Deterministic approach for solving multi-objective non-smooth Environmental and Economic dispatch problem

• Autor(a) principal: Gonçalves, Elis [UNESP]
• Data de Publicação: 2019
• Outros Autores: Balbo, Antonio Roberto [UNESP], da Silva, Diego Nunes, Nepomuceno, Leonardo [UNESP], Baptista, Edméa Cássia [UNESP], Soler, Edilaine Martins [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1016/j.ijepes.2018.07.056; http://hdl.handle.net/11449/188091
• topic: Economic/environmental dispatch; Interior/exterior-point; Modified logarithmic barrier function; Multi-objective programming; Smoothing functions methods

The Environmental and Economic Dispatch Problem with Valve-Point loading effect representation (EEDP-VP) is a multi-objective, nonconvex and non-differentiable optimization problem. Due to these difficulties, it has been solved in the literature mainly by heuristic approaches, while deterministic approaches are scarce. Therefore, the main objectives of this paper are to propose a deterministic approach for solving this problem and compare its solutions with the ones obtained by some heuristic and deterministic approaches. The deterministic approach proposed has the following features: the multi-objective nature of the problem is handled by the Progressive Bounded Constraints (PBC) strategy, while the modified logarithmic barrier function method is used to solve the subproblems resulting from the PBC strategy; a smoothing technique is used to handle non-differentiability issues, while the inertia correction strategy is used so that only descent directions are generated. The methodology is applied to five generation systems and the results show that the Pareto-curve is obtained more efficiently when compared to other heuristic and deterministic optimization approaches.