Empowered Cat and Mouse Optimizer for Exploring Quasi Optimal Solutions in Optimization Problems

Abstract

Population-based optimization algorithms are essential for solving a wide range of optimization problems encountered in various scientific fields and real-world applications. This study introduces a modified version of the Cat and Mouse Based Optimizer (CMBO) known as Empowered CMBO (E-CMBO), aimed at global optimization. By considering the population proportions of mice and cats and incorporating additional formulas for mouse movements to evade cat attacks, this adjustment significantly reduces the execution time of the CMBO algorithm. Experimental results demonstrate the effectiveness of E-CMBO in addressing diverse optimization problems, particularly those posed by unimodal and high-dimensional multimodal objective functions. E-CMBO consistently outperforms CMBO by producing fitness values closer to the minimum and achieving faster convergence rates across different scenarios. While it may exhibit slightly reduced efficacy in fixed-dimensional multimodal functions in one scenario compared to CMBO, overall, E-CMBO demonstrates robust performance across various problem landscapes. The advantages of E-CMBO lie in its simplicity, high robustness, fast convergence, and fewer parameters. This study highlights E-CMBO as a promising optimization tool, offering significant advantages in convergence speed and solution quality, particularly for complex optimization problems.