Real Vehicle Test Research Articles

Path tracking of autonomous tractor based on model-free adaptive control

Autonomous tractors, with path tracking as one of the key technologies, have become a research hotspot in recent years. However, the nonlinearity of and uncertainties in the autonomous tractor path tracking process can lead to challenges in terms of control. Although the model-based control method has been widely used in path tracking and achieved good results, only a limited number of model-free adaptive control (MFAC) methods have been developed for path tracking. This paper proposes an MFAC path tracking controller for autonomous tractors based on the use of a preview heading deviation method in which the preview heading angle is used as an input to control the steering, enabling the controller to achieve a small lateral error and stable control. In the simulation, the tracking effects of the designed controller, proportional integral differential (PID) and pure pursuit controllers at different speeds were compared. The results show that the proposed controller has good tracking ability and adaptability and can achieve high tracking accuracy. Finally, the tracking errors of the designed and PID controllers were compared in a real vehicle test, with the results verifying the effectiveness and practicality of the designed controller.

Research on Tire–Road Parameters Estimation Algorithm for Skid-Steered Wheeled Unmanned Ground Vehicle

Skid-steered wheeled vehicles can be applied in military, agricultural, and other fields because of their flexible layout structure and strong passability. The research and application of vehicles are developing towards the direction of “intelligent” and “unmanned”. As essential parts of unmanned vehicles, the motion planning and control systems are increasingly demanding for model and road parameters. In this paper, an estimation method for tire and road parameters is proposed by combining offline and online identification. Firstly, a 3-DOF nonlinear dynamic model is established, and the interaction between tire and road is described by the Brush nonlinear tire model. Then, the horizontal and longitudinal stiffness of the tire is identified offline using the particle swarm optimization (PSO) algorithm with adaptive inertia weight. Referring to the Burckhardt adhesion coefficient formula, the extended forgetting factor recursive least-squares (EFRLS) method is applied to identify the road adhesion coefficient online. Finally, the validity of the proposed identification algorithm is verified by TruckSim simulation and real vehicle tests. Results show that the relative error of the proposed algorithm can be well controlled within 5%.

Vibration fatigue analysis of the main pulley frame of a rotary drilling rig based on a real vehicle test

The pulley frame of a rotary drilling rig is located at the top of the mast and plays the role of changing the direction of wire rope movement. For a certain type of rotary drilling rig, the front cover plate of the main pulley frame was cracked and damaged during the working process, and the vibration signal of the pulley frame was tested under the relevant working conditions using the test method. 3D computer-aided design (CAD) modelling of the pulley frame was performed using the Creo software, from PTC [1]. The CAD model was used to create a finite element model using ANSYS [2] software. An analysis was performed using ANSYS Workbench and the nCode Designlife [3] software to provide a fatigue life estimation of the main pulley frame, focusing upon a random vibration fatigue life analysis approach. This paper describes an approach for industry and academia to analyze, design and optimize such pulley frames for rotary drilling rigs.

Human-Like Trajectory Planning for Autonomous Vehicles Based on Spatiotemporal Geometric Transformation

Human-driven vehicles and different levels of autonomous vehicles are expected to coexist on roads in the future. However, autonomous systems behave differently from their human-driver counterparts, these two behaviors are incompatible with one another, thereby negatively impacting traffic efficiency and safety. Herein, we present the construction of human-like trajectories for use in autonomous vehicles as a possible solution to this issue. We present a trajectory planning method based on the spatiotemporal geometric transformation of driving scenarios to generate human-like trajectories. Speed and safety redundancy data were collected through driving tests to understand human driving behaviors. Self-driving scenarios were abstracted as a Lorentz coordinate system under a three-dimensional Minkowski space-time. A surrounding-manifold tensor equation was established using differential geometry theory to depict the relationship between the trajectory constraints and the geometric spatiotemporal background. A metric tensor field can be solved from the equation to construct the corresponding “volcano space-time,” which is a three-dimensional general Riemannian space for placing the subject vehicle and the surroundings. The geodesics of the volcano space-time are solved using the geodesic equation and are projected back into the three-dimensional Minkowski space-time. Geodesic trajectories were fitted as Bézier curves in this study and corrected according to the vehicle dynamics constraints for trackability. In simulation and real vehicle tests, trajectories generated using the proposed algorithm exhibited collision avoidance and trackability, and the algorithm offered behaviors that were similar to those of human drivers under the same scenarios.

