Mimic

摘要

Advanced robotic arms have started seeking applications way beyond industrial use in the recent times. Use of these arms in exacting areas like military and medicine will not only extend precision and accuracy to various practices, but also help in saving more lives. This paper is an attempt to propose an algorithm to control robotic arms that are able to mimic the movement of the human arm and also add more precision to a movement by eliminating natural unsteadiness and external noise. The proposed method has been tested on a robotic arm using servo motors which work in accordance with a sensor setup worn on the human arm. The algorithm can be broken down into two steps. In the first step, the readings from the sensors are converted to movement in the robotic arm using an estimation algorithm that uses polynomial regression to predict motion on the basis of certain predefined set of values. The removal of noise and unsteadiness is carried out using a refinement algorithm to eliminate errors of various kinds in the second step.