The shape memory alloy (SMA)-based actuators have been increasingly used in different domains, such as automotive, aerospace, robotic and biomedical applications, for their unique properties. However, the precision control of such SMA-based actuators is still a problem. Most traditional control methods use the force/displacement signals of the actuator as feedback signals, which may increase the volume and weight of the entire system due to the additional force/displacement sensors. The resistance of the SMA, as an inherent property of the actuator, is a dependent variable which varies in accordance with its macroscopic strain or stress. It can be obtained by the voltage and the current imposed on the SMA with no additional measuring devices. Therefore, using the resistance of the SMA as feedback in the closed-loop control is quite promising for lightweight SMA-driven systems. This paper investigates the resistance characteristics of the SMA actuator in its actuation process. Three factors, i.e., the resistivity, the length, and the cross-sectional area, which affect the change of resistance were analyzed. The mechanical and electrical parameters of SMA were obtained using experiments. Numerical simulations were performed by using the resistance characteristic model. The simulation results reveal the change rules of the resistance corresponding to the strain of SMA and demonstrate the possibility of using the resistance for feedback control of SMA.
Shape memory alloy (SMA) has been widely used in different applications due to its unique shape memory property. However, when used as an actuator, it exhibits a hysteresis behavior in its relation between temperature and strain, which is highly nonlinear and difficult to control. Although studies have been conducted on establishing various constitutive models of SMA, it is still difficult to achieve the precise control of the SMA wire with the existing models. In this work, a new promising approach regarding the SMA control task as a reinforcement learning (RL) problem is proposed to address this issue, which does not require accurate mathematical models. Both RL and an improved method named deep reinforcement learning (DRL) are used to solve the problem of precise control of a 1-D SMA wire actuator, respectively. The simulation results indicate that with the DRL method, the agent can precisely control the output deformation of the SMA wire after only ten episodes of training. Compared with the DRL method, the RL agent can also achieve the same training target but with hundreds of training.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.