A Multiobjective Evolutionary Programming Algorithm and Its Applications to Power Generation Expansion Planning

Jose L. Ceciliano Meza,Mehmet Bayram Yildirim,Abu S. M. Masud,J.L.C. Meza,M.B. Yildirim,A.S.M. Masud
DOI: https://doi.org/10.1109/tsmca.2009.2025868
2009-09-01
Abstract:The generation expansion planning (GEP) problem is defined as the problem of determining WHAT, WHEN, and WHERE new generation units should be installed over a planning horizon to satisfy the expected energy demand. This paper presents a framework to determine the number of new generating units (e.g., conventional steam units, coal units, combined cycle modules, nuclear plants, gas turbines, wind farms, and geothermal and hydro units), power generation capacity for those units, number of new circuits on the network, the voltage phase angle at each node, and the amount of required imported fuel for a single-period generation expansion plan. The resulting mathematical program is a mixed-integer bilinear multiobjective GEP model. The proposed framework includes a multiobjective evolutionary programming algorithm to obtain an approximation of the Pareto front for the multiobjective optimization problem and analytical hierarchy process to select the best alternative. A Mexican power system case study is utilized to illustrate the proposed framework. Results show coherent decisions given the objectives and scenarios considered. Some sensitivity analysis is presented when considering different fuel price scenarios.
What problem does this paper attempt to address?