Single Pedal Control of Battery Electric Vehicle by Pedal Torque Demand With Dynamic Zero Position

Intelligent pedal control is a new development trend of vehicle intelligence, an important application of which is the single pedal control being the driving and braking inputs for energy management of battery electric vehicle. Usually, a constant pedal angle is used as the zero position that is a boundary between the driving and the braking. Based on the dynamic zero position that is a variable pedal angle depicting the direction switch point during pedal rotation, we show a single pedal control approach to operate the driving and the braking of battery electric vehicle, which can make full use of the pedal stroke and improve the control performance. Pedal torque demand is observed by a new pedal dynamics model that depicts the dynamic variation of pedal position. A nonlinear model predictive controller is designed to realize optimal-based energy management, satisfy dynamic demands, improve ride quality, and ensure wheel anti-slip safety. To validate the effect of the single pedal control approach, both simulations and real vehicle tests are carried out. The results illustrate that the energy consumption of single pedal control can be reduced than the dual pedal control in several representative driving cycles benefited from the more efficient use of regenerative braking.

Study on Detection and Recognition of Traffic Lights Based on Improved YOLOv4.

To resolve the issues of a deep backbone network, a large model, slow reasoning speed on a mobile terminal, low detection accuracy for small targets and difficulties detecting and recognizing traffic lights in real time and accurately with YOLOv4, a traffic lights recognition method based on improved YOLOv4 is proposed. The lightweight ShuffleNetv2 network is utilized to replace the CSPDarkNet53 network of YOLOv4 to satisfy the requirements of a mobile terminal. The reformed k-means clustering algorithm is applied to generate anchor boxes for avoiding the sensitivity issue of outliers and initial values. A novel attention mechanism named CS2A is added to enhance the extraction capability of effective features. Multiple data augmentation methods are combined to improve the generalization ability of the model. Ultimately, the detection and recognition of traffic lights can be realized. The S2TLD dataset is selected for training and testing, and it can be proved that the recognition accuracy and model size are greatly optimized. Meanwhile, a self-made dataset is selected for training and testing. Compared with the conventional YOLOv4, the recognition accuracy of the proposed algorithm for traffic lights’ state information increases by 1.79%, and the model size decreases by 81.97%. Appropriate scenes are selected for real-vehicle testing and the results demonstrate that the detection speed of the presented algorithm increases by 16%, and the recognition effect for small targets increases by 37% in comparison with conventional YOLOv4.

Cooperative Adaptive Cruise Algorithm Based on Trajectory Prediction for Driverless Buses

Cooperative adaptive cruise control (CACC) technology offers a proven solution to the current traffic congestion problems caused by the yearly growth of car ownership. Coping with random lane changes of bypass vehicles under the condition of traffic congestion is a challenge for urban driverless vehicles. In this paper, to meet the demand for high comfort driverless buses driving on urban roads, an active anti-disturbance following control method for driverless buses based on bystander vehicle intention recognition and trajectory prediction is proposed for the scenario of bystander vehicle cut-in during driving to alleviate the disturbance caused by bystander vehicles, improve passenger comfort, and suppress multi-vehicle oscillation. The simulation results show that the intelligent prediction system-based queue reduces the traffic oscillation rate by an average of 9.8% and improves the comfort level by an average of 11% under side-car insertion conditions. The results of the real vehicle test show that the vehicles based on the intelligent prediction algorithm have a 25.5% reduction in maximum speed adjustment, 14.5 m average reduction in following distance, 6% improvement in comfort, and 27% improvement in rear vehicle comfort.

