Deep Reinforcement Learning Based Decision‑Making Strategy of Autonomous Vehicle in Highway Uncertain Driving Environments

Uncertain environment on multi-lane highway,e.g.,the stochastic lane-change maneuver of surrounding vehicles,is a big challenge for achieving safe automated highway driving.To improve the driving safety,a heuristic reinforcement learning decision-making framework with integrated risk assessment is proposed.First,the framework includes a long short-term memory model to predict the trajectory of surrounding vehicles and a future integrated risk assessment model to estimate the possible driving risk.Second,a heuristic decaying state entropy deep reinforcement learning algorithm is introduced to address the exploration and exploitation dilemma of reinforcement learning.Finally,the framework also includes a rule-based vehicle decision model for interaction decision problems with surrounding vehicles.The proposed framework is validated in both low-density and high-density traffic scenarios.The results show that the traffic efficiency and vehicle safety are both improved compared to the common dueling double deep Q-Network method and rule-based method.

Biogenic volatile organic compound emission patterns and secondary pollutant formation potentials of dominant greening trees in Chengdu,southwest China

Integral to the urban ecosystem,greening trees provide many ecological benefits,but the active biogenic volatile organic compounds(BVOCs)they release contribute to the production of ozone and secondary organic aerosols,which harm ambient air quality.It is,therefore,necessary to understand the BVOC emission characteristics of dominant greening tree species and their relative contribution to secondary pollutants in various urban contexts.Consequently,this study utilized a dynamic enclosure system to collect BVOC samples of seven dominant greening tree species in urban Chengdu,Southwest China.Gas chromatography/mass spectrometry was used to analyze the BVOC components and standardized BVOC emission rates of each tree species were then calculated to assess their relative potential to form secondary pollutants.We found obvious differences in the composition of BVOCs emitted by each species.Ficus virens displayed a high isoprene emission rate at31.472μgC/(gdw(g dry weight)·hr),while Cinnamomum camphora emitted high volumes of D-Limonene at 93.574μgC/(gdw·hr).In terms of the BVOC emission rates by leaf area,C.camphora had the highest emission rate of total BVOCs at 13,782.59μgC/(m^(2)·hr),followed by Cedrus deodara with 5466.86μgC/(m^(2)·hr).Ginkgo biloba and Osmanthus fragrans mainly emitted oxygenated VOCs with lower overall emission rates.The high BVOC emitters like F.virens,C.camphora,and Magnolia grandiflora have high potential for significantly contributing to environmental secondary pollutants,so should be cautiously considered for future planting.This study provides important implications for improving urban greening efforts for subtropical Chinese urban contexts,like Chengdu.