Abstract
The chest compression is one of the most important actions for Cardio Pulmonary Resuscitation (CPR). Because many people will be survived by the CPR in the early-stage of cardio-respiratory arrest, it is expected the chest compression become widely known through training. The chest compression is accompanied with an internal force and it is difficult for trainees to understand the required force only from the demonstrations. In this paper, we develop an enhanced motion that enables the trainees to yield the required force for the chest compression. The skilled motion is captured and modeled by an autonomous control system containing the information of human motion. Based on the decomposition of the controller, Pseudo-reference is designed. Through the training with/without Pseudo-reference, the effectiveness of the enhanced motion is evaluated.
