STEP PLANNING ALGORITHM FOR N-LEGGED WALKING ROBOT WITH SMOOTH DIRECTION CONTROL

Abstract

Precise step-planning algorithm is important when we want to drive an n-legged walking robot. Probably the most widely used precise control system is based on neural networks, but those require a lot of computing power. The main goal of this paper is to propose a computationally efficient method for precise and smooth movement of a robot. An algorithm capable of computing end-step position based on movement around a circle was developed. It can be driven by control vector or (with minor modification) by a radius and a tangent. It is simple to implement, adjust and debug, and needs only a small amount of computational power. The proposed algorithm only generates the end points of each step. It can be used as a basis for more complex solution and can be also used in a system with a large number of legs. The movement of body, the kinematics of mechanical parts and the manipulation of center of mass must be solved by other algorithms.