Recent advances in nonlinear control technologies for shape memory alloy actuators

Abstract

This paper reviews recent developments in nonlinear control technologies for shape memory alloy (SMA) actuators in robotics and their related applications. SMA possesses large hysteresis, low bandwidth, slow response, and non-linear behavior, which make them difficult to control. The fast response of the SMA actuator mostly depends upon, (1) type of controller, (2) rate of addition and removal of heat, and (3) shape or form of the actuator. Though linear controllers are more desirable than nonlinear ones, the review of literature shows that the results obtained using nonlinear controllers were far better than the former one.\ud Therefore, more emphasis is made on the nonlinear control technologies taking into account the intelligent controllers. Various forms of SMA actuator along with different heating and cooling methods are presented in this review, followed by the nonlinear control methods and the control problems encountered by the researchers

Nonlinear dynamics of shape memory alloy oscillators in tuning structural vibration frequencies

“…Furthermore, the material itself is inherently dissipative due to the hysteretic stress-strain relation, which makes it very useful for vibration suppression. Practically, the material temperature can be increased relatively fast via one of the available heating technologies, but the cooling rate is problematic for fast response [3,11,27]. Depending on the size and design of the devices, the material can be cooled by using heat sinking, or forced air/liquid, but its response will still be limited within a few Hertz [27].…”

Time-delayed dynamic neural network-based model for hysteresis behavior of shape-memory alloys

“…developed a promising Sarsa-based reinforcement learning algorithm for learning a voltage-strain control policy which once the RL agent learns the most appropriate voltage required to achieve each goals train from each initial strain, it can be used thereafter to control the length of an SMA wire in real time [32]. On the other hand, as far as we know, there are a few reports on updating hysteresis model parameters online to adapt the change of input frequency [33]. Gédouin et al [34] proposed a real-time hysteresis model which is purely a numerical model by using derivative estimations.…”

Design and Implementation of Gains Scheduled PI Control System for Shape Memory Alloy Actuator

“…Utilizing SMA actuators in real applications is, thus, constrained by the nonlinear behavior which requires complex control strategies [5]. In recent years, many control strategies ranging from simple controllers to highly sophisticated nonlinear controllers are applied to improve the performance of SMA actuators.…”

“…Sreekumar et al [5] of compensators and error governors [6]. Yee and Roy [7] presented architecture for force control using differential controller, which consists of a PID controller, a dynamic saturation block, and an anti-windup circuit for the integrator.…”