A Hybrid Spiral-Genetic Algorithm for Global Optimization

Abstract

Genetic algorithm (GA) is a well-known population-based optimization algorithm. GA utilizes a random approach in its strategy which inspired from a biological process of a chromosome alteration. Chromosomes which consists of several genes are randomly self-altered their own structure and also randomly combined their structure with other chromosomes. The unique biological process has inspired many researchers to develop an optimization algorithm. Yet, the algorithm still popular and is adopted as a tool to solve many complex problems. On the other hand, Spiral Dynamic Algorithm (SDA) is a relatively new population-based algorithm inspired by a natural spiral phenomenon. It utilizes a deterministic approach in its strategy. Movement of a search point from one location to another in a form of a spiral trajectory and relies on pre-defined parameters. However, both algorithms suffer a pre-matured convergence and tend to trap into a local optima solution. This paper presents an improved algorithm called a Hybrid Spiral-Genetic Algorithm. The algorithm is developed based on a combination of the wellknown GA and the SDA. The spiral equation of the SDA is adopted into the GA to enhance both exploration and exploitation of the original GA. The algorithm is tested with several benchmark functions of a single-objective algorithm and compared with the original SDA and GA. The result of the test shows that the proposed algorithm outperformed its predecessor algorithms significantly.