A New Hybrid Approach for Unit Commitment Using Lagrangian Relaxation Combined with Evolutionary and Quadratic Programming

This paper proposes an approach which combines Lagrangian relaxation principle with evolutionary programming and quadratic programming (LREQP) for short-term thermal unit commitment. Unit commitment is a complex combinatorial optimization problem which is difficult to be solved for large-scale power systems. This paper presents a hybrid method for ramp rate constrained unit commitment (RUC) problem. Hybrid LREQP minimizes the total supply cost subject to the power balance, generation ramp limit constraints, on/off line minimum level constraints, minimum up and new down time constraints and generator operating constraints. The proposed solution method solves unit commitment problem with two coordination procedures. In the first procedure, an evolutionary programming algorithm is used to improve a solution obtained by the Lagrangian relaxation method: Lagrangian relaxation gives the starting point for a evolutionary programming procedure. In the second procedure, economic dispatch (ED) by quadratic programming (QP) is performed 24 hours to simultaneously dispatch output subject to all constraints and unit decommitment is carried out. Hybrid LREQP is tested on the 26 unit IEEE reliability test system. The proposed algorithm takes the advantages of both methods and therefore it can search a better solution within short computation time.