摘要
系统采用分层的状态同步方式,模拟了以往较难实现的无人驾驶多车辆同步实时交互场景.在终端内部,车辆与场景通过交换位置数据和速度数据的方式实现同步;在网络上,各终端都通过与服务器交换数据同步到场景中来,实现跨终端同步仿真,并可弹性扩展至大规模车辆仿真.通过同步多个仿真环境中的交通参与者行为、车辆运动学观测以及时钟信号,使需要多车配合才能实现的场景成为可能.
The system adopts layered state synchronization mode and simulates the real-time interactive scenario of unmanned multi-vehicle synchronization which is difficult to realize in the past. Inside the terminal, the vehicle and the scene are synchronized through the exchange of location data and speed data. On the network, each terminal synchronizes to the scene by exchanging data with the server, and realizes the synchronous simulation across the terminal, which can be extended to large-scale vehicle simulation with elasticity. By synchronizing the behavior of traffic participants, vehicle kinematics observations, and clock signals in multiple simulation environments, it is possible to realize a scenario that requires the cooperation of multiple vehicles.
引文
[1] 刘欣,梅涛,张涛,等.车辆智能行为验证与评估的虚拟场景研究[J].系统仿真学报,2009,21(24):7821-7823.
[2] 王姗姗.PanoSim:新一代先进汽车智能驾驶模拟仿真系统[J].科技纵览,2015(10):68-71.
[3] FLETCHER L,TELLER S,OLSON E.The MIT-Cornell collision and why it happened[J].Journal of Field Robotics,2008,25(10):1-1.
[4] 陈嘉敏.DARPA无人驾驶挑战赛,我们还要经历多少个十年?[EB-OL].(2017-10-23) [2017-11-24].https://www.iyiou.com/p/58040.
[5] JONES M T.开源的物理引擎[EB-OL].(2011-09-07) [2011-10-07].https://www.ibm.com/developerworks/cn/opensource/os-physicsengines/index.html.
[6] 德国BOSCH公司.BOSCH汽车工程手册[M].3版.魏春源,译.北京:北京理工大学出版社,2009.
[7] 赫尔菲什.数字校色[M].黄裕成,周一楠,译.北京:人民邮电出版社.2010.
[8] 李军.城市设计理论与方法[M].武汉:武汉大学出版社,2010.