Dynamic Actuator Selection and Robust State-Feedback Control of\n Networked Soft Actuators

The design of robots that are light, soft, powerful is a grand challenge.\nSince they can easily adapt to dynamic environments, soft robotic systems have\nthe potential of changing the status-quo of bulky robotics. A crucial component\nof soft robotics is a soft actuator that is activated by external stimuli to\ngenerate desired motions. Unfortunately, there is a lack of powerful soft\nactuators that operate through lightweight power sources. To that end, we\nrecently designed a highly scalable, flexible, biocompatible Electromagnetic\nSoft Actuator (ESA). With ESAs, artificial muscles can be designed by\nintegrating a network of ESAs. The main research gap addressed in this work is\nin the absence of system-theoretic understanding of the impact of the realtime\ncontrol and actuator selection algorithms on the performance of networked\nsoft-body actuators and ESAs. The objective of this paper is to establish a\nframework that guides the analysis and robust control of networked ESAs. A\nnovel ESA is described, and a configuration of soft actuator matrix to resemble\nartificial muscle fiber is presented. A mathematical model which depicts the\nphysical network is derived, considering the disturbances due to external\nforces and linearization errors as an integral part of this model. Then, a\nrobust control and minimal actuator selection problem with logistic constraints\nand control input bounds is formulated, and tractable computational routines\nare proposed with numerical case studies.\n