Demand-Response-Based Distributed Preventive Control to Improve Short-Term Voltage Stability

Abstract

Short-term voltage stability (STVS) issues pose a significant threat to modern power systems. Based on extensive observations of STVS events, their primary predictors can be summarized as summer peak load levels with a high percentage of air conditioners, and occurrence of faults caused by thunderstorms or other emergencies. The predictors can be used as the triggering signals for preventive voltage control (PVC). Previous design of PVC dominantly focuses on the improvement of long-term voltage stability, and demand response (DR) resources are not fully utilized. In this paper, a DR-based PVC method is proposed to improve STVS if the day-ahead forecast indicates that the system will experience high STVS risk. We formulate PVC as an optimal power flow problem and solve it in a distributed manner, where the central controller (CC) and the local controllers jointly compute an optimal schedule. Thus, the computational burden of the CC is reduced compared with the existing centralized algorithms, and customer privacy is well protected. The simulation results not only verify the advantage of the improved index, but also demonstrate the fast convergence and effectiveness of the distributed PVC, which, therefore, offers an alternative approach to improve STVS.