LiDAR point cloud semantic segmentation algorithm is crucial to the environmental understanding of unmanned vehicles. At this stage, in autonomous vehicles, effectively integrating the complementary information of LiDAR and camera has become the focus of research. In this work, a network framework (called PI-Seg) for LiDAR point clouds semantic segmentation by fusing appearance features of RGB images is proposed. In this paper, the perspective projection module is introduced to align and synchronize point clouds with images to reduce appearance information loss. And an efficient and concise dual-flow feature extraction network is designed, a fusion module based on a continuous convolution structure is used for feature fusion, which effectively reduces the amount of parameters and runtime performance, and more suitable for autonomous driving scenarios. Finally, the fused features are added to the LiDAR point cloud features as the final output features, and the point cloud category label prediction is realized through the MLP network. The experimental results demonstrate that PI-Seg has a 5.3% higher mIoU score than SalsaNext, which is also a projection-based method, and still has a 1.4% performance improvement compared with the latest Cylinder3D algorithm, and in quantitative analyses the mAP value also has the best performance, showing that PI-Seg is better than other existing methods.

for wireless power transfer system (WPT), the delay of communication between primary side and secondary side influences the robustness of system. This paper proposes a bilateral control strategy based on the LCL-S compensation network wireless charging system without communication. The proposed bilateral control strategy adopts disturbance control to optimize efficiency on primary side, and proportional-integral (PI) control to adjust the charging current/voltage on secondary side, which realizes the separation of control between the primary and secondary side. The feasibility of proposed bilateral control strategy is verified through simulation and experiment. Based on the realization of CC mode and CV mode, the efficiency of the system is improved remarkably.

To evaluate the Influence of polyoxymethylene dimethyl ethers (PODE) on particulate matter (PM) chemical features, PM samples emitted from diesel fuel and PODE/diesel blends at volume ratios of 10 %, 20 %, and 30 % (P10, P20, and P30) were characterized using gas chromatography-mass spectrometry, X-ray photoelectron spectroscopy (XPS), and Fourier transform infrared spectroscopic (FT-IR). Results showed that adding PODE in diesel fuel could increase the proportions of the lower carbon-atom-number components and the contents of oxygen-containing compounds in the soluble organic fractions extracted from PM samples. The ratio of oxygen to carbon (O/C), the functional groups, and the nanostructure of dry soot obtained from XPS showed that the O/C rose as the PODE volume ratio increased. The graphitization degree of dry soot decreased in the order: diesel fuel > P10 > P30 > P20. The relative content of hydroxyl functional groups exhibited the same trend, while the relative content of carbonyl functional groups exhibited an opposite trend with the graphitization degree. Moreover, according to FT-IR, both the branching degree and the relative content of hydrocarbon functional groups of aliphatics are influenced by the graphitization degree of dry soot. A turning point at P20 observed by analysis results above indicated that the chemical characteristics of PM could be affected not only by fuel properties but also by the process of fuel combustion and PM formation.

In congested street or highway scenarios, such as lane-change in dynamically changing traffic flow situations, the planning and tracking of trajectories for connected and automated vehicles (CAVs) represent some of the most challenging tasks. As an introduction to automated lane-change in CAVs, this paper presents a control Lyapunov function (CLF) approach that follows an optimal desired trajectory while observing the constraints imposed by the control barrier function (CBF) in order to avoid collision with surrounding vehicles. By combining CLF and CBF within the framework of quadratic programs (QP), it allows the implementation to simultaneously track the control objective (represented by the CLF) and satisfy the constraints of the desired state of the system (represented by the CBF). Therefore, it provides the possibility of tracking simultaneously the path of the unmanned vehicle, within the constraints of the surrounding vehicles as well as the surrounding environment. From simulations and comparison results, the controller presented here can perform collision avoidance well and can be used on a real traffic system, which has the advantage of providing faster and more precise lane-change results than another work.

