An embedded self-sensing motion control system for a strip-shaped dielectric elastomer actuators

Abstract

When dielectric elastomer actuators (DEAs) are actuated via high voltage, their electrical capacitance changes according to the geometry. Therefore, displacement of the actuator can be correlated to the change in capacitance, thus opening up the possibility of self-sensing DEA devices. Self-sensing can be exploited to achieve a sensorless closed loop DEA system, which is attractive from size, weight, and cost perspectives. This research work presents an embedded control system, which enables self-sensing closed loop position control of a DEA. The proposed architecture is cost effective, compact in size, easy to integrate as well as to reprogram in comparison to previous self-sensing implementations relying on FPGA systems. In the developed setup, the online self-sensing algorithm is used for estimation of displacement in a spring-biased strip DEA. For this system, understanding and mapping the correlation between estimated capacitance, applied voltage, and resulting displacement is essential for achieving an accurate DEA position reconstruction. An experimental setup is developed, and used to test a spring-biased DEA system. Self-sensing based feedback control is then used to achieve a tight regulation of the actuator displacement. To verify the effectiveness of the sensorless closed loop control system, its performance is finally compared to sensor-based feedback architectures.