Search

PDF file - 41K

PDF file - 212K

Lateral velocity is an important parameter to characterize vehicle stability. The acquisition of lateral velocity is of great significance to vehicle stability control and the trajectory following control of autonomous vehicles. Aiming to resolve the problems of poor estimation accuracy caused by the insufficient modeling of traditional model-based methods and significant decline in performance in the case of a change in road friction coefficient, a deep learning method for lateral velocity estimation using an LSTM, long-term and short-term memory network, is designed. LSTM can well reflect the inertial characteristics of vehicles. The training data set contains sensor data under various working conditions and roads. The simulation results show that the prediction model has high accuracy in general and robustness to the change of road friction coefficient.

This paper aims to solve the problem of optimal design on a comfortable human-machine arrangement of car drivers with different physical signs under dynamic manipulation. Based on the biomechanical characteristic of human skeletal muscle and Hill muscle mechanics model, this paper constructs the human seat musculoskeletal model of 5th, 50th, and 95th percentile physical signs of Chinese car drivers under dynamic manipulation. The six-degree-of-freedom flexible test bench was set up and the center composite method was used to optimize the number of experiments. The consistency and relevance analysis of the actual measurement and dynamic manipulation simulation was carried out to comprehensively analyze the human-machine arrangement parameters such as vehicle seat, pedal, and steering wheel, so as to realize the optimization design of the hard point size of comfortable driving posture and verify the rationality and applicability of the test results through the vehicle road test.

Lithium‐ion batteries (LIBs) have become the main choice for electric vehicles (EVs). However, the thermal runaway problems of LIBs largely limit the wider promotion of EVs. To provide background and insight for the improvement of battery safety, the general working mechanism of LIBs is described in this review, followed by a discussion of the thermal runaway process, including the trigger conditions and material factors. Moreover, advances made to improve battery safety are examined from the perspective of battery materials and management systems. Thus, this review provides a general picture of the thermal runaway risks of LIBs and corresponding solutions with the aim of facilitating safer battery designs.

To ensure that autonomous vehicles satisfy the requirements of the traffic environment, vehicle driving ability, and desired driver experience during obstacle avoidance, this paper proposes a trajectory planner that considers three aspects: driving passability, regional safety, and driving acceptance. Multiresolution state lattices and Bézier curve fitters are applied to a state lattice framework to generate candidate obstacle avoidance trajectory clusters. Trajectory evaluation is then carried out in the above three aspects by using trajectory passability, safety and driver behavior proximity, and a trajectory evaluation function is designed to evaluate and screen trajectory clusters. The trajectory passability is checked by the vehicle motion capability set, which is established based on the vehicle dynamics model. The trajectory safety is evaluated by the potential field function between the fitted trajectory and the vehicle driving environment boundary with consideration of the inevitable collision state. The parameters of the vehicle motion state for the fitted trajectory are matched with the driving data of real drivers with different driving styles to evaluate the proximity between the trajectory and driver behavior. The rationality and effectiveness of different driving styles of trajectory planners are verified by vehicle tests under different vehicle velocities and different obstacle disturbances.

To meet the higher requirements intelligent and electric trend of automobile puts forward for brake-by-wire system, the master cylinder pressure estimation and control is the core technology that need to be researched and improved. Most of the existing research about master cylinder pressure control is based on pressure sensor equipped in it. In this scheme, the failure of pressure sensor will seriously affect the reliability and it is difficult to ensure pressure tracking speed and accuracy at the same time. This article proposes an adaptative pressure estimation and control architecture which is based on estimated pressure instead of pressure sensor for integrated electro-hydraulic brake system. About the master cylinder pressure estimation, we propose an adaptative fusion method which calculates the weights of estimator based on vehicle mode and electro-hydraulic mode according to master cylinder pressure, longitudinal speed and ABS status, the estimated pressure from fusion method finally is used to update the hydraulic mode of integrated electro-hydraulic brake system. According to the updated hydraulic mode, we can derive the final estimated pressure and update the coefficient of the friction mode. In addition, this article proposes an adaptative LQR controller for master cylinder pressure control which uses fuzzy-logic controller to adjust the weights of LQR controller based on target pressure and difference compared with actual pressure. At last, through Mode-in-Loop and Hardware-in-Loop tests, we compare the estimated pressure and actual pressure, pressure tracking results of Piecewise PID controller and that of adaptative LQR controller in ramp, step and sinusoidal response. Based on the analysis of experiment results, we can conclude that the proposed adaptative pressure estimation and control architecture can estimate MC pressure effectively in any scene and ensure the pressure tracking speed and accuracy at the same time.