This study investigates the fuel atomization effect of alternating magnetic field (AMF)-based preprocessing equipment that promotes the atomization of automobile fuel. Two coils are connected to the output terminal of the preprocessing equipment and wound on an STS 316 L fuel pipe in the form of a solenoid. Positive sine waves and negative sine waves are output to the two coils, and the AMF in the solenoids causes atomization of the fuel flowing inside the pipe. A spray visualization experiment was conducted with fine particles sprayed from the injector, and the spray arrival distance, spray angle, and spray area were recorded. The information was analyzed to digitize the spray penetration, angle, and area at 2 ms after firing the injector. The fuel atomization effect of the AMF-based preprocessing equipment was also investigated Gasoline was used as the fuel, and the fuel atomization effect was analyzed using the number of windings around the STS 316 L fuel pipe, the frequency applied to the coil, the distance between the two solenoid coils, and the fuel injection pressure as variables.

The channel structure of a proton-exchange membrane fuel cell (PEMFC) is very important for sustaining high performance with a reliable lifetime. There have been numerous computational studies to investigate the physics of electrochemical reactions with various flow field phenomena. Because straight channels reduce manufacturing costs, a straight channel was selected as the subject of a numerical study to investigate water removal and oxygen starvation in the channel. In this study, the liquid saturation distribution caused by the baffled obstacle geometry of a PEMFC with channels that were straight and parallel was analyzed in a computational study. Because baffled obstacles generate a vortex in the channel, the flow field structure improves the supply of reactants. When the operating pressure is reduced from 3 atm to 1 atm, the baffled obstacle structure interrupts the flow of reactant and product water so that the accumulation of liquid water increases from 0.75 to 0.86 in catalyst layer. As the number of baffled obstacles was increased from 13 to 19, the current density improved from 0.772 to 0.773 A/cm2 in the reference condition. The baffled obstacle geometry also affects the liquid water accumulation flow field. Results show that the design of the baffled obstacles in a channel should consider the number of baffled obstacles along with the baffled obstacle geometry such that water removal is accelerated.

Controller area network (CAN) bandwidths are currently limited when accommodating increased traffic volume due to the increase in vehicle electronic control units (ECUs). Research is underway to develop a lossless CAN data compression algorithm that uses existing bandwidth networks to reduce the size of transmitted data while minimizing cost and reliability issues without hardware changes. In this study, we propose the optimization of quotient remainder compression (QRC) parameters to improve the compression ratio. The relationships between the divisor, quotient remainder (QR) parameters, and the amount of change in the data were analyzed, and various divisor and QR parameter values were applied to determine the optimal parameter values that provide the best compression results. Experimental results obtained using actual vehicle driving data showed that the compression rate was improved through QRC parameter optimization compared to using the existing QRC system. The proposed approach can be applied to CAN, CAN-FD, FlexRay, and Ethernet systems to reduce the bus load rate and improve the network bandwidth availability.

This paper presents the model predictive control (MPC) based multifunctional advanced driver-assistance system (MADAS) that is optimized for rear-end collision avoidance. First, the system’s operation is judged by considering the driver’s intention of avoidance and the possibility of avoiding obstacle vehicles. Once the system is activated, the lateral tire force corresponding to the driver’s steering input, which is essential for collision avoidance, is realized with the highest priority. The use of each tire friction circle is then maximized by utilizing available tire forces for braking through quadratic programming. While the MADAS ensures the lateral maneuver and deceleration of the vehicle, the system still can generate additional yaw moment calculated from the MPC, the upper level controller, to track the driver’s desired yaw rate or prevent the vehicle from becoming unstable. The nonlinearity inevitably encountered in maximizing tire forces is reflected through the extended bicycle model and the combined brushed tire model. The proposed system is verified by the vehicle dynamics software CarSim, and the simulation results show that the MADAS performs better in rear-end collision avoidance situations than conventional advanced driver-assistance systems (ADAS).

