A Simplification Method for Robust Optimization of Power System Based on LMP

Abstract

The problem of robust DCOPF with the uncertainty from load and renewable power is considered. Intuitively, high load power and low renewable power output tend to increase the cost of controllable generators. Thus in most cases, the worst case of load power takes its upper bound and the worst case of renewable power takes its lower bound. As a result, the number of uncertain variables can be reduced. However, there are also counter-examples where the worst case is the opposite case. To study which uncertain variables can be reduced to deterministic ones, we summarize the connection between locational marginal price (LMP) and the worst case of uncertain variables. A simplification method based on LMP is proposed to reduce the number of uncertain variables. Simulations on 14bus, 118-bus, and 300-bus test systems have demonstrated the effectiveness of the proposed approach in reducing computation time.