A Fast Calculation Method for N-1 Security-Constrained Economic Dispatch via Low-Rank Approximation Surrogate Model

Abstract

With the increase in the proportion of renewable energy connected to the grid, a large number of N-1 security constraints should be considered in the security-constrained economic dispatch (SCED) to ensure the reliable operation of the power system, which causes a great computational burden. However, only a small number of constraints in the N-1 security constraints are contributing factors. Therefore, the elimination of redundant constraints can significantly improve the solution efficiency of the SCED model. First, a new SCED model considering wind power and photovoltaic is established, and a low-rank approximation (LRA) surrogate model is constructed based on the historical operation information of the SCED model. Then, the active constraints are identified based on the estimation results of the LRA surrogate model and the active constraint set is constructed. Next, an iterative solution process of the SCED model based on LRA surrogate model is proposed. Finally, the simulations are conducted on an IEEE 39-bus system. The simulation results show that the error between the solution results of the LRA surrogate and the SCED models is less than 10%, the active constraint identification accuracy is raised, and the average iterative solution time of the proposed solution process is reduced by more than 50%, which indicates the improvement of the solution efficiency of the SCED model.

Keywords

• security-constrained economic dispatch (sced)
• low-rank approximation (lra)
• constraints identification
• active constraint set