensor errors was verified by simulating sensor zero-point offset and long-term output variations. The algorithm's ability to maintain stable output was evaluated by observing the difference between the slope estimation results under fault conditions and those under normal conditions.

[0071] To address signal disconnection faults, the algorithm's fault tolerance capability in the event of missing key sensor data was verified by simulating IMU data interruption. Simultaneously, through simulations of intelligent tire data failure scenarios, the ability of the slope estimation algorithm incorporating intelligent tire information to perform reasonable data degradation processing when some data sources are unavailable was verified.

[0072] To address scenarios involving abnormal suspension signals, the stability of the IMU-based and suspension information fusion-based side slope estimation algorithm under such conditions is verified by simulating suspension height signal deviations and abnormal changes. Through tests with different fault type combinations, the algorithm's adaptability to multi-source sensor anomalies can be analyzed, and the effectiveness of the algorithm's redundancy degradation logic can be verified.

[0073] In terms of automated data evaluation, the testing platform can automatically collect various types of data during the testing process, including true slope data output from the HIL real-time simulation environment, estimated data output from the slope estimation algorithm, and multi-domain control feedback data. By synchronously collecting and correlating multi-source data, it achieves complete data recording and result traceability during the testing process.

[0074] The true value data serves as the algorithm evaluation benchmark, used to calculate the error between the true value data and the algorithm's estimated data. By calculating the difference between the two, the estimation accuracy of the slope estimation algorithm under different operating conditions is evaluated. The algorithm's estimated data is used to analyze the algorithm's output change trend, dynamic response characteristics, and stability under abnormal operating conditions. The multi-domain control feedback data is used to evaluate the actual control effect of the slope estimation results applied to the chassis domain controller, intelligent driving domain controller, and body domain controller.

[0075] The testing platform can automatically calculate error indicators based on the collected data and generate corresponding accuracy reports. These reports record algorithm performance under different slope angles, vehicle speeds, road adhesion conditions, and coupling scenarios, quantifying algorithm performance through statistical methods. Simultaneously, for fault injection test results, the platform can automatically generate robustness evaluation reports, comprehensively assessing the algorithm's output stability, degradation capability, and fault adaptability under different abnormal conditions.

[0076] Through automated test scripts, fault injection, and automated data analysis, a complete testing process can be achieved, from normal operating condition testing to abnormal operating condition verification. Automated testing improves the efficiency of slope estimation algorithm verification, reduces testing bias caused by manual operation, and enhances the consistency and repeatability of the testing process. Fault injection enhances the algorithm's boundary condition verification capability, enabling early detection of problems under extreme conditions. Automated data evaluation enables quantitative analysis of test results, providing standardized data support for slope estimation algorithm optimization, parameter adjustment, and mass production verification.

[0077] The vehicle side slope estimation closed-loop test bench consists of a core attitude simulation unit, a HIL simulation unit, a domain control hardware unit, and a data acquisition unit. The units establish a complete hardware connection relationship through real-time communication and data synchronization to realize road side slope environment simulation, IMU physical attitude excitation, vehicle domain controller closed-loop interaction, and test data synchronous acquisition.

[0078] The core attitude simulation unit employs a high-precision lateral rotary table as the actuator for simulating vehicle attitude changes. The rotary table is fixedly mounted on the test base to ensure mechanical stability and attitude output accuracy during operation. Production-ready IMU sensors are rigidly mounted on the rotary table surface, with a rigid connection ensuring synchronization between the IMU sensors and the rotary table's attitude changes, avoiding attitude transmission errors caused by installation gaps, structural deformation, or other factors. The rotary table can perform single-axis lateral rotation according to test requirements, simulating lateral attitude changes of the vehicle in a side-slope road environment by adjusting the rotation angle.

[0079] The rotary table establishes a real-time communication connection with the HIL real-time system via serial communication, receiving dynamic angle control commands output by the HIL real-time system and adjusting its attitude in real time according to vehicle operating conditions and road slope changes. In a fixed slope test environment, the rotary table maintains the corresponding attitude based on the target slope angle. In dynamic gradual slope, bumpy coupled slope, and steering coupled slope test environments, the rotary table dynamically adjusts according to real-time angle commands, enabling the IMU sensors mounted on the rotary table to output raw measurement data corresponding to the vehicle's attitude changes. Through real-time interaction between the rotary table and the HIL real-time system, synchronous correspondence between the virtual road environment and physical attitude changes can be achieved, improving the realism of sensor data during testing.

