Global optimisation, Differential evolution, Control parameters, CoDE, Real-world problems, Experimental comparison

An adaptive variant of CoDE algorithm uses three couples of settings of two control parameters. These combinations provide well performance when solving a various type of optimisation problems. The aim of the paper is to replace original values of control parameters in CoDE to achieve better efficiency in real-world problems. Two different variants of enhanced CoDE algorithm are proposed and compared with the original CoDE variant. The new combinations of F and CR parameters are selected from results provided in a preliminary study where 441 various combinations of these parameters were evaluated. The results show that newly proposed CoDE variants (CoDE FCR1 and CoDE FCR2) perform better than the original CoDE in most of 22 real-world problems.