Generalized pigeon-inspired optimization algorithm for balancing exploration and exploitation

Abstract

<p indent="0mm">A generalized pigeon-inspired optimization (GPIO) algorithm for balancing the exploration and exploitation abilities is proposed herein. The traditional pigeon-inspired optimization algorithm includes two operators, namely the map and compass operator and the landmark operator. These two operators are implemented only for one round at a single run. In the GPIO algorithm, the search process is divided into multiple stages, and two operators are implemented in each stage. These two operators are implemented for multiple rounds at one single run. The map and compass operator focuses on the exploration ability, while the landmark operator focuses on the exploitation ability. The GPIO algorithm changes the execution order of the two operators without additional objective function evaluation. Moreover, the structure of the solutions and the parameter settings are extended in the GPIO algorithm, which is beneficial to search quality improvement. The simulation results show that the GPIO algorithm improves the search efficiency and the search results of the algorithm.