Real Vehicle Test Research Articles

Hierarchical hybrid trajectory planning for autonomous vehicle considering multiple road types

This paper combines A* algorithm and quintic polynomial to propose a novel hybrid trajectory planning method that can be used on both structured and unstructured roads for autonomous vehicle. The architecture of our approach can be generally divided into two layers. In the upper level, an improved A* algorithm is adopted to search for a feasible path as a reference based on the indication of global path, static and dynamic obstacles, road conditions and traffic rules. Particularly, relating to the road type, the evaluation functions are appropriately selected. According to this reference path, the relative position of the vehicle to the obstacles is evaluated. Hence, the target vehicle velocity and related position can be determined, and the desired velocity curve is then generated by a cubic polynomial. In the lower level, a segmented quintic polynomial is utilized for smooth the trajectory. Finally, the proposed method has been validated via Hardware-in-the-loop (HIL) and real-vehicle tests. The experimental results show that it can not only ensure a safe, curvature-continuous and kinodynamically feasible trajectory planning, but also own good adaptability to complex scenarios.

Modeling analysis of automatic parking path planning for wheeled AGV

Path planning is an important part of vehicle automatic parking. In this paper, the vehicle kinematics model is built based on the Ackerman steering principle, and the automatic parking path planning is explored. The starting conditions of one-step vertical parking are determined, the kinematic constraints are summarized, the parking tracking scheme is designed, and the parking path planning model is established. The parking path is fitted by B-spline curve to obtain the path optimization function. The parking space coordinate system is established according to the vertical parking space, and the parking path data obtained by simulation are input into the vehicle controller for one-step vertical parking test. The test results show that the error between the real vehicle test parking trajectory and the set trajectory is within the allowable range by using the parking path planning algorithm, indicating that the parking path planning strategy is effective.

A Real-Time Recognition System of Driving Propensity Based on AutoNavi Navigation Data.

Driving propensity is the driver’s attitude towards the actual traffic situation and the corresponding decision-making or behavior during the driving process. It is of great significance to improve the accuracy of safety early warning and reduce traffic accidents. In this paper, a real-time identification system of driving propensity based on AutoNavi navigation data is proposed. The main work includes: (1) A dynamic data acquisition method of AutoNavi navigation is proposed to obtain the time, speed and acceleration of the driver during the navigation process. (2) The dynamic data collection method of AutoNavi navigation is analyzed and verified through the dynamic data obtained in the real vehicle experiment. The principal component analysis method is used to process the experimental data to extract the driving propensity characteristics variables. (3) The fruit fly optimization algorithm combined with GRNN (generalized neural network) and the feature variable set are used to build a FOA-GRNN-based model. The results show that the overall accuracy of the model can reach 94.17%. (4) A driving propensity identification system is constructed. The system has been verified through real vehicle test experiments. This paper provides a novel and convenient method for building personalized intelligent driver assistance systems in practical applications.

Machine-learning-based hybrid recognition approach for longitudinal driving behavior in noisy environment

Driving behavior recognition has attracted wide attention as it can act as an important reference input of many vehicle intelligent control systems. In this paper, a real-time recognition of driver’s longitudinal driving behavior is investigated by proposing a hybrid adaptive pattern recognition method. Firstly, a framework of integrated behavior recognition model is established consisting of two sub models to cluster and label the sample data, respectively. Secondly, a new fast high-stability clustering method is proposed to solve the problem of clustering samples with non-negligible noise. And the test results show the clustering process can be completed within a short time lower than 0.1 s with different number of cluster centers. Then, support vector machine and artificial neural network are employed and trained to construct a heuristic self-labeling approach to label the clustered sample automatically in real time with high accuracy (92.9%) under an experimental driving cycle generated from our real-vehicle test bench. Subsequently, the two established modules are integrated and offline trained by 96 thousand historical data, and employed to a 400-second-long online application under a smoothly-varying driving condition, and three further applications under extreme driving cycles with about 1000 s, respectively. Simulation results show that the proposed integrated model is capable of eliminating the interference of noise, and has a relatively stable performance of recognition for different driving cycles with different driving behaviors (92.7% of overall performance for commonly cycles, and greater than 91.2% for extreme cycles), enhancing the capacity for online application.

