Multi heap-based optimizer

This paper introduces the Multi Heap-Based Optimizer (MHBO), an enhanced version of the Heap-Based Optimizer (HBO), which is inspired by corporate rank hierarchies. While HBO employs a single heap to organize solutions and update their positions based on interactions with immediate bosses and colleagues, MHBO incorporates multiple heaps to improve exploration and convergence. Key enhancements in MHBO include a novel position-updating mechanism for the root solutions of each heap, allowing inter-heap interaction, and a heap-switching phase that facilitates solution exchange among heaps to diversify the search process. MHBO is evaluated comprehensively on 24 unimodal and 25 multimodal benchmark functions, originally used in the HBO studies. These benchmarks test its ability to navigate both simple and complex search spaces with multiple local optima. Results demonstrate that MHBO consistently outperforms HBO in terms of convergence accuracy and efficiency, highlighting its ability to exploit well and balance exploration and exploitation. The proposed framework offers a promising and adaptable solution for optimization problems, especially those with complex landscapes.