Multi-stage thermal-economical optimization of compact heat exchangers: a new evolutionary-based design approach for real-world problems

Abstract

The complicated task of design optimization of compact heat exchangers (CHEs) have been effectively performed by using evolutionary algorithms (EAs) in the recent years. However, mainly due to difficulties of handling extra variables, the design approach has been based on constant rates of heat duty in the available literature. In this paper, a new design strategy is presented where variable operating conditions, which better represent real-world problems, are considered. The proposed strategy is illustrated using a case study for design of a plate-fin heat exchanger though it can be employed for all types of heat exchangers without much change. Learning automata based particle swarm optimization (LAPSO), is employed for handling nine design variables while satisfying various equality and inequality constraints. For handling the constraints, a novel feasibility based ranking strategy (FBRS) is introduced. The numerical results indicate that the design based on variable heat duties yields in more cost savings and superior thermodynamics efficiency comparing to a conventional design approach. Furthermore, the proposed algorithm has shown a superior performance in finding the near-optimum solution for this task when it is compared to the most popular evolutionary algorithms in engineering applications, i.e. genetic algorithm (GA) and particle swarm optimization (PSO).