Vehicle state estimation based on extended Kalman filter and radial basis function neural networks

To improve the reliability of vehicle state parameter estimation, a vehicle state fusion estimation method based on dichotomy is proposed. An extended Kalman filter algorithm is designed based on the vehicle 3 degrees of freedom dynamic model. Meanwhile, considering the influence of dynamic model and sensor noise and its coefficient selection on the estimation results, a radial basis function neural network estimation algorithm is designed. To further improve the reliability of the estimation algorithm, a method of estimation algorithm fusion is proposed based on the idea of mutual compensation between model- and data-driven estimation algorithms. The weights of the estimation results of different algorithms are assigned through the dichotomy. The redundancy and fusion of estimation algorithms can improve estimation performance. The effectiveness of the fusion method is verified by the co-simulation of MATLAB/Simulink and CarSim, and the real vehicle test. The results show that the change trend of the estimation result is consistent with the actual state parameters change trend, and the estimation accuracy after algorithm fusion is significantly improved compared to a single extended Kalman filter or radial basis function.

Visualization pipeline of autonomous driving scenes based on FCCR-3D reconstruction

Visual sensor-based human–computer interaction can provide environmental perception data of traffic road scenes for mechatronic intelligent vehicles. The existing three-dimensional (3D) reconstruction visualization schemes are mainly based on structured scene reconstruction with multiview geometry and are offline operations. It is difficult for large-scale autonomous driving scenes with wide baselines to ensure consistent reconstruction results in the wide field-of-view area and the center of the light-of-sight area. It is also difficult to achieve uniform reconstruction results for regions with drastic depth changes and inconsistent texture fineness. Therefore, we propose a 3D reconstruction-based visualization pipeline for autonomous driving scenes of wide baseline. First, the stereo geometric model is initialized. The parameters of distortion correction and polar calibration are calibrated. Second, the proposed feature chain code resampling (FCCR) is used to uniformly extract the features for regions with inconsistent texture fineness and depth variation. The filtered 3D data are obtained based on the stereo geometric model. Finally, the texture mapping of the mesh skeleton of 3D data is carried out by the bilinear combination of surrounding multilevel gradual textures. The visualization results of scene rendering and scene roaming are realized. The pipeline has been effectively verified on the simulated experimental platform, real vehicle test, and public dataset. It can meet the real-time 3D reconstruction visualization of wide-baseline scenes. The reconstruction texture is uniform, and the detail response is rich. We can present the intuitive, 3D, real-time, graphical physical quantities of scene data for drivers, intelligent vehicles, AR-HUD, and control systems. The method we present has research value and application significance for the practical engineering of electronic images.

Pure Electric Sweeper Performance Analysis and Test Verification of Dust Extraction Port

Purely electric sweepers are widely used in the urban sanitation industry due to their emission-free nature and ease of miniaturisation. The dust suction port is the key to the dust suction system of the sweeper, and improving the design level of the dust suction port of the sweeper can effectively improve the operational performance of the sweeper. Using the company’s self-developed Ruiqing S26 pure electric sweeper as the research object, a CFD (Computational Fluid Dynamics) method was used to analyse the influence of the dust suction port structure parameters (front baffle tilt angle, outlet diameter) and sweeper operation parameters (driving speed, operating pressure) on the dust suction effect of the sweeper, and was verified through real vehicle tests. The results of the study show that changing the angle and outlet diameter results in a change in the flow field characteristics and, consequently, the same change in the removal efficiency, with 65° and 160 mm being the optimum angle and outlet diameter, respectively. The tests investigated the flow field characteristics of the dust extraction opening and the removal efficiency. This study can provide theoretical reference for performance optimisation and parameter matching of the sweeper.

