Bioinspired deformation computational design method for muscle-driven soft robots using MPM

Abstract

In order to adapt to complex and changing environments, animals have a wide variety of locomotor forms, which has inspired the investigation of their deformation and driving mechanisms. In this paper, we propose a computational design method for muscle-driven soft robots to satisfy desired deformations, aiming to mimic the deformation behavior of muscle-driven animals in nature. In this paper, we generate the ideal muscle-driven layout for the soft robot by inputting an initial shape and a desired shape, so that it can closely achieve the desired deformation. The material point method is utilized to simulate the soft medium so as to achieve the effect of coupling and coordinated deformation of arbitrary shapes. Our method efficiently searches for muscle layouts corresponding to various deformations and realizes the deformation behaviors of a variety of bio-inspired robots, including soft robots such as bionic snakes, frogs, and human faces. Experimental results show that for both the bionic snake and frog soft robots, the overlap of the geometric contour regions between the actual and simulated deformations is more than 90%, which validates the effectiveness of the method. In addition, the global muscle distributions of the bionic snake and human face soft robots during motion are generated and validated by effective simulation.