Being one of the core power units of electric vehicles, the lithium-ion batteries (LIBs) are broadly concerned. However, in the cases of abuses, LIBs may counter thermal runaway, threatening the personal and property safety of users. In order to avoid the occurrence of thermal runaway, the battery thermal management system (BTMS) has been introduced to improve the safety, optimize the efficiency and prolong the service life of lithium-ion batteries. In this review, feasible thermal management schemes of LIBs system were summarized chronically, different thermal management schemes were evaluated, and case studies were made. The schemes of controlling the internal reaction control in the battery are highlighted as well. This review offers a comprehensive view of BTMS and proposes a promising future for the employment of lithium-ion batteries.

Drifting enables the vehicle to quickly adjust its posture and velocity, with the potential to enhance vehicle safety and maneuverability in extreme operating conditions. To improve the handling stability and tracking accuracy of drift control, a controller framework is presented based on model predictive control theory for electronic-two-rear-wheel drive (e-2RWD) automobiles. Primarily, the non-linear bicycle dynamics model is constructed, and in which the motion relationship between the vehicle and the specified trajectory is revealed.Then, a set of reference states are solved for the specified trajectory and desired sideslip. Finally, the MPC drift controller is used to work in conjunction with the wheel speed closed-loop controller to provide coordinated control of the steering angle and both rear wheel drive torques. The performance of the proposed drift controller is assessed in software simulations in two high sideslip driving scenarios with different trajectories. The simulation results show that the proposed controller is able to rapidly enter the steady-state drifting and effectively reduce the fluctuations of vehicle states, striking a reasonable balance between satisficing the accuracy demands and keeping vehicle handling stability. The major contribution of this paper is the first application of the MPC and wheel speed controller combined operation mode to drift control. Due to the advantage of proposed method predicting the vehicle's future dynamic behavior in advance, the controller outputs can be effectively constrained and optimized.

Trajectory prediction and behavioral decision-making are two important tasks for autonomous vehicles that require good understanding of the environmental context; behavioral decisions are better made by referring to the outputs of trajectory predictions. However, most current solutions perform these two tasks separately. Therefore, a joint neural network that combines multiple cues is proposed and named as the holistic transformer to predict trajectories and make behavioral decisions simultaneously. To better explore the intrinsic relationships between cues, the network uses existing knowledge and adopts three kinds of attention mechanisms: the sparse multi-head type for reducing noise impact, feature selection sparse type for optimally using partial prior knowledge, and multi-head with sigmoid activation type for optimally using posteriori knowledge. Compared with other trajectory prediction models, the proposed model has better comprehensive performance and good interpretability. Perceptual noise robustness experiments demonstrate that the proposed model has good noise robustness. Thus, simultaneous trajectory prediction and behavioral decision-making combining multiple cues can reduce computational costs and enhance semantic relationships between scenes and agents., 26 pages, 6 figures

Nowadays, with more and more attention being paid to the characteristics and experience of drivers, a large number of driver classification algorithms have emerged. However, these methods basically cannot be adjusted independently to each driver. Therefore, this paper proposes a self-learning lane change motion planning system considering the driver’s personality. Firstly, the method of driver data acquisition and processing is determined to obtain and extract the lane change data. Then, the planning system built in this paper is explained from two aspects: lane change trigger and lane change trajectory. According to the artificial potential field theory, an obstacle driving risk field is established to evaluate the acceptance of environmental risks of different drivers, and to achieve personalized lane change triggers through online statistics. At the same time, the safety of lane change is ensured by establishing the safety distance model of the target lane. On the other hand, the driver characteristic coefficient Jc and the driver reaction and operation time td are introduced into the traditional Gaussian-distributed model to establish a personalized lane change trajectory planning model, in which the parameters are obtained from offline and online learning. Offline learning is based on DTW for trajectory matching, and uses AP clustering to obtain the generalized parameters; Online learning uses LSTM to achieve personalized updates. Finally, this paper selected 15 drivers for verification, and the results show that the motion planning system can well reproduce the lane change behavior of the driver.

