Multi-strategy synergy-based backtracking search optimization algorithm

Abstract

In order to solve the problems of backtracking search optimization algorithm with slow convergence speed and easy to fall into local optimum, an improved backtracking search optimization algorithm based on multi-strategy synergy is proposed. Foremost, a combination mutation strategy based on chaotic map and gamma distribution is introduced. The poor individuals are mutated to generate better quality individuals under a certain probability. Next, the global optimal individual information is introduced into the cross equation to guide the population update. Last but not least, a small habitat displacement method based on simulated annealing is designed. The poor individual is found by the niche radius, and the rich individuals are reconstructed by the global optimal individual information and the Gaussian distribution random function, and the convergence speed of the algorithm is improved. The simulated annealing algorithm is integrated on the niche technology to ensure the diversity of the new population and improve the convergence speed of the algorithm. In this paper, some standard test functions are selected, and numerical simulations are carried out in low-dimensional and high-dimensional states, compared with seven well-performing algorithms. The improved algorithm was analyzed by complexity, T test and ANOVA test. Simulation experiments on 20 standard test functions show that the improved algorithm has a faster convergence speed and higher convergence accuracy. Even in a high-dimensional multi-peak function, the convergence accuracy of the improved algorithm after the same number of iterations is 15 times higher than the original algorithm above the magnitude.