Vehicle Motion Planning With Joint Cartesian-Frenét MPC

The Frenét frame is commonly used in motion planning for its superiority of reshaping nonconvex curving boundaries and decoupling lateral and longitudinal behaviors. Nevertheless, such frame is not friendly for describing vehicle-related characteristics such as body shape and classical dynamic models. By contrast, the Cartesian frame facilitates straightforward modeling of vehicle dynamics, but confronts nonconvex constraints on curving lanes. For purpose of combining the complementary superiority of both frames, this letter proposes a joint Cartesian-Frenét MPC planning method based on a locally linear intermediary connection which is updated along with solving iterations. The augmented Cartesian-Frenét state space is constrained by a Cartesian vehicle control model that stands for tracking feasibility, a set of Frenét road constraints that account for driving safety, and a Cartesian-Frenét transformation that connects both frames. The final optimization is solved iteratively, where each iteration formulates into a quadratic programming problem with help of the apriori trajectory solved by the previous iteration. A comparative simulation and a 5-month-long real-vehicle test prove that the proposed method can produce a feasible and smooth trajectory within a safe area efficiently in complex real-world scenarios.

Modeling, Simulation and Implementation of All Terrain Adaptive Five DOF Robot.

The ability of an off-road robot to traverse obstacles determines whether the robot can complete complex environmental tasks. In order to improve the off-road ability of off-road robots, this paper proposes a new design idea, in which four hub motors are the power system of the robot, the steering system of the robot is composed of a steering machine and a stepping motor, and a five degree of freedom robot model is established. The body structure is designed according to the characteristics of arthropods. The body structure is divided into three modules, and the connecting rod is used as the joint system of the robot to connect the three parts. The body can deform when facing complex obstacles, so as to adapt to different terrains. Then the body structure is simplified, and a mathematical model is established to describe the mathematical relationship between body joint changes. In order to verify the ability of the adaptive all-terrain cross-country robot to traverse obstacles, the load-bearing experiment and obstacle-crossing simulation experiment were carried out through Adams software, and the continuous traversing performance at low obstacles and the ability to break through high obstacles were tested, respectively. The experimental results prove that the designed adaptive all-terrain off-road robot is feasible, has good carrying capacity, and has good passability in the face of low obstacles and high obstacles. Using Ansys software to perform finite element analysis on the wheel connection, the experimental results show that the strength meets the material strength requirements. Finally, a real vehicle test is carried out to verify the correctness of the simulation results.

A new preprocessment method for road peak adhesion coefficient fusion estimation

Accurate estimation of road peak adhesion coefficient is of great significance in vehicle active safety systems. In order to prevent the vehicle from running out of control in actual driving process, the estimation algorithm should be able to adapt to various working conditions, and transmit estimated information of road peak adhesion coefficient to Electronic Control Unit (ECU) of the vehicle timely and accurately. Therefore, a new preprocessment method for the fusion estimation algorithm of road adhesion coefficient is proposed. The core of this method is the equal ratio relationship between the longitudinal, lateral peak adhesion coefficients and the utilization adhesion coefficient under adjacent typical roads. According to this relationship, this method accomplishes normalization of the tyre model in the form of introducing parameters from the outside, so that the tyre model can be combined with the filtering algorithm to be applied to estimation, which solves the problem that the precise tyre model cannot be used for road adhesion coefficient estimation due to its complex construction. This method can adapt to most of tyre models in the field of vehicle dynamics. In addition, the existing estimation algorithms need sufficient excitation (when the comprehensive slip ratio is 0.15–0.20) to estimate the accurate road peak adhesion coefficient. This method can greatly reduce the system error caused by insufficient road excitation, and can also obtain accurate estimation results when the excitation is insufficient. Finally, in order to verify the preprocessment method, the magic formula tyre model is used to describe the tyre characteristics. After processing by the preprocessment method, the road peak adhesion coefficient is estimated in combination with Unscented Kalman Filter (UKF). The high accuracy and timeliness of the estimated results in simulation and real vehicle tests verify the effectiveness of the preprocessment method.

