Design optimization of process plants using real coded genetic algorithm

The application of artificial intelligence based search and optimization algorithms is an active research area in many engineering fields. In this work, real coded genetic algorithm, a modified genetic algorithm is used for the optimal design of industrial process plants. This technique has been implemented for the optimal design of reactor network, Williams-Otto process plant and multiproduct process plant. The modification to simple genetic algorithm, using real coded representation of variables along with specifically designed mutation and crossover operator, yields accurate results for the problems considered in this paper, at a lesser computational effort. The results have been compared with the conventional and global optimization techniques. The simulation study clearly demonstrates that the proposed real coded genetic algorithm is practical, robust, and reliable optimization technique for the design of industrial process plants.