Reinforcement Learning with Adaptive Curriculum Dynamics Randomization for Fault-Tolerant Robot Control
This addresses fault tolerance for robots in remote or extreme environments, but it is incremental as it builds on existing curriculum learning and dynamics randomization techniques.
The study tackled actuator failure in quadruped robots by developing an adaptive curriculum reinforcement learning algorithm with dynamics randomization (ACDR), which achieved higher average rewards and walking distances compared to conventional methods.
This study is aimed at addressing the problem of fault tolerance of quadruped robots to actuator failure, which is critical for robots operating in remote or extreme environments. In particular, an adaptive curriculum reinforcement learning algorithm with dynamics randomization (ACDR) is established. The ACDR algorithm can adaptively train a quadruped robot in random actuator failure conditions and formulate a single robust policy for fault-tolerant robot control. It is noted that the hard2easy curriculum is more effective than the easy2hard curriculum for quadruped robot locomotion. The ACDR algorithm can be used to build a robot system that does not require additional modules for detecting actuator failures and switching policies. Experimental results show that the ACDR algorithm outperforms conventional algorithms in terms of the average reward and walking distance.