Multi-agent Reinforcement Learning with
Hybrid Action Space for Free Gait Motion Planning of Hexapod Robots
Training process
E1: Random plane plum-blossom piles
E2: Random height plum-blossom piles
E3: Random stair plum-blossom piles
E4: Simplified version of E3 in real world
If you have trouble opening videos, please open this link via Chrome instead of Firefox.