Parallel optimization route promoted by accepting imperfect solutions for the global optimization of heat exchanger networks

The global optimization of heat exchanger network synthesis remains a hotspot and challenge in the field of chemical process integration. The random walk algorithm with compulsive evolution has shown good performance in avoiding entrapment in the local optima by accepting imperfect solutions with a certain mutation probability. This work first investigates the detailed effects of accepting imperfect solutions on the continuity of optimization and the efficiency of global optimization by analyzing the imperfect solutions generated in the optimization process. Then, a novel parallel optimization route is established to balance the global and local search ability by combing basic and fine-search optimization levels. Additionally, enhancing strategies promoted by accepting imperfect solutions and large step lengths are integrated into the parallel optimization route to further improve the efficiency of structure evolution and global optimization. Finally, three case studies are presented to verify the effectiveness of the proposed method and discuss the roles of each module and different module combinations in facilitating the global search. Many promising results with more structure possibilities are obtained in each case study, with the obtained optimal solutions being lower than most reported in the literature. This indicates the effectiveness of the proposed method in facilitating the structure evolution and global optimization for heat exchanger network synthesis.