查看更多>>摘要：Risk assessment is a crucial component of collision warning and avoidance systems for intelligent vehi-cles.Reachability-based formal approaches have been developed to ensure driving safety to accurately detect potential vehicle collisions.However,they suffer from over-conservatism,potentially resulting in false-positive risk events in complicated real-world applications.In this paper,we combine two reach-ability analysis techniques,a backward reachable set(BRS)and a stochastic forward reachable set(FRS),and propose an integrated probabilistic collision-detection framework for highway driving.Within this framework,we can first use a BRS to formally check whether a two-vehicle interaction is safe;otherwise,a prediction-based stochastic FRS is employed to estimate the collision probability at each future time step.Thus,the framework can not only identify non-risky events with guaranteed safety but also provide accurate collision risk estimation in safety-critical events.To construct the stochastic FRS,we develop a neural network-based acceleration model for surrounding vehicles and further incorporate a confidence-aware dynamic belief to improve the prediction accuracy.Extensive experiments were conducted to val-idate the performance of the acceleration prediction model based on naturalistic highway driving data.The efficiency and effectiveness of the framework with infused confidence beliefs were tested in both naturalistic and simulated highway scenarios.The proposed risk assessment framework is promising for real-world applications.
查看更多>>摘要：Decision-making and motion planning are extremely important in autonomous driving to ensure safe driving in a real-world environment.This study proposes an online evolutionary decision-making and motion planning framework for autonomous driving based on a hybrid data-and model-driven method.First,a data-driven decision-making module based on deep reinforcement learning(DRL)is developed to pursue a rational driving performance as much as possible.Then,model predictive control(MPC)is employed to execute both longitudinal and lateral motion planning tasks.Multiple constraints are defined according to the vehicle's physical limit to meet the driving task requirements.Finally,two prin-ciples of safety and rationality for the self-evolution of autonomous driving are proposed.A motion envelope is established and embedded into a rational exploration and exploitation scheme,which filters out unreasonable experiences by masking unsafe actions so as to collect high-quality training data for the DRL agent.Experiments with a high-fidelity vehicle model and MATLAB/Simulink co-simulation environ-ment are conducted,and the results show that the proposed online-evolution framework is able to gen-erate safer,more rational,and more efficient driving action in a real-world environment.
查看更多>>摘要：Human agency has become increasingly limited in complex systems with increasingly automated decision-making capabilities.For instance,human occupants are passengers and do not have direct vehi-cle control in fully automated cars(i.e.,driverless cars).An interesting question is whether users are responsible for the accidents of these cars.Normative ethical and legal analyses frequently argue that individuals should not bear responsibility for harm beyond their control.Here,we consider human judg-ment of responsibility for accidents involving fully automated cars through three studies with seven experiments(N=2668).We compared the responsibility attributed to the occupants in three conditions:an owner in his private fully automated car,a passenger in a driverless robotaxi,and a passenger in a con-ventional taxi,where none of these three occupants have direct vehicle control over the involved vehicles that cause identical pedestrian injury.In contrast to normative analyses,we show that the occupants of driverless cars(private cars and robotaxis)are attributed more responsibility than conventional taxi pas-sengers.This dilemma is robust across different contexts(e.g.,participants from China vs the Republic of Korea,participants with first-vs third-person perspectives,and occupant presence vs absence).Furthermore,we observe that this is not due to the perception that these occupants have greater control over driving but because they are more expected to foresee the potential consequences of using driverless cars.Our findings suggest that when driverless vehicles(private cars and taxis)cause harm,their users may face more social pressure,which public discourse and legal regulations should manage appropriately.
