A Partition-Based Random Search Method for Multimodal Optimization

Practical optimization problems are often too complex to be formulated exactly. Knowing multiple good alternatives can help decision-makers easily switch solutions when needed, such as faced with unforeseen constraints. A multimodal task aims find global optima well high-quality local of an problem. Evolutionary algorithms niching techniques commonly used for problems, where a rough estimate the number is required determine population size. In this paper, partition-based random search method proposed, in which entire feasible domain partitioned into smaller and subregions iteratively. Promising regions faster than unpromising regions, thus, promising areas will exploited earlier areas. All parallel, allows found single run. The proposed does not require prior knowledge about it sensitive distance parameter. By cooperating refine obtained solutions, demonstrates performance many benchmark functions optima. addition, numerous optima, captured low-quality