Thermal analysis and performance testing of heavy load truck composite brake linings

Abstract One of the most important variables that determine the braking performance of heavy-duty trucks is the high temperature increases in the brake components. In this study, the effects of high temperatures in brake lining compositions were investigated using thermogravimetric analysis (TGA). Furthermore, dynamometer tests and SEM analysis were performed. Three different compositions, which are called standard, fly ash and sisal fiber mixture, were compared with existing commercial heavy-duty truck brake lining. The thermal behavior of samples analyzed in TGA machine compared to the weight losses with increasing temperatures until 900 °C. Dynamometer tests which simulated the real part production and vehicle tests were realized in brake lining supplier. Finally, with the SEM analysis, both normal and worn surfaces were analyzed and reasons for wear surfaces were investigated. The findings suggest that there is a high potential in fly ash mixture to apply in commercial trucks applications which occurs at very high temperatures and is very suitable for environment and cost reduction points.

Improved Traffic Sign Detection Algorithm Based on Faster R-CNN

The traffic sign detection algorithm based on Faster Region-Based Convolutional Neural Network (R-CNN) has been applied to various intelligent-vehicles driving scenarios. However, the model of the current detection algorithm has certain shortcomings, which include the influence of weather and light, the detection of distance traffic signs, and the detection of similar traffic signs. To solve these problems, this paper proposes an improved traffic sign detection method based on Faster R-CNN. First, we propose a fusion method that fuses the feature pyramid into the Faster R-CNN algorithm. This fusion method can extract object features with precision and decrease the influence of weather and light. Second, a deformable convolution (DCN) which can train the algorithm to identify traffic signs with precision and make similar signs more distinguishable, and in particular make it work better with distorted images, is added to the backbone network. Lastly, we apply ROI align to replace the ROI pooling, which can avoid the distant traffic sign detail loss caused by pooling and increase the detection precision of distant traffic signs. The experimental results on both the TT100k dataset and real intelligent vehicle tests demonstrate that the algorithm is superior to the original Faster R-CNN algorithm and four other state-of-the-art methods in traffic sign detection, specifically in small-target traffic sign detection and low-intensity environments such as sunset time and rainy days. Therefore, the proposed method is helpful to improve the traffic sign detection performance in extreme environments (low-light intensity or rainy weather).

Research on Direct Yaw Moment Control of Electric Vehicles Based on Electrohydraulic Joint Action

To solve the problem of lateral instability of the vehicle caused by insufficient lateral force of the tires due to the insufficient torque provided by the motor to the tire when the vehicle turns sharply or avoids obstacles in an emergency, a layered control method is used to design a lateral stability control system. The upper decision layer selects the yaw rate and the sideslip angle of the center of mass as the control variables and uses the joint state deviation of the yaw rate and the sideslip angle of the center of mass and the rate of change of the deviation as the input of the sliding mode variable structure controller to calculate the additional yaw moment required to maintain vehicle stability. The lower torque distribution layer realizes the distribution of torque through the electro-hydraulic coordinated control method: the torque distribution rule based on real-time load transfer calculates the torque corresponding to the control wheel and generates the torque through the hub motor and transmits it to the wheel. When the torque output from the motor cannot provide sufficient torque for the vehicle, hydraulic braking is used as a compensating control, and the difference between the required yaw torque and the motor-generated yaw torque is used as the required torque for hydraulic control to calculate the wheel cylinder pressure required to brake the wheels. Based on the joint simulation model of MATLAB/Simulink and Carsim, the sine and double shift line working condition are selected for stability simulation experiments. From the simulation results, it can be seen that the yaw rate and sideslip angle of the center of mass of the vehicle with sliding mode control and electro-hydraulic coordinated control almost coincide with the ideal value curve, which are both smaller than the output parameters of the uncontrolled vehicle. From the perspective of the motor output torque, compared with pure motor control, the effect of electro-hydraulic coordinated control is better, and the hydraulic system can compensate for the braking torque in time and enhance the lateral stability of the vehicle. The designed control strategy can make the yaw rate and the sideslip angle of the center of mass of the vehicle follow the reference value better, which can effectively avoid the vehicle sideslip and instability and improve the vehicle yaw stability and driving safety. However, due to the limitations of experimental equipment, the proposed method could not be applied to the real vehicle test. The real vehicle test can better test the control effect of the proposed method.