查看更多>>摘要：The forward design of trajectory planning strategies requires preset trajectory optimization functions,resulting in poor adaptability of the strategy and an inability to accurately generate obstacle avoidance trajectories that conform to real driver behavior habits.In addition,owing to the strong time-varying dynamic characteristics of obstacle avoidance scenarios,it is necessary to design numerous trajectory optimization functions and adjust the corresponding parameters.Therefore,an anthropomorphic obstacle-avoidance trajectory planning strategy for adaptive driving scenarios is proposed.First,numer-ous expert-demonstrated trajectories are extracted from the HighD natural driving dataset.Subsequently,a trajectory expectation feature-matching algorithm is proposed that uses maximum entropy inverse reinforcement learning theory to learn the extracted expert-demonstrated trajectories and achieve automatic acquisition of the optimization function of the expert-demonstrated trajectory.Furthermore,a mapping model is constructed by combining the key driving scenario information that affects vehicle obstacle avoidance with the weight of the optimization function,and an anthropomorphic obstacle avoidance trajectory planning strategy for adaptive driving scenarios is proposed.Finally,the proposed strategy is verified based on real driving scenarios.The results show that the strategy can adjust the weight distribution of the trajectory optimization function in real time according to the"emergency degree"of obstacle avoidance and the state of the vehicle.Moreover,this strategy can generate anthro-pomorphic trajectories that are similar to expert-demonstrated trajectories,effectively improving the adaptability and acceptability of trajectories in driving scenarios.
查看更多>>摘要：Model mismatches can cause multi-dimensional uncertainties for the receding horizon control strategies of automated vehicles(AVs).The uncertainties may lead to potentially hazardous behaviors when the AV tracks ideal trajectories that are individually optimized by the AV's planning layer.To address this issue,this study proposes a safe motion planning and control(SMPAC)framework for AVs.For the control layer,a dynamic model including multi-dimensional uncertainties is established.A zonotopic tube-based robust model predictive control scheme is proposed to constrain the uncertain system in a bounded min-imum robust positive invariant set.A flexible tube with varying cross-sections is constructed to reduce the controller conservatism.For the planning layer,a concept of safety sets,representing the geometric boundaries of the ego vehicle and obstacles under uncertainties,is proposed.The safety sets provide the basis for the subsequent evaluation and ranking of the generated trajectories.An efficient collision avoid-ance algorithm decides the desired trajectory through the intersection detection of the safety sets between the ego vehicle and obstacles.A numerical simulation and hardware-in-the-loop experiment validate the effectiveness and real-time performance of the SMPAC.The result of two driving scenarios indicates that the SMPAC can guarantee the safety of automated driving under multi-dimensional uncertainties.
查看更多>>摘要：Active suspension systems(ASSs)have been proposed and developed for a few decades,and have now once again become a thriving topic in both academia and industry,due to the high demand for driving comfort and safety and the compatibility of ASSs with vehicle electrification and autonomy.Existing review papers on ASSs mainly cover dynamics modeling and robust control;however,the gap between academic research outcomes and industrial application requirements has not yet been bridged,hindering most ASS research knowledge from being transferred to vehicle companies.This paper comprehensively reviews advances in ASSs for road vehicles,with a focus on hardware structures and control strategies.In particular,state-of-the-art ASSs that have been recently adopted in production cars are discussed in detail,including the representative solutions of Mercedes active body control(ABC)and Audi predictive active suspension;novel concepts that could become alternative candidates are also introduced,includ-ing series active variable geometry suspension,and the active wheel-alignment system.ASSs with com-pact structure,small mass increment,low power consumption,high-frequency response,acceptable economic costs,and high reliability are more likely to be adopted by car manufacturers.In terms of con-trol strategies,the development of future ASSs aims not only to stabilize the chassis attitude and atten-uate the chassis vibration,but also to enable ASSs to cooperate with other modules(e.g.,steering and braking)and sensors(e.g.,cameras)within a car,and even with high-level decision-making(e.g.,refer-ence driving speed)in the overall transportation system-strategies that will be compatible with the rapidly developing electric and autonomous vehicles.