Autonomous vehicles (AVs) need to adequately predict the trajectory space of surrounding vehicles (SVs) in order to make reasonable decision-making and improve driving safety. In this paper, we build the driving behavior intention recognition module and traffic vehicle expected trajectory prediction module by deep learning. On the one hand, the driving behavior intention recognition module identifies the probabilities of lane keeping, left lane changing, right lane changing, left acceleration lane changing, and right acceleration lane changing of the predicted vehicle. On the other hand, the expected trajectory prediction module adopts an encoder-decoder architecture, in which the encoder encodes the historical environment information of the surrounding agents as a context vector, and the decoder and MDN network combine the context vector and the identified driving behavior intention to predict the probability distribution of future trajectories. Additionally, our model produces the multiple behaviors and trajectories that may occur in the next 6 s for the predicted vehicle (PV). The proposed model is trained, validated and tested with the HighD dataset. The experimental results show that the constructed probabilistic multi-modal expected trajectory prediction possesses high accuracy in the intention recognition module with full consideration of interactive information. At the same time, the multi-modal probability distribution generated by the anticipated trajectory prediction model is more consistent with the real trajectories, which significantly improves the trajectory prediction accuracy compared with other approaches and has apparent advantages in predicting long-term domain trajectories.

Three-dimensional object detection plays an important role in autonomous driving because it provides the 3-D locations of objects for subsequent use in decision-making modules. A novel method is proposed using a monocular image and cascade geometric constraints to achieve robust 3-D vehicle detection. The framework is divided into two stages. In the first stage, the monocular image input is processed using a heatmap-based detection network with five branches to regress the orientation, dimension, center projection of the bottom face, viewpoint classification, and 2-D bounding box. In the second stage, the intersection-over-union threshold is increased to filter out imprecise 2-D bounding boxes. Thereafter, cascade geometric constraints are used to obtain the final 3-D box output, which improves the detection performance under truncation and occlusion conditions. The proposed method is tested on the KITTI-3-D benchmark and is shown to be effective and efficient. The proposed framework does not depend on external sources or subnetworks and can be trained in an end-to-end manner.

This study proposes a deep learning framework for driver identity identification by extracting information from the vehicular controller area network (CAN) bus signals. First, naturalistic driving data of 20 drivers were collected under a fixed testing route with different road types and different traffic conditions. Then, a one-dimensional convolutional neural network was constructed for driver identification, which consists of two convolutional-pooling layers, a fully connected layer, and a SoftMax layer. Model optimisation algorithms were applied to improve accuracy and speed up the training process. Also, the model parameters were optimised by evaluating their influences on the model results. Furthermore, the performance of the proposed algorithm was compared with that of the K-nearest neighbour, support vector machine, multi-layer perceptron, and long short-term memory model. The authors used the M a c r o F 1 score as an evaluation criterion and the identification score of the authors' proposed model reaches 99.10% under 20 testing subjects where the data time window size is one second and the sample data overlap is 80%. The results show that the model's performance is significantly better than the other algorithms, which can effectively identify driver identities with stability and robustness.

As the energy density of lithium-ion batteries continues to increase, battery safety issues characterized by thermal runaway have become increasingly severe. Battery safety issues have severely restricted the large-scale application of power batteries. Among them, the flammable liquid organic electrolyte is one of the main reasons for the safety hazards of battery thermal runaway. This article reviews the flame-retardant mechanism and research progress of phosphorus-based flame-retardant additives, nitrogen-based flame-retardant additives, and halogen-based flame-retardant additives. The design strategies of conventional flame-retardant additives and intelligent flame-retardant additives in lithium-ion batteries are summarized. Finally, a development direction and research prospects of flame-retardant additives in lithium-ion battery electrolytes are prospected.

Objective：The moral and ethical issue is a great challenge to the development of autonomous vehicles. There may be distinctions between the choices made by an observer and a participant. The paper is designed to investigate whether drivers will sacrifice the fewest people to save more people in social dilemma, and whether human drivers would give priority to protecting pedestrians or self-protection in an emergency. Methodology: The experiment was conducted with a total of 50 participants assigned to three groups. Three experimental scenarios were designed and each of them contained a social dilemma. A driving simulator was used in this study to explore the choices of human drivers in social dilemma. In addition, the simulator results were compared with those of questionnaire survey. Result: In study 1, 73% of 22 participants swerved into the right lane to hit only one pedestrian for the safety of other five. In study 2 and 3, more participants chose to hit the barrier to protect the pedestrian. Conclusion: A conclusion can be drawn from the second and third group of experiments that most drivers consider not only their own safety, but the safety of pedestrians. Most of the participants intended to minimize the total amount of harm in social dilemma. The choice of crashing into barriers to protect a pedestrian can also be seen as a way to minimize the total amount of harm.

