Chaotic aquila optimizer enhanced with elite opposite-based learning and variable search strategies

The Aquila Optimizer (AO) is a recently developed metaheuristic algorithm that has shown promising performance across various optimization problems due to its strong exploration ability. However, its limited exploitation capability can lead to premature convergence, particularly in complex and high-dimensional search spaces. This study proposes an Enhanced Chaotic Aquila Optimizer (ECAO) to address these limitations and enhance the robustness and adaptability of AO. The proposed improvements involve three key modifications: (1) the integration of chaotic maps into the position updating mechanism to promote search diversity and avoid local optima; (2) the application of elite opposition-based learning to enhance solution quality; and (3) a variable search strategy to strengthen the exploitation phase. Initially, five AO variants incorporating different chaotic maps are assessed using classical benchmark functions to identify the most effective configuration. The most successful version is then extended with the two additional strategies to construct the final ECAO. Comprehensive experiments are conducted on the CEC2019 and CEC2020 benchmark suites, as well as on the real-world problems of CEC2011. Comparative analyses with standard AO, state-of-the-art, and mainstream algorithms, supported by statistical tests and convergence graphics, demonstrate that ECAO consistently achieves superior accuracy, faster convergence, and greater robustness. These results highlight the innovation and practical value of ECAO for solving diverse and challenging optimization tasks. Although the ECAO algorithm requires relatively higher computational time, this additional cost remains at an acceptable level given the complexity of the optimization problems and the significant performance improvements achieved.