期刊名称：IOP Conference Series: Earth and Environmental Science
印刷版ISSN：1755-1307
电子版ISSN：1755-1315
出版年度：2019
卷号：304
期号：2
页码：1-8
DOI：10.1088/1755-1315/304/2/022062
出版社：IOP Publishing
摘要：In order to solve the multi-objective joint dispatching problem of watershed and reservoir, and overcome the shortcoming of single element heuristic intelligent algorithm, based on the analysis of the convergence, stability and applicability of the algorithm, an Improved Flower Pollination Algorithm (IFPA) combined with the reality of basin reservoir group is proposed. In IFPA, the elements in traditional Flower Pollination Algorithm (FPA) and the basic elements in game theory are mapped one by one. Two strategies (initializing the pollen operator with chaotic sequence and inverse learning transforming the population) are introduced into traditional FPA respectively and are gamed to optimize the problem. Function tests show that the convergence speed and optimization accuracy of IFPA are 15% and 10% higher than those of the traditional FPA respectively. IFPA is applied to the basin reservoir group dispatching and the actual dispatching results are consistent with the related literature. The study shows that IFPA is superior to the traditional FPA and can be used in multi-objective problem group joint dispatching of basin reservoir group.