Energy Consumption Analysis of High-Speed Trains under Real Vehicle Test Conditions

The study collected statistics on the actual operation of national railway electric multiple units (EMUs) and compared the energy consumption of different EMU models at different speed levels. An important method for studying the relationship between speed and energy consumption of EMUs has been constructed based on group method of data handling to reflect how the energy consumption of different EMU models changes with speed. The energy consumption of CRH2 and CRH380A EMUs on flat and sloping lines was compared. Moreover, the start-up energy consumption of CRH2 and CRH380A EMUs was compared. The effect of the number of stops of CRH2 and CRH3 EMUs on energy consumption was analyzed. Furthermore, an idea for improving the general expression and calculation methods of traction energy consumption of EMUs was proposed. Finally, suggestions on the construction of a traction energy consumption information system were provided, and the selection of different high-speed EMU models and the reasonable determination of the operating speed were discussed.

Research on Low-Frequency Noise Control of Automobiles Based on Acoustic Metamaterial.

With the transformation of the trend of vehicle electrification, the overall noise level in the vehicle is gradually reduced. The problem of low-frequency noise in the vehicle, which was previously ignored, is becoming more and more prominent. To solve the vehicle low-frequency noise problem, a combination of real-vehicle tests and simulation analysis is carried out. During the test, the driver and passengers feel that there is a relatively obvious low-frequency roar noise in the car, which results from the structural radiation noise of the trunk door vibration. Therefore, to solve this problem, we design an acoustic metamaterial with lightweight and miniaturized features based on the local resonance principle of phononic crystals. Firstly, the selection of the resonant unit configuration and the design of the band gap are implemented. Then, the layout planning of the whole vehicle, the layout of the resonance unit and the design of the base frame are implemented. The actual vehicle test results show that: after attaching the designed acoustic metamaterial, the low-frequency noise sound pressure levels in the front and rear of the vehicle were reduced by 2.0 dB (A) and 2.3 dB (A), respectively, meanwhile, the interior noise sound quality was improved. The sound pressure level at the driver’s right ear in the car has an abnormal peak of around 35Hz. In addition, the driver and passengers feel that there is a relatively obvious low-frequency roar noise in the car, and through low-pass filtering of the collected signals, it is confirmed that the peak frequency is the main cause of the low-frequency roar in the car. The low-frequency steady-state noise of the car is generally considered to be the low-frequency vibration of the body panel and the radiation occurs. Through the finite element simulation analysis (Grid Participation Analysis) of the abnormal peak frequency, the results show that the low-frequency roar is caused by the low-frequency vibration of the tailgate sheet metal, and the problem peak frequency is not coupled with the acoustic cavity mode. Facing the problem of the low-frequency roar radiated into the car by the vibration of the tailgate sheet metal parts, based on the local resonance band gap theory, we developed a design to suppress the 35 Hz vibration of the tailgate sheet metal parts and meet the characteristics of lightweight and miniaturization. By attaching the acoustic metamaterial to the tailgate and performing CAE simulation of the whole vehicle, it is determined that the structure can indeed reduce the 35 Hz noise in the car and the peak value of the tailgate sheet metal vibration.

Real-time Acquisition and Analysis Test Data of Commercial Vehicle Handling Stability Based on LABVIEW Virtual Instrument

Aiming at the problems of complex time-consuming and heavy data processing workload in the field test of commercial vehicle handling stability, a vehicle handling stability test solution based on virtual instrument technology is proposed. This solution is based on the existing test equipment and conditions, using LABVIEW virtual instrument, integrated application of GPS, AHRS and steering force angle tester to form a set of road test system solutions that can realize vehicle handling and stability. The system can meet the requirements of the national standard Controllability and Stability Test Procedure for Automobile (GB / T 6323-2014) to realize the vehicle handling stability test, and the system test is verified by the real vehicle test data, which provides a more efficient and convenient test method for the study of vehicle handling stability test.

