Table of Content

25 July 2022, Volume 44 Issue 7 Previous Issue    Next Issue

Select

Reservation-based Vehicle Platoon Control at Unsignalized Intersections Under Mixed Traffic Condition Yihe Chen,Weiwei Kong,Jie Yu,Keqiang Li,Yugong Luo 2022, 44 (7):  953-959.  doi: 10.19562/j.chinasae.qcgc.2022.07.001 Abstract ( 279 )   HTML ( 30 )   PDF (1799KB) ( 296 )   Save The most current researches on the coordination control for intelligent and connected vehicle （ICV） at unsignalized urban intersections assume a 100% market penetration rate of ICV without consideration of the effects of human driven vehicle （HDV） on the system. This paper aims at resolving the coordinative control problem at unsignalized intersections under mixed traffic condition with HDVs and ICVs coexist， and exploring the effects of ICV penetration rate on system performance. Firstly， a reservation-based hierarchical platoon control framework is proposed， in which， the top layer assigns the intersection crossing moment for each vehicle and the bottom layer is in charge of execution and speed trajectory planning. Then the control and scheduling strategies of ICV and HDV are formulated respectively based on platoon reservation way. Finally， a comparative simulation is performed under different ICV penetration rates and traffic flows. The results show that under an ICV penetration rate of 10% to 100% and a traffic flow of 400 to 1 000 vehicles per hour， the proposed method can effectively enhance the traffic efficiency and reduce the average fuel consumption of vehicles， compared with actuated traffic light control strategy. Figures and Tables | References | Related Articles | Metrics

Select

Eco-driving Control at Signalized Intersections with Consideration of Time Delay Xinyu Chen,Lijun Qian,Qidong Wang 2022, 44 (7):  960-968.  doi: 10.19562/j.chinasae.qcgc.2022.07.002 Abstract ( 171 )   HTML ( 10 )   PDF (3112KB) ( 229 )   Save In view of the eco-driving problem at signalized intersections， a robust model predictive control method with considerations of time-varying and time delay is proposed in this paper. Firstly， an optimal target speed calculation method is given based on the traffic signal timing information， the optimal cruising speed and the front vehicle state information， and a discrete nonlinear system model with time-varying time delay is established. Then Lyapunov function is constructed by using the upper and lower bound of time delay， and the feedback control law is solved by utilizing linear matrix inequality to ensure the robust stability of the system and the optimization of target performance. Finally， simulations and intelligent vehicle tests are conducted for verification. The results show that the control method proposed can ensure platoon safety and improve fuel economy and traffic smoothness. Figures and Tables | References | Related Articles | Metrics

Select

A Decision-making Method for Longitudinal Autonomous Driving Based on Inverse Reinforcement Learning Zhenhai Gao,Xiangtong Yan,Fei Gao 2022, 44 (7):  969-975.  doi: 10.19562/j.chinasae.qcgc.2022.07.003 Abstract ( 313 )   HTML ( 18 )   PDF (3153KB) ( 373 )   Save Obtaining autonomous driving decision-making strategies by using human driver data is a hot spot in current research on autonomous driving technology. Most of the classic reinforcement learning decision-making methods artificially construct reward functions by designing formulas related to safety， comfort， and economy， which leads to a big gap between decision-making strategies and human drivers. This paper uses the maximum margin inverse reinforcement learning algorithm. Taking the driver’s driving data as expert demonstration data， a reward function is established， and the longitudinal autonomous driving decision-making by imitating the driver is realized. The simulation test results show that compared with the reinforcement learning method， the reward function of the inverse reinforcement learning method is automatically extracted from the driver's data， which reduces the difficulty of establishing the reward function， and the obtained decision-making strategy has higher consistency with the driver’s behavior. Figures and Tables | References | Related Articles | Metrics

Select

Automatic Driving Edge Scene Generation Method Based on Scene Dynamics and Reinforcement Learning Jiangkun Li,Weiwen Deng,Bingtao Ren,Wenqi Wang,Juan Ding 2022, 44 (7):  976-986.  doi: 10.19562/j.chinasae.qcgc.2022.07.004 Abstract ( 381 )   HTML ( 21 )   PDF (3967KB) ( 412 )   Save For solving the problem of low-probability and high-risk edge test scenes， an edge scene reinforcement generation method based on scene dynamics and reinforcement learning is proposed to fulfill the automatic generation of edge scenes and simulate the features of confrontation and game behavior between vehicles in the real world. Firstly， the scene models dynamically changing with time is described by a set of differential equations as a scene dynamic system. Then， neural network is used as a general function approximator， to construct the scene black-box controller for fulfilling the optimization solving of edge scene controller based on reinforcement learning. Finally， with the cut-in scene for overtaking as an example， a verification simulation is performed with Matlab/Simulink software. The results show that the edge scene models generated by reinforcement learning exhibit an excellent performance in terms of scene interactive gaming， scene coverage and repeatable test. Figures and Tables | References | Related Articles | Metrics

