Transient Stability Emergency Control Using Asynchronous Parallel Mixed-Integer Pattern Search

Special protection schemes (SPSs) are designed to monitor system conditions and apply control to avoid system collapse when subjected to a disturbance. Transient stability emergency control (TSEC) is a practical and effective SPS mechanism that utilizes generator tripping and load shedding as control actions to maintain system synchronism during severe contingencies, but faces significant challenges when it comes to a real-world grid of significant scale and complexity. In this paper, considering the characteristic of different control measures, TSEC problem is modeled as a mixed-integer nonlinear optimization problem with differential algebraic equation constraints. The proposed mixed-integer pattern search approach is able to obtain an optimal strategy without explicit derivative information of various complex component models. Furthermore, asynchronous parallelism is exploited and implemented to improve work-load balance and accelerate search procedure. Case studies on three different scale systems, including a real-world electric grid in China demonstrate that the proposed approach is promising to obtain TSEC strategy for large-scale power systems.