Gas diffusion layers (GDLs) are the most rigid layers in a 5‐layer or 7‐layer membrane electrode assembly (MEA) of a proton exchange membrane fuel cell. Therefore, in the fuel cell analysis, the mechanical properties of GDLs have a great impact on the stress distribution of the membrane as well as the performance of the whole cell. However, the mechanical properties of GDLs are not sufficiently studied. Nonlinear behavior of GDLs under cyclic compression is discussed rarely in literature. The existing model takes both constraints into consideration, but due to a geometrical oversimplification of the carbon paper microstructure, it shows some deviation, which cannot be overcome by selecting appropriate parameters. In this paper, the geometry of carbon paper microstructure has been reanalyzed. Based on the improved geometry formula, a modified model of carbon paper GDL is presented and verified by experimental data. Furthermore, both the experimental mechanical characteristics are achieved as support data for the improved model.

For the different degrees of danger caused by different driving habits, a lane departure warning algorithm based on probability statistics of driving habits is proposed in this paper. According to the different driving habits of different drivers, the early warning mechanism can be adaptively adjusted through the method of probability statistics to make lane departure warning more targeted and accurate. Firstly, each frame of image is preprocessed, including gray treatment, edge detection and binarization. Then, Canny operator is used to detect the edge, and Hough transform is applied to detect the lines. And the lane median line equation for the detection and identification of lane also can be calculated. After that, the image coordinate system is transformed into the world coordinate system by means of the formula and matrix of coordinate conversion. According to the theory of Kalman filter, the statistics of lateral acceleration and lateral velocity are updated continuously, and the position of the vehicle in the next moment is predicted by the state transition equation and the forecast equation. From the results of experiments and the comparison with exhaustive algorithms, the advantages of using Kalman filter to predict the location of vehicles and the improved time-to-lane-crossing combined with probabilistic statistics to warning are illustrated clearly.

Regarding changes of subjective and objective parameters of comfort during short and long-term driving, the arbitrariness of traditional subjective evaluation methods, as well as the defects of objective method where issues can be qualitatively but not quantitatively analyzed. By measuring pressure distribution and electromyography during short and long-term driving respectively, the article obtained pressure indicators, electromyography characteristic parameters and the corresponding subjective evaluations of drivers. The variation of subjective and objective parameters of comfort, the difference in terms of comfort and the fatigable body parts were tested. By means of correlation analysis completed the indicators screen. Proposed a new comprehensive weighting method-AHP to limit entropy method, established the mapping relations between subjective comfort and objective indicators, which based on pressure distribution and physiological information during short-term and long-term driving. Obtained a quantitative evaluation method that applies pressure distribution, physiological information and subjective evaluation to effectively evaluate driving comfort, and thus provided a theoretical basis for evaluating driving comfort.

In a crash situation, drivers typically make evasive maneuvers before an upcoming impact, which can affect the kinematics and injury during impact. The purpose of the current study was to investigate the response and effect of drivers’ cervical muscles in a frontal impact. A crash scenario was developed using a vehicle driving simulator, and 10 volunteers were employed to drive the simulator at 20 km/h, 50 km/h, 80 km/h and 100 km/h. Electromyography (EMG) was recorded from the sternocleidomastoideus (SCM), splenius cervicis (SPL) and trapezium (TRP) muscles using a data acquisition system, and the level of muscle activation was calculated. A numerical study was conducted using data collected in the experiment. The results revealed that the cervical muscles were activated during drivers’ protective action. EMG activity of cervical muscles before impact was greater than that during normal driving. EMG activity increased with driving speed, with the SCM and TRP exhibiting larger increases than the SPL. The kinematics and load of the driver were influenced by muscle activation. Before the collision, the head of an active model stretched backward, while the passive model kept the head upright. In low-speed impact, the torque and shear of the cervical muscle in the active model were much lower than those in the passive model, while the tension of the cervical muscle was higher in the active model compared with the passive model. The results indicated that the incidence of cervical injury in high-speed impact is complex.

