Enhancing Chimp Optimization Algorithm Using Local Search Capabilities and Machine Learning for Real Engineering Problems

The Chimp Optimization Algorithm (ChOA) has emerged as a highly efficient optimization technique, demonstrating its prowess across diverse problem domains.However, its reliance on local search methods presents vulnerabilities, such as diminished diversity, susceptibility to premature convergence, and local minima.Thus, this study proposes two versions of enhancement the basic version of ChOA.The first version called ChOAO, integrates Opposition-based learning (OBL) to foster superior solution selection.The second version called ChOAORL, utilizes the concept of Reinforcement Learning (RL) to enhance the local search capabilities of ChOAO.It also effectively mitigates the risk of trapping the algorithm in local optima.The proposed versions are assessed utilizing the Friedman rank test on two sets of benchmark functions CEC 2017 and real-world problems CEC 2011.The results illustrate that ChOAORL achieved the best rank using CEC 2017 in both dimensions, 10 with a 1.48 mean rank and 30 with a 1.42 mean rank.Also, it outperformed other similar algorithms in terms of convergence precision and stability in all CEC 2011 real problems.