Resumen de Application of the bio-inspired internal model concept to control of a bipedal robot
E.A. Turowska, A. Forner Cordero, Juan Camilo Moreno Sastoque, Juan A. Gallego Juárez, A. Montellano López, José Luis Pons Rovira
The stability of gait is one of the most important issues in humans and in biped robots. In this respect bioinspiration can be of use to improve bipedal robot design. Here we test if the internal model concept, borrowed from Biology, can be applied to control a biped walker, the ESBiRRo. Two models of di®erent complexity were tested in order to obtain the minimal complexity model that still can provide results with suffcient accuracy. These models are tested in the prediction of the configuration of the biped in the next Double Stance phase after a perturbation occurrence, in order to select the adequate recovery strategy. It is shown that the predictions obtained for the simpler model with 3 links were precise enough. This justifies the choice of using this simpler model in order to minimize the computational calculation costs.