Traffic state recognitions can provide a strategic support for control and management of urban traffic, which is crucial to ease traffic congestion, reduce road accidents, and ensure road traffic efficiency. This paper proposes an effective traffic density estimation method based on image processing. In the beginning, a whole image is divided into several cells, and then a region of interest (ROI) is extracted based on calculating varieties of pixel values in a temporal sequence of each cell. Then a texture feature descriptor, a histogram of multi-scale block local binary pattern (HMBLBP) is proposed for local feature representation. The HMBLBP of all cells in the ROI are concatenated as a global feature. Furthermore, principle component analysis is performed for dimensionality reduction to save computational cost. At last, the method proposed is tested with two datasets captured from real-world traffic scenarios. By using the support vector machine (SVM) classifier, traffic states are classified into heavy, medium and light densities. Reliable performances are shown in the experimental tests.

Clear cell Renal Cell Carcinoma (ccRCC) is characterized in most cases (>80%) by the inactivation of von Hippel-Lindau (VHL) tumor suppressor. VHL loss leads to HIF2α accumulation and constitutive activation of its downstream genes important in carcinogenesis. Preclinical and clinical evidence demonstrate that HIF2α inhibition by the first-in-class belzutifan (MK-6482) offers an effective treatment for ccRCC. Here we discovered a novel, potent, selective, and orally available small molecule HIF2α inhibitor (NKT2152) through rational drug design. Cellular thermal shift assay demonstrated NKT2152 engaged HIF2α and stabilized it against heat-induced denaturation in ccRCC 786-O cells. NKT2152 disrupted HIF2α/HIF1β complex in a dose-dependent manner, resulting in increased cytosol localization and reduced nucleus localization of HIF2α in 786-O cells. In vitro, NKT2152 specifically suppressed the expression of HIF2α target genes including vascular endothelial growth factor A, cyclin D1, and glucose transporter 1 in 786-O cells. Pharmacodynamic analysis of NKT2152 in the 786-O xenograft tumor model confirmed that NKT2152 potently inhibited HIF2α as demonstrated by the repression of HIF2α-regulated genes at both mRNA and protein levels. Moreover, NKT2152 treatment, administrated twice daily by oral gavage, led to a dose-dependent tumor growth inhibition or tumor regression in both ccRCC cell line-derived (A498 and 786-O) and patient-derived xenograft tumor models. More pronounced tumor growth inhibition was observed when NKT2152 was combined with a VEGFR inhibitor or a CDK4/6 inhibitor in the 786-O xenograft tumor model. In addition to directly promoting tumor progression by regulating gene expression in tumor cells, HIF2α has also been proposed to play an integral role in shaping the tumor microenvironment under hypoxia. Given hypoxia is a hallmark of most solid tumors and is associated with poor survival in a wide range of cancer types, we tested the hypothesis that HIF2α inhibition by NKT2152 may exert broader anti-tumor activity in other solid tumors that lack a VHL gene deficiency. Indeed, we demonstrated here that NKT2152 caused significant tumor growth inhibition in multiple solid tumor xenograft models, including but not limited to hepatocellular carcinoma. NKT2152 had excellent oral bioavailability and achieved high exposures in mouse, rat and dog. These data not only support clinical development of NKT2152 in ccRCC, but also provide exciting opportunities for expanding to other solid tumor indications with tremendous unmet medical needs. Clinical evaluation of NKT2152 for the treatment of ccRCC is ongoing. Citation Format: Jing Lu, Hairong Wei, Wenfeng Sun, Jianlin Geng, Ke Liu, Jessica Liu, Zhihong Liu, Jiping Fu, Yigang He, Keshi Wang, Yan Lou, Zhenhai Gao. NKT2152: A highly potent HIF2α inhibitor and its therapeutic potential in solid tumors beyond ccRCC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 6330.

In order to improve the ride comfort and safety of the traditional adaptive cruise control (ACC) system when the preceding vehicle changes lanes, this paper proposes a target vehicle selection algorithm based on the prediction of the lane-changing intention of the preceding vehicle. First, NGSIM dataset is used to train a lane-changing intention prediction algorithm based on sliding window SVM, and the lane-changing intent of the preceding vehicle in the current lane can be identified by lateral position offset. Secondly, according to the lane-changing intention and the collision threat of the preceding vehicle, the target vehicle selection algorithm is studied under three different conditions: safe lane-changing condition, dangerous lane-changing condition, and lane-changing cancellation condition. Finally, the effectiveness of the algorithm proposed in this paper is verified in the co-simulation platform. The simulation results show that the target vehicle selection algorithm proposed in this paper can ensure the smooth transfer of the target vehicle and effectively reduce the longitudinal acceleration fluctuation of the subject vehicle when the preceding vehicle changes lanes safely or cancels the lane change. In the case of a dangerous lane change, the target vehicle selection algorithm proposed in this paper can respond to the dangerous lane change in advance compared with the target vehicle selection method of the traditional ACC system, which can effectively avoid collisions and improve the safety of the subject vehicle.

