System Identification and Optimal PI Control for Nylon Smart Actuators

Twisted and Coiled Nylons are smart and environmentally friendly actuators that can be used for either exerting force or displacement. For the best usage of this actuator, it is necessary to model and control its behavior. In this paper, we estimate a model that governs the dynamics of Nylons and then design an optimal controller with its help. The model describes the relationship between three key elements in Nylons; Force, Current, and Voltage. By designing a standard test-bed, we extract data on Force, Current, and Voltage in two-by-two pairs. Then we apply different system identification techniques to find a suitable model. The model is then verified. With a legit mode, we put design requirements (agility, cost-efficiency, and tracking) compacted in a cost-function and put the model and PI controller form as constraints. Then by utilizing the Genetic Algorithm, we find the optimal gains for the controller. The PI control form guarantees zero tracking error and we show its generality through following different reference forces. The controller presents both majestic precision and agility at the same time. At last, the legitimacy of the model is reinforced through the plots that resemble the Nylons behavior perfectly.

Keywords:

Optimal PI Control , Smart Materials , Nylon Actuator , system identification