A solution for unit commitment using Lagrangian relaxation combined with evolutionary programming

This paper proposes an approach which combines Lagrangian relaxation princi ple and evolutionary programming for short-term thermal unit commitment. Un it commitment is a complex combinatorial optimization problem which is diff icult to be solved for large-scale power systems. Up to now, the Lagrangian relaxation is considered the best to deal with large-scale unit commitment although it cannot guarantee the optimal solution. In this paper, an evolu tionary programming algorithm is used to improve a solution obtained by the Lagrangian relaxation method: Lagrangian relaxation gives the starting poi nt for a evolutionary programming procedure. The proposed algorithm takes t he advantages of both methods and therefore it can search a better solution within short computation time. Numerical simulations have been carried out on two test systems of 30 and 90 thermal units power systems over a 24-hou r periods. (C) 1999 Elsevier Science S.A. All rights reserved.