Corrective model-predictive control in large electric power systems

Abstract

Enhanced control capabilities are required to coordinate the response of increasingly diverse controllable resources, including FACTS devices, energy storage, demand response and fast-acting generation. Model-predictive control (MPC) has shown great promise for accommodating these devices in a corrective control framework that exploits the thermal overload capability of transmission lines and limits detrimental effects of contingencies. This work expands upon earlier implementations by incorporating voltage magnitudes and reactive power into the system model utilized by MPC. These improvements provide a more accurate prediction of system behavior and enable more effective control decisions. Performance of this enhanced MPC strategy is demonstrated using a model of the Californian power system containing 4259 buses. Sparsity in modelling and control actions must be exploited for implementation on large networks. A method is developed for identifying the set of controls that is most effective for a given contingency. The proposed MPC corrective control algorithm fits naturally within energy management systems where it can provide feedback control or act as a guide for system operators by identifying beneficial control actions across a wide range of devices.