EXPLORING TWISTED STRING ACTUATION FOR APPLICATION IN SOFT ROBOTICS
Loading...
Authors
Konda, Revanth
Issue Date
2023
Type
Dissertation
Language
Keywords
Adaptive Control , Hysteresis , Modeling and Control , Robot Design , Soft Robotics , Twisted String Actuators
Alternative Title
Abstract
To make robots more ubiquitous, robotic structures that are compliant, lightweight, and low-cost are desired. Soft robots, which are fabricated using compliant materials and soft actuators, are ideal for such applications. However, producing high-performance soft robots is challenging due to the lack of systematic procedures and inherent limitations of the actuators which drive them. Despite many advances in the field of soft actuators and artificial muscles, the quest for an ideal soft actuator which exhibits sufficient contraction and force generation, while being compact is still an ongoing effort. In an attempt to explore novel actuation techniques for soft robotics, in this dissertation, the employment of twisted string actuation for soft robotic applications is investigated and potential limitations which arise as a consequence are addressed. Due to their tendon-based (muscle-like) linear actuation, high force generation, and high operational bandwidth, twisted string actuators (TSAs) are highly suitable to actuate soft robotic devices with advantages over other soft actuators. TSAs typically generate strains of 30--40% of their untwisted length, exhibit energy efficiency of 75--80%, and can exert more stress than skeletal muscles. Although TSAs have been widely adopted in multiple robotic applications, their inclusion in soft robots has been limited.In this dissertation, the application of TSAs in soft robotic manipulation is explored, in that a soft robotic manipulator is presented. The design of the robot is discussed. A physics-based model is developed to predict the manipulator's kinematic motion. An inverse model is derived to realize open-loop control. The proposed modeling and control approaches were experimentally verified to be effective. While the proposed soft robot exhibited promising performance, there were a few notable limitations due to the soft material and actuation mechanism: Firstly, the usage of soft material resulted in the nonlinear behavior called hysteresis. In addition, termed as lonely stroke, the first input-output cycle of the soft robots was inconsistent with subsequent cycles that were repeatable and exhibited hysteresis. The lonely stroke not only affected the behavior of the hysteretic system, but also presented coupling with the subsequent repeatable hysteresis cycles. As a part of this dissertation, a model to capture and compensate for the hysteresis with lonely stroke property is proposed. A modeling approach is developed through the expansion of the input range of the Preisach operator, a widely adopted hysteresis model, to physically infeasible region. The effectiveness of the proposed scheme was validated by simulation and experimental results.Secondly, most existing studies on control of TSAs assume that the external force applied to the TSA is measurable or predictable. Furthermore, existing studies also assume that all the TSA parameters such as the motor properties and string stiffnesses are accurately known. However, the system parameters could be difficult to measure, could change over time due to general wear and tear, and creep. The external forces applied to the TSA could be difficult to predict or measure. To address these issues, parameter estimation and control strategies were developed for TSAs, assuming {little or} no knowledge about the system parameters. A parameter estimation strategy which utilized the least squares algorithm and gradient algorithm is presented. An adaptive control strategy based on model reference control with feedback linearization is proposed. The developed estimation and control strategies were tested to be effective through simulation.Lastly, the limited strain exhibited by TSAs adversely affected the design and performance of the soft robot. Therefore, it is strongly desirable but challenging to enhance the TSAs' consistent strain generation while maintaining compliance. Existing studies predominantly considered coiling after the regular twisting stage to be undesirable—non-uniform and unpredictable knots, entanglements, and coils formed to create an unstable and failure-prone structure. Coiling would work well for TSAs when uniform coils can be consistently formed. As a part of this dissertation, we realize uniform and consistent coil formation in coiled TSAs, which greatly increases their strain. Furthermore, we investigate methods for enabling uniform coil formation upon coiling the strings in a TSA and present a procedure to systematically “train” the strings. Coiling resulted in approximately 70% strain in stiff TSAs and approximately 60% strain in compliant TSAs. TSAs capable of operating in the coiling phase were termed as twisted and coiled string (TCS) actuators.To further establish the coiling mechanism as a reliable actuation mode for TCS actuators, the behavior of the TCS actuators in the coiling phase was experimentally characterized. The non-smooth behavior was investigated by using sequences of input motor turns with different frequencies. The hysteretic behavior and the properties of transitioning conditions were examined by applying input cycles with different bandwidths and under multiple loading conditions. Secondly, a physics-based kinematic modeling strategy which utilized the geometry of the coiled strings were developed to capture the behavior of the coiling-induced non-smooth behavior of the TCS actuators. Lastly, the kinematic model was inverted to realize open-loop control of TCS muscles through inverse compensation. The proposed modeling and control strategies were experimentally validated to be efficient.