**Author(s): ** Hernando Castañeda M. |

Wladimir Rodríguez G. |

Eliézer Colina M.**Journal: ** Avances en Sistemas e Informática ISSN 1657-7663

**Volume: ** 4;

**Issue: ** 1;

**Start page: ** 73;

**Date: ** 2007;

VIEW PDF DOWNLOAD PDF Original page**Keywords: ** Dynamic System |

Fuzzy Logic |

Application of Intelligent System**ABSTRACT**

In Molecular Dynamic, successive configurations are generated by integrating Newton’s law of motion; the resulting trajectory specifies how the position and velocities of the particles in the system move with time. The expensive part is the calculation of forces on each particles from current positions, based on the force field, when MD using simples models all collisions are perfectly elastic and occur when the separation between the centers of the particles equal to point of discontinuity in the potential. When is using with continuouspotentials, the force on each particle will change whenever theparticle change its position or whenever any of the other particleswith which it interacts change position, the motions of all particles are coupled together, giving rise to a many body problem that cannot be solved analytically; finite difference methods has to be used. The main task in the proposed method is to generate a group of data of training by means of the accumulation of the functions of potential of Lennard Jones, to extract the structural features starting from its trajectories and a number of dynamic objects segment in a small clusters number, in such a way that the objects in each cluster are in the most possible thing similar and the objects in different clusters are the less similar ones, allowing to predict the behavior of the exit variables. The goals is build an automated system to capture important events such as defect disintegration and defect amalgamation but goals initial is to understand the interaction among defect usingdynamic fuzzy pattern recognition. The task of clustering methods is to partition a number of objects into small numbers of homogeneous clusters so that objects belonging to any one the clusters would be as similar as possible and object of different clusters as dissimilar possible. The most important problem arising in this context is the choice of a relevant similarity measure, which is use for definition of the clustering criterion.

## You may be interested in: