Perfumer Optimization Algorithm: A Novel Human-Inspired Metaheuristic for Solving Optimization Tasks

In this paper, a new metaheuristic algorithm called Perfumer Optimization Algorithm (POA) is introduced that mimics the behavior of a perfumer when making a perfume.The main idea of POA is derived from two basic perfumer strategies (i) making extensive changes to the perfume ingredients to achieve the desired fragrance and (ii) making small and precise changes to the perfume based on creativity and attention to the specific details of the perfume.The theory of POA is stated and then mathematically modeled in two phases of exploration and exploitation.The efficiency of POA to handle optimization applications is challenged to handle twenty-nine standard benchmark functions from the CEC 2017 test suite.The optimization results show that POA has been able to provide acceptable performance for optimizing the CEC 2017 test suite by balancing exploration and exploitation during the search process.In order to analyze the ability of POA in optimization applications, the obtained results are compared with the performance of nine well-known metaheuristic algorithms.Comparison of simulation results shows that POA has achieved better results in most benchmark functions, providing more successful performance compared to the compared algorithms.These results show that POA has a special efficiency for addressing optimization challenges in various sciences.