Modeling a quasi-passive bipedal walker: when and where to kick

Abstract

Walking is one of the most complex tasks a human being can perform. It takes several years to develop the patterns that enable a person to walk and run. Although it is now possible to replicate these patterns in machines such as robots, actuators are needed to control the pace and compensate for the energy loss of walking. However, in small robots (in the scale of centimeters) the size and weight of the actuators could be a limiting factor for their use. To solve this problem, quasi-passive walkers have been recently proposed. These walkers consist of several coupled rigid bodies, usually forming a biped, that move on flat surfaces in the presence of a gravitational field. It is well known that these bipeds can descend on small slope planes without the need for actuators. Walking on horizontal planes requires the presence of actuators to initiate and maintain the pace, however, these actuators could be small enough to ”kick” the walker at the right point at the right time. In this work, we present a physical-mathematical model of a bipedal walker composed of five rigid bodies (two legs, two feet, and the hip) moving in a horizontal plane. The model focuses on the natural motion of the biped under certain initial conditions with the ultimate goal of determining the best position of the actuators and the actuation time to maintain the gait. This model could be used as a guide for the construction of small laboratory-scale walkers and as a teaching tool in biomedical engineering courses.