Select

Point Cloud Segmentation Algorithm Based on Adaptive Threshold DBSCAN for Roadside LiDAR Lisheng Jin,Yang He,Huanhuan Wang,Zhen Huo,Xianyi Xie,Baicang Guo 2022, 44 (7):  987-996.  doi: 10.19562/j.chinasae.qcgc.2022.07.005 Abstract ( 278 )   HTML ( 12 )   PDF (5167KB) ( 212 )   Save Aiming at the defect of lidar point-cloud data collected from roadside， that the point-cloud in the same target is divided into multiple targets caused by the reduction of density due to the increase of distance， a roadside point-cloud segmentation algorithm based on DBSCAN with adaptive threshold is proposed. Firstly， the collected roadside point-cloud is filtered by using improved GPF and straight-through filter， and the non-ground point-cloud is extracted from road area. Then， the adaptive coefficient function is constructed based on the effective distance and sigmoid function， and the selection rule of radius threshold in vicinal point search during clan grow in DBSCAN clustering algorithm is optimized. Finally， the non-ground points are clustered using the DBSCAN clustering algorithm with adaptive threshold， with the point-cloud subordinating to single target obtained. The continuous data of 1 055 frames of real scenes are collected for testing， and the results show that the average C-H coefficient increases by around three times， the average D-B coefficient rises by 4.52%， and the average contour coefficient is raised by 77.78%， indicating that the segmentation algorithm based on DBSCAN with adaptive threshold can enhance the intra-class consistency and inter-class difference of the point-cloud cluster， effectively reducing the over-segmentation phenomenon of roadside point-cloud with a high engineering application value. Figures and Tables | References | Related Articles | Metrics

Select

Traffic Vehicles Intention Recognition Method Driven by Data and Mechanism Hybrid Jian Zhao,Dongjian Song,Bing Zhu,Hangzhe Wu,column：Han Jiayi,Yuxiang Liu 2022, 44 (7):  997-1008.  doi: 10.19562/j.chinasae.qcgc.2022.07.006 Abstract ( 195 )   HTML ( 11 )   PDF (3744KB) ( 152 )   Save Traffic vehicle intention recognition is of great significance to improve the performance of intelligent vehicle decision-making and planning. This paper analyzes each stage of the driver’s lane changing process from the perspective of the driving behavior generation mechanism， and establishes the driver’s intention prediction model based on Markov decision process （MDP）， the lane changing feasibility analysis model based on the dynamic safety field， and the vehicle behavior recognition model based on bi-directional long short-term memory（Bi-LSTM）. By combining the above-mentioned models with a clear temporal relationship， a traffic vehicle intention recognition method driven by data and mechanism hybrid is proposed， and the NGSIM data set is used for model training and verification. The results show that the recognition accuracy of the proposed method is over 90% at 1.8 s before the traffic vehicle reaching the lane changing point， and the accuracy is 97.88% at the lane changing point， which proves high recognition accuracy and long advance recognition time. Figures and Tables | References | Related Articles | Metrics

Select

Vision and Single RSU Assisted Vehicle Positioning Method Shuxuan Sheng,Chongbo Jing,Chaoyang Jiang 2022, 44 (7):  1009-1017.  doi: 10.19562/j.chinasae.qcgc.2022.07.007 Abstract ( 200 )   HTML ( 13 )   PDF (3694KB) ( 249 )   Save To accurately acquire the vehicle location in the GNSS restricted environment such as urban canyon， a visual vehicle locating method is proposed with single road side unit （RSU） assisted. A camera is used to measure the lateral distance from the vehicle to the lane lines， and the RSU and vehicle are interacted for the distance measurement and communications. The information of GNSS， IMU， RSU and camera observed are fused through the error state Kalman filtering algorithm to fulfill the accurate estimation of the location and orientation of vehicle. Real vehicle tests are conducted to analyze the effects of the ranging information of single RSU and the lateral distance measured on positioning results. The results show that the ranging information of single RSU can effectively reduce the longitudinal positioning error， but its correction effects of lateral positioning error gradually reduce with the increase of the distance from RSU. This defect can be effectively remedied by lateral distance measurement， and complementing both advantages results in a root mean square error in horizontal positioning less than 10 cm， verifying the effectiveness of the method proposed. Figures and Tables | References | Related Articles | Metrics

Select

Intelligent Vehicle Positioning by Fusing LiDAR and Double-layer Map Model Zewu Deng,Zhaozheng Hu,Zhe Zhou, LiuYulin,Chao Peng 2022, 44 (7):  1018-1026.  doi: 10.19562/j.chinasae.qcgc.2022.07.008 Abstract ( 168 )   HTML ( 6 )   PDF (2227KB) ( 156 )   Save In order to enhance the positioning accuracy of intelligent vehicles， a method fusing LiDAR and double-layer map model is proposed， in which the double-layer map model is created by adding laser point-cloud-based sparse feature map on the top of lane map， and the sparse feature map consists of the position and azimuth of vehicles， 2D intensity features and 3D features. The sparse feature map can provide an accurate position reference for intelligent vehicle positioning， effectively reducing accumulative positioning error. In addition， the lane lines are extracted from the LiDAR intensity data to provide highly accurate and linear lateral position constraints. During positioning， a Kalman filter framework is introduced to fulfill the effective fusion of LiDAR and double-layer map， in which the process of state prediction utilizes the motion constraints of vehicle to construct the short-time and constant-speed movement model and to observe the variables including the results of laser odometer positioning， the lateral position constraints based on lane map layer and the positioning based on sparse feature map layer. Tests and measurements are conducted on both campus and urban road environment to verify the effectiveness of the proposed algorithm. The results show that the fusion positioning algorithm proposed can reduce the positioning error by 40%~60% under different environments， with a relative positioning error less than 0.3%. Figures and Tables | References | Related Articles | Metrics

Select

Control Method of Path Following for Automatic Parking System Shukai Zhang,Hui Chen,Meicen Liu 2022, 44 (7):  1027-1039.  doi: 10.19562/j.chinasae.qcgc.2022.07.009 Abstract ( 243 )   HTML ( 10 )   PDF (4640KB) ( 197 )   Save In this paper， the parking path following method based on the differential flatness theory decouples the error between the vehicle and the path reference point horizontally and longitudinally， and uses the controller corresponding to the horizontal and longitudinal error to control， so that the vehicle can follow the reference point moving with the path more accurately and finally realize more accurate following of the whole parking path. For longitudinal control， integration of effective distance is proposed and the longitudinal error correction rules are set， so as to ensure that the error between the reference point of the vehicle is the transverse error that can be controlled laterally to the largest extent. For lateral control， linear quadratic regulator （LQR） is used. And the relationship between curvature changing rate and decay factor α， which can accelerate the damping of lateral error， is established to make the vehicle follow the reference point accurately when the curvature of the path changes or not. Simulations and real-vehicle tests show that the improved transverse and longitudinal path following control strategy can make the whole process follow the reference point more accurately and realize more accurate path following. Figures and Tables | References | Related Articles | Metrics

Select

Research on Parallel Parking Path Planning Method for Narrow Parking Space Jie Hu,Linglei Zhu,Ruinan Chen,Xinkai Zhong,Wencai Xu,Minchao Zhang 2022, 44 (7):  1040-1048.  doi: 10.19562/j.chinasae.qcgc.2022.07.010 Abstract ( 229 )   HTML ( 7 )   PDF (2423KB) ( 202 )   Save Aiming at the parallel parking scenes for narrow parking space， a parking planning method based on curve combination and numerical optimization is proposed in this paper. Firstly， the parking process is reversed and divided into two stages： adjustment planning and berth entering planning. The adjustment planning is devised by establishing a constraint optimization model based on clothoid-arc combination， to guide the vehicle to adjust its position and posture in the berth to find the key points of leaving， while the berth entering planning is devised by setting up a constraint optimization model based on a clothoid-arc-line combination and a quintic polynomial to guide the vehicle to find the best parking point. Then， a simulation is carried out on different working conditions with Matlab to validate the adaptability of the algorithm. Finally， real vehicle test is conducted to verify the effectiveness of the planned path. The results of simulation and real vehicle test show that the parallel parking path planning method proposed can meet the parking requirements in a narrow parking space with strong adaptability. Figures and Tables | References | Related Articles | Metrics

