An Auto-tuning Method of the Population Size in Differential Evolution for Engineering Optimization Problems
Se Jung Lee · Korean Journal of Computational Design and Engineering · 2020
Differential Evolution (DE) is a widely used evolutionary algorithm among many metaheuristic algorithms to solve global optimization problems. In this study, an auto-tuning method of the population size in DE was proposed and adopted the previous methods for the crossover rate and the scale factor based on DE to solve engineering optimization problems with constraints. Five engineering optimization problems were selected from literature and the proposed algorithm was evaluated while comparing with the fixed population size. The proposed method shows both comparable convergence performance and possibility of generalization of DE method to handle various problems with ease.