Proposing an Improved Version of the Bat Algorithm

The bat algorithm is an example of meta-heuristic algorithms from the collective swarm intelligence, which is based on the echolocation behavior of bats. This algorithm preserves the diversity of the solution by using a frequency tuning method that can quickly and efficiently shift from exploration to exploitation. Therefore, when a fast and accurate solution is needed, this algorithm becomes an efficient optimizer for any application. Although the bat algorithm has many practical benefits, it also has some disadvantages. One of these disadvantages that reduces its efficiency is being trapped in the local optimum. To solve the mentioned problem in this research, the position and speed of the initial population is updated in three ways with different formulas, this makes the final answer of the problem not trapped in the local optimum and diversity occurs in the population. In this article, the performance of the improved bat algorithm on 11 sample objective functions has been investigated and compared with other similar algorithms, and finally the results show the superiority and accuracy of this algorithm compared to similar samples.