No abstract

Self-sensing actuation of shape memory alloy (SMA) means to sense both mechanical and thermal properties/variables through the measurement of any internally changing electrical property such as resistance/inductance/capacitance/phase/frequency of an actuating material under actuation. The main contribution of this paper is to obtain the stiffness from the measurement of electrical resistance of a shape memory coil during variable stiffness actuation thereby, simulating its self-sensing characteristics by developing a Support Vector Machine (SVM) regression and nonlinear regression model. Experimental evaluation of the stiffness of a passive biased shape memory coil (SMC) in antagonistic connection, for different electrical (like activation current, excitation frequency, and duty cycle) and mechanical input conditions (for example, the operating condition pre-stress) is done in terms of change in electrical resistance through the measurement of the instantaneous value. The stiffness is then calculated from force and displacement, while by this scheme it is sensed from the electrical resistance. To fulfill the deficiency of a dedicated physical stiffness sensor, self-sensing stiffness by a Soft Sensor (equivalently SVM) is a boon for variable stiffness actuation. A simple and well-proven voltage division method is used for indirect stiffness sensing; wherein, voltages across the shape memory coil and series resistance provide the electrical resistance. The predicted stiffness of SVM matches well with the experimental stiffness and this is validated by evaluating the performances such as root mean squared error (RMSE), the goodness of fit and correlation coefficient. This self-sensing variable stiffness actuation (SSVSA) provides several advantages in applications of SMA: sensor-less systems, miniaturized systems, simplified control systems and possible stiffness feedback control.

No abstract

No abstract

Self-sensing actuation (SSA) assists in sensing the vital property of the shape memory coil which can be used to monitor and control the actuation. The stiffness characteristic of the shape memory coil is sensed during actuation which plays a significant role in development of Intelligent Robotics in defense systems. The electrical property of shape memory coil such as electrical resistance changes due to martensitic phase transformation which is further used to sense the mechanical properties such as strain, stress, temperature, length, and force. Nowadays electrical properties are used to sense the stiffness of the shape memory coil. As of now, there is no well-established analytical model to predict the stiffness of sensing during actuation accurately. Therefore, Machine Learning (ML) based data-driven intelligent model is proposed in this paper for auto-sensing of the stiffness. The experimental facility has been developed for the collection of data with respect to diverse Joule heating currents. To determine the experimental data values of stiffness and electrical resistance of shape memory coil is a cumbersome task. Hence we have proposed an automated method to predict the stiffness of the shape memory coil using ML methods. The Classical Polynomial and Feedforward Neural Network (FFNN) models are developed for analyzing the stiffness of the shape memory coil. It is found that FFNN model outperforms the other ML based model by attaining 95.2650 % accuracy. The FFNN model is also able to explain almost all the predicted stiffness values which are experimentally recorded. The FVU (Fraction Variance Unexplained) statistical parameter explains the prediction of FFNN with the value of 0.0842. The great advantage of the ML model is to replace two sensors (Force and displacement sensors) with one soft sensor (ML model). It will be useful in the controlling robotics and other devices which require high precision in data generated by the sensors.