Online Estimation of the Temperature of Shape Memory Alloy Wire Using Extended Kalman Filter

In order to use and control Shape Memory Alloy (SMA) actuators, it is essential to measure its state variables to be used as the feedback in the control loop. The wire temperature is one of critical state variables need to be fed back. However, measuring this variable is difficult and usually contains some noises and delay. Therefore, it is desirable to estimate this variable instead of measuring it. Thermoelectric model is one of the most common models used to estimate the SMA wire temperature. This model calculates the SMA wire temperature based on its input electric current. In this paper, first three unknown parameters of thermoelectric model are estimated using Extended Kalman filter (EKF) and the wire temperature is calculated based on the identified model. The parameter estimation and temperature calculation are performed on a practical SMA actuator. Then, in order to eliminate the effects of environmental disturbances and the thermoelectric model inaccuracies, the temperature is estimated using EKF. In this method, all measurable data such as the input current, the strain and stress of the SMA wire are used in the temperature estimation. The estimator combines the information obtained from both thermoelectric and Brinson models and the measurement data. This method is used for online temperature estimation of the SMA wire on a practical SMA actuator. The results show that the estimated temperature matches the actual wire temperature with high precision. Furthermore, the temperature estimation using EKF is more accurate than the estimates of the thermoelectric model.