期刊文献+

驾驶行为机器人仿真环境与拥堵算法的研究

Research on Simulation Environment and Congestion Algorithm of Driving Robot
下载PDF
导出
摘要 目的 强化学习具有较强的感知能力,但是缺乏一定的决策、能力,而强化学习具有决策能力,对感知问题束手无策。因此,将两者结合起来,即为深度强化学习,优势互补,为复杂系统的感知决策问题提供了解决思路。在现实生活中,我们出行会受到多种因素的影响,人类通过多次的驾驶,驾驶水平会越来越高,机器人也是同样如此。方法 该文对用户驾驶行为机器人进行研究,构建一个机器人驾驶仿真环境,伴随着拥堵情况,让机器人自己不断地去学习,如何更快、更安全到达目的地。结果 在该仿真环境下,机器人通过学习,使其能够更快、更安全到达目的地。结论 在设置好拥堵及仿真环境后,通过训练能够使机器人更快、更安全到达目的地。 Objective Reinforcement learning has strong perceptive ability, but it lacks certain decision-making ability, while reinforcement learning has decision-making ability and is helpless to perception problems. Therefore,the combination of the two can provide deep reinforcement learning with complementary advantages and provide a solution to the perceptual decision-making problem of complex systems. In real life, our travel will be affected by many factors. Through many times of driving, the driving level will become higher and higher, and so will robots.Methods In this paper, the user driving behavior robot is studied, and a robot driving simulation environment is built. With the traffic jam, the robot can learn how to reach the destination faster and more safely. Results In this simulation environment, the robot can reach the destination faster and more safely through learning. Conclusions After setting the congestion and simulation environment, the robot can reach the destination faster and more safely through training.
作者 李雨松 徐野 哈乐 LI Yusong;XU Ye;HA Le(School of Automation and Electrical Engineering,Shenyang Ligong University,Shenyang,Liaoning Province,110159 China;Department of Medical Engineering,North General Hospital,Shenyang,Liaoning Province,110000 China)
出处 《科技资讯》 2022年第20期40-43,共4页 Science & Technology Information
关键词 强化学习 用户驾驶行为 深度强化学习 自动驾驶 Reinforcement learning User driving behavior Deep reinforcement learning Auto pilot
  • 相关文献

参考文献3

二级参考文献6

共引文献14

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部