Key words
global optimization, random search, function, simulation, search strategies
articles
The author:
Gribanova E. Published in
№ 2(68)
29 april 2017 year
Rubric: Algorithmic efficiency
Rubric: Algorithmic efficiency
A stochastic algorithm for finding the global minimum of the function
Optimization problems have received attention in different research areas. This article provides an
overview of the methods of searching for the global minimum. The paper presents a new algorithm for
solving one-dimensional function optimization problem. The algorithm employs random variables and
events. The algorithm generates two types of elements: search and exploration. The elements of the
first type strive to explore the unexplored areas. The probability of failing into each interval depends
on the distance between the points. The elements of the second type examine the areas where the objective
function takes the best value. The probability of failing into each interval depends on the objective
function value. The ratio of searching elements depends on the task. If we need to find all local
minima, it is better to use elements of the search. To obtain a more accurate solution, the number
of exploration elements is increased. Solution to two examples described. Results of computational
experiments comparing the presented algorithm with other known algorithms are presented. The result
showed that the solution could be found for a smaller number of steps compared with the simple
random search algorithm.
The author:
Gribanova E.
Degree:
PhD in Technique, Tomsk State University of Control System and Radio Electronics
Location:
Tomsk