Anti-jerking and traction torque compensation strategy for P3 + hybrid electric vehicle during power upshift

In this study, an anti-jerking and traction torque compensation strategy for P3 + hybrid electric vehicle (HEV) equipped with hybrid automated mechanical transmission (HAMT) during power upshift is proposed. Firstly, to investigate the power upshift characteristics, a powertrain model with three torque ports and flexible drive shaft is established. Secondly, based on the established powertrain model, a hierarchical structure optimization strategy for power upshift is developed. The optimization strategy makes full use of the fast and accurate torque response of the traction motor, aims at simultaneously suppressing the longitudinal jerk and compensating the traction torque, solves the optimal traction motor torque to improve the power upshift performance of P3 + HEV. Finally, a layer-by-layer simulation and a real vehicle test of the power upshift process from first gear to second gear are implemented to verify the effectiveness of the proposed optimization strategy. The verification results show that the proposed strategy can realize smooth and traction torque uninterrupted power upshift process at the same time. Compared with the traditional PI controller the proposed optimization strategy can achieve smaller longitudinal jerk.

Research on Vehicle Rollover Warning and Braking Control System Based on Secondary Predictive Zero-Moment Point Position

<div class="section abstract"><div class="htmlview paragraph">To solve the contradiction between model complexity and the warning accuracy of the algorithm of the vehicle rollover warning, a rollover state warning method based on the secondary predictive zero-moment point position for vehicles is proposed herein. Taking a sport utility vehicle(SUV) as the research object, a linear three-degrees-of-freedom vehicle rollover dynamics model is established. On the basis of the model, the lateral position of the zero-moment point and its primary and secondary rates of change are calculated. Then, the theoretical solution of time-to-rollover of the vehicles is deduced from the lateral position of the secondary predictive zero-moment point. When the rollover warning index, the lateral position of the zero-moment point, is greater than the set threshold, the active anti-rollover control system will be triggered. The active anti-rollover braking control system adopts a hierarchical control strategy. Taking the rollover warning index as the control target, the upper controller calculates the braking torque required by the front-outer wheel based on the Fuzzy Proportional Integral Derivative (PID) algorithm and the differential braking control strategy. The lower executive controller calculates the corresponding brake wheel cylinder pressure and inputs it to the corresponding wheel through the Electrical Hydraulic Brake(EHB), so that a certain additional yaw moment can be formed to prevent the vehicle from rolling over. Under some typical driving conditions, Carsim-Matlab co-simulation and real vehicle tests are carried out. The comparative test results show that the proposed rollover warning method based on the secondary predictive zero-moment point position can accurately and effectively predict the roll-over state of the vehicle, and the active anti-rollover braking control system can control the vehicle in time and effectively to prevent the vehicle from rollover.</div></div>

A novel objective evaluation method of drivability for passenger cars considering subjective and objective consistency

In this paper, aiming at the problems that the random mixed working conditions are difficult to identify, the insufficiency of drivability evaluation indicators system and the lack of robustness for the evaluation model, a novel objective evaluation method of drivability for passenger cars is proposed. First, combining the sliding time window method and Naive Bayesian model (NBM), a hybrid working condition recognition model of multi-source sensor information fusion is constructed, and the real vehicle test under random mixed working conditions shows that the accuracy of working condition identification can reach 95.6%. Then, the objective evaluation index for crawling conditions, starting conditions and tip-in conditions are studied according to the suggestions of subjective evaluation engineers, an objective indicator evaluation system considering expert knowledge and information redundancy between indicators is constructed, among them, R-type clustering and principal component analysis (PCA) are used to eliminate the information redundancy between objective indicators. Finally, based on the back propagation (BP) neural network to optimize the fuzzy membership function, an improved fuzzy comprehensive evaluation (Improve-FCE) method considering subjective and objective consistency is designed. Real vehicle test and analysis results demonstrate that the development of intelligent drivability objective evaluation tool (I-DOET) has good reliability and stability, the maximum subjective and objective relative error of the improved FCE model is 5.13%, and the pass rate reaches 93.3%.

A Duration-distance Based Correction Strategy to Improve Cut-in Scenario Experience for Mass Production Adaptive Cruise Systems

The adaptive cruise system needs to meet the acceptance conditions of the vehicle manufacturer before mass production. One of the crucial evaluation scenarios is cut-in, especially when the distance between the target vehicle and the ego vehicle is close and the relative speed has little difference, or the vehicle in front suddenly brakes. In this paper, the desired acceleration and jerk of the adaptive cruise system are calculated using kinematic equations. The ego vehicle reaches the state of the target vehicle after a period of predicted time, where the predicted time is determined by the relative speed and distance between the two vehicles. To improve the subjective experience of the driver in the cut-in scene, the authors propose a fuzzy logic correction strategy based on the duration-distance of the target vehicle, which changes the trend of acceleration by reshaping the jerk. Cut-in simulation and real-vehicle test results show that this strategy can reduce the discomfort caused by rapid acceleration changes to the driver under the premise of ensuring the safety of the ego vehicle.

