Control of diversity in genetic algorithms using multimodal strategies

Abstract

An optimization process is a kind of process that systematically comes up with solutions that are better than a previous solution used before. Optimization algorithms are used to find solutions which are optimal or near-optimal with respect to some goals, to evaluate design tradeoffs, to assess control systems, to find patterns in data, and to find the optimum values (local or global) of mathematical functions. A genetic algorithm is one of the optimization techniques. In this way, a heuristic search that is inspired by Charles Darwin’s theory of natural evolution. This algorithm reflects the process of natural selection where the fittest individuals are selected for reproduction in order to produce offspring of the next generation which are population algorithms that emulate behavior similar to Darwinian natural selection. Taking into account these issues, this article shows the performance of a genetic algorithm designed, which allows to find several minimums within a function from the control of population diversity. To perform the tests, the algorithm with four different functions was used, with the particularity of having several minima with the same value. Proposed strategy was compared with a conventional genetic algorithm, the result was the conventional one can only find some of the minimums of the function and sometimes only one, while the proposal finds most of the minimums