Multiobjective Optimization of Vehicle Handling and Stability Based on ADAMS

In order to improve the steering characteristics, roll characteristics, and lateral characteristics of heavy commercial vehicles, a multiobjective handling stability comprehensive score optimization model (MHSCS optimization model) was proposed in this study. In this study, the main factors affecting the steering characteristics, roll characteristics, and lateral characteristics are determined by combining the road test method and ADAMS software simulation analysis method. Based on the above road tests and ADAMS software simulation analysis, the MHSCS optimization model was proposed with “understeering degree,” “vehicle roll angle”, and “rearward amplification (RWA)” as evaluation indexes, and the response surface method combined with genetic algorithm was used to carry out multiobjective optimization of heavy commercial vehicle. ADAMS simulation results show that the comprehensive improvement degree of steering characteristics, roll characteristics, and lateral characteristics of a heavy commercial vehicle after optimization is 15.26%. Finally, the field road test results show that the scoring error between the comprehensive scoring optimization model and the real vehicle test was controlled at 0.4%, which proves the accuracy of the optimization model established in this study and effectively improves the handling stability of heavy commercial vehicles.

Optimal Speed Model of Urban Underwater Tunnel Based on CO2 Emissions Factor

This study aims to reduce air pollution caused by vehicle emissions in confined spaces and realize low-carbon travel in urban underwater tunnels. Based on the MEET (Methodologies for Estimating Air Pollutant Emissions from Transport) and COPERT (Computer Programme to Calculate Emissions from Road Transport) models, combined with real vehicle test data, an urban underwater tunnel speed–CO2 emissions factor model was constructed. Results show that: Different working conditions have a great impact on the MEET model; load and slope factors expand the actual CO2 emissions factor, which is different from the actual situation. The CO2 emissions factor in the COPERT model is negatively correlated with the speed, and there are fewer variables in the model, so the parameters are more controllable and more in line with the actual situation. According to the vehicle gasoline consumption and taking CO2, i > GC as the judgment index, the optimal limit speed of the ramp is calculated to be 40 km/h, while the main line maintains the existing state of 60 km/h. The model is simple and easy to operate, can be applied to estimate vehicle CO2 emissions factor at underwater tunnels in other cities, providing a basis for traffic management and effectively realizing low-carbon travel.

Research on driving visual comfort of two‐lane mountainous road based on traffic safety evaluation

AbstractIn order to obtain the quantitative relationship between multisource data from road landscape and driving visual comfort in two‐lane mountainous road, we collected the driver's eye movement data and road environment data through real vehicle tests. The threshold segmentation method and the target landscape area ratio method are used to process the visual image, and the spatial enclosure degree is used to express the driver's sense of spatial closure. In addition, an HSV color model is established to characterize the driver's visual perception of the mountain highway landscape environment. Combining the target landscape area ratio of road landscape with the amount of color information, the calculation method of road landscape information is obtained. According to the characteristics of normal distribution of pupil area ratio, five grades of driver's visual comfort level are proposed, and the regression relationship between comfort level and landscape information is established. The study found that there is a significant negative correlation between the amount of road landscape information and the driver's visual comfort. When the amount of road landscape information is less than 0.338, the driving comfort is higher, which is conducive to driving safety. However, when the amount of landscape information in the road area exceeds 0.644, the driving comfort decreases significantly, which is detrimental to driving safety.

Millimeter-Wave Radar in-the-Loop Testing for Intelligent Vehicles

Intelligent vehicle testing and validation technology is a hotspot in the field of intelligent vehicles. Millimeter-wave radar has been widely used in intelligent vehicles and advanced driving assistant systems (ADASs). Owing to low efficiency, high maintenance cost and limited scenario types, the traditional real-vehicle-testing can make it difficult to meet the requirements of intelligent vehicle testing. Also, the degree of reality of millimeter-wave radar sensor models in the virtual test platform is not convincing enough. To solve the mentioned problems, the millimeter-wave radar in-the-loop (mmRil) testing for intelligent vehicles is proposed in this study. First, the radar echo simulation method is developed. Next, based on the millimeter-wave flight time, Doppler frequency shift and the radar power attenuation equation, the mmRil test system is built. A geometric model of a radar target detection and a power attenuation model under bad weather conditions are established. Then, the accuracy of the testbench is verified. Finally, a typical intelligent vehicle control strategy is used for testing under various working conditions and bad weather using the mmRil testbench. The test results show that the mmRil testing has higher authenticity and accuracy than the simulation testing and can be used to test the intelligent vehicle algorithms under severe weather conditions. Compared with the real vehicle testing, the mmRil testing has higher efficiency, repeatability and safety. Therefore, the mmRil testing represents an indispensable testing method in the field of intelligent vehicle testing.

