Multi-strategy improved African Vulture Optimization Algorithm for global optimization and engineering design problems

In this paper, a multi-strategy improved African Vulture Optimization Algorithm (MIAVOA) is proposed address the drawback of premature convergence of the African Vulture Optimization Algorithm (AVOA). Firstly, the gaussian quasi reflection-based learning strategy is introduced, which improves the vulture initial population’s randomness and diversity. Then, the adaptive control strategy is used to enhance the search ability of the algorithm and avoid premature convergence. Furthermore, the elite candidate pooling strategy is designed in the exploitation phase, which expands the discovery fields for the optimal solution and reinforces the ability to escape from local optima. Finally, the formula of starvation factor is modified to balance the exploitative and explorative abilities of algorithm. MIAVOA is compared with seven state-of-the-art meta-heuristics on CEC 2022 and 23 classical test functions. It is observed that the proposed algorithm significant outperforms the other compared algorithms in terms of convergence and accuracy on the majority of benchmark functions. In addition, four engineering design problems and mobile robot path planning problem are utilized to evaluate the performance of MIAVOA. The experimental results demonstrate MIAVOA is effective and can achieve better applicability in real-world scenarios.