سنتز شبکه مبدل حرارتی چند مرحلهای کارآمد با استفاده از رویکرد متا-اکتشافی

Multiperiod heat exchanger networks (HEN) are required in plants with seasonal alterations to operating conditions. Like for single-period HEN, the synthesis of multiperiod HEN can be formulated as a mathematical programming optimization problem. However, since the network needs to feasibly perform heat integration under different process conditions, additional constraints are required and problem complexity is increased. Studies on the subject based on mathematical programming often use deterministic approaches and rely on commercial solvers. In this work, a meta-heuristic two-level method based on Simulated Annealing and Rocket Fireworks Optimization (SA-RFO), originally developed for single-period HEN synthesis, is adapted to handle multiperiod HEN optimization. A new post-optimization (PO) strategy is coupled with the main method in order to improve the results. Four case studies are investigated and results are compared to the literature. The solutions achieved presented lower total annual costs (TAC) than those obtained by other methods and the new PO scheme was able to significantly improve the results.