Comparação de abordagens MOPSO no planejamento da operação de sistemas hidrotérmicos
Salvato in:
| Autore principale: | |
|---|---|
| Data di pubblicazione: | 2014 |
| Natura: | Master thesis |
| Lingua: | por |
| Fonte: | Repositório Institucional da UFG |
| Download full: | http://repositorio.bc.ufg.br/tede/handle/tede/3977 |
Riassunto: | The operation planning of hydrothermal systems is a complex, dynamic, stochastic, nonlinear and interconnected problem. In this study, we consider that this problem must tackle two objectives simultaneously: minimize thermoelectric generation (by maximizing the use of hydroelectric plants) and maximize water reservoirs’ level of hydroelectric plants. This dissertation presents the application of some multiobjective meta-heuristics, using a set of eight actual plants from Brazilian interconnected system in three periods of medium-term planning. The algorithms used were of two types: those based on particle swarms (MOPSO , MOPSO-TVAC , SMPSO, MOPSO-CDR and MOPSO-DFR) and evolutionary algorithms (SPEA2 and MOEAD/DRA). The results from previous studies, made with single objective techniques, were inserted in the initial population of the algorithms and compared with those simulations with normal initialization. We observed that MOPSO-CDR outperformed the other algorithms in the test scenarios while, in some cases, MOPSO has also generated competitive results. |
Lascia un commento!