Interaction force transfer for characteristic evaluation of touch motion

The prediction and transfer of interaction force between operator and environment is an interesting and important issue. This paper proposed an interaction force of touch motion transfer method from operator to observer. Compared with the conventional implementation of force sensor, the desired force is predicted by electromyography signals detected from the operator. Given the different muscle-tendon condition, the experiments are divided into vertical gesture part and horizontal gesture part. Three muscles in the forearm are selected to record the EMG signals. A forearm muscle-skeleton model is developed and a dynamic equation is defined to predict the interaction force. To calibrate the coefficients of the dynamic equation, a least-squares-like method is implemented to find a local optimal solution. Compared with the other complicated parameter calibration methods, this method can provide sufficient prediction results for our particular case. The predicted interaction force is repeated by Phantom Premium on the observer´s side. The proposed interaction force transfer method is aimed for the purpose of characteristic evaluation of touch motion.