Аңдатпа
A new population-based metaheuristic algorithm named Urutau Optimization Algorithm (UOA), inspired by the natural behaviour of the Urutau bird in the wild, is introduced in this paper.First, the ethology of this bird is investigated, and its characteristic behaviours are extracted to design the algorithm structure.Then, the theory of the UOA algorithm is fully explained and its mathematical structure is modelled in three main parts including initialization, exploration phase, and exploitation phase.The exploration phase in UOA is designed based on the simulation of the multidirectional environmental monitoring behaviour of the Urutau; a behaviour in which the bird identifies potential opportunities by examining different areas of the environment.Also, the exploitation phase is designed based on the simulation of the precise and targeted attack behaviour of this bird on the prey to perform the local search process with high accuracy.The efficiency of the proposed UOA algorithm in optimization applications is evaluated on 23 standard benchmark functions including unimodal and multimodal functions.The results of the unimodal function optimization show that UOA has a very high ability in local search, exploitation, and accurate convergence towards the optimal solution.Also, the results of the high-dimensional multimodal test functions show that the algorithm can provide successful performance in global search by maintaining population diversity and preventing premature convergence.In addition, the results of the fixed-dimensional multimodal test functions show that an effective balance between exploration and exploitation is established and that UOA can provide stable and competitive performance in complex multi-vertex problems.The results are compared with 9 well-known metaheuristic algorithms to evaluate the performance quality of UOA.The simulation results show that UOA has achieved the first best optimizer rank in 19 out of a total of 23 benchmark functions, providing a successful performance of 82.60% in competition with the compared algorithms.Also, qualitative analysis of convergence curves shows that UOA has a fast, stable, and uniform convergence trend and can avoid getting stuck in local optima during the search process.The findings of this study show that the UOA algorithm, with its high ability in exploration, exploitation, and balancing between the two in the search process, provides successful and acceptable performance in various optimization applications and can be used as an effective tool for solving complex optimization problems.