A heuristic algorithm with a dynamic generation strategy for optimizing energy systems

Abstract

The generation of new heat exchangers during optimization is the main driving force for the structural evolution of the heat exchanger networks. However, the new structure formed by randomly generating heat exchanger using a basic algorithm has only a limited reduction in the target cost, and this aspect has rarely been considered in previous studies. Therefore, this study proposes a new heuristic algorithm with a dynamic generation strategy, to improve the structure evolution capability of heat exchanger networks. The split ratio accompanying the generation strategy based on the concept of the dynamic equilibrium point is established, which improves the generation performance of new heat exchanger. To address the problem of heat exchangers with coupling relationship that are easy to be eliminated in the later stage of optimization due to relatively small heat load, an incentive heat load strategy is designed to coordinate with the former method to provide a certain opportunity for the new heat exchanger to be allocated a large heat load. Compared to the optimal results listed in the literatures, the results of three industrial network cases obtained by the proposed method are 1,711,806 $/yr, 6,665,342 $/yr, and 1,385,391 $/yr. This demonstrates that the proposed scheme improves the global search ability of the basic algorithm, promoting individual structure disturbance and competitiveness of new heat exchange units.