2023年第2期周苏1,2王若伊1张友发2张岗1(1.同济大学汽车学院,上海201804;2.同济大学中德学院,上海201804)【摘要】自动驾驶汽车混合A*算法运行迅速,但是难以保证曲率的连续性和汽车舒适度,并且不能保证优化后的结果无碰撞。针对以上问题,以提高汽车舒适性为目的,基于最小冲击度对A*算法进行优化,在MATLAB和栅格地图中进行了算法预研,解决了低速场景下自动驾驶汽车运动过程中舒适度不足的问题。然后,针对优化算法可能导致轨迹碰撞的问题,对优化算法进行了改进,保证了汽车的安全性,并给出了“走廊”自动调整的策略。仿真结果表明:优化后的轨迹无碰撞且曲率更小,加速度不超出正常范围,可以满足正常驾驶需求。主题词:混合A*算法轨迹优化算法贝塞尔曲线自动驾驶中图分类号:U463.6;TP18文献标识码:ADOI:10.19822/j.cnki.1671-6329.20220270HybridA*AlgorithmBasedTrajectoryGenerationforAutomatedDrivingVehicleZhouSu1,2,WangRuoyi1,ZhangYoufa2,ZhangGang1(1.SchoolofAutomotiveStudies,TongjiUniversity,Shanghai201804;2.Sino-GermanCollege,TongjiUniversity,Shanghai201804)【Abstract】ThehybridA*algorithmofautomateddrivingvehiclesrunsquickly,butitisdifficulttoensurethecontinuityofcurvatureandcomfortofvehicles,andcannotguaranteetheoptimizedresultswithoutcollision.Aimingattheaboveproblems,thispaperaimstoimprovethecomfortofvehicles,andoptimizeA*algorithmbasedonminimumimpact.Algorithmpre-studyinMATLABandrastermapisconductedtosolvetheproblemofinsufficientcomfortduringthemotionofautomateddrivingvehiclesinlow-speedscenarios.Then,theoptimizationalgorithmisimprovedtoensurethesafetyofvehiclesfortheproblemthattheoptimizationalgorithmmayleadtotrajectorycollision,andthestrategyofautomaticadjustmentofthe“corridor”isgiven.Thesimulationresultsshowthattheoptimizedtrajectoryiscollision-freeandthecurvatureofthetrajectoryissmaller,andtheaccelerationdoesnotexceedthenormalrange,whichcanmeetthenormaldrivingrequirements.Keywords:HybridA*algorithm,Trajectoryoptimizationalgorithm,Besselcurve,Automateddriving基于混合A*算法的自动驾驶汽车轨迹生成【欢迎引用】周苏,王若伊,张友发,等.基于混合A*算法的自动驾驶汽车轨迹生成[J].汽车文摘,2023(2):44-54.【Cite...