Brain-like intelligent decision-making is a prevailing trend in today&rsquo, s world. However, inspired by bionics and computer science, the linear neural network has become one of the main means to realize human-like decision-making and control. This paper proposes a method for classifying drivers&rsquo, driving behaviors based on the fuzzy algorithm and establish a brain-inspired decision-making linear neural network. Firstly, different driver experimental data samples were obtained through the driving simulator. Then, an objective fuzzy classification algorithm was designed to distinguish different driving behaviors in terms of experimental data. In addition, a brain-inspired linear neural network was established to realize human-like decision-making and control. Finally, the accuracy of the proposed method was verified by training and testing. This study extracts the driving characteristics of drivers through driving simulator tests, which provides a driving behavior reference for the human-like decision-making of an intelligent vehicle.

According to the thermal ignition theory, the fire triangle (i.e. combustible, oxidizer, and ignition source) of lithium-ion battery (LIB) eruption gases, which are caused by thermal runaway, is key to causing LIB fires. The three fire boundaries corresponding to the fire triangle are the minimum battery eruption gases (BEG) concentration (cBEG, ignition), minimum oxygen concentration (cO2, ignition) and lowest temperature required for ignition. The purpose of this paper is to theoretically clarify the three fire boundaries of BEG. The BEG identification results of 29 thermal runaway tests in an inert atmosphere in open literature were summarized, and a BEG time-sequence diagram was drawn. According to Le Chatelier's mixing rule, the physical properties of gases and the thermal ignition theory, the three fire boundaries were theoretically analyzed for batteries with four types of cathode materials (i.e. LixCoO2 (LCO), LixFePO4 (LFP), LixNiyCozAl1-y-zO2 (NCA), and LixNiyCozMn1-y-zO2 (NMC)) at different state of charge (SOC) (0%~143%). The results showed that both of cBEG, ignition and cO2, ignition had different trends for different types of batteries with the increase in the SOC values at discharged states, but they were almost unchanged in the full and overcharged states. For the four types of batteries, from low to high, cBEG, ignition was NMC < LCO < NCA < LFP, and cO2, ignition was NCA < LFP < NMC < LCO. By controlling the SOC and/or selecting a reasonable battery type, cBEG, ignition and/or cO2, ignition could be changed, thereby changing the probability of battery fire. Batteries were prone to forced-ignition with forced-ignition sources regardless of the BEG temperature. When the eruption gases of a battery are at the eruption temperature without forced-ignition sources, the battery is prone to auto-ignition. However, the battery is hard to auto-ignited when the BEG temperature is cooled below the minimum self-ignition point (about 260 ℃) of the BEG components. The BEG ignition mode can be controlled by changing its temperature and ignition sources.These results can be used to guide battery pack design, thermal management system design and fire safety protection.

At a signalized intersection with a permitted left-turning phase, left-turning and opposing through vehicles always compete for priority according to actual traffic conditions instead of fo...

SUMMARY Using computer-aided engineering (CAE) in the concept design stage of automobiles has become a hotspot in human factor engineering research. Based on human musculoskeletal biomechanical computational software, a seated human-body musculoskeletal model was built to describe the natural sitting posture of a driver. The interaction between the driver and car in various combinations of seat-pan/back-rest inclination angles was analyzed using an inverse-dynamics approach. In order to find out the “most comfortable” driving posture of the seat-pan/back-rest, the effect of seat-pan/back-rest inclination angles on the muscle activity degree, and the intradiscal L4-L5 compression force were investigated. The results showed that a much larger back-rest inclination angle, approximately 15°, and a slight backward seat-pan, about 7°, may relieve muscle fatigue and provide more comfort while driving. Subsequently, according to the findings above, a preliminary driving-comfort function was constructed. RESUMO O uso de engenharia assistida por computador (CAE) na fase de projeto do conceito do automóvel tornou-se um ponto de acesso na pesquisa de fatores humanos. Com base no software computacional biomecânico musculoesquelético humano, foi construído um modelo musculoesquelético sentado para descrever a postura sentada natural de um condutor. A interação entre um motorista e um carro em várias combinações de ângulos de inclinação do assento-pan/encosto foi analisada usando uma abordagem dinâmica do verso. A fim de descobrir a postura de condução “mais confortável” do assento-pan/encosto, o efeito dos ângulos de inclinação do assento-pan/dorso sobre o grau de atividade muscular e a força de compressão intradiscal L4-L5 foi investigado. Os resultados mostraram que um ângulo de inclinação para trás muito maior, aproximadamente 15°, e um ligeiro assento-pan para trás, cerca de 7°, pode aliviar a fadiga muscular e levar a dirigir em uma posição confortável. Posteriormente, de acordo com as conclusões acima expostas, foi construída uma função preliminar de conforto ao dirigir.