The tire building process is a key part of tire manufacturing, serving as a bridge between the construction design and the finished tire, and the bladder is the core component to complete the shaping process. With the help of finite element method, the tire building process can be effectively reproduced, which help to carry out targeted problem solving and solution design. In this paper, the numerical simulation method is used to study the 205/55R16 radial tire building process, and the reliability of the simulation method is verified by comparing simulated green tire and its test section; then the shaping simulation model incorporating the bladder is established and compared with the finished tire parts to illustrate the reliability; On this basis, the influence of bladder parameters was analyzed by orthogonal design of experiment and simulated annealing optimization algorithm, the sensitivity of different parameters was obtained, the bladder parameters were optimized; compared with the bladder with original parameters, the optimized bladder effectively reduced the stress by 44.67 % and 55.54 %, while significantly improving the contact force between the bladders and the green tire. The results of the study provide a good methodological basis and theoretical guidance for tire design, manufacturing and bladder optimization.

The dual-fuel combustion strategy using compressed natural gas (CNG) and diesel exhibits immense potential for improving conventional diesel engines. To reduce CO2 emissions, diesel should be substituted with CNG to the maximum extent possible. In this study, a numerical investigation was conducted using a three-dimensional computational model for cases with 97 % CNG substitution (CNG97), which has not been explored through experiments previously owing to the technical limitations of diesel injection systems. The simulation model was validated through experiments conducted under operating conditions similar to a diesel engine; the results revealed that ignition was initiated by the diesel fuel regardless of the CNG substitution rate. In the CNG97 case, ignition occurred near the center of the cylinder owing to short spray penetration. Consequently, the CNG remained inside the piston bowl and top crevice area. In addition, the small quantity of diesel weakened the initial stage of combustion. These two effects retarded combustion significantly, which caused substantial phasing loss in CNG97. Based on these findings, it was concluded that the diesel injector should be optimized for a high CNG substitution rate to reduce CO2 emissions effectively.

The current study aimed to develop a method for estimating whole-body Injury metric values (WBIMs), which are widely used in epidemiological studies, to determine airbag deployment threshold during frontal crashes using a computational human surrogate model and real-world crash data. To this end, a finite element human body model was instrumented to predict the risk of the injuries. The whole body was divided into 22 body regions. Then, body region-specific injury pattern databases were constructed for these body regions using the NASS-CDS database. Monte Carlo Sampling was performed to calculate WBIMs, such as the probability of death and lost years of life. A series of frontal crash simulations was performed for various Delta-V with and without deploying pyrotechnic restraint systems. Lastly, the WBIM values obtained from the proposed method were compared to those obtained from the NASS-CDS. From the Delta-V of 25 km/h, the airbag-deployed conditions demonstrated a protective effect compared to the non-deployed conditions. The predicted WBIM values using the proposed method demonstrated a similar trend to that presented in the field data. The proposed method to estimate WBIM values can be used to evaluate various occupant protection systems.

The recent proliferation of intelligent technologies has promoted autonomous driving. The autonomous parking system has become a popular feature in autonomous driving. Hybrid A-star algorithm is a commonly used path planning algorithm for its simplicity to deploy and the good characteristics of the generated paths in the practical engineering. To further enhance the path safety and efficiency of path planning in the autonomous parking system, this paper proposes an improved hybrid A-star algorithm through the safety-enhanced design and the efficiency-enhanced design. The safety-enhanced design integrates the Voronoi field potential into the path searching stage to take more account of path safety. The efficiency-enhanced design proposes a multi-stage dynamic optimization strategy which divides the path planning into multiple stages and performs dynamic optimization in each stage. Through simulation experiments, it is verified that the proposed improved algorithm not only generates a much safer path which stays farther from the obstacles but also significantly improves the searching efficiency in terms of time and space, merely at a finite cost of pre-processing work which can also be repeatedly utilized. We hope this paper will promote relative research on path planning in autonomous parking and serve as a reference for the practical engineering.

