Two-stagemodelpredictivecontrolforvoltagecollapse.PDFVIP

1 Two-Stage Model Predictive Control for Voltage Collapse Prevention Bo Gong Ian A. Hiskens Abstract—The paper proposes a two-stage model predictive are either insuf?cient to stabilize the system or that have quite control (MPC) strategy for alleviating voltage collapse. The ?rst large social cost. De?ciencies in various static approaches stage uses a static load shedding algorithm to obtain stable were discussed in [5]. MPC prediction simulations. The second stage uses the MPC simulations, and associated trajectory sensitivities, to establish In recent years, industry has begun adopting online dynamic a linear program (LP) that optimizes the control action. This security assessment (DSA) tools, with explicit use of dynamic LP builds on trajectory approximations that are generated from models and time domain simulation. These tools provide the sensitivities. The ?rst stage prestabilizing process improves a good foundation for integrating online dynamic control the accuracy of the trajectory approximations, with consequent applications. Based on the DSA framework, online controls improvement in the LP optimization results. The control action determined by the LP is subsequently applied to the actual can be sought, coordinated and justi?ed for their longer term system. The paper discusses the integration of this MPC algo- impact to the systems. rithm into an energy management system (EMS) environment. Power systems are large scale nonlinear systems with nu- A standard 10 bus voltage collapse case is used to illustrate the merous physical limits (or constraints). Theoretically comput- performance of the overall control strategy. ing a closed loop optimal control strategy for these systems is impossible. Model predict