Real-time modeling and prediction of physiological hand tremor

This paper provides results of our investigation of adaptive algorithms for real-time modeling and prediction of physiological tremor signals. The goal of the investigation is to find effective methods which are suitable for small embedded systems used in real-time signal processing of physiological hand tremor. Our investigation shows that physiological hand tremor can be represented with an AR(3) process. Based on this fact, it is possible to identify the model of the physiological hand tremor with simple adaptive algorithms. On the other hand an adaptive linear predictor can also be realized, based on the same adaptive algorithms, which is significant for the suppression of physiological hand tremor in hand-held tools.