Tapered composite energy-absorbing components have superior advantages in weight reduction and crashworthiness improvement subjected to oblique compressions compared to straight structures. This study investigated the crashworthiness characteristics of uni-directional carbon fiber reinforced plastic (UD-CFRP) tubes under various loading angles, and further provided the guidance on multi-objectives optimization for UD-CFRP tubes accounting for multiple loading cases. The crashworthiness characteristics of straight UD-CFRP tubes subjected to three compressive angles (0°, 10° and 20°) were firstly explored experimentally, and results indicated that energy-absorbing capacity of samples decreased with the loading angles increasing due to changes in deformation behaviors. The multi-layer finite element modes (FEMs) were developed and validated, and simulations found that internal energy (IE) of intra-CFRP layer and inter-cohesive layer, friction energy decreased with the increase in loading angles. Parametric studies indicated crashworthy performances of UD-CFRP samples under multiple loading angles can be further improved by adjusting the tapered angle or wall thickness. Consequently, the synthetic special energy absorption (SEAβ), synthetic peak crushing force (PCFβ) and mass of tapered tube were optimized accounting for three different loading groups. Compared to baseline sample, the SEAβ was improved by 14 %, while the PCFβ and mass were reduced by 30.2 % and 19 %, respectively.

This study uses deep reinforcement learning (DRL) combined with computer vision technology to investigate vehicle fuel economy. In a driving cycle with car-following and traffic light scenarios, the vehicle fuel-saving control strategy based on DRL can realize the cooperative control of the engine and continuously variable transmission. The visual processing method of the convolutional neural network is used to extract available visual information from an on-board camera, and other types of information are obtained through the vehicle’s inherent sensor. The various detected types of information are further used as the state of DRL, and the fuel-saving control strategy is built. A Carla–Simulink co-simulation model is established to evaluate the proposed strategy. An urban road driving cycle and highway road driving cycle model with visual information is built in Carla, and the vehicle power system is constructed in Simulink. Results show that the fuel-saving control strategy based on DRL and computer vision achieves improved fuel economy. In addition, in the Carla–Simulink co-simulation model, the fuel-saving control strategy based on DRL and computer vision consumes an average time of 17.55 ms to output control actions, indicating its potential for use in real-time applications.

In this study, the floating liner method is utilized to measure the friction generated in the piston assembly of a single-cylinder gasoline engine. The piston assembly is subjected to combustion pressure, lubricating friction, and asperity friction during engine operation and reciprocates within the cylinder. Therefore, the main goal of this study is to investigate the effect of engine combustion and lubricant conditions on piston friction. First, we analyze how the combustion pressure, obtained by changing the combustion load and the ignition timing, affects piston friction. Second, by adjusting engine speed and coolant/oil temperature, we analyze how each condition affects piston friction. Through the experiments for each case, it was confirmed that the friction increased as the combustion pressure increased under the same lubrication conditions. It was also confirmed that the piston friction was significantly measured due to the increase in lubrication friction as the engine speed increased. Finally, it was confirmed that as the temperature decreased, the oil viscosity increased, resulting in a large friction loss.

The functions of the tire can be fully performed only under the appropriate inflation pressure. Tire inflation pressure loss rate (IPLR) is an important indicator used to evaluate tire inflation pressure retention performance. When the tire is under a rolling state, the temperature will increase leading to an accelerated oxidation process. Standard tire IPLR test is costly and time-consuming. In order to simulate the tire IPLR under the steady rolling state, the permeability parameters of rubbers and rubber-cord components (belts & carcass) are tested. Moreover, the oxidation reaction is coupling with the IPLR model based on the basic autoxidation scheme (BAS). The simulation results show that the IPLR of the tire under rolling state increases obviously. A novel calculation method of rubber oxygen consumption is proposed and applied to the oxygen consumption of different parts of the tire. Two approaches are adopted to reduce tire IPLR based on the innerliner structure design. This simulation method can predict tire IPLR not only under steady rolling state but also under static state and even under tire aging condition in oven. It also provides an important model basis for tire aging and life prediction research in the future.

