Partition-Based Cooperative Decision-Making for High-Frequency Generator Tripping Control

To address the decline in system inertia and the risk of frequency instability resulting from high penetration of renewable energy, and to overcome the limitations of centralized control—such as high computational burden and slow response—as well as the lack of global coordination in decentralized control, this study proposes a cooperative decision-making strategy, based on system partitioning, for high-frequency generator tripping control. The method first combines spectral clustering with nodal frequency response correlation analysis to achieve dynamic system partitioning and the selection of characteristic monitoring nodes. A two-layer cooperative architecture consisting of zone controllers and a central controller is then established, in which the zone controllers are responsible for aggregating local information, while the central controller dynamically generates a zonal generator tripping priority sequence based on four indicators: regional power surplus ratio, equivalent inertia, frequency deviation, and Rate of Change of Frequency. A simulation study conducted on an actual provincial power grid in the northwest region with a renewable energy penetration rate of 31.1% showed that compared with the existing decentralized control strategy of the power grid, this method not only achieved better frequency recovery, but also reduced the generator tripping capacity by 6.29%. Compared with advanced centralized strategies, it reduces recovery time by 10% while achieving good frequency recovery. By aggregating regional information to reduce central computing load and implementing coordinated control between regions through a global optimization mechanism, this strategy provides an effective method for high-frequency safety control in power grids with high penetration rates of renewable energy.