人形机器人踢足球：每秒输出50次关节角指令，训练了140个模拟年
机器人踢足球的视频，难点讲得挺清楚：每秒50次关节指令，还在仿真里自己跟自己踢了140年才练出来。
视频展示人形机器人踢足球，需要在同一时间处理平衡、控球、碰撞恢复、对手建模和策略等多项任务。机器人的控制方式是以每秒50次的频率直接输出关节角度指令，而不是从离散动作空间里做选择。训练方面，策略在仿真环境中累计运行了140个模拟年，打了数百万场加速的虚拟足球赛。训练方法是反复与自身历史版本的对局策略比赛，从得分有效的动作中学习。
This is just so tough for robots. The balance, ball control, collision recovery, opponent modelling, locomotion, and strategy all have to run at once.
here the humanoid is producing joint-angle commands 50 times a second, not choosing from a neat discrete action space.
The training setup is even more interesting, it basically spent 140 simulated years getting cooked by itself. it played millions of accelerated virtual soccer games, repeatedly competing against saved earlier versions of its own policy and learning from which actions helped it score.