Loading...
Welcome to Journal of Automotive Safety and Energy,
汽车安全与节能学报

Current Issue

  • 2019, Vol. 10 No. 4 Published on:31 December 2019 Previous issue    Next issue
    Progress & Prospects
    Key techniques of semantic analysis of driving behavior in decision making of autonomous vehicles
    LI Guofa，CHEN Yaoyu，Lü Chen，TAO Da，CAO Dongpu，CHENG Bo
    2019, 10(4):  391-412.  doi:10.3969/j.issn.1674-8484.2019.04.001
    Abstract ( 534 )   PDF (1772KB) ( 1670 )  
    It is one of the key technologies in intelligent vehicle development to precisely decode the hidden semantic characteristics from driving signals along time sequences. This paper systematically proposed a framework of semantic analysis in driving behavior research for decision making in autonomous driving. The 4 stages of semantic analysis were summarized including car-following and lane change model, driving maneuver offline recognition, driver intention online prediction, and unsupervised sematic analysis. The state-of-the-art development of the 4 key techniques in semantic analysis was introduced in detail at home and abroad. The key techniques include maneuver recognition and prediction, smart decision making, driving style estimation, and behavior segmentation. The learning algorithms in semantic analysis for autonomous driving development were introduced in detail, including Bayesian network, hidden Markov model, deep neural network, and reinforcement learning. Finally, the applications of semantic analysis in driver assistance, intelligent decision making, and path planning were discussed for better development of autonomous driving technologies.
    Related Articles | Metrics
    Automotive Safety
    Trajectory prediction of dynamic vehicles in structured roads
    XIE Hui, GAO Bin, XIONG Shuo, WANG Yue
    2019, 10(4):  413-422.  doi:10.3969/j.issn.1674-8484.2019.04.002
    Abstract ( 213 )   PDF (1947KB) ( 1295 )  
    A trajectory prediction strategy based on traffic scene feature identification was proposed to improve the accuracy of dynamic vehicle trajectory prediction around autonomous cars in structured roads. The surrounding vehicles were detected, tracked and localized by lidars and integrated navigation system. The driving intention estimation model was used to judge the vehicle driving behavior and classify the traffic scenes. The vehicle motion trajectory was predicted based on vehicle motion model and Markov decision process for lane keeping and lane changing scenes, respectively. The results showed that the average prediction error based on the vehicle motion model is about 0.12 m for the lane keeping scene, when the prediction time domain is 4 s for the lane change scene, the average prediction error based on the Markov decision process is about 0.17 m when the prediction time domain is 8 s. Therefore, the strategy can accurately predict the trajectory of dynamic vehicles in real time.
    Related Articles | Metrics
    Differential steering control and its applicability of distributed electric vehicle
    PEI Xiaofei, LIU Zhiting, CHEN Zhenfu, CHEN Keji, YANG Bo,
    2019, 10(4):  423-432.  doi:10.3969/j.issn.1674-8484.2019.04.003
    Abstract ( 240 )   PDF (5585KB) ( 340 )  
    A hierarchical controller of differential steering was proposed to improve the steering redundancy and maneuverability. The desired wheel-speeds were determined by the proportion integration differentiation (PID) control and the look-up table model. The required torque for differential steering were calculated by the adaptive discrete sliding mode control. The boundary condition of differential steering was defined by simulation experiments. The influence of vehicle velocity and adhesion coefficient on the boundary condition was analyzed, and the steering performance of vehicles was compared under differential assisted steering control and pure steering wheel control. The results show that the proposed differential steering control strategy can be integrated with the traditional steering system to assist steering, which can effectively reduce the turning radius and improve the steering performance at medium and low vehicle velocity, and the effectiveness and real-time of the differential steering control strategy is also verified. 
    Related Articles | Metrics
    Differences of the motion sickness associated brain activity regions based on the driving simulator and fNIRS
    ZHANG Chenyang, LI Shuguang, LI Yaohua, YAO Jin, ZHU Lei, WANG Min
    2019, 10(4):  433-442.  doi:10.3969/j.issn.1674-8484.2019.04.004
    Abstract ( 216 )   PDF (2200KB) ( 351 )  
    The pathogenesis of motion sickness which appears on automated vehicles and electric vehicles needs to be explored to develop autonomous driving. The driving data, including driving time and distance, and cerebral oxygen exchange from 52 participants under the straight driving were collected and their differences were analyzed, based on the six-degree-of freedom driving simulator platform and noninvasive Functional NearInfrared Spectroscopy (fNIRS). The results show that the BA6, BA17 and BA18 areas of brain have the highest degree of activity, followed by the BA1, BA2, BA3 and BA4 areas, while the BA8, BA10 and BA40 areas are in the decreasing order. Brain Brodmann areas differ which are associated with motion sickness under the straight driving. These results will be helpful to have a better understood of the motion sickness from the perspective of the brain.
