Autotunned Fuzzy Based PID Deformation Control Of a Shape Memory Alloy Actuated Cantilever Beam

Abstract

Shape Memory Alloy based actuators have been used in a wide variety of applications along the past decades. In fact, due to their good force to mass ratio, along with their shape recovery capability, even after being subjected to relatively high mechanical stresses, this kind of material has been employed in applications that range from biomedical engineering to robotic joints, and they are always associated with smooth and quiet displacements. Besides, when it comes to linear displacements actuation, shape memory alloy actuators surpass even conventional electrical machines in many aspects. However, one major drawback appears when dealing with such class of actuators: accurate position control of mechanical systems actuated by shape memory alloys is non-trivial, due mainly to the non-linearities associated with the actuator`s phase transition model, which presents a hysteric behavior with respect to its temperature, and the variance of the thermal parameters with respect to time and temperature. In this context, a model independent control approach is proposed in this paper for controlling the deformation of a cantilever beam actuated by a shape memory alloy wire. This control technique relies on an on-line tunned PID controller based on Fuzzy logic developed around a classical Ziegler-Nichols tunned PID controller. At the end, the hybrid Fuzzy-PID controller is compared with the conventional PID controllers previously implemented.