Deceleration Convergence Strategy for Evolved Bat Algorithm

Evolved Bat Algorithm (EBA) is one of the optimization method in swarm intelligence published in recent years. However, the searching ability of the artificial agents are sometimes limited from its original design. To overcome this drawback, a mixture signal composed of a periodical signal and a level linearly decreased Direct Current (DC) signal is led into the process of the conventional EBA. The newly involved signal provides larger chance for the artificial agents to circle back to where it came from and exploit the region, again. In order to test the accuracy on finding the near best solutions, two test functions in four dimensional conditions with known global optimum are used in the experiments. The experimental results indicate that our proposed strategy improves the searching result of the conventional EBA about 54.11 percent in average.