Anti-Rollover Control of a Counterbalance Forklift Based on Roll Energy Classification

This study proposes an energy stability index suitable for a counterbalance forklift, aiming to solve the problem that the traditional vehicle stability index load transfer rate is difficult to accurately judge the stability state of such a forklift. The analysis of the structure of the counterbalance forklift and contact relationship between the body and rear axle are used as bases to establish a two-stage dynamic model and derive real-time instability energy of a forklift truck in the rollover process under different conditions. Additionally, this research proposes an anti-rollover control strategy of the counterbalance forklift based on roll energy classification. The corresponding calculation formula of the forklift energy stability index is selected according to different stability conditions to perform roll energy classification. An anti-rollover oil cylinder and active rear wheel steering are used as an anti-rollover control actuator, and an anti-rollover control is implemented based on model predictive control. Last, a simulation based on European standard operating conditions and real vehicle test verification is conducted. Results show that the anti-rollover control strategy based on roll energy classification can effectively reduce the risk of forklift rollover and improve the stability and active safety of a counterbalance forklift.

Research on the Identification of Tyre-Road Peak Friction Coefficient under Full Slip Rate Range Based on Normalized Tyre Model

The accurate estimation of the tyre-road peak friction coefficient is the key basis for the normal operation of the vehicle active safety control system. The estimation algorithm needs to be able to adapt to various conditions encountered in the actual driving process of the vehicle and obtain the estimation results timely and accurately. Therefore, a new normalized strategy is proposed in this paper. The core is the equal ratio between the peak friction coefficient and the utilization friction coefficient between adjacent typical roads. This strategy can establish the direct connection (normalization) between tyre force and tyre-road peak friction coefficient through most tyre models in the field of vehicle dynamics and accomplish estimation by combining with the filtering algorithm. In addition, most of the vehicle dynamic estimation algorithms are limited by road excitation, and it is difficult to obtain satisfactory estimation results. This strategy can greatly reduce the system error caused by insufficient road excitation (slip rate is not 0.15–0.20) and improve the applicability of the estimation algorithm to the actual driving process of the vehicle. Finally, the magic formula (MF) tyre model is selected to describe the tyre characteristics after treatment of the normalized strategy; the tyre-road peak friction coefficient is estimated by combining the extended Kalman filter and vehicle dynamics model. Satisfactory estimation results are obtained in both simulation and real vehicle tests, which verifies the effectiveness of the proposed normalized strategy.

Time-Domain Signal Reconstruction of Vehicle Interior Noise Based on Data-Driven Method

Background: In the process of high-speed driving, there are many source signals that affect the ear noise of passengers in the car. It is important to obtain the reference signal of Active Noise Control (ANC) of the vehicle at high speed condition. Objective: This paper introduces a method to study the time-domain signal reconstruction of interior noise based on data-driven method. Methods: Based on the noise signal collected in a car, the key point signals affecting the interior noise are determined by the acoustic transfer path analysis method. Considering the time-varying characteristics of the noise signal and the complex nonlinear relationship of interior noise, a noise reconstruction model based on wavelet decomposition Radial Basis Function (RBF) neural network is established. And the BP neural network noise reconstruction model is set up to compare the reconstruction effect. Results: According to the reconstruction comparison, the average absolute error (0.0072) of the proposed algorithm model is smaller than the average absolute error of the noise reconstruction BP network model based on wavelet decomposition (0.0280), and the accuracy is improved by 74.29%. The average absolute error between the reconstructed value of RBF neural network and the real-value is smaller than that of BP neural network, and the error of the proposed model is less than 0.01. The method proposed in this paper can reconstruct the interior noise signal of vehicle accurately and effectively. Conclusion: This paper proposes a reconstruction model of vehicle interior noise signal based on wavelet decomposition RBF neural network algorithm driven by data-driven, and verifies the effectiveness of the algorithm with the real vehicle test data. The reconstruction method of RBF neural network based on data-driven wavelet decomposition provides a certain reference value for ANC to obtain high-precision reference signal.

