Self-Sensing and Control of Soft Electrothermal Actuator

Abstract

Soft electrothermal actuators have been of particular interest to researchers in soft mechatronic/robotic systems due to their large deformation, lightweight, and low power actuation. Closed-loop control of soft actuators is critical for high-precision applications, due to the nonlinear relationship between actuation voltage and output bending motion. In this article, a resistive self-sensing approach was developed for the soft electrothermal actuator, which enables the closed-loop control of the actuator without external sensors. The self-sensing of the actuator's deformation is achieved by measuring the resistance change of the embedded microfilament heater of the soft electrothermal actuator. When the bimorph electrothermal actuator deflects under the temperature change, its deflection can be detected by the temperature-incurred resistance change of the embedded heater. A closed-loop controller was designed using the self-sensed deflection signal. To handle the different responses of the actuator in the heating and cooling stages, a switching proportional-integral-derivative control algorithm was designed. Specifically, two sets of control parameters were tuned and used for the heating and cooling stages, respectively. The performance of the designed control algorithm was evaluated for step response, tracking of complex wave signals, and disturbance rejection. Compared with the open-loop operation, the closed-loop controlled actuator demonstrated a much more rapid and accurate response (the rising time and settling time were reduced over 80%) and excellent tracking and disturbance rejection capabilities.