Simultaneous localization and mapping have become a basic requirement for most automatic moving robots. However, the LiDAR scan suffers from skewing caused by high-acceleration motion that reduces the precision in the latter mapping or classification process. In this study, we improve the quality of mapping results through a de-skewing LiDAR scan. By integrating high-sampling frequency IMU (inertial measurement unit) measurements and establishing a motion equation for time, we can get the pose of every point in this scan&rsquo, s frame. Then, all points in this scan are corrected and transformed into the frame of the first point. We expand the scope of optimization range from the current scan to a local range of point clouds that not only considers the motion of LiDAR but also takes advantage of the neighboring LiDAR scans. Finally, we validate the performance of our algorithm in indoor and outdoor experiments to compare the mapping results before and after de-skewing. Experimental results show that our method smooths the scan skewing on each channel and improves the mapping accuracy.

Clustering is a technique for dividing a network into different group of nodes and managing the transmission of data among the interacting nodes, to improve the effectiveness and safety of information transfer. Clustering have been well studied and applied in traditional mobile networks. However, vehicle networks have short connection time, frequently changing topology, and other unique properties that conventional clustering cannot transfer well. The vehicle nodes in Vehicle Ad-hoc Networks (VANETs) are most directly affected by the surrounding vehicle nodes and exchanged information with them. However, this will cause network congestion or even the spread of malicious messages. The inclusion of vehicle’s (driver’s) social relationships in vehicle communication clustering will increase the degree of trust between vehicle nodes, making communication more purposeful and accurate. This study proposed a new clustering for vehicle networks that is based on drivers’ social relationship combined with the instantaneous position and speed of the vehicle node. Simulation results showed that this clustering method can improve the effectiveness of information transmission and increase the utilization of the application layer.

The temperature of the battery jet is one of the key basic parameters for the design of battery thermal management system for vehicles, with sufficient results under combustion conditions in the presence of air. However, the original temperature distribution of battery jet in an inert atmosphere and its variation with the state of charge are not very clear for prismatic Ni-rich automotive batteries. This is closer to the real inside environment of the battery pack. In this work, a 50 Ah commercial prismatic cell with a Li(Ni0.6Mn0.2Co0.2)O2 cathode is triggered to a thermal runaway using external heating in a sealed chamber with a nitrogen atmosphere to avoid combustion caused by oxygen from the outside. The results show that the farther away from the safety valve, the lower the temperature of the jet. The jet temperature and its rise rate show an increasing trend with the maximum value of 701℃ and 173℃/s detected with increasing states of charge. Therefore, the BTMS design needs to take into account the high thermal load and high thermal shock caused by thermal runaway even in the absence of external air to participate in the combustion.

The study on the mechanism of failure and thermal runaway of lithium-ion battery (LIB) induced by mechanical deformation has received considerable attention. LIBs connected in series are easily overcharged in practical applications. However, the influence of overcharging on the mechanical response of LIBs remains unclear. Thus, we investigated the lateral compression performance of cylindrical batteries before and after short-term cycles at various overcharge states. The onset of short circuits in compression tests for all the batteries before and after cycling at 4.2 and 4.3 V occurred at their modulus peaks, while that of the batteries after cycling at 4.4 and 4.5 V occurred at either the modulus fluctuation points or the first major modulus peaks. Thermal runaway accidents occurred on the batteries at all overcharge states after the short circuits were triggered. Moreover, thermal runaway would occur on the batteries charged at 4.2–4.4 V, when their anode tabs are located in the compression area. The thermal runaway risks of the test batteries would reach 100% when the voltages of these batteries exceeded 4.4 V. Results obtained by using a thermal camera revealed that the highest surface temperatures of all the batteries without thermal runaway were lower than 85 °C during the compression processes, whereas those of the batteries with thermal runaway were between 200 °C and 600 °C. Further analysis of the data indicated that the batteries before and after cycling at high overcharge voltages failed at minimal moduli and stresses, and this trend became obvious with the cycling of batteries.