Front vehicle detection based on improved fusion method for lidar and visual image

In order to ensure the detection accuracy, an improved adaptive weighted (IAW) method is proposed in this paper to fuse the data of images and lidar sensors for the vehicle object’s detection. Firstly, the IAW method is proposed in this paper and the first simulation is conducted. The unification of two sensors’ time and space should be completed at first. The traditional adaptive weighted average method (AWA) will amplify the noise in the fusion process, so the data filtered with Kalman Filter (KF) algorithm instead of with the AWA method. The proposed IAW method is compared with the AWA method and the Distributed Weighted fusion KF algorithm in the data fusion simulation to verify the superiority of the proposed algorithm. Secondly, the second simulation is conducted to verify the robustness and accuracy of the IAW algorithm. In the two experimental scenarios of sparse and dense vehicles, the vehicle detection based on image and lidar is completed, respectively. The detection data is correlated and merged through the IAW method, and the results show that the IAW method can correctly associate and fuse the data of the two sensors. Finally, the real vehicle test of object vehicle detection in different environments is carried out. The IAW method, the KF algorithm, and the Distributed Weighted fusion KF algorithm are used to complete the target vehicle detection in the real vehicle, respectively. The advantages of the two sensors can give full play, and the misdetection of the target objects can be reduced with proposed method. It has great potential in the application of object acquisition.

Delay-Dependent Sliding Mode Variable Structure Control of Vehicle Magneto-Rheological Semi-Active Suspension

The vehicle semi-active suspension with Magneto-Rheological Damper (MRD) has been a hot research topic of this decade, featuring the challenging task of the robust control with actuator time delay considerations. In this study, a delay dependent sliding mode variable structure control, based on the Linear Matrix Inequality (LMI), is proposed to suppress the vibration of the Magneto-Rheological Semi-Active Suspension (MRSS) control system. In accordance with the nonlinear characteristics of MRD, a dynamic model of automotive semi-active suspension system, considering time delay, is established. By defining a parameter-dependent Lyapunov switching functional, the conditions for asymptotic stability of closed-loop time delay system are derived, while the sliding mode variable structure control with reduced conservatism is designed. According to the method of LMI, the asymptotic stability problem of sliding mode is transformed into a feasibility problem, which can be solved by the solver ‘feasp’ in LMI toolbox. In addition, the calculation of the critical time delay of MRSS is expressed as a generalized eigenvalue optimization problem. For comparison purposes, three representative controllers, including a conventional sliding mode controller, a delay dependent controller, and a smith compensation, are studied. Simulation and real vehicle testing on bump and random road responses show that, the designed delay dependent controller can ensure the stability of the suspension system, weaken the influence of time delay on the control performance and effectively improve the ride comfort of the vehicle.

Retracted] Study on Driver Gaze Characteristics in Sight Distance Limited Section of Mountain Highway Based on Visual Information

In order to study the gaze behavior characteristics of drivers in mountainous road sections with limited sight distance, the real vehicle test is carried out by using Smart Eye Pro 5.7 noninvasive eye tracker. Combined with the sight distance change rate theory, 6 typical test representative mountainous sections are selected to study the gaze distribution law and gaze duration of drivers in different mountainous sections. The research shows that when the driver drives on the test section with the most unfavorable sight distance of 44 m, 50 m, and 56 m, the fixation characteristics of “from far to near” are significant, and the long fixation duration accounts for a large proportion of the driver. When the driver drives on the section with the most unfavorable sight distance of more than 70 m, i.e., the sight distance change rate of less than 1.33, the fixation characteristics of “from far to near” disappear. The driver’s fixation stability increases, the fixation freedom increases, and the proportion of medium and long fixation duration decreases. The data analysis provides a theoretical basis for drivers to pass safely in mountainous sections.