Multiple phases modified termite life cycle optimizer for data clustering and engineering optimization

Termite life cycle optimizer (TLCO) is a recently introduced swarm intelligence metaheuristic algorithm inspired by the intricate behaviors exhibited by termite colonies, demonstrating competitive performance compared to other state-of-the-art algorithms. However, the original TLCO encounters challenges related to unbalanced exploration and exploitation, low convergence accuracy, and premature stagnation of iterations in high-dimensional complex applications. To address these issues, this study introduces an enhanced variant of the termite life cycle optimizer, referred to as the modified termite life cycle optimizer (MTLCO). The MTLCO introduces multiple novel phases for optimizing complex computational problems. Key among these are the Best Agent Guide Phase and the transition factor (TF) phase, which aim to steer the optimization in a more directed manner. Additionally, the Control Randomization Value and Direction serve to introduce randomness with precision, ensuring diversity in the solution space. The Phasor Operator Phase, an innovative mechanism, aids in further improving the convergence rate. Taking cues from other meta-heuristic paradigms, a new phase rooted in the Whale optimization algorithm (WOA) strategy has been incorporated, providing a more adaptive search mechanism. To counter stagnation and premature convergence, a Restart Strategy has been devised, facilitating a fresh search whenever required. For comprehensive validation, the enhanced MTLCO has been integrated into the Bonobo Optimizer and evaluated using the CEC’20 benchmark functions. Furthermore, its effectiveness is underscored by promising outcomes in five intricate engineering problems and 15 data clustering challenges. Comparative analyses with conventional methods affirm the superior performance, robustness, and versatility of the proposed MTLCO.