[0080] The HIL simulation unit uses a real-time processor to run the vehicle dynamics model and the side slope scenario library, performing real-time calculations on vehicle driving status, road environment status, and side slope change processes. The vehicle dynamics model is used to simulate the dynamic response of the vehicle under different road conditions, speed states, and motion states. The side slope scenario library is used to store and call test scenarios such as fixed side slope, dynamic gradual side slope, bumpy coupled side slope, and steering coupled side slope.

[0081] The HIL simulation unit can generate corresponding vehicle operating conditions according to test requirements, including vehicle speed, steering status, road conditions, and true values ​​of side slope, and send test data to each control unit of the vehicle through the communication interface. The HIL simulation unit interacts with each domain controller in real time via CANFD and in-vehicle Ethernet communication, achieving high-speed transmission of vehicle status data, control commands, and feedback information. Specifically, CANFD communication is used to transmit vehicle control-related data, while in-vehicle Ethernet enables higher bandwidth data exchange, meeting the real-time communication needs between multiple domain controllers.

[0082] During testing, the HIL simulation unit sends vehicle operating condition commands to each domain controller and simultaneously uploads standard side slope ground truth data, providing a benchmark reference for evaluating the accuracy of the side slope estimation algorithm. The high-speed computing power of the real-time processor ensures that vehicle model operation, road environment simulation, and controller interaction remain synchronized in real time, providing a stable data foundation for closed-loop testing of vehicle side slope estimation.

[0083] The domain control hardware unit adopts the hardware control equipment actually used in mass-produced vehicles, including the chassis domain controller (CDC), the intelligent driving domain controller (ADC), and the body domain controller (BDC). Each domain controller is connected according to the actual vehicle's interconnections, making the test environment more closely resemble the real vehicle's operating state. The chassis domain controller, as the core control unit for the slope estimation algorithm and chassis control execution, receives raw IMU data, suspension height signals, and vehicle state information output from the HIL simulation environment, executes the slope estimation algorithm, and generates corresponding control strategies based on the slope estimation results.

[0084] The intelligent driving domain controller receives information related to vehicle posture and road conditions, and verifies intelligent driving function logic such as lane keeping compensation, slope speed limit, and rollover risk assessment under different side slope angles. The body domain controller verifies vehicle comfort-related control functions, including seat posture adjustment and related body control strategies under steep side slope conditions.

[0085] The domain control hardware unit is further integrated with the air suspension supply system, suspension height simulation module, and intelligent tire signal simulator to form a real-vehicle-level hardware closed-loop environment. The air suspension supply system is used to simulate the actual vehicle air suspension adjustment process, providing an execution environment for left and right suspension height adjustment and air supply control under side slope conditions; the suspension height simulation module is used to output vehicle suspension height change signals, providing data input for the side slope estimation algorithm based on IMU and suspension information fusion; the intelligent tire signal simulator is used to output dynamic tire lateral force, ground pressure, and tread deformation data, providing test conditions for the multi-source fusion side slope estimation algorithm with intelligent tire configuration.

[0086] Through the aforementioned hardware connection methods, the chassis domain controller, intelligent driving domain controller, and body domain controller can perform data interaction and control feedback in a real hardware environment, realizing a complete closed-loop verification from slope perception and algorithm calculation to vehicle function response.

[0087] The data acquisition unit is used to synchronously acquire and uniformly manage multi-source data during testing, including turntable attitude data, IMU raw data, algorithm output slope data, and control execution data from various domains. During testing, the data acquisition unit records the actual output attitude of the turntable in real time to analyze changes in physical attitude; it synchronously acquires raw data such as acceleration and angular velocity output by the IMU sensors to analyze the dynamic response characteristics of the sensors; it acquires the output data of the slope estimation algorithm to compare with the true value data output by the HIL to evaluate the accuracy of the algorithm; and it simultaneously acquires control execution data from the chassis domain, intelligent driving domain, and body domain to analyze the impact of the slope estimation results on the overall vehicle control function.

