Bat algorithm with triangle-flipping strategy for numerical optimization

Bat algorithm (BA) is a novel population-based evolutionary algorithm inspired by echolocation behavior. Due to its simple concept, BA has been widely applied to various engineering applications. As an optimization approach, the global search characteristics determine the optimization performance and convergence speed. In BA, the global search capability is dominated by the velocity updating. How to update the velocity of bats may seriously affect the performance of BA. In this paper, we propose a triangle-flipping strategy to update the velocity of bats. Three different triangle-flipping strategies with five different designs are introduced. The optimization performance is verified by CEC2013 benchmarks in those designs against the standard BA. Simulation results show that the hybrid triangle-flipping strategy is superior to other algorithms.