A Time Decomposition and Coordination Strategy for Power System Multi-Interval Operation

Abstract

This paper presents a time decomposition strategy to reduce the computational complexity of power system multi-interval operation problems. We focus on the economic dispatch problem. The considered scheduling horizon is decomposed into multiple smaller sub-horizons. The first time interval of each sub-horizon is modeled as the coupling interval between two consecutive sub-horizons. The interdependencies between the sub-horizons are mathematically modeled using ramp rates of generating units. A distributed coordination strategy, which is based on auxiliary problem principle, is developed to coordinate the economic dispatch solutions of the sub-horizons to find an optimal solution for the whole operation horizon. We also propose an initializing technique to start the iterative coordination algorithm from a good-enough point. This technique enhances the convergence rate significantly. The proposed algorithm is deployed to solve a week-ahead economic dispatch problem on the IEEE 118-bus system, and promising results are obtained.