Pivot method for global optimization

PABLO SERRA; A.F. STANTON; SABRE KAIS

PHYSICAL REVIEW E

AMER PHYSICAL SOC

Lugar: New York; Año: 1997 vol. 55 p. 1162 - 1165

1539-3755

A pivot algorithm for the location of a global minimum of a multiple-minimum problem is presented. Thepivot method uses a series of randomly placed probes in phase space, moving the worst probes to be near betterprobes iteratively until the system converges. The approach chooses nearest-neighbor pivot probes to searchthe entire phase space by using a nonlocal distribution for the placement of the relocated probes. To test thealgorithm, a standard suite of functions is given, as well as the energies and geometric structures of Lennard-Jones clusters, demonstrating the extreme efficiency of the method. Significant improvement over previousmethods for high-dimensional systems is shown.