[0088] To ensure the correspondence between data from different sources, the data acquisition unit synchronizes the timestamps of multi-source data, enabling the rotation table attitude data, IMU data, algorithm calculation data, and control feedback data to be correlated and analyzed under a unified time reference. Through this time synchronization mechanism, the dynamic relationships between vehicle attitude changes, sensor responses, algorithm calculations, and control execution during the test can be accurately reconstructed, improving the accuracy and reliability of test result analysis.

[0089] Through the collaborative connection between the core attitude simulation unit, HIL simulation unit, domain control hardware unit, and data acquisition unit, a complete closed-loop test hardware architecture for vehicle side slope estimation is formed. This architecture can simultaneously realize road side slope environment simulation, real IMU physical attitude excitation, multi-domain controller hardware interaction, and synchronous analysis of test data, providing high-fidelity verification conditions for vehicle side slope estimation algorithms that closely approximate the actual vehicle operating environment.

[0090] The vehicle side slope estimation closed-loop test process is used to comprehensively verify the performance of the side slope estimation algorithm under different road environments, vehicle motion states, sensor configurations, and abnormal operating conditions. The test process relies on the HIL real-time simulation environment, high-precision attitude rotary table, vehicle domain control hardware, and data acquisition and analysis environment. By setting different test scenarios, collecting multi-source data, and evaluating the results, the test achieves complete testing of the side slope estimation algorithm from basic accuracy verification, dynamic operating condition verification, configuration compatibility verification, cross-domain control verification to fault robustness verification.

[0091] In the basic static accuracy test, it is mainly used to evaluate the basic estimation capability of the slope estimation algorithm under fixed road slope conditions. Before the test, a fixed slope condition is configured through the HIL real-time simulation environment, with slope angles set to 0°, 3°, 6°, 9°, 12°, and 15°, forming a static slope test environment covering different slope ranges. The HIL real-time simulation environment outputs the corresponding true slope data according to the set slope angle and sends the target attitude angle control command to the attitude simulation unit.

[0092] After receiving the angle control command, the high-precision attitude rotary table adjusts its own rotation angle to ensure that the mass-produced vehicle IMU sensor mounted on the rotary table surface is in an attitude state consistent with the target side slope angle. Since the IMU sensor is directly mounted on the rotary table structure, it can accurately sense the attitude changes generated by the rotary table and output corresponding raw IMU measurement data. The test platform simultaneously acquires the IMU output data and the side slope estimate calculated by the side slope estimation algorithm, and compares the algorithm output results point-by-point with the standard true side slope values ​​provided by the HIL real-time simulation environment.

[0093] Data analysis under different fixed slope angle conditions allows us to obtain the estimation error of the algorithm at various slope points. The testing process focuses on statistically analyzing static estimation errors, including the deviation between the algorithm's output slope angle and the standard slope angle, the error variation trend under different slope ranges, and the stability of the algorithm's output. Through continuous testing at multiple static slope angle points, the basic accuracy calibration of the slope estimation algorithm can be completed, determining the algorithm's estimation capability under basic road slope conditions, and providing a fundamental evaluation basis for subsequent dynamic working condition testing.

[0094] The dynamic side slope coupling test is mainly used to verify the dynamic response and decoupling capabilities of the side slope estimation algorithm when facing complex attitude changes during actual vehicle driving. The test environment is configured with dynamic gradual side slope conditions, straight bumpy side slope conditions, and turning coupled side slope conditions to simulate the complex side slope changes that vehicles may encounter in real road environments.

[0095] In dynamic gradient slope testing, the HIL real-time simulation environment generates continuously changing slope angles based on a preset road model. It controls a high-precision attitude rotary table to dynamically oscillate in real time according to the slope change trend, enabling the IMU sensor to continuously output dynamic data generated by attitude changes. By analyzing the following relationship between the slope estimation algorithm output and the HIL true value curve, the algorithm's response speed and tracking accuracy in continuously changing slope environments are evaluated.

[0096] In the straight-line bumpy side slope test, a road model was used to simulate the vehicle's driving state under conditions of side slope characteristics and road bumps. The rotating platform was dynamically adjusted according to changes in vehicle attitude, ensuring that the IMU data simultaneously included the effects of the road side slope and the attitude disturbances caused by vehicle bumps. During the test, the algorithm's estimation results were compared with the actual side slope changes to analyze whether the algorithm could effectively suppress the effects of vehicle vertical vibration and attitude disturbances, thus verifying the algorithm's adaptability to complex dynamic environments.

[0097] In the cornering coupled side slope test, the lateral motion of the vehicle during cornering is simulated, causing the vehicle's attitude change to be simultaneously affected by the road side slope and vehicle roll. The rotary table generates corresponding dynamic attitude changes based on the vehicle's motion, and the IMU output includes data reflecting the combined effects of vehicle roll and road side slope. Through test analysis, the side slope estimation algorithm is verified to effectively distinguish between vehicle roll and road side slope changes, achieving side slope decoupling under dynamic attitude disturbance conditions and improving the algorithm's reliability under actual vehicle motion conditions.

[0098] The dual-version algorithm compatibility test verifies the test platform's adaptability to different vehicle configurations and different side slope estimation algorithm schemes. During the test, the test mode is switched between a test mode without intelligent tires and a test mode with intelligent tires, depending on the vehicle configuration requirements, to separately verify the basic algorithm and the multi-source fusion algorithm.

[0099] In low-spec mode, the intelligent tire simulation signal is disabled, and only IMU sensor data and suspension height signal are retained as inputs for the slope estimation algorithm. The HIL real-time simulation environment outputs vehicle speed, steering status, and road condition information, the high-precision attitude rotary table outputs IMU dynamic attitude data, and the suspension simulation module outputs the vehicle suspension height difference signal. The slope estimation algorithm estimates the slope based on the IMU attitude information and the relationship between suspension geometric changes. The estimation accuracy and stability of the algorithm in the absence of intelligent tire information are analyzed through test results.

[0100] The low-spec mode is mainly used to verify the reliability of the side slope estimation algorithm under the basic vehicle configuration conditions, and to evaluate whether the algorithm can only... This invention constructs a test architecture that combines a real-time HIL simulation environment with physical rotary table IMU attitude simulation. It abandons the traditional method of generating virtual IMU data purely through software, instead utilizing a high-precision physical rotary table to directly drive mass-produced vehicle IMU sensors to generate realistic attitude responses. This ensures that the IMU output data includes attitude changes, sensor noise, and response hysteresis characteristics from actual operation. By using physical attitude excitation, the consistency between the laboratory test environment and the real vehicle road operating environment is improved, effectively reducing the deviation between simulation test results and real-vehicle test results, and enhancing the authenticity and reliability of the side slope estimation algorithm verification results.

[0101] This invention utilizes the HIL real-time simulation environment, a road slope scenario library, and automated test scripts to achieve comprehensive verification of static slopes, dynamic gradual slopes, bumpy coupled slopes, steering coupled slopes, and boundary fault conditions. During testing, it can simulate different slope angles, vehicle speeds, road surface adhesion states, and vehicle motion states, solving the problems of traditional test benches' limited testing scenarios, lack of dynamic conditions, and difficulty in reproducing extreme road environments. This allows the slope estimation algorithm to be verified under various conditions closely resembling real-world vehicle operating environments, improving the algorithm's testing coverage and meeting the needs of mass-production-level algorithm development and verification.

[0102] This invention enables a single test bench to be compatible with different vehicle configurations and different side slope estimation algorithms by setting up a test mode without intelligent tires and a test mode with intelligent tires. In the test mode without intelligent tires, the accuracy and stability of the basic side slope estimation algorithm can be verified based on IMU data and suspension signals. In the test mode with intelligent tires, data such as intelligent tire lateral force, ground pressure, and tread deformation can be introduced to verify the performance of the multi-source information fusion side slope estimation algorithm. Through the dual-mode compatible design, it can adapt to the development needs of high and low configuration vehicle platforms, reduce the cost of repeatedly building test equipment for different algorithm versions, and improve the versatility and utilization efficiency of the test platform.

[0103] This invention constructs a full-link closed-loop verification system consisting of attitude simulation, slope perception, chassis control, intelligent driving decision-making, and vehicle comfort control. This system not only evaluates the accuracy of the slope estimation algorithm output but also further verifies the practical application effects of the slope estimation results in different control domains of the vehicle. Through data interaction between the chassis domain controller, intelligent driving domain controller, and vehicle domain controller, the impact of the slope estimation results on functions such as air suspension leveling, centralized air supply control, rear-wheel steering compensation, lane keeping compensation, slope speed limiting, rollover risk assessment, and seat posture adjustment can be verified. This achieves complete testing from perception algorithm verification to cross-domain collaborative control verification of the entire vehicle, meeting the vehicle-level verification requirements under the intelligent vehicle domain control architecture.

