Precise position and trajectory control of pneumatic soft-actuators for assistance robots and motion therapy devices

The goal of this work is the development and performance analysis of control algorithms for new soft fluidic actuators with rotary elastic chambers (REC-actuators). Due to their inherent compliancy these actuators fulfill the requisites for building intrinsically safe mechanisms as assistance robots and motion therapy devices, working in direct physical contact with humans. Besides the difficulties common for control design of pneumatic systems, these actuators itself posses several nonlinear characteristics causing specific problems in their modeling and control. In this work the decentralized joint control scheme is implemented, where the position controller has a cascade structure with a non-linear model based pressure control in the inner loop. Two different position control approaches, which require minimal information on the dynamics of the actuator mechanical subsystem, were investigated and tested: sliding mode control with time delay estimation as well as a fuzzy control with parameter optimization based on genetic algorithms.