Optimization of numerical and engineering problems using altered differential evolution algorithm

Abstract

In this study, an altered differential evolution (ADE) is presented for numerical and engineering problem optimization. It incorporates innovative mutation strategy with new control parameters using the perception of particle swarm optimization (PSO) process, to enhance exploration and exploitation activities extra profusely and increase the global search capacity. Also, a new crossover rate is employed in ADE, to attain higher convergence accuracy and quality optimal solutions. Finally, a novel selection strategy is introduced in ADE, to facilitate information sharing as well as for escaping local minima and keeps progressing. To investigate the suggested ADE performance, a collection of 13 classical benchmark functions, CEC2014 and CEC2017 benchmark suite are solved. Furthermore, the superiority and applicability of the ADE algorithm are further demonstrated through experimentation on six famous real-life engineering problems. The experimental and statistical test outcomes, collectively indicate that compared to other modern optimization algorithms, overall ADE exhibits superior performance. Also, comparison results show that ADE has powerful exploration and exploitation capabilities, excellent convergence performance, and strong ability for gaining high quality solution.