Adaptive Omni-directional Walking Method with Fuzzy Interpolation for Biped Robots

Abstract

Omni-directional walking is an important part for biped robots. Whereas, the problems of complex model, low stability, low flexibility, etc., existing in walking planning of biped robots, attract interests of many researchers. This paper proposes an adaptive omni-directional walking method with fuzzy interpolation (OW-FI) applied to a fast-stable omni-directional walking. A new separated omni-directional walking method consisting of straight walking, based on the improve Hermite interpolation, and rotation is designed to simplify the walking model and achieve a good key frame data selection. The mapping relationship between poses and actuators' angular speeds is then constructed by walking transforming model based on kinematics. In order to implement a stable and adaptive walking, a fuzzy approach to walking parameters adjustment is presented, where the step length and rotational speed of omni-directional walking are controlled fuzzily. The results of simulations and experiments demonstrate that the proposed method has better properties in adaptability and stability than the competing methods.

A survey of research on several problems in the RoboCup3D simulation environment

“…By designing the endpoint of the trajectory in advance [40][41][42], cubic spline interpolation can also be used to plan kicking action trajectories. Other interpolation methods such as Hermite interpolation have also been used in action development in RoboCup3D [43,44].…”