Adaptive Improved Flower Pollination Algorithm for Global Optimization

In the last few years, meta-heuristic-driven optimization algorithms have been employed to solve several problems since they can provide simple and elegant solutions. In this work, we introduced an improved adaptive version of the Flower Pollination Algorithm, which can dynamically change its parameter setting throughout the convergence process, as well as it keeps track of the best solutions. The effectiveness of the proposed approach is compared against with Bat Algorithm and Particle Swarm Optimization, as well as the naive version of the Flower Pollination Algorithm. The experimental results were carried out in nine benchmark functions available in literature and demonstrated to outperform the other techniques with faster convergence rate.