A novel perspective for equivalent aggregation of wind farm: Measuring the dynamic similarity between output time-series

This paper proposes a time-series similarity-based equivalent modeling approach for large-scale wind farms, designed to improve accuracy while balance computational efficiency. The proposed method utilizes time-delay embedding to extract nonlinear temporal features from wind turbine output time-series data, achieving a dynamic similarity-based clustering process without the need for predefined indicators and metrics. To enhance model accuracy and robustness, an improved Geometric Template Matching algorithm is employed, incorporating the Frobenius norm for refined similarity measurement, particularly in sparse data scenarios. The proposed approach is validated through comparative studies against the Single-Unit Model and the Detailed Model using real-world data from three wind farms. Model performance is assessed, including accuracy, robustness, and generalization capability. Results demonstrate that the proposed method achieves superior performance and accurate equivalent modeling, maintaining reliability across diverse wind conditions. This work offers a scalable and efficient solution for wind farm aggregation, contributing to improved power system analysis and operational decision-making.