TY - JOUR
T1 - Dynamic Compensation of Load-Dependent Creep in Dielectric Electro-Active Polymer Actuators
AU - Jiang, Zhaoguo
AU - Li, Yuan
AU - Wang, Qinglin
N1 - Publisher Copyright:
© 2001-2012 IEEE.
PY - 2022/2/1
Y1 - 2022/2/1
N2 - Dielectric Electro-Active Polymer (DEAP) has been generally used to fabricate smart actuators or sensors in recent years because the DEAP-based actuators or sensors are widely considered to be potential for many applications, especially in intelligent bio-inspired robotics. However, except for the hysteresis nonlinearity which can lead to reduce tracking accuracy of actuator, the occurrence of creep also makes the response of DEAP-based actuator unstable and difficult to control, especially for tasks requiring sustained actuation where the strain is to be held constant for longer periods. Furthermore, in practical operating conditions, the external mechanical loads put on the DEAP-based actuators strongly affect the creep response induced by high voltage. Although a closed-loop strategy can be applied to eliminate the creep easily, it requires many displacement sensors which are usually expensive and difficult to install in some specific applications. Therefore, the aim of this paper is to develop a feedforward compensation scheme for dynamical cancelling the creep phenomena, considering the DEAP-based actuator is working in practical operating conditions with different input control voltages and external mechanical loads. First, a relative creep model which is only related to the properties of material is developed from five-parameters Voigt-Kevin viscoelastic model. Then, a creep compensator is obtained that can compensate the creep phenomena without inversion of creep model. Moreover, in order to fitting the compensator to varied external load situations, a neural network is used to extend the parameter of compensator as a function of input control voltages and external loads. Eventually, the proposed feedforward creep compensator has two inputs and one output. Comparative experiments have been implemented and the results clearly shown that the proposed compensator significantly reduces the creep phenomena induced by varied input control voltages and loads.
AB - Dielectric Electro-Active Polymer (DEAP) has been generally used to fabricate smart actuators or sensors in recent years because the DEAP-based actuators or sensors are widely considered to be potential for many applications, especially in intelligent bio-inspired robotics. However, except for the hysteresis nonlinearity which can lead to reduce tracking accuracy of actuator, the occurrence of creep also makes the response of DEAP-based actuator unstable and difficult to control, especially for tasks requiring sustained actuation where the strain is to be held constant for longer periods. Furthermore, in practical operating conditions, the external mechanical loads put on the DEAP-based actuators strongly affect the creep response induced by high voltage. Although a closed-loop strategy can be applied to eliminate the creep easily, it requires many displacement sensors which are usually expensive and difficult to install in some specific applications. Therefore, the aim of this paper is to develop a feedforward compensation scheme for dynamical cancelling the creep phenomena, considering the DEAP-based actuator is working in practical operating conditions with different input control voltages and external mechanical loads. First, a relative creep model which is only related to the properties of material is developed from five-parameters Voigt-Kevin viscoelastic model. Then, a creep compensator is obtained that can compensate the creep phenomena without inversion of creep model. Moreover, in order to fitting the compensator to varied external load situations, a neural network is used to extend the parameter of compensator as a function of input control voltages and external loads. Eventually, the proposed feedforward creep compensator has two inputs and one output. Comparative experiments have been implemented and the results clearly shown that the proposed compensator significantly reduces the creep phenomena induced by varied input control voltages and loads.
KW - Creep phenomena
KW - Dielectric electro-active polymer actuator
KW - Load-dependent
KW - Neural network
UR - http://www.scopus.com/inward/record.url?scp=85121837627&partnerID=8YFLogxK
U2 - 10.1109/JSEN.2021.3135551
DO - 10.1109/JSEN.2021.3135551
M3 - Article
AN - SCOPUS:85121837627
SN - 1530-437X
VL - 22
SP - 1948
EP - 1955
JO - IEEE Sensors Journal
JF - IEEE Sensors Journal
IS - 3
ER -