[0104] This invention utilizes a built-in, fully automated test script to achieve automatic configuration of test scenarios, automatic traversal of test conditions, automatic data acquisition, and automatic analysis of test results. During testing, it can automatically execute tests based on preset conditions, combining different side slope angles, vehicle speeds, road adhesion states, and vehicle motion states, while simultaneously acquiring true side slope data, algorithm estimation data, and control feedback data from each domain. By automatically calculating error indicators, generating accuracy reports, and outputting robustness evaluation reports, it reduces manual operation and data processing, improves test execution efficiency, shortens the development and verification cycle of the side slope estimation algorithm, and meets the rapid iterative development needs of intelligent vehicle algorithms.

[0105] This invention outputs the true value of the standard slope through a real-time simulation environment using HIL (High-Intensity Logic) and simultaneously records rotary table attitude data, IMU raw data, algorithm-output slope data, and multi-domain control execution data using a data acquisition unit, achieving data correlation and result traceability throughout the testing process. By comparing and analyzing the true value data with the algorithm-estimated data, the slope estimation error can be accurately quantified, obtaining algorithm performance indicators under different working conditions. Simultaneously, the test results provide standardized data for algorithm parameter adjustment, accuracy calibration, fault optimization, and mass production acceptance, making the slope estimation algorithm testing process measurable, evaluable, reproducible, and traceable, thus improving the quality of algorithm development and verification.

[0106] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system, characterized in that, Includes the following steps: A closed-loop test bench for vehicle side slope estimation was constructed, and a HIL real-time simulation environment, a high-precision attitude rotary table, and a vehicle domain control hardware interaction environment were configured. The vehicle dynamics model, road side slope scenario library, road surface adhesion model, and vehicle driving condition model were established through the HIL real-time simulation environment to generate true side slope data corresponding to fixed side slope, dynamic gradual side slope, bumpy coupled side slope, and steering coupled side slope. Dynamic attitude adjustment is achieved by controlling a high-precision attitude rotary table based on true side slope data. The vehicle's IMU sensor is installed on the high-precision attitude rotary table, which simulates the vehicle's side slope driving attitude change. Raw IMU data containing noise and hysteresis characteristics is collected and output to the vehicle side slope estimation algorithm for side slope estimation. Based on the raw IMU data and the output of the slope estimation algorithm, the slope estimation closed-loop verification was carried out through the vehicle domain control hardware interaction environment. The chassis control response, intelligent driving decision response and vehicle comfort control response were verified respectively. The data fusion effect of different slope estimation algorithms was verified by switching between test modes without intelligent tires and test modes with intelligent tires. The closed-loop test process for vehicle side slope estimation is automated with traversal of operating conditions and fault injection testing. Test data is collected according to preset side slope angle, vehicle speed, road surface adhesion status and vehicle motion status. Error calculation and robustness evaluation are performed on the side slope estimation results under IMU anomaly, signal drift, signal disconnection, intelligent tire data anomaly and suspension signal anomaly.

2. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, The HIL real-time simulation environment includes a real-time processor, a vehicle dynamics model, a road side slope scenario library, a road surface adhesion model, and a vehicle driving condition model. The vehicle dynamics model is used to simulate the motion response of the vehicle under different speeds, steering states, and road conditions. The road side slope scenario library is used to configure test scenarios for fixed side slopes, dynamic gradual side slopes, bumpy coupled side slopes, and steering coupled side slopes. The road surface adhesion model is used to simulate different road adhesion states. The vehicle driving condition model is used to generate vehicle operating state data and output the corresponding side slope ground truth data as a benchmark for evaluating the accuracy of the side slope estimation algorithm.

3. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, A high-precision attitude rotary table is fixedly mounted on the test base. The actual vehicle-produced IMU sensor is rigidly mounted on the table surface of the high-precision attitude rotary table. The high-precision attitude rotary table supports single-axis lateral rotation and continuous angle adjustment, with an angle control accuracy of no more than 0.05°. The high-precision attitude rotary table interacts with the HIL real-time simulation environment in real time through a communication interface. It receives dynamic angle control commands based on the true side slope data and drives the actual vehicle-produced IMU sensor to generate corresponding attitude changes, so as to output raw IMU data containing sensor noise characteristics and response hysteresis characteristics.

4. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, The vehicle domain control hardware interaction environment includes a chassis domain controller, an intelligent driving domain controller, a body domain controller, an intelligent tire simulator, an air suspension centralized air supply module, and a suspension height signal simulation module. Among them, the chassis domain controller is used to execute the side slope estimation algorithm and output chassis control commands, the intelligent driving domain controller is used to execute driving decision control based on the side slope state, the body domain controller is used to execute posture comfort control, and the air suspension centralized air supply module and the suspension height signal simulation module are used to simulate the vehicle suspension control process, forming a hardware closed-loop verification environment for vehicle side slope estimation and vehicle control feedback.

5. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, In the non-intelligent tire test mode, the intelligent tire simulation signal input is turned off. The vehicle speed signal, steering signal and road condition signal are output through the HIL real-time simulation environment. The IMU dynamic attitude data is output through the high-precision attitude rotary table. The vehicle suspension height difference signal is output through the suspension height signal simulation module. The side slope estimation algorithm is based on the IMU data and the suspension height difference signal to estimate the side slope, which is used to verify the accuracy and stability of the side slope estimation algorithm based on the IMU and suspension geometry decoupling method.

6. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, In the intelligent tire test mode, the intelligent tire simulator is activated, outputting dynamic tire lateral force, ground pressure, and tread deformation data. The intelligent tire simulation data is then fused with IMU raw data, suspension height data, and visual simulation data. The performance of the multi-domain fusion side slope estimation algorithm is verified through multi-source information fusion. The intelligent tire simulator adjusts the output data according to the vehicle's operating status and the road side slope status to simulate the changes in tire force under different road conditions.

7. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, The closed-loop verification of slope estimation includes verification at the perception layer, chassis control layer, intelligent driving decision layer, and vehicle comfort layer. Specifically, the perception layer calculates static accuracy, dynamic following accuracy, steady-state error, and transient response error by comparing the output value of the slope estimation algorithm with the true slope value output by the HIL real-time simulation environment. The chassis control layer verifies the response of the air suspension to independent left and right leveling, centralized high and low pressure air supply switching, and rear-wheel reverse steering control triggered by the slope estimation results. The intelligent driving decision layer verifies lane keeping compensation, slope speed limiting, and rollover risk assessment logic. The vehicle comfort layer verifies seat posture adjustment and low-pressure air source adaptation control logic.

8. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, The automated test cycle includes automatically executing tests according to preset side slope angles, vehicle speeds, road surface adhesion states, and vehicle motion states. The side slope angle ranges from 0° to 15°, and the test conditions include different gradient side slopes, high and low vehicle speeds, different road surfaces with different adhesion, and coupled conditions of turning, straight driving, and bumpy driving. During the test, the HIL real-time simulation environment, high-precision attitude rotary table, and vehicle domain control hardware interaction environment are automatically controlled to operate in coordination, realizing continuous testing of the side slope estimation algorithm under multiple conditions.

9. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, Fault injection testing includes injecting noise into IMU sensor data, simulating IMU signal drift and signal disconnection, and simulating smart tire data failure and suspension signal abnormalities. Under fault conditions, the output results of the slope estimation algorithm are compared with the true slope values ​​output by the HIL real-time simulation environment to verify the data fault tolerance capability, adaptive weighted degradation capability, and estimation stability of the slope estimation algorithm under multi-source signal anomaly conditions.

10. The closed-loop test method for vehicle side slope estimation based on rotary table IMU simulation and HIL system according to claim 1, characterized in that, Test data acquisition includes synchronously acquiring high-precision attitude table attitude data, IMU raw data, slope estimation algorithm output data, and execution feedback data from each domain controller, and achieving multi-source data synchronization through a unified timestamp; the slope estimation error is automatically calculated based on the acquired data, and accuracy error curves, working condition pass rates, and fault robustness evaluation reports are generated for slope estimation algorithm parameter calibration, performance optimization, and mass production testing and acceptance.

Images