Position Control of a Shape-Memory Alloy Actuator Using a Preisach- Model- Based Inverse-Temperature Method

Tensioned wires made of shape-memory alloys (SMAs) exhibit strain-temperature hysteresis when their crystal structures undergo cyclic phase transformations due to periodic temperature variations. To compensate for this highly nonlinear behavior during operation, we introduce an inverse control method which uses an experimentally-identified temperature-stress-strain Preisach model. The phase transitions of SMAs are induced both mechanically, according to the superelasticity effect (SE), and thermally, according to the shape-memory effect (SME); most existing Preisach-model-based control schemes for SMAs, however, employ mappings between an exciting electrical current and the resulting strain output because the most common form of thermal SMA excitation is Joule heating. In the proposed approach, we first perform a system identification procedure to find a relationship between temperature, stress and strain; then, the identified model is employed to develop a numerical inversion algorithm which is the key element of the proposed controller. Through this technique, the inverse control scheme computes the temperature-reference signal required to generate a desired strain output when an SMA wire operates as an actuator under a constant stress. The main advantages of this method are its modularity, as it can be integrated into a feedforward control structure to modulate the actuator's output, and its computational efficiency. To test and validate the suitability, efficacy and accuracy of the resulting controller, we employ data from both real-time simulations and experiments. These results indicate that the introduced approach can be potentially adopted to synthesize and implement position controllers for SMA-based actuators driven by methods other than direct electricity; for example, catalytic chemical reactions.