Application of evolutionary computation techniques for generation expansion planning

This paper presents the application of evolutionary computation (EC) techniques, improved evolutionary strategies (IES), improved evolutionary programming (IEP) and improved genetic algorithm (IGA) to least cost generation expansion planning (GEP) problem. Least-cost GEP problem is a highly constrained nonlinear discrete dynamic optimization problem. Several conventional non-linear optimization methods have been used to solve the GEP problem. These methods may fail to provide global optimal due to involvement of discrete variables in the constraints. Recently EC techniques are used to solve the combinatorial optimization GEP problems, due to its global search characteristics. The GEP problem is illustrated for a synthetic reliable test system with 4, 6 and 14 years planning horizon. The results obtained using IES, EEP and IGA are verified using dynamic programming (DP) for 4 and 6 year planning horizon. The problem with 14 year planning horizon is simulated using IES, IEP and IGA.