An integrated random walk algorithm with compulsive evolution and fine-search strategy for heat exchanger network synthesis

Highlights: • Fine-search strategy was proposed to improve the performance of RWCE. • Fine-search strategy can enhance the global search ability of continuous variables. • FS-RWCE could refine and enhance the evolution of integer variables in HENS. • FS-RWCE was demonstrated reliable and applicable to large-sized HENS problems. - Abstract: Heat exchanger network synthesis has been extensively studied in process system engineering for its complexity and difficulty resulting from stream matches and the nonlinearity of continuous variables. Stochastic methods have difficulties in finding the precise optimum solution on the near optimal regions and expanding the integer variables optimization in the late evolution. Therefore, a novel fine-search strategy was established on the basis of the evolutionary mechanism of random walk algorithm with compulsive evolution. The fine-search strategy was efficient in achieving the accuracy of solutions for a certain heat exchanger network structure. Then, the fine-search strategy and Random Walk algorithm with Compulsive Evolution were integrated to enhance and refine the optimization for continuous and integer variables in heat exchanger networks synthesis simultaneously. The integrated method could satisfy the needs of global and local search abilities for heat exchanger network synthesis. Finally, the proposed method was applied in three different-sized cases and more economical in contrast to the best results with no splits published thus far were obtained.