Globally Multi-Objective Optimization of Shell and Tube Heat Exchangers

This study presents the development of a multi-objective adaptation of the set trimming method (multi-objective set trimming [MOST]), designed as a reliable and globally efficient optimization approach. Then, its application to shell and tube heat exchangers is presented, and the results are compared with those of the fast and elitist nondominated sorting genetic algorithm (nondominated sorting genetic algorithm II [NSGA-II]) and exhaustive search in terms of both computational time and convergence. Effectiveness and annual cost are considered as the fitness functions, and the heat exchanger geometric parameters, including eight design parameters, are selected. The optimization results revealed that the optimum Pareto fronts in the case of NSGA-II are totally dominated by the results obtained using MOST in all studied cases. In addition, using MOST, the effectiveness of the final optimum solution is improved by 0.94–2.10% as compared with NSGA-II. On the other hand, as well as effectiveness, the annual cost is also improved simultaneously as compared with NSGA-II, with the percentage in the range of 1.07–5.41%.