An effective and efficient fruit fly optimization algorithm with level probability policy and its applications

• A level probability policy and new mutation parameter are used for proposed FOA.
• Results of 29 test functions show LP–FOA outperforms the existing FOAs, DE and PSO.
• A delicate LP–FOA based coding is used to solve joint replenishment problem (JRP).
• LP–FOA is better than the current best intelligent algorithm for JRPs.

An improved fruit fly optimization algorithm (FOA) is proposed for optimizing continuous function problems and handling joint replenishment problems (JRPs). In the proposed FOA, a level probability policy and a new mutation parameter are developed to balance the population diversity and stability. Twenty-nine complex continuous benchmark functions are used to verify the performance of the FOA with level probability policy (LP–FOA). Numerical results show that the proposed LP–FOA outperforms two state-of-the-art variants of FOA, the differential evolution algorithm and particle swarm optimization algorithm, in terms of the median and standard deviations. The LP–FOA with a new and delicate coding style is also used to handle the classic JRP, which is a complex combinatorial optimization problem. Experimental results reveal that LP–FOA is better than the current best intelligent algorithm, particularly for large-scale JRPs. Thus, the proposed LP–FOA is a potential tool for various complex optimization problems.