    Related Articles | Metrics
    Tensile and wear resistance of carbon fiber/carbon nanotube composite tread adhesive
    MA Yunhai, CHEN chuangfa, DUANMU Lingjian, WEI Yintao
    2019, 10(4):  443-450.  doi:10.3969/j.issn.1674-8484.2019.04.005
    Abstract ( 241 )   PDF (1025KB) ( 309 )  
    The effect of the amount of carbon fiber and carbon nanotubes treated with silane coupling agent and dopamine on the wear resistance and tensile strength of the compound tread adhesive was studied to improve the wear resistance and mechanical properties of the tire tread to extend its servise life. The experimental scheme was designed and arranged by the method of quadratic regression orthogonal design. Mechanical properties and tribological properties of rubber materials were measured by universal testing machine and DIN wear meter. An improved multi-objective optimization algorithm based on NSGA-II was used to optimize the test indexes. The response surface was drawn and the best combination of parameters was obtained by the Pareto optimal solution set of the original model. The results show that the abrasion loss is 139.7 mm3, and the fracture strength is 20.8 MPa by 82.66 phr （parts per hundred） for natural rubber, 7.43 phr for carbon nanotubes and 6 phr for carbon fiber, showing that the carbon fiber/carbon nanotube composite tread adhesive has good wear resistance and tensile properties.
    Related Articles | Metrics
    Vehicle object detection method based on data fusion of LADAR points and image
    HU Yuanzhi，LIU Junsheng，HE Jia，XIAO Hang，SONG Jia
    2019, 10(4):  451-458.  doi:10.3969/j.issn.1674-8484.2019.04.006
    Abstract ( 384 )   PDF (1230KB) ( 749 )  
     A fusion scheme with 4 lines LADAR (laser detection and ranging) sensor and camera was adopted to provide more precise detection for traffic, for an intelligent vehicle. Firstly, by using deep learning technique to detect image. Then, mapping LADAR data to image through a space transfer matrix. Finally, by using an R-Tree algorithm to quickly match LADAR points and corresponding detection boxes. The traffic’s real location was calculated easily by laser’s ranging. The proposed fusion frame was tested by images and point cloud data collected from real motorway scenes. The results show that the false negative (FN) of the fusion frame method is 8.03%, which is lower than that of 14.86% come from the Mask R-CNN method. Therefore, the fusion data could decrease probability of the FN compare with single data.
    Related Articles | Metrics
    Detection and tracking method of lane line for intelligent vehicles under complex illumination condition
    JIN Zhilin, HE Linxuan, ZHAO Wanzhong
    2019, 10(4):  459-466.  doi:10.3969/j.issn.1674-8484.2019.04.007
    Abstract ( 257 )   PDF (2155KB) ( 338 )  
    An improved lane line detection and tracking method was proposed for the driving safety of intelligent vehicles under complex illuminations. Gray-level transformation was used for image preprocessing to increase the contrast between lane lines and roads. Lane lines in binary images were extracted by using a progressive probabilistic Hough transform, and were fitted by using a least square method. Dynamic region of interest was established based on Kalman-filter according to the results of the previous image to track the lane lines. The lane line detection and tracking experiments were carried out under different illumination conditions, such as night light, weak light, normal light and strong light. The results show that this method has an accuracy of 97.53% and a real-time performance of 60 ms in the detection and tracking of lane lines, and has robustness against the interference of road shadow, traffic signs, vehicles and other factors.
    Related Articles | Metrics
    Improvement of collision safety design of the protection devices for a vehicle cylindrical container
    DONG Yuzhe, WANG Bo, ZHANG Jinhuan
    2019, 10(4):  467-473.  doi:10.3969/j.issn.1674-8484.2019.04.008
    Abstract ( 255 )   PDF (1006KB) ( 272 )  
    The motion of a truck-mounted cylindrical-box and the failure of protection devices under frontal impact condition were investigated by means of finite element simulation. The protection devices were improved with changing the fixed position of the belt and the structure of the retractor. The results show that the previous protection device has a certain restraint effect on the cylindrical container, which suffers less mechanical impact in the frontal collision process at the initial velocity of 50 km/h and below. The previous constraint system has a poor constraint effect, and the cylindrical box has a secondary collision in the frontal collision condition at the initial speed of 70 km/h. The improved constraint system can ensure that no secondary collision occurs in the frontal collision condition of the cylindrical container at 70 km/h. The peak acceleration of the cylindrical box drops to 25g from 60g. Therefore, the safety performance of the improved constraint system is improved.