Select

An Overview of Drive Control Technology of Ultra-High-Speed Permanent Magnet Motors for Vehicles Cheng Lin,Yao Xu,Jilei Xing,Xingming Zhuang,Xiongwei Jiang 2022, 44 (7):  1049-1058.  doi: 10.19562/j.chinasae.qcgc.2022.07.011 Abstract ( 339 )   HTML ( 30 )   PDF (2640KB) ( 195 )   Save Ultra-high-speed permanent magnet motor （UhsPM） has the advantages of small size， high efficiency and high power density， and is widely used in automotive application fields such as fuel cell air compressors and electrically assisted turbochargers. The features of small inductance and high fundamental frequency make its drive control more difficult than that of normal-speed permanent magnet motor. In this paper， the research status quo of automotive UhsPM drive control technology in terms of circuit topology， voltage modulation strategy matching and position-sensor-less control is elaborated， the research highlights of various technologies are summarized， and the evaluations by advantage and disadvantage comparison are given. Finally， the future development trends are forecasted. Figures and Tables | References | Related Articles | Metrics

Select

Research on Driving Force Distribution Control Method of Distributed Electric Vehicles Xiaoyan Peng,Xingfei Xing,Qingjia Cui,Jing Huang 2022, 44 (7):  1059-1068.  doi: 10.19562/j.chinasae.qcgc.2022.07.012 Abstract ( 293 )   HTML ( 18 )   PDF (2153KB) ( 276 )   Save Aiming at the handling and stability problems of distributed electric vehicles under normal and faulty driving motor conditions， a driving force distribution control method combining front wheel steering and driving force reconstruction is proposed. Firstly， a sliding mode weighted controller is designed based on yaw rate and the side slip angle of mass center to calculate the additional yaw moment required； and the optimal distribution model of driving force for the normal and faulty conditions of drive motor is established respectively， in which， the yaw moment is compensated by coordinating the steering of front wheels for limiting the output capacity of drive motor in faulty conditions. Then the optimal driving force distribution value is solved out based on quadratic programming theory. Finally， a Carsim/Simulink joint simulation is conducted to verify the effectiveness of the coordinated control method proposed. The results show that this method can make full use of the redundant characteristics of distributed drives to ensure the distributed electric vehicles can meet the requirements of handling stability under normal and faulty driving motor conditions. Figures and Tables | References | Related Articles | Metrics

Select

Analysis on Influence of Fiber Pore Characteristics on Liquid Water Transmission of Gas Diffusion Layer in PEMFC Qingshan Liu,Fengchong Lan,Jiqing Chen,Junfeng Wang,Changjing Zeng 2022, 44 (7):  1069-1080.  doi: 10.19562/j.chinasae.qcgc.2022.07.013 Abstract ( 144 )   HTML ( 5 )   PDF (4849KB) ( 141 )   Save The characteristics of the fiber pores inside the gas diffusion layer （GDL） of proton exchange membrane fuel cell （PEMFC） have significant effects on the liquid water transmission capability， and even the performance and service life of the fuel cell. In this paper， a gas-liquid two-phase model， which can capture the moving behavior of liquid water through the fiber pores of GDL， is established based on the fluid volume method， and the effects of the pore fraction， the cross section shape， the ratio of longitudinal spacing over lateral spacing of fibers and contact angle on the liquid water transmission capability inside GDL are analyzed. The results show that the ratio of longitudinal spacing over lateral spacing of fiber decides the moving direction of liquid water through its effects on the longitudinal and lateral capillary force， and the increase of contact angle along longitudinal direction can accelerate the climbing of liquid water， hence significantly enhancing the water drainage capacity of GDL. Figures and Tables | References | Related Articles | Metrics

Select

An AUKF-Based SOC Estimation Method for Lithium-ion Battery Ping Wang,Qingrui Gong,Ze Cheng,Ji’ang Zhang 2022, 44 (7):  1080-1088.  doi: 10.19562/j.chinasae.qcgc.2022.07.014 Abstract ( 278 )   HTML ( 13 )   PDF (2304KB) ( 146 )   Save A state of charge （SOC） estimation method of lithium-ion battery based on adaptive unscented Kalman filter （AUKF） is proposed in this paper. Firstly， the second-order RC equivalent circuit model of battery is established with its parameters identified. Then， aiming at the deficiency of unscented Kalman filter （UKF） algorithm， the convergence criterion for general filter is introduced， and the UKF algorithm is improved by the adaptive adjustment of measurement noise and process noise and the correction of Kalman gain， forming an AUKF-based SOC estimation method. Finally， verifications are performed with test data and public battery dataset， and the results show that the method proposed has fast convergence speed and high estimation accuracy. Figures and Tables | References | Related Articles | Metrics

