A bacterial gene recombination algorithm for solving constrained optimization problems

Creature evolution manifests itself in the improved ability of species to adapt to their surroundings. Swarm intelligence and gene optimization are found in the population of interacting agents that are able to self-organize and self-strengthen. In this study, a new gene-based algorithm for constrained optimization problems is proposed and called ''bac-terial gene recombination algorithm'' (BGRA). BGRA is inspired by the intelligent behavior of gene recombination in bacterial swarming. By referring to each phase of the well-regulated gene evolutionary process of bacteria, BGRA enables the exploration of problems and solution exploitation. For illustration, a set of constrained optimization problems are taken from the literature for testing purposes. In addition, satisfactory feasible solutions for constrained optimization problems were obtained by combining the BGRA and a penalty function method. Experimental results show that the proposed algorithm can yield near-optimal and stable solutions compared to the relevant literature, and thus, it can be an efficient alternative in the solving of constrained optimization problems. The collective intelligent behavior of insect or animal groups in nature has attracted the attention of researchers. Entomologists have studied these behaviors to model biological swarms, and engineers have applied these models as a framework for solving complex real-world problems. Several heuristic algorithms have been developed for solving optimization problems. Genetic Algorithm (GA) is a typical heuristic algorithm developed some time ago. GA was first established on a sound theoretical basis by Holland [1], who attempted to simulate the phenomenon of natural evolution. In natural evolution, each species searches for beneficial adaptations in an ever-changing environment. Another algorithm, ''particle swarm optimization'' (PSO), which simulates the social behavior of bird flocking or fish schooling was introduced by Eberhart and Kennedy in 1995 [2]. One more evolutionary algorithm introduced recently is called the ''differential evolution (DE) algorithm [3,4]''. DE has been proposed specifically for numerical optimization problems. The DE algorithm is a population-based algorithm and, like genetic algorithms, uses similar operators; crossover, mutation and selection. The main difference in constructing improved solutions is that genetic algorithms rely on crossover while DE relies on mutation operation. In this study, a bacterial gene recombination algorithm (BGRA) is proposed to solve constrained optimization problems. The proposed algorithm can be generally classified into two parts; natural gene combination and artificial combination. The former can enhance the convergence by promoting the ability of solution exploitation; the latter focuses on the capability of 0096-3003/$-see front matter Ó 2014 Elsevier …