S. Kai, G.B. Shrestha, and L. Goel (Singapore)
Competitive power market, Market clearing, Scheduling and dispatching, Optimization, Linear Programming, Interior point method, Primal affine scaling algorithm.
Optimizing the operation of a deregulated power system on the basis of the bids submitted by the power producers and the customers involves price clearing in the market, which effectively determines scheduling and dispatch in the system. Adopting simplified models, all these constraints can be represented by linear relationships and therefore the scheduling at a particular hour can be optimized using linear programming techniques. Most markets are a day-ahead market and the participants submit the bids for 24 hours. Therefore the optimization over a period of time is a coupled LP optimization problem, which is very time consuming. Improvements in the primal affine scaling method to solve LP problem with bounded variables in two respects are presented in this paper. Mathematical formulations to simplify the Phase–I algorithm to obtain an initial feasible interior point as well as the potential push technique to obtain faster convergence are presented. The proposed techniques are applied to the scheduling problem in the IEEE 30-bus system. By adjusting the numbers of power producers (ie, generators), customers (ie, loads), and network branches in the system, the performance of the proposed improved techniques is evaluated for different sizes of the system. The computational time for one stage LP optimization with and without the proposed modification is compared and it is shown that the reduction in computational time increases with the in crease in the size of the power system. It is also shown that the computation time for coupled 8-hour scheduling using dynamic programming with restricted search path is also significantly reduced when the proposed improvements are adopted.
Important Links:
Go Back