Select

Effect of EGR Driven by VNT on Combustion and Emission of Light-Duty Diesel Engine at Different Altitudes Jun Wang,Lizhong Shen,Yuhua Bi,Jilin Lei 2022, 44 (7):  1088-1097.  doi: 10.19562/j.chinasae.qcgc.2022.07.015 Abstract ( 160 )   HTML ( 3 )   PDF (2898KB) ( 89 )   Save Variable nozzle turbocharging （VNT） technology and exhaust gas recirculation （EGR） are ideal technical measures to reduce in-cylinder emission for light-duty diesel engine. The implementation of China VI light-duty vehicle emission regulations has drawn more attention to the pollutant emission of light-duty vehicles in plateau areas. Therefore， the effect of EGR driven by VNT on combustion and emission performance of a light-duty high-speed direct injection diesel engine is experimentally studied at altitudes of 0， 1 000， and 1 960 m by using a plateau environment simulation device. The results show that when EGR is driven by VNT， the variation range of EGR rate is almost the same within the same VNT nozzle opening range at different altitudes. As EGR rate increases， maximum mean temperature decreases at different altitudes.At the maximum torque condition， the peak cylinder pressure decreases， and the start of combustion is delayed， and combustion duration is prolonged， however， at the rated power condition， the peak cylinder pressure increases， the start of combustion is advanced， and combustion duration is shortened. EGR can still effectively reduce NO x emissions at different altitudes， and the decreasing amplitude of brake specific NO x emission increases as altitude increases. When driving EGR based on VNT at different altitudes， with the increase of EGR rate， brake specific CO emission and smoke both increase at the maximum torque condition， while brake specific CO emission and smoke both decrease accordingly at the rated power condition. Furthermore， EGR has a more significant effect on brake specific CO emission and smoke in plateau areas. Figures and Tables | References | Related Articles | Metrics

Select

Trajectory Tracking Control Algorithm of Emergency Collision Avoidance for Tractor Semi-trailer Combination Daofei Li,Anfei Zha,Biao Xu,Jiajie Zhang 2022, 44 (7):  1098-1106.  doi: 10.19562/j.chinasae.qcgc.2022.07.016 Abstract ( 254 )   HTML ( 9 )   PDF (3619KB) ( 256 )   Save To enhance the emergency collision avoidance performance of commercial vehicles， the way of steering / braking joint collision avoidance is adopted. In view of that tractor semi-trailer combination is prone to instability in emergency steering， a nonlinear model predictive controller with consideration of rollover and yaw instability prevention is designed and validated by simulations under different load and speed conditions in emergency collision avoidance scene. Accounting for the inadequate real time performance of the controller， with which as the datum of tracking performance a linear time-varying model predictive control algorithm is designed for improving its real time performance in a condition of meeting the requirements on the tracking accuracy of collision avoidance trajectory. Finally， a small scaled test vehicle of tractor semi-trailer combination is produced to carry out trajectory tracking verification in emergency collision avoidance condition. The results show that the control algorithm designed can accurately follow the target trajectory， meeting the requirements of practical application. Figures and Tables | References | Related Articles | Metrics

Select

Failure Load Analysis and Thickness Matching Optimization of Weld Seam on Front Axle of a Light Truck Jinzhi Feng,Chenglin Yuan,Jiawei Yu,Xinrong Liu,Lihui Zhao 2022, 44 (7):  1107-1116.  doi: 10.19562/j.chinasae.qcgc.2022.07.017 Abstract ( 158 )   HTML ( 10 )   PDF (3726KB) ( 125 )   Save In view of the weld seam failure in the front axle of a light truck， the load spectra of vehicle durability test are collected in proving ground， and the virtual iteration method is used to get the load time history of each connecting point of front axle. The finite element model of front axle is built and a fatigue life simulation on front axle’s weld seam is carried out using inertia relief method. The analysis results are consistent with the failure features of road test in proving ground， verifying the accuracy of the model. Furthermore， the local stress state at the weld seam is obtained by means of virtual strain gauge， the failure dominant load of the weld seam under the random load of proving ground is determined by studying the damage direction distribution， uniaxial damage comparison and principal stress direction distribution， and the effects of the weld seam thickness at three localities in failed zone on fatigue life are analyzed. Finally， a matching optimization is conducted on the three thickness parameters by applying adaptive response surface technique and the results show that the fatigue life of the weld seam zone is increased by nearly 8 times， and the road test in proving ground also meets the durability requirement of 10 000 km driving without failure. Figures and Tables | References | Related Articles | Metrics