    Related Articles | Metrics
    Multilevel hierarchical control of anti-rollover for heavy-duty tractor-semitrailer
    ZHAO Shu’en，ZHANG Xiong，TANG Juntao
    2019, 10(4):  474-482.  doi:10.3969/j.issn.1674-8484.2019.04.009
    Abstract ( 181 )   PDF (3018KB) ( 266 )  
     A 5 degree-of-freedom dynamic model was established for tractor and trailer lateral with yaw and roll motion for the rollover stability of heavy-duty semi-trailer trains under extreme conditions. A model predictive control method was used to obtain the yaw moments of tractor and trailer with an anti-rollover target controls. Two subsystems of active steering and direct yaw moment control were coordinated and controlled according to the vehicle real-time yaw rate and the tire load transfer rate. The joint simulation analysis was done by using TruckSim/Simulink pair for a layered hierarchical control strategy of semi-trailer. The results show that the maximum lateral accelerations are improved with the rates of 7.76% and 12.29% for the tractors and the trailers respectively under same conditions, the maximum absolutes of yaw angula-rate are improved with the rates of 23.88% and 3.39% respectively, and the maximum articulation angle of the car train is reduced by 19.87%. Therefore, the layered hierarchical control improves the anti-rollover stability of the vehicles due to active steering and direct yaw moment control compared with the traditional heavy-duty semi-trailer train.
    Related Articles | Metrics
    Automotive Energy Efficiency & Environment Protection
    Impedance Characteristics of Cell-to-Cell Inconsistency of Polymer Electrolyte Membrane Fuel Cells
    SUN Weihua, XU Liangfei, HU Zunyan, JIANG Hongliang, LI Jianqiu, OUYANG Minggao
    2019, 10(4):  483-491.  doi:10.3969/j.issn.1674-8484.2019.04.010
    Abstract ( 229 )   PDF (2329KB) ( 277 )  
     A parameter extraction and consistency analysis of a polymer electrolyte membrane fuel cell (PEMFC) was proposed based on electrochemical impedance spectroscopy (EIS) to clarify the consistency of the PEMFC stacks under different conditions.  An EIS model, which is derived from the reaction mechanism in the cathode catalyst layer (CCL), was introduced followed by an introduction to the parameter extraction workflow based on the genetic algorithm (GA) and the nonlinear least squares (NLS) algorithm. Five parameters (i.e., the Tafel slope, CCL proton conductivity, double layer capacity, and the oxygen diffusion coefficients in the gas diffusion layer and CCL) were identified using this workflow. Experiments under different dew point temperatures of the membrane humidifier on a short fuel cell stack with 50 single cells were conducted. The impedances of 6 cells in the stack were measured, and the parameters of the 6 cells were extracted. The inconsistency within the stack was then evaluated. The results show that the parameter extraction method based on the GA and NLS is effective for identifying the parameters of the fuel cell; cell performance inconsistencies can be clearly identified using EIS parameters and normalized parameters; and the consistency of the stack improves when the provided gas is moister under low current density.
    Related Articles | Metrics
    Active disturbance rejection control of e-Turbo diesel engine intake oxygen concentration
    CHEN Tao, LI Shaohua, SONG Kang, XIE Hui
    2019, 10(4):  492-501.  doi:10.3969/j.issn.1674-8484.2019.04.011
    Abstract ( 158 )   PDF (2010KB) ( 420 )  
    An active disturbance rejection control algorithm was designed to have a better control ability with faster response speed to intake oxygen concentration of electrically- assisted-turbocharger (e-Turbo) diesel engine. An observer of intake and exhaust oxygen concentration for control was established being used as a virtual oxygen sensor. The control algorithm was tested and evaluated on a precisely calibrated GT-Power diesel engine model. The results show that the average response time decreases by 0.86 s comparing with traditional proportion-integral-differential (PID) control during the changes of the target intake oxygen concentration in the simulation. The maximum relative error to the target is 2.33% during the transient process at engine tests. Therefore, the active disturbance rejection control algorithm achieves a rapid control of intake oxygen concentration.
    Related Articles | Metrics
    Distribution and characteristics of degradation of lithium ion batteries cycled at low temperature
    LIU Zhongxiao, GE Hao, LI Zhe, ZHANG Yakun, ZHANG Jianbo
    2019, 10(4):  502-510.  doi:10.3969/j.issn.1674-8484.2019.04.012
    Abstract ( 204 )   PDF (908KB) ( 430 )  
    Orthogonal experiments were utilized to explore the characteristics and distribution of degradation of lithium ion batteries cycled at low temperature. The main aging factors and internal lithium plating distribution were investigated with the statistical method and post-mortem physical-chemical analysis, respectively. The results showed that the order of the 3 important aging factors are the ambient temperature > charge current > charge cut-off voltage. The amount of lithium plated in center electrodes is greater than that in the outer electrodes. The amount of lithium plated in central area and the area near the negative tab is greater than that on other areas in the same electrode layer. The main reason for such lithium plating distribution is attributed to the non-uniformity in the current density in the battery, as the voltage at areas with larger current density reaches the lithium plating potential easier. Therefore, to alleviate degradation from lithium plating, an effective measure is to design battery with more uniform current density distribution.
    Related Articles | Metrics
    Preparation, processing and properties of hemp stalk powder/ polylactic acid composites for automobiles
    WANG Chunhong, WANG Lijian, ZUO Hengfeng, WANG Ni, Noor Intan Saffinaz Anuar, Peyman Servati, Frank Ko
    2019, 10(4):  511-517.  doi:10.3969/j.issn.1674-8484.2019.04.013
    Abstract ( 213 )   PDF (1025KB) ( 602 )  
    A fully degradable wood-plastic composite material was prepared for automotive interiors to reduce the mass of automobiles and protect the environment. The composite was prepared from hemp stalk powder and degradable polylactic acid (PLA) by using a twin-screw extruder for pelletization and compression molding. An optimal process for press molding was obtained: 160 ℃, 7.5 MPa, 5 min for the three influencing factors of temperature, pressure, and time, after orthogonal test analysis of 4 mechanical performance indicators. The results show that the composite has its highest mechanical properties in tensile strength, bending strength, tensile modulus and bending modulus, which are 40.26, 73.84 MPa and 2.06, 4.35 GPa for the hemp stalk powder with a mass fraction of 30% and a diameter of 40 mesh.
    Related Articles | Metrics
    Control algorithm with differential prior PID for variable valve timing of engines
    CHEN Zhijun， YUAN Qiang， ZHAO Yuhan
    2019, 10(4):  518-524.  doi:10.3969/j.issn.1674-8484.2019.04.014
    Abstract ( 247 )   PDF (1077KB) ( 386 )  
    This paper proposed a differential prior proportional differential derivative (PID) control algorithm to optimize the original control of the variable valve timing (VVT) valve of the engine. Moreover, analyzed the working principle and mechanical structure of the VVT and determined key parameters of the VVT before established a mathematical model and Simulink simulation model, by means of a Matlab/Simulink platform. Using a real machine platform and the simulation platform to verify the proposed control algorithm. The results show that the overshoot of the differential prior PID algorithm proposed in this paper is reduced by 0.8% with the static error reduced by 0.15%, in the step test; and the overshoot is reduced by 1% with the static error is reduced by 0.15% in the tracking test in the VVT valve control, in the face of the fluctuation of the target expectation value, compared with the traditional PID algorithm, Therefore, the algorithm completes the optimal control of the VVT valve.
    Related Articles | Metrics
    Battery equalization strategy based on K nearest neighbor genetic algorithm
    LI Jun，HUANG Zhixiang，TANG Shuang
    2019, 10(4):  525-530.  doi:10.3969/j.issn.1674-8484.2019.04.015
    Abstract ( 217 )   PDF (815KB) ( 245 )  
     The inductor was used as the energy storage component and the equalization circuit was built based on the principle of step-up and step-down chopping to reduce the current equilibrium time and the balance energy loss. The K-nearest neighbor genetic algorithm was employed to realize the equalization control of the battery pack and it was compared with the mean equalization strategy. Simulations were performed in MATLAB/Simulink of self-equalization, charging and discharging in 3 balanced conditions. The results show that the equilibrium speed under the 3 conditions is improved by 26.70%, 22.70% and 24.80%, respectively. The reduction of switching operations can reduce the energy loss of the battery pack, and the energy conversion rate is increased by 5.06%. Therefore, the proposed equilibrium strategy has a good balance effect, which can effectively improve the inconsistency among battery packs and reduce the equilibrium time and the balance energg loss.
    Related Articles | Metrics
    Cooperated control of signal and vehicle trajectory under the autonomous vehicle environment
    DAI Rongjian，DING Chuan，LU Yingrong，ZHAO Fuquan
    2019, 10(4):  531-539.  doi:10.3969/j.issn.1674-8484.2019.04.016
    Abstract ( 318 )   PDF (1358KB) ( 368 )  
     A cooperated control method of signal and vehicle trajectory was proposed for isolate signal intersection based on the technologies of cooperative vehicle and infrastructure system (CVIS) and autonomous vehicle (AV) to improve the traffic efficiency and reduce the stop times of vehicles at the signal intersection under the autonomous vehicle environment. The method was simulated using the multi-agent technology. The signals in this control method were optimized according to the information of AVs, such as position, speed, etc. The optimized signal-timing plan was used to control trajectories of AVs. A multi-agent simulation environment was developed by NETLOGO, by which simulation experiments were conducted. The results indicate that this control method reduces about 53.4% travel time and 61.5% stop times of vehicles compared with traditional signal control strategy. Therefore, the method improves traffic efficiency and reducts stop-times.
    Related Articles | Metrics