Improved Parallel Bat Algorithm Based on a Communication Strategy.

Parallel processing is an effective measure of inefficient and effective computations of function optimization. Now we propose a new communication strategy for the parallel Bat Algorithm to solve numerical optimization problems. Firstly, the population of bats is divided into several independent groups, and they are independent. With every fixed number of iterations, different groups will exchange information and update. We use benchmark functions to test accuracy convergence behavior. From the analysis and summary of the experimental results, we get the following conclusions. The communication strategy improves the accuracy of BA in finding the best solution. The algorithm has improved significantly.