Robots collision avoidance using learning through imitation

This paper deals with the collision avoidance of the cooperative robots using the learning through imitation. Each physical robot acts fully independently, communicating with corresponding virtual prototype and imitating her behavior. Each physical robot reproduces the motion of her virtual prototype. The estimation of the collision-free actions of the virtual cooperative robots and the transfer of the virtual joint trajectories to the physical robots who imitate there virtual prototypes, are the original ideas. We tested the present strategy on several simulation scenarios; involving two virtual robots that each must cooperate with other and estimating collision-free actions. The effectiveness of the proposed strategy is discussed by theoretical considerations and illustrated by simulation of the motion of two cooperative manipulators. It is shown that the proposed collision-free strategy, while tracking the end-effector trajectory, is efficient and practical.