Design and implementation of a 2DOF human finger using shape memory alloys with computer-based online force control

In this paper, using shape memory alloy (SMA) wires, a model-based controller for force tracking of fingertips of an artificial hand finger is developed. Different aspects of modeling including static analysis, heat transfer analysis and also investigation of SMA wires, which are used as actuators, are studied. In order to parametrize the SMA wires a modified Brinson’s model is used. Brinson’s original model was not capable of accurately predicting all loading conditions thus some modifications are utilized. In order to control the applied potential difference of memory wires and consequently to control the electrical current to these wires and based on the model of shape memory wires a controller is designed. The main goal of presented controller is force controlling of a two DOF hand finger in order to grasp objects. This model contains SMA model and for compensation of system uncertainties, a proportional integrator derivative (PID) controller has been included in the closed-loop system. The effect of compensator will act only on the derivative type states and this is a new approach compared to similar literature. Simulation results of tracking a reference signal is reported which confirm that the model is in good agreement with experimental tests. The analysis of the relative tracking error for an arbitrary reference signal is 18% for the maximum normalized overshoot in experimental tests and 5% in the simulation.