Real-time robot motion control with circulatory fields

Abstract

This paper introduces a new feedback algorithm for steering a point robot through an obstacle field. The key innovation is the use of a circulatory field to rotate the robot path around the obstacles instead of the common potential field which repels the robot. This idea is motivated by a charged particle in a magnetic field generated by a current flowing around the obstacle. In constrast, the potential field approach is associated with a repulsive static electric field generated by charges of the same polarity as the robot, on the obstacle. The circulatory field does not generate any spurious local minimum as it does not change the total energy of the system. By combining with an attractive potential field associated with the desired destination, this method achieves global convergence while avoiding collisions with obstacles.

Key Aspects of PSO-Type Swarm Robotic Search: Signals Fusion and Path Planning

“…Aiming at the problem solving, some modified APF methods are proposed to overcome local minimum. The key ideas may be fallen into two categories: one establishes new potential functions with a few or even no local minima (Ge & Cui, 2000;Warren, 1990); the other uses some techniques to escape from local minima, including random walk (Janabi-Sharifi & Vinke, 1993), wall following (Borenstein & Koren, 1989), and other heuristic methods (Singh et al, 1996). Except for these efforts, Jugh et al apply the PSO algorithm to path optimization of multiple robots ).…”

Equilibria and steering laws for planar formations

“…(A preliminary formulation of a continuum model appears in [8].) We are also studying obstacle avoidance, where our approach to formations based on gyroscopic forces leads to a natural approach to obstacle avoidance based on gyroscopic forces [2,15,21].…”

Solving the potential field local minimum problem using internal agent states

“…Although the resulting potential field again does not have local minima, off-line computation is required to provide a solution to the Laplace equation. The solenoidal field method [19] replaces repulsive forces by magnetic-type forces that lead the agent to a path along which it can follow the obstacle boundaries.…”