Cultural Bat Optimization Algorithm: An Improved Variant of Bat Algorithm

Abstract

This paper presents a novel variant of the BAT algorithm called the Cultural Bat Algorithm CABAT. The new variants were achieved using situational and normative knowledge inherent in the cultural algorithm. The idea of the Bat algorithm is inspired by the intelligent echolocation behaviors of microbats. These intelligent behaviors are codified as an optimization agent for solving global optimization problems. Although this algorithm has been widely accepted due to its robustness, simplicity, and flexibility, it still suffers from an imbalance between exploration and exploitation. This research developed improved variants of the original Bat algorithm using cultural evolution processes to address these challenges Four new variants of the Bat algorithm were eventually developed using cultural situational and normative knowledge. The performance of the new variants was tested against the standard Bat algorithm using eleven commonly used benchmark functions. An extensive comparative study was conducted using statistical metrics computed over thirty independent runs. Results showed that all the CABAT variants performed better than the original bat algorithm.