An Improved Flower Pollination Algorithm with Three Strategies and Its Applications

摘要

The flower pollination algorithm is a recently presented meta-heuristic algorithm, but limited in searching precision and convergence rate when solving some complex problems. In order to enhance its performance, this paper proposes an improved flower pollination algorithm, combined with three strategies, i.e., a new double-direction learning strategy to advance the local searching ability, a new greedy strategy to strengthen the diversity of population and a new dynamic switching probability strategy to balance global and local searching. These strategies can increase searching precision and make solution more accurate. Then 12 standard test functions and two structural design examples are selected to appraise the performance of the newly proposed algorithm. The results show that our new algorithm has outstanding performance, such as high accuracy, fast convergence speed and strong stability on solving some complex optimization problems.