Evaluation of position tracking control performance of a variable stiffness ankle exoskeleton robot with various controller types

The ankle musculoskeletal system continuously changes the joint stiffness during walking. Inspired by this biomechanical behavior, a variable stiffness ankle exoskeleton robot called as VS-AnkleExo was introduced. Although a variable stiffness ankle joint design provides a flexible, safe and biomimetic, the difficulty of stable and precise position tracking under external disturbances is the biggest problem of such soft actuators. In this paper, the position control performance and sensitivity to external disturbances of hybrid controllers created by combining different types of fuzzy logic theory and classical Proportional Integral Derivative (PID) control method implemented on VS-AnkleExo were evaluated experimentally. VS-AnkleExo will be used for applications such as rehabilitation and walking assistance. In this study, firstly, PID, Fuzzy PD, Fuzzy PD+I, Fuzzy PD+PID and ZTS (Z, T and S type membership functions) Fuzzy PD controllers position tracking performances were evaluated for different references without disturbances. In the continuation of the study, the experiments of step and sine inputs response with disturbances were carried out to test the sensitivity of the proposed controllers against the disturbing forces. The test results showed that the Fuzzy PD+PID controller could reduce position tracking errors more effectively than other controllers and was more robust against externally applied disturbances.