Most traffic accidents are caused by driver-related factors such as poor perception, aggressive decision-making, or improper maneuvering. Therefore, it is critical to evaluate and predict driving risks to provide drivers with timely feedback. However, risk assessment involves challenges related to a lack of labeled driving data and the presence of data imbalance in the description of different driving risk levels. To address these challenges, a cost-sensitive semi-supervised deep learning method is developed to obtain driving risk scores based on naturalistic vehicle trajectories. A convolutional neural network and a long short-term memory encoder/decoder network are embedded into a semi-supervised framework that uses only a small labeled dataset to label the remaining unlabeled data and produce a trained network model. As fixed weights cannot adapt to changes in the degree of class imbalance that occur over progressive semi-supervised learning iterations, an adaptive over-balanced cross-entropy loss function is developed to adaptively maintain an over-balanced state for the high-risk class to achieve cost-sensitive learning. The experimental results indicate that the accuracy of the proposed method in determining the current and future 2 s risk scores is 96.63% and 92.06%, respectively, thereby constituting the best comprehensive performance among existing machine learning methods. Moreover, the method is verified using a spatio-temporal diagram of driving risk-trajectory and a current–future risk score diagram. The findings demonstrate that the proposed method can be used to assess driving risks in a reliable and robust manner.

This work studies the impact protection properties of both martensite NiTi wires and superelastic NiTi wires, with 0.5 mm diameter, at different initial impact energies using a new-designed impact test plant. In particular, impact experiments under a specific temperature range of A s - A f , were performed for the martensite NiTi wires. The test results show that both the martensite NiTi wires and superelastic NiTi wires have excellent impact protection properties and can dissipate a significant amount impact energy compared to other more metal materials such as Cu, Fe, and Al. Otherwise, the martensite NiTi wires can dissipate more impact energy in a shorter period of time than superelastic NiTi wires while generating a smaller impact force, F max , due to their larger plastic deformation energy. Moreover, the dissipated energy, W d , of martensite wires increases from 23.8 MJ/m3 to 53.6 MJ/m3, and the impact force, F max , increases from 42.96 N to 95.9 N, while the temperature rises from 35 °C to 65 °C. That means the impact protection properties of martensite NiTi wires are temperature dependent in a specific temperature range A s - A f . Those results in this study also reveal that the impact protection properties of both martensite NiTi wires and superelastic NiTi wires represent the basis for new shape memory damping element applications, particularly in automotive safety systems.

Zn2SnO4-based anode materials have recently attracted considerable attention due to their high capacity and low price for lithium-ion batteries. However, their performance is affected by temperature and temperature-dependent characters have not been investigated sufficiently. In this regard, we tested the electrochemistry performance of Co-doped Zn2SnO4–graphene–carbon (Co–ZTO–G–C) nanocomposite anode at various temperatures (− 25, 25 and 60 °C) and analyzed the main limitations and improvements of its low- and high-temperature behavior. Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) results demonstrated that severe concentration polarization, the absence of Zn2SnO4/Zn(Sn) redox couple and large charge-transfer resistance Rct limited its low-temperature performance. Further electrochemical performance analysis indicated that the doped Co could effectively decrease Rct of the nanocomposite and improve its capacity at low temperature. It also suggested that graphene and carbon layer contributed to maintaining its capacity during high-temperature cycles. Field emission scanning electron microscopy (FESEM) and X-ray diffraction (XRD) results revealed that the performance degeneration of the nanocomposite at elevated temperature was mainly attributed to severe volume expansion/contraction of Zn2SnO4 nanoparticles and destruction of Zn2SnO4 cubic structure. The XRD results also showed that the cubic structures of Zn2SnO4 at all temperatures were destroyed after cycling, which led to cyclic performance degeneration of the Co–ZTO–G–C nanocomposite.

Early studies on intelligent control methods for automatic starting control for vehicles mainly focus on traditional parametric adjustment. However, attempts toward the combination of learning algorithms and models are rare. When drivers’ starting behavior is not considered and tedious parameters are merely used for drive control, the effects result in discomfort for drivers. Therefore, to imitate drivers’ starting behavior when dealing with automatic drive control in vehicles, we must first develop an acceleration fitting based on DFT by analyzing the starting characteristics. Then, we design an iterative learning algorithm to achieve automatic starting control of vehicles. Finally, a simulation test is conducted based on CARSIM to verify the validity and feasibility of the proposed method.