查看更多>>摘要：High-efficiency and low-cost knowledge sharing can improve the decision-making ability of autonomous vehicles by mining knowledge from the Internet of Vehicles(IoVs).However,it is challenging to ensure high efficiency of local data learning models while preventing privacy leakage in a high mobility environ-ment.In order to protect data privacy and improve data learning efficiency in knowledge sharing,we pro-pose an asynchronous federated broad learning(FBL)framework that integrates broad learning(BL)into federated learning(FL).In FBL,we design a broad fully connected model(BFCM)as a local model for train-ing client data.To enhance the wireless channel quality for knowledge sharing and reduce the commu-nication and computation cost of participating clients,we construct a joint resource allocation and reconfigurable intelligent surface(RIS)configuration optimization framework for FBL.The problem is decoupled into two convex subproblems.Aiming to improve the resource scheduling efficiency in FBL,a double Davidon-Fletcher-Powell(DDFP)algorithm is presented to solve the time slot allocation and RIS configuration problem.Based on the results of resource scheduling,we design a reward-allocation algorithm based on federated incentive learning(FIL)in FBL to compensate clients for their costs.The simulation results show that the proposed FBL framework achieves better performance than the compar-ison models in terms of efficiency,accuracy,and cost for knowledge sharing in the IoV.
查看更多>>摘要：Ultrafine-grained pure metals and their alloys have high strength and low ductility.In this study,cry-orolling under different strains followed by low-temperature short-time annealing was used to fabricate pure nickel sheets combining high strength with good ductility.The results show that,for different cry-orolling strains,the uniform elongation was greatly increased without sacrificing the strength after annealing.A yield strength of 607 MPa and a uniform elongation of 11.7％were obtained after annealing at a small cryorolling strain(e=0.22),while annealing at a large cryorolling strain(ε=1.6)resulted in a yield strength of 990 MPa and a uniform elongation of 6.4％.X-ray diffraction(XRD),transmission elec-tron microscopy(TEM),scanning electron microscopy(SEM),and electron backscattered diffraction(EBSD)were used to characterize the microstructure of the specimens and showed that the high strength could be attributed to strain hardening during cryorolling,with an additional contribution from grain refinement and the formation of dislocation walls.The high ductility could be attributed to annealing twins and micro-shear bands during stretching,which improved the strain hardening capacity.The results show that the synergistic effect of strength and ductility can be regulated through low-temperature short-time annealing with different cryorolling strains,which provides a new reference for the design of future thermo-mechanical processes.
查看更多>>摘要：Blue energy,which includes rainfall,tidal current,wave,and water-flow energy,is a promising renewable resource,although its exploitation is limited by current technologies and thus remains low.This form of energy is mainly harvested by electromagnetic generators(EMGs),which generate electricity via Lorenz force-driven electron flows.Triboelectric nanogenerators(TENGs)and TENG networks exhibit superiority over EMGs in low-frequency and high-entropy energy harvesting as a new approach for blue energy har-vesting.A TENG produces electrical outputs by adopting the mechanism of Maxwell's displacement cur-rent.To date,a series of research efforts have been made to optimize the structure and performance of TENGs for effective blue energy harvesting and marine environmental applications.Despite the great pro-gress that has been achieved in the use of TENGs in this context so far,continuous exploration is required in energy conversion,device durability,power management,and environmental applications.This review reports on advances in TENGs for blue energy harvesting and marine environmental monitoring.It intro-duces the theoretical foundations of TENGs and discusses advanced TENG prototypes for blue energy har-vesting,including TENG structures that function in freestanding and contact-separation modes.Performance enhancement strategies for TENGs intended for blue energy harvesting are also summa-rized.Finally,marine environmental applications of TENGs based on blue energy harvesting are discussed.
查看更多>>摘要：Identifying workers'construction activities or behaviors can enable managers to better monitor labor efficiency and construction progress.However,current activity analysis methods for construction work-ers rely solely on manual observations and recordings,which consumes considerable time and has high labor costs.Researchers have focused on monitoring on-site construction activities of workers.However,when multiple workers are working together,current research cannot accurately and automatically iden-tify the construction activity.This research proposes a deep learning framework for the automated anal-ysis of the construction activities of multiple workers.In this framework,multiple deep neural network models are designed and used to complete worker key point extraction,worker tracking,and worker con-struction activity analysis.The designed framework was tested at an actual construction site,and activity recognition for multiple workers was performed,indicating the feasibility of the framework for the auto-mated monitoring of work efficiency.
NETL
NSTL
万方数据