Enhanced techniques on sequential unit commitment with interchange transactions

This paper presents two enhanced techniques for improving the solution optimality and computation performance of the sequential unit commitment (SC) method with interchange transactions. The conventional SC method, although often presenting superior performance over other methods, can lead to nonnear-optimal solutions in some circumstances due to the use of a local decision scheme to identify the best unit to be committed at each stage. The proposed technique, instead, uses a global-like decision scheme. It defines a small set of locally advantageous units which are individually examined globally by generating tentative commitment schedules to identify the globally best unit to commit at each stage. Studies have shown that the global-like decision scheme can effectively improve the solution optimality. Meanwhile, while an interchange transaction is incorporated with the unit commitment study, the constant transaction price often causes solution oscillation during iterations. A varying-/spl lambda/ technique is proposed in this paper. This technique properly models the impact of the interchange transactions on the power system hourly energy prices and, hence, successfully overcomes the oscillation problem such that the loading level of a transaction can be optimally determined similarly as for a generating unit.