Most applicable metaheuristic algorithms require additional control parameters except for the terminal condition and population size. Adjusting these control parameters to obtain the best possible answers for unknown optimization methods is a major challenge. The present work introduces the comprehensive learning Rao algorithm (CLRAO), a novel metaheuristic method. This is a new version Rao algorithm that uses a comprehensive learning method to improve the global search capabilities of Rao algorithms and increase the convergence speed. Proposed algorithm uses three basic candidate operators to update individual positions. Comprehensive learning Rao algorithm's performance is studied using twenty-five standard benchmark problems and four engineering optimization tasks: compression or tension spring, I-beam, gear train and inverse kinematics of inchworm robot. The Friedman test is applied to validate the effectiveness of the suggested algorithms. The suggested algorithms are more successful and resilient than the other optimization methods examined by earlier researchers to tackle standard benchmark functions and complicated engineering design problems based on the comparison of results.
展开▼