Quadrupedal mammal locomotion dynamics 2D model

摘要

Making a quadrupedal robotic machine walk is a problem with an infinity of solutions. Nevertheless, this question seems to have been solved successfully by mammals. Hence a particular approach in the design of walking robots has been considered, based on the animals movement characteristics. Indeed, studying animal walking can help produce a class of solutions towards the definition of quadrupedal robots. From this starting point, a model can be created according to a reduced number of criteria, notably the shape of the trajectory followed by the ankle or wrist during motion relatively to the hip or to the shoulder respectively. The problem lies in the difficulty to extract significant and valuable information from the mere observation of animals walking. This paper presents an experiment designed to extract angular variations data from the observation of a walking hedgehog and the use of these data to create a 2D dynamical simulation model to be used as an analysis basis.