6. Conclusion

This document presents methods for the simulation and control of soft robots. These methods rely on a FEM approach to handle the mechanical deformations of the robot. Thanks to a set of Lagrange multipliers defined on the boundary conditions, actuators, effectors and contacts are modeled accurately. These models allow the direct simulation of the robot and its interaction with its environment. Moreover, the mechanical representation can be used as an inverse problem optimization that automatically computes the actuation to obtain control in the task space, even in situation of contact with the environment. The capabilities of these methods are illustrated with several experiments showing that a reasonable accuracy between simulated and real soft robots can be obtained.

The results we present in this document are open-loop experiments. There is no feedback signal...