A NEW PROBABILISTIC ALGORITHM FOR SOLVING A CLASS OF SINGLE OR MULTI-OBJECTIVE OPTIMAL PROBLEMS

Abstract

We consider a class of single-objective optimization problems which haves  the character: there is a fixed number k (1≤kO_k. In this paper, we propose a new numerical optimization technique, Search Via Probability (SVP) algorithm, for solving single-objective optimization problems of the class O_k. The SVP algorithm uses probabilities to control the process of searching for optimal solutions. We calculate probabilities of the appearance of a better solution than the current one on each of iterations, and on the performance of SVP algorithm we create good conditions for its appearance. We tested this approach by implementing the SVP algorithm on some test single-objective and multi-objective optimization problems, and we found good and very stable results.