Improved Model Predictive Control Path Tracking Strategy Based an Online Updating Algorithm With Cosine Similarity and a Horizon Factor

Path tracking plays an essential role in autonomous vehicles. To ensure tracking accuracy and improve tracking adaptability in different velocities, a path tracking strategy based on an improved model predictive control (MPC) method is presented in this research. First, a path tracking controller based on improved MPC based an online Updating algorithm is constructed. The update mechanism is triggered by using the cosine similarity, when the cosine similarity is lower than the predefined threshold value, making the state space and cost function of MPC match real-time conditions to rectify the sensitivity of MPC to vehicle speed. Additionally, to further enhance the controller’s performance, a fuzzy control is employed to determine the horizon factor for optimizing the prediction horizon and control horizon online. Also, the weighting factors of the prediction horizon and control horizon are discussed to improve the adaptability at varying velocities. Next, the improved MPC controller is compared with the traditional MPC controller for a double lane change maneuver. The validation results demonstrate that the proposed strategy achieves good adaptability with satisfactory tracking accuracy at various velocities. Finally, the feasibility of the proposed strategy is verified in a real prototype vehicle test.

Research on 3D Point Cloud Data Preprocessing and Clustering Algorithm of Obstacles for Intelligent Vehicle

Environment perception is the foundation of the intelligent driving system and is a prerequisite for achieving path planning and vehicle control. Among them, obstacle detection is the key to environment perception. In order to solve the problems of difficult-to-distinguish adjacent obstacles and easy-to-split distant obstacles in the traditional obstacle detection algorithm, this study firstly designed a 3D point cloud data filtering algorithm, completed the point cloud data removal of vehicle body points and noise points, and designed the point cloud down-sampling method. Then a ground segmentation method based on the Ray Ground Filter algorithm was designed to solve the under-segmentation problem in ground segmentation, while ensuring real time. Furthermore, an improved DBSCAN (Density-Based Spatial Clustering of Application with Noise) clustering algorithm was proposed, and the L-shaped fitting method was used to complete the 3D bounding box fitting of the point cloud, thus solving the problems that it is difficult to distinguish adjacent obstacles at close distances caused by the fixed parameter thresholds and it is easy for obstacles at long distances to split into multiple obstacles; thus, the real-time performance of the algorithm was improved. Finally, a real vehicle test was conducted, and the test results show that the proposed obstacle detection algorithm in this paper has improved the accuracy by 6.1% and the real-time performance by 13.2% compared with the traditional algorithm.

Research on Fuzzy Control Method of Automobile Antilock Braking System

The automobile Antilock Braking System (ABS) can prevent the wheel from locking by automatically adjusting the brake pressure at a high speed during emergency braking. While improving the braking effect of the car, it can also keep the steering ability of the car and ensure the safety of passengers. The automobile braking process has a strong nonlinearity and uncertainty, so exploring a simple and reliable control strategy is the focus of current research. Based on this, this paper proposes a variable-domain fuzzy proportional integral differential (PID) control strategy using a particle swarm optimization (PSO) algorithm to iteratively find the optimal theoretical domain. First, the PSO strategy is used to obtain the optimal regulation parameters. Then, the dynamic antiinterference ability of the control system is guaranteed by the variable theory domain fuzzy PID control, and the PID parameters and variable theory domain expansion factor are optimized by the PSO to increase the utilization degree of fuzzy rules initially set. Finally, compared with the traditional control strategy, the simulation and real vehicle test prove that the proposed system can significantly improve the control accuracy of abs. The proposed system has the advantages of small overshoot, short adjustment time, strong antiinterference ability, and practicability, which greatly improves the tracking performance of the ABS.