Design and implementation of neural controllers for shape memory alloy–actuated manipulator

Abstract

This article presents a new setup and investigates neural control approaches developed to control a 1-degree-of-freedom shape memory alloy–actuated manipulator. Shape memory alloy–actuated manipulators have drawbacks such as hysteretic behavior and parameter uncertainty. The suggested control approaches should deal with these problems and make the manipulator to track the desired position. First, a mathematical model of the manipulator is presented. The control methods are then developed and applied to the experimental setup. The high performances of the proposed controllers are indicated by experimental results. A comparison between the applied control approaches is made based on the step response main factors and the control signal of each method. Finally, the frequency range for which the suggested methods are appropriate is confirmed.