Advancements in robotic manipulation have led to the development of Variable Stiffness Actuators (VSAs), which have the potential to revolutionize the field by endowing manipulators with high levels of compliant actuation. VSAs are known to provide robustness and flexibility, and hence, they are ideal for tasks requiring variable stiffness, especially in soft robotics and shock-absorbing applications by efficiently harnessing potential energy for repetitive movements. The current research focuses on developing an SMA-based agonist-antagonistic VSA, which follows a non-linear force-displacement relationship. The prototype has been developed with SMA coils in bipenniform configuration with a rotary end effector coupled with an optical encoder to measure the angular displacement. The experiments were conducted on an SMA coil with bias weights, and a regression model was trained for temperature variation with input voltage. ANN (Artificial neural network) was deployed for training the model, achieving an accuracy of 89.12%. Further, an LSTM (Long short-term memory)-based RL (reinforcement learning) model is proposed, that can be integrated with the SMA-based VSA. This architecture defines the change in the state of the current angular displacement depending upon the history of actions. The actions signify the input voltages sampled at regular time intervals during the experiment. Thus, the developed SMA-based VSA system promises to elevate the degree of automation and broaden robotics applications in compliance, adaptability, and efficient energy utilization.
Actuators regulate motion in manufacturing and industrial automation by applying an excitation force or torque. Conventional actuators do have their advantages; however, they have multiple components (prone to wear and tear), are expensive during maintenance, bulky, and suffer from backlashes. Therefore, smart-material-based actuators have been increasingly proposed to overcome such shortcomings. Shape memory alloy (SMA) is generally considered for such applications due to its high power-to-weight ratio, noise-free, energy-efficient operation, and facilitating miniaturization. The current research exploits the advantages of the pennate musculature with the properties of SMA to develop a bipennate SMA-based rotary actuator. Pennate muscle fibers are aligned obliquely to the muscle line of action, enabling fiber force to be coupled to macro-level muscle force, resulting in increased force output. The study presents an ergonomic-design-integration-framework of an SMA-driven rotary actuator. The lightweight gearless actuator has drivability without backlash, compatible with a rhombus-based-compliant power transmission system. An analytical model of the bipennate SMA-based rotary actuator has been developed and experimentally validated. The new actuator delivers at least twice the actuation torque (2.1 N-m) compared to the SMA-based rotary actuators reported in the literature. The actuator also delivers a high associated angular displacement ranging from 60°-70°. The actuator design parameters have been optimized by implementing a constrained gradient descent algorithm such that the output torque, stroke, and efficiency of the actuator system can be tailored as per the requirement and application. The actuator has varied applications, from healthcare devices to next-generation space robots.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.