Partitioning method of reserve capacity based on spectral clustering considering wind power

With the rapid development of the power system, the complexity of power system continues to increase. The large-scale of clean energy increases the complexity of the power system structure, which makes the power system's operating status complex and changeable. To ensure the safety, stability, and economic operation of the power system, the power system needs to maintain a certain spinning reserve capacity and ensure that the reserve is available. The spinning reserve optimization method based on fixed partitions cannot effectively deal with the source-net-load volatility and the line congestion problems caused by wind power. To this aim, this paper proposes a power system reserve capacity partitioning algorithm based on the spectral clustering algorithm considering wind power. First, the energy-reserve joint optimization unit commitment problem is built. Secondly, considering the uncertainty of wind power output and the line N - 1 failure, a risk assessment method for line congestion is established; then the power transfer distribution factor (PTDF) of the congested line is considered as the similarity measure. A method based on the iterative partition method based on spectral clustering is proposed. Finally, the reserve capacity configuration based on the dynamic partition results is determined. The IEEE-118 bus system is used in this paper to verify the proposed method. The results show that the reserve capacity dynamic partition method proposed in this paper determines the number of partitions automatically. Reserve capacity partitions are built reasonably. After the reserve capacity partitions, the availability of spinning reserve capacity is effectively guaranteed.