Improved Flower Pollination Algorithm Based on Mutation Strategy

A new metaheuristic intelligent algorithms-flower pollination algorithm (FPA) and a novel differential evolution mutation strategy-Target Mutation(TM) strategy are introduced. An improved FPA based on mutation strategy-MFPA algorithm, is proposed to overcome the low accuracy computation, low speed convergence, and easy to fall into local optimization. The MFPA improves the TM, and introduces the strategy to FPA local search process to enhance local development ability of the algorithm, and introduces the random mutation operator to FPA global searching process to enhance the global exploring ability of the algorithm. Finally through the test functions to test, the results show that MFPA algorithm optimization ability is superior to the original FPA, particle swarm optimization algorithm and bat algorithm.