Thermal Unit Commitment Using Hybrid Particle Swarm Optimization

Abstract

In operational planning of power system, unit commitment problem (UCP) is important optimizing task. The UCP means to determine the optimal schedule of start-up or shutdown of generating units for scheduled period to meet the load demand. Unit commitment problem is two decision process of determining the commit/decommit of generating units and economic dispatch of electric power providing the satisfaction of all operating constraints i.e. power equality and inequality constraints and all other operating constraints so that operating cost be minimized. The unit commitment problem is multimodal and non-convex problem. Various search technique has been applied to find optimal solution. In this thesis work hybrid particle swarm optimization has been applied to solve UCP. The particle swarm optimization has number of advantage but still it shows slow convergence characteristic near global optimal solution. In the proposed HPSO the property of simulated annealing for hill climbing has been hybridized with PSO. The proposed algorithm helps to explore the search area more effectively. Finally the HPSO algorithm is tested on two UCP and results has been compared with results reported in the literature.