Application of Comprehensive Learning Particle Swarm Optimization to Least Cost Generation Expansion Planning Problem with Solar Plant

Generation Expansion Planning (GEP) is a challenging problem as both supply and demand for energy have temporal and spatial variations. It also involves integration of system elements with a complex mix of alternative candidate plants having differing physical and production capabilities and characteristics. The integration of all such elements in a system framework makes the GEP a large-scale, long-term, non-linear, mixed-variable mathematical modeling problem. The accurate solution of such realistic models is essential to create an efficient and economic power system.In recent years a number of solution methodologies have been proposed to solve the models and to address the efficiency issue. The Comprehensive Learning Particle Swarm Optimization (CL-PSO) algorithm is drawing more and more attention of the researchers and it has been applied in the solution of GEP in many recent studies. The aim of this study is the GEP for the candidate system, integrating all critical system elements leading to the formulation of a realistic mathematical system and the employment of GEP in the model solutions. It also demonstrates the effectiveness of CL-PSO algorithm in finding efficient solutions to the identified problem. The test system planning is carried out for two different planning horizons of 6 and 14 years respectively. As the system is expected to get an increasing proportion of solar power plants, in future, a special focus is given to study the impact of such increase. The resulting variations in different cost components and the variations in reliability indices are also reported.