A Multi-Strategy Improved Butterfly Optimization Algorithm

Butterfly optimization algorithm (BOA) is an excellent swarm intelligence algorithm that simulates the foraging behavior of butterfly populations, but it has such defects as poor accuracy, easy to stagnate at the local optimum, insufficient convergence and no balance between exploitation and exploration. Hence this paper proposes a multi-strategy improved butterfly optimization algorithm (MIBOA). By introducing four optimization strategies, including Bernoulli chaos map, dynamic switch probability, adaptive inertia weight and opposite-based learning, the above problems are effectively improved. And MIBOA is compared with other swarm intelligence algorithms on eight classical benchmark functions. The results demonstrate its excellent optimization performance and robustness.