Random exploration approach for automatic chamferless insertion

The stochastic approach strategy to realize the robotized insertion of low-clearance, chamferless parts is studied in both the analytical and experimental contexts. The analytical approach is discussed in terms of stochastic differential equations that involve Gaussian white and colored noises processes to model a planar radom search. Special attention is devoted to characterize the time required for the insertion, a random variable whose first moment calculation (i.e., the mean) is dealt with. In the mathematical modelization context adopted, it is remarkable that the calculated mean mating time grows slowly (i.e., logarithmically), with the precision required to perform an insertion. The theoretical results are validataed on a robotized assembly system, also presented in this article. In this experimental system, the random movements are generated by pseudorandom binary sequences that, for the time scale considered to substain the logarithmic behavior obtained analytically. Hence, in addition to its simiplicity and flexibility, the random strategy approach appears to be very efficient when high mating precision is required.


Published in:
International Journal of Robotics Research, 14, 2, 161-173
Year:
1995
Publisher:
MIT Press
ISSN:
02783649
Keywords:
Note:
Cited By (since 1996): 1
Export Date: 6 December 2012
Source: Scopus
CODEN: IJRRE
Language of Original Document: English
Correspondence Address: Hongler, M.O.
Other identifiers:
Scopus: 2-s2.0-0029293848
Laboratories:




 Record created 2013-01-07, last modified 2018-04-20


Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)