Model reference adaptive control of a soft bending actuator with input constraints and parametric uncertainties

Abstract

In this article, model reference adaptive control of a pneumatically actuated soft robot has been studied in detail. To deal with the effects of system uncertainties, in the proposed control scheme, parametric uncertainties and input constraints are taken into account. To design such a controller, based on experimental analysis, the robot has been modeled as a second-order Linear Parameter Varying (LPV) system. Then, the dominant dynamics are presented as a Linear Time-Invariant (LTI) system, while uni-directional input constraint has been considered as a critical issue in the control scheme design. Furthermore, to compensate parametric uncertainties as well as unmodeled dynamics, adaptive laws are modified. Finally, the effectiveness is studied in different scenarios on an experimental platform to validate our claims. Moreover, to show the proposed approach capabilities and performance, the proposed controller has been compared with a PID and a recent sophisticated robust-adaptive controller, which presented a new formulation to achieve a better tracking performance with guaranteed stability in the presence of different constraints and unmodeled dynamics.