🚀 go-pugleaf

In article <102392@netnews.upenn.edu> sarkar@grip.cis.upenn.edu (Nilanjan Sarkar) writes:
>I am looking for pointers towards merging algorithms of mobile robots.

Your problem seems to involve reactive path planning, so I'll give a
pointer to some of the work going on at this lab:

  Connolly CI, Grupen RA, "Harmonic Control", Proc. 1992 International 
  Symposium on Intelligent Control, August 1992, IEEE

  Connolly CI, "Applications of Harmonic Functions to Robotics",
  Proc. 1992 International Symposium on Intelligent Control,
  August 1992, IEEE

There are others, but these are the best to start with.  In robotics
parlance, it's a "potential field" approach -- but it's based on the
classical potential (Laplace's equation).  The basic idea is an appeal
to incompressible fluid flow -- hence no local minima.  The digital
implementations we use are adequate for real-time planning for a
Denning robot, and for low-resolution control/planning for a P-50 arm.

The method can also be implemented using a resistive network.  This
sort of an implementation can provide reaction to dynamic obstacles on
the order of milliseconds, if not microseconds.  Another convenient
aspect of the approach is that the resulting paths are almost always
smooth.  The two papers above describe how the method can be used with
a dynamic environment map, and they contain other pointers to the
literature.
-- 
	-	-	-	-	-	-	-
Christopher Ian Connolly			connolly@cs.umass.edu
Laboratory for Perceptual Robotics		wa2ifi
University of Massachusetts at Amherst		Amherst, MA 01003