An Improved Flower Pollination Algorithm Based on Logistic Chaotic Mapping and Natural Mutation
Zongliang Guo, Runze Suo, Xinming Zhang · 2022
Function optimization has always been a hotspot in the field of computing. It plays a crucial part in engineering application today. In recent years, many swarm intelligence optimization algorithms have been proposed and applied to solve function optimization problems. The traditional swarm intelligence optimization algorithm, taking flower pollination algorithm (FPA) as an example, has the problems of low accuracy, slow later-iteration convergence rate, and easy to fall into local optimal solutions. To solve these problems, a new flower pollination algorithm based on logistic chaotic mapping and natural mutation (LNFPA) is proposed in this paper. LNFPA uses logistic chaotic mapping to generate the initial population, cross operator to perform local search, and natural mutation operation to boost the search after each iteration. A large number of benchmark function tests have demonstrated that LNFPA has a more stable performance, faster iteration speed, and higher accuracy than FPA. In addition, this paper also analyses some shortcomings of LNFPA, and possible improvement ways in the future.