Modeling and control of hysteresis for DEAP actuator

Hysteresis phenomenon and nonlinear characteristics presented in dielectric electro active polymers (DEAP) materials show the large influences in the control of these smart materials. These disadvantages can limit the efficiency in the use of the DEAP materials. Therefore, an internal model control (IMC) strategy was presented by using the combination of the Preisach type non-linear auto regressive exogenous (NARX) fuzzy-based adaptive particle swarm optimization (APSO) algorithm forward hysteresis model and the approximated inverse hysteresis model in order to control the position of a DEAP actuator. At first, the forward hysteresis model was proposed combining the Preisach model and the dynamic NARX fuzzy structure. The recursive least squares method and the APSO algorithm were then applied to estimate and identify the weighting mass of the Preisach operators and the NARX fuzzy parameters. Secondly, the approximated inverse hysteresis compensation is studied to model the inverse hysteresis of the system. Finally, the real-time experiments demonstrate the effectiveness in tracking performance and hysteresis compensation of the IMC strategy with the accuracy about 97.5%.