In the previous research, we had developed a Nitrogen Oxides emissions (NOx) prediction model using deep learning methodology. However, it is necessary to develop a lightweight model for practical application. We developed a machine learning model for predicting NOx using Random Forest method to reduce the negligible input features. For comparison, a Base model was developed with all features. The Shapley Additive Explanations method, which can show the influence of the input features, and the Pearson Correlation Coefficient method, which can show the proportional relationship between the output and the input values, were used to choose the dominant features. The final model was determined using the Shapley Additive Explanations method. The final model shows similar prediction performance to the base model, though it has only 11 features (30 % of the Base model). The final model shows 15.76 for the root mean square error and 0.965 for R2 with the test data. By extracting the dominant features influencing NOx prediction, we could develop a lightweight and accurate NOx prediction model.

In order to improve the zero-pressure driving performance of the inserts supporting run-flat tire and solve the problems of heavy weight and large moment of inertia, the topology optimization theory based on variable density method was used to optimize the inserts structure of run-flat tire. The mechanical model of zero-pressure driving and the contact finite element model between insert, tire and ground was established. Further, stiffness, strength and modal characteristics of the inserts structure under zero-pressure driving condition before and after optimization were compared and simulated. The results show that under the maximum load of the tire, the optimized inserts structure can reduce the weight by 19.4 % with the requirements of the stiffness and strength, and the load bearing performance of the inserts structure is optimized, and the accuracy of the design is improved. On the premise of satisfying the design requirements of the tire dynamic characteristics, the natural frequency of each order for the optimized inserts structure is reduced. The static load characteristics of the tires under zero-pressure conditions before and after optimization were verified based on the tire load characteristics test bench. The research results provide a reference for the design and optimization of the inserts supporting run-flat tire.

Hydrogen is advantageous for use in internal combustion engines (ICEs) because of its high laminar flame speed and carbon-neutral characteristics. However, because of its low minimum ignition energy, abnormal combustions such as back-fire, pre-ignition, and knocking hinder its efficient supply as a fuel to ICEs under near stoichiometric conditions. Hence, in this study, the effect of various excessive air ratios (λ) on the performance of a hydrogen-fueled spark ignition engine using port fuel injection (PFI) was evaluated under wide-open throttle conditions. After determining the highest limit for the maximum brake torque, a hydrogen direct injection (DI) system was applied under the same λ and abnormal combustion limits to improve the brake power (BP). Results show that at 1,500 rpm, a maximum BP of 14.7 kW was achieved under a λ value of 1.4 with PFI; however, it can be increased up to 21.1 kW using DI with the same brake thermal efficiency under a compression ratio of 14.

Advanced driver assistance systems are an important step on the way towards the autonomous driving. However, there are new challenges in the release of increasingly complex systems. For the testing of those systems many test kilometers are necessary to represent sufficient diversity. Hence, the virtual testing of driver assistance systems brings new opportunities. In virtual environments, it is possible to run a much higher distance in a short time. Simultaneously, the complexity of the environment and the test scenarios are individually adjustable. It is possible to test scenarios that are not feasible in a real environment due to a risk of injury. A big challenge is the physical correct implementation of real vehicles and their components into the Virtual Reality. To enable a realistic virtual testing the vehicles surrounding sensors need to be modeled adequately. Thus, this paper presents an approach for the implementation of a Lidar model into a Virtual Reality. A physical Lidar model is combined with a real-time capable vehicle dynamics model to investigate the influence of vehicle movements to the sensor measurements. The models are implemented into a highly realistic virtual city environment. Finally, a test campaign shows the influence of the Lidars physics and the vehicle dynamics on the detection results.