Reactive obstacle avoidance and control action continuity. application to the ND algorithm

Abstract

Endowing a physical agent with the capability of moving in a safe and agile manner is an issue of capital importance during the development of these systems. In this context, our work is related to obstacle avoidance in general, and specifically to ND algorithms. Contrary to many previous methods, the ND approach is not aimed at devising a general motion law, but operates over a reduced set of possible situations that are treated by its particular motion law. The big earning of this idea is that it eases the design of control, as now motion laws are specific to every identifiable situation. However, it also raises new issues as nothing guarantees the smoothness of motion commands when the diagnostic changes. Extensive experimentation with an implementation of this method has revealed that often negligible changes in the perceived environment induce changes in the identification of the situation that have as a consequence large modifications of commanded velocities. These unnecessary and drastic changes in the commanded velocities seriously degrade the coherence and agility of the agent movement. In this paper a modification of the ND approach is presented